Sleisenger and Fordtran's Gastrointestinal and Liver Disease, 9th Edition

Sleisenger and Fordtran’s

Ninth Edition

Gastrointestinal and Liver Disease

pathophysiology/diagnosis/management Mark Feldman, MD

William O. Tschumy Jr., MD, Chair of Internal Medicine Director, Internal Medicine Residency Program Medical Director, Research Services Texas Health Presbyterian Hospital Dallas Clinical Professor of Internal Medicine University of Texas Southwestern Medical School Dallas, Texas

Lawrence S. Friedman, MD Professor of Medicine Harvard Medical School Professor of Medicine Tufts University School of Medicine Chair, Department of Medicine Newton-Wellesley Hospital Assistant Chief of Medicine Massachusetts General Hospital Boston, Massachusetts

Lawrence J. Brandt, MD

Professor of Medicine and Surgery Albert Einstein College of Medicine Emeritus Chief, Division of Gastroenterology Montefiore Medical Center Bronx, New York

1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899

SLEISENGER AND FORDTRAN’S GASTROINTESTINAL AND LIVER DISEASE: PATHOPHYSIOLOGY/ DIAGNOSIS/MANAGEMENT

ISBN: 978-1-4160-6189-2

Copyright © 2010, 2006, 2002, 1998, 1993, 1989, 1983, 1978, 1973 by Saunders, an imprint of Elsevier Inc. Chapters 32 and 106 are in the public domain. Mayo Clinic reserves all rights to original Mayo drawings in Chapter 120. All rights reserved.  No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permissions may be sought directly from Elsevier’s Rights Department: phone: (+1) 215 239 3804 (US) or (+44) 1865 843830 (UK); fax: (+44) 1865 853333; e-mail: [email protected]. You may also complete your request on-line via the Elsevier website at http://www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data Sleisenger and Fordtran’s gastrointestinal and liver disease : pathophysiology, diagnosis, management / [edited by] Mark Feldman, Lawrence S. Friedman, Lawrence J. Brandt.—9th ed.    p. ; cm.   Rev. ed. of: Sleisenger & Fordtran’s gastrointestinal and liver disease / edited by Mark Feldman, Lawrence S. Friedman, Lawrence J. Brandt. 8th ed. c2006.   Includes bibliographical references and index.   ISBN 978-1-4160-6189-2   1.  Gastrointestinal system—Diseases.  2.  Liver—Diseases.  I.  Sleisenger, Marvin H.  II.  Feldman, Mark, 1947-  III.  Friedman, Lawrence S. (Lawrence Samuel), 1953-  IV.  Brandt, Lawrence J.  V.  Sleisenger & Fordtran’s gastrointestinal and liver disease.  VI.  Title: Gastrointestinal and liver disease.   [DNLM: 1. Gastrointestinal Diseases. 2. Liver Diseases. WI 140 S632 2010]   RC801.G384 2010   616.3′3—dc22 2009014222 Acquisitions Editor: Druanne Martin Developmental Editor: Anne Snyder Publishing Services Manager: Frank Polizzano Project Manager: Jeff Gunning Project Management Assistance: Lee Ann Draud, Robin Hayward Design Direction: Steve Stave Printed in Canada Last digit is the print number:  9  8  7  6  5  4  3  2  1

Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org

This edition is dedicated to our grandchildren Noah and Jordan Feldman, Shayna and Olivia Feldgus, Christopher Friedman, Zachary and Noah Fishkind, and Chloe Jane King

Contributors Julian A. Abrams, MD

Assistant Professor of Clinical Medicine, Department of Medicine, Columbia University College of Physicians and Surgeons; Division of Digestive and Liver Diseases, Columbia University Medical Center; Cancer Epide­ miology Program, Herbert Irving Comprehensive Cancer Center, New York, New York Adenocarcinoma and Other Tumors of the Stomach

Nezam H. Afdhal, MD

Associate Professor of Medicine, Harvard Medical School; Attending Physician, Beth Israel Deaconess Medical Center, Boston, Massachusetts Gallstone Disease

Rakesh Aggarwal, MD, DM

Additional Professor, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India Hepatitis E

Karin L. Andersson, MD, MPH

Instructor in Medicine, Harvard Medical School; Staff Hepatologist, Gastrointestinal Unit, Massachusetts General Hospital, Boston, Massachusetts Acalculous Biliary Pain, Acalculous Cholecystitis, Cholesterolosis, Adenomyomatosis, and Polyps of the Gallbladder

Jane M. Andrews, MBBS, PhD, FRACP

Clinical Associate Professor, Department of Medicine, University of Adelaide Faculty of Health Sciences; Senior Consultant in Gastroenterology, Department of Gastroenterology and Hepatology, Royal Adelaide Hospital, Adelaide, South Australia, Australia Small Intestinal Motor and Sensory Function and Dysfunction

Paul Angulo, MD

Professor of Medicine and Section Chief, Hepatology, Division of Digestive Diseases and Nutrition, University of Kentucky Medical Center, Lexington, Kentucky Primary Biliary Cirrhosis

Fernando Azpiroz, MD, PhD

Professor of Medicine, Autonomous University of Barcelona; Chief, Department of Gastroenterology, University Hospital Vall d’Hebron, Barcelona, Spain Intestinal Gas

Bruce R. Bacon, MD

James F. King, MD, Endowed Chair in Gastroenterology and Professor of Internal Medicine, Saint Louis Univer­ sity School of Medicine; Director, Division of Gastro­ enterology and Hepatology, Saint Louis University Hospital, St. Louis, Missouri Hemochromatosis

Christina Wood Baker, PhD

Instructor in Psychiatry, Harvard Medical School; Depart­ ment of Psychiatry, Massachusetts General Hospital, Boston, Massachusetts Eating Disorders

William F. Balistreri, MD

Dorothy M. M. Kersten Professor of Pediatrics and Asso­ ciate Chair, Subspecialty Education, Department of Pediatrics, University of Cincinnati College of Medicine; Director Emeritus, Pediatric Liver Care Center, and Medical Director Emeritus, Liver Transplantation; Program Director, Fellowship in Transplant Hepatology, Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Other Inherited Metabolic Disorders of the Liver

Todd H. Baron, MD

Professor of Medicine, Mayo Clinic College of Medicine; Consultant, Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota Endoscopic Treatment of Pancreatic Disease; Endoscopic and Radiologic Treatment of Biliary Disease

Bradley A. Barth, MD, MPH

Assistant Professor of Pediatrics, University of Texas Southwestern Medical School; Attending Physician, Pediatric Gastroenterology, Children’s Medical Center, Dallas, Texas Anatomy, Histology, Embryology, and Developmental Anomalies of the Pancreas

Anne E. Becker, MD, PhD, ScM

Associate Professor of Psychiatry and Associate Professor of Medical Anthropology, Harvard Medical School; Director, Eating Disorders Clinical and Research Program, Massachusetts General Hospital, Boston, Massachusetts Eating Disorders

Alex S. Befeler, MD

Associate Professor of Internal Medicine and Medical Director for Liver Transplantation, Saint Louis Univer­ sity School of Medicine, St. Louis, Missouri Tumors and Cysts of the Liver

Kfir Ben-David, MD

Assistant Professor, University of Florida College of Medicine; Chief of Minimally Invasive Gastroeso­ phageal and Bariatric Service, Shands at University of Florida; Director of Bariatric Surgery, Malcolm Randall VA Medical Center, Gainesville, Florida Appendicitis

L. Ashley Blackshaw, PhD

Affiliate Professor of Medicine, University of Adelaide Faculty of Health Sciences; Principal Research Fellow, Royal Adelaide Hospital, Adelaide, South Australia, Australia Small Intestinal Motor and Sensory Function and Dysfunction

Boris Blechacz, MD, PhD

Instructor in Medicine, Mayo Clinic College of Medicine, Rochester, Minnesota Tumors of the Bile Ducts, Gallbladder, and Ampulla

vii

viii

Contributors Lawrence J. Brandt, MD

Professor of Medicine and Surgery, Albert Einstein College of Medicine; Emeritus Chief, Division of Gastroen­ terology, Montefiore Medical Center, Bronx, New York Vascular Lesions of the Gastrointestinal Tract; Intestinal Ischemia

George A. Bray, MD

Boyd Professor, Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, Louisiana Obesity

Robert S. Bresalier, MD, FACP

Professor of Medicine, Birdie J. and Lydia J. Resoft Distin­ guished Professor in Gastrointestinal Oncology, and Director, Gastrointestinal Cancer Research Laboratory, Department of Gastroenterology, Hepatology, and Nutri­ tion, University of Texas M. D. Anderson Cancer Center, Houston, Texas Colorectal Cancer

Robert S. Britton, PhD

Associate Research Professor, Division of Gastroenterol­ ogy and Hepatology, Saint Louis University School of Medicine; Saint Louis University Hospital, St. Louis, Missouri Hemochromatosis

Simon J. Brookes, MD

Professor of Human Physiology, Department of Human Physiology and Center for Neuroscience, Flinders Uni­ versity School of Medicine, Adelaide, South Australia, Australia Colonic Motor and Sensory Function and Dysfunction

Alan L. Buchman, MD, MSPH

Professor of Medicine and Surgery and Medical Director of Intestinal Rehabilitation/Transplant Center, Division of Gastroenterology, Northwestern University Feinberg School of Medicine, Chicago, Illinois Short Bowel Syndrome

J. Steven Burdick, MD

Associate Professor of Medicine, University of Texas Southwestern Medical School; Staff Physician and Director of Endoscopy, Parkland Health and Hospital System; Staff Physician, Zale Lipshy University Hospi­ tal and St. Paul University Hospital, Dallas, Texas Anatomy, Histology, Embryology, and Developmental Anomalies of the Pancreas

Robert L. Carithers, Jr., MD

Professor of Medicine, University of Washington School of Medicine; Director, Liver Care Line, and Medical Director, Liver Transplant Program, University of Washington Medical Center, Seattle, Washington Alcoholic Liver Disease

Julie G. Champine, MD

Professor of Radiology, University of Texas Southwestern Medical School; Chief of Radiology, Parkland Health and Hospital System, Dallas, Texas Abdominal Abscesses and Gastrointestinal Fistulas

Francis K. L. Chan, MD

Professor of Medicine, Chinese University of Hong Kong; Chief of Gastroenterology and Hepatology, Department of Medicine and Therapeutics, Prince of Wales Hospital, Hong Kong, China Treatment of Peptic Ulcer Disease

Joseph G. Cheatham, MD

Instructor in Clinical Medicine, Uniformed Services Uni­ versity for Health Sciences, Bethesda, Maryland; Fellow in Gastroenterology, Walter Reed Army Medical Center, Washington, DC Hepatitis A

Shivakumar Chitturi, MD, FRACP

Senior Staff Specialist in Gastroenterology and Hepatol­ ogy, Canberra Hospital, Canberra, Australian Capital Territory, Australia Liver Disease Caused by Drugs

Daniel C. Chung, MD

Associate Professor of Medicine, Harvard Medical School; Director, Gastrointestinal Cancer Genetics Service, Massachusetts General Hospital, Boston, Massachusetts Cellular Growth and Neoplasia

Raymond T. Chung, MD

Associate Professor of Medicine, Harvard Medical School; Director of Hepatology and Medical Director, Liver Transplant Program, Massachusetts General Hospital, Boston, Massachusetts Bacterial, Parasitic, and Fungal Infections of the Liver, Including Liver Abscess

Robert R. Cima, MD

Associate Professor of Surgery, Mayo Clinic College of Medicine; Consultant in Colon and Rectal Surgery, Mayo Clinic, Rochester, Minnesota Ileostomy, Colostomy, and Pouches

Robert H. Collins, Jr., MD

Professor of Internal Medicine and Medical Oncology, University of Texas Southwestern Medical Center, Dallas, Texas Gastrointestinal Lymphomas

Ian J. Cook, MD, FRACP

Conjoint Professor of Medicine, Department of Medicine, University of New South Wales Faculty of Medicine; Senior Staff Specialist in Gastroenterology and Director, Gastroenterology Department, St. George Hospital, Sydney, New South Wales, Australia Colonic Motor and Sensory Function and Dysfunction

Diane W. Cox, PhD, FCCMG, FRSC

Professor of Medical Genetics, University of Alberta Faculty of Medicine, Edmonton, Alberta, Canada Wilson Disease

Sheila E. Crowe, MD

Professor of Medicine, Division of Gastroenterology and Hepatology, University of Virginia School of Medicine, Charlottesville, Virginia Helicobacter pylori

Albert J. Czaja, MD

Professor of Medicine, Mayo Clinic College of Medicine, Rochester, Minnesota Autoimmune Hepatitis

Brian G. Czito, MD

Associate Professor of Radiation Oncology, Duke Univer­ sity School of Medicine, Durham, North Carolina Radiation Injury

Contributors Ananya Das, MD, DM, FASGE

Professor of Medicine, Mayo Clinic College of Medicine; Associate Chair and Director of Endoscopy, Department of Medicine, Mayo Clinic Arizona, Scottsdale, Arizona Tumors of the Esophagus

Fredric Daum, MD

Professor of Pediatrics and Clinical Scholar of Medicine, Stony Brook University Medical Center School of Medicine, Stony Brook; Chief, Division of Pediatric Gastroenterology, Hepatology, and Nutrition, Winthrop University Hospital, Mineola, New York Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine

Gary L. Davis, MD

Director, General and Transplant Hepatology, Baylor University Medical Center, Dallas, Texas Hepatitis C

Paul A. Dawson, PhD

Professor, Department of Internal Medicine, Section of Gastroenterology, Wake Forest University School of Medicine, Winston-Salem, North Carolina Bile Secretion and the Enterohepatic Circulation

Mark H. DeLegge, MD

Professor of Medicine, Medical University of South Carolina, Charleston, South Carolina Nutrition in Gastrointestinal Diseases

George D. Demetri, MD

Associate Professor of Medicine, Harvard Medical School; Director, Ludwig Center at Dana-Farber/Harvard Cancer Center, Boston, Massachusetts Gastrointestinal Stromal Tumors (GISTs)

Kenneth R. DeVault, MD

Professor of Medicine, Mayo Clinic College of Medicine; Chair, Division of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, Florida Symptoms of Esophageal Disease

Adrian M. Di Bisceglie, MD, FACP

Professor of Internal Medicine, Saint Louis University School of Medicine, St. Louis, Missouri Tumors and Cysts of the Liver

Philip G. Dinning, PhD

Research Fellow, Department of Medicine, University of New South Wales Faculty of Medicine, Sydney, New South Wales, Australia Colonic Motor and Sensory Function and Dysfunction

Iris Dotan, MD

Lecturer, Sackler School of Medicine; Head, IBD Center, Department of Gastroenterology and Liver Diseases, Sovrasky Medical Center, Tel Aviv, Israel Mucosal Immunity

Douglas A. Drossman, MD

Professor of Medicine and Psychiatry and Co-Director, UNC Center for Functional GI and Motility Disorders, Division of Gastroenterology and Hepatology, Univer­ sity of North Carolina School of Medicine; Attending Physician, UNC Hospitals, Chapel Hill, North Carolina Biopsychosocial Issues in Gastroenterology

David E. Elliott, MD, PhD

Professor and Director, Division of Gastroenterology and Hepatology, University of Iowa Carver College of Medicine, Iowa City, Iowa Intestinal Infections by Parasitic Worms

B. Joseph Elmunzer, MD

Clinical Lecturer in Internal Medicine, University of Michigan Medical School, Ann Arbor, Michigan Biliary Tract Motor Function and Dysfunction

Grace H. Elta, MD

Professor of Medicine, University of Michigan Medical School, Ann Arbor, Michigan Biliary Tract Motor Function and Dysfunction

Silvia Degli Esposti, MD

Associate Clinical Professor of Medicine and Director, Center for Gastrointestinal Services, Alpert Medical School of Brown University, Providence, Rhode Island Gastrointestinal and Hepatic Disorders in the Pregnant Patient

Michael B. Fallon, MD

Professor of Medicine, Dan and Lillie Sterling Professor of Gastroenterology, and Director, Division of Gastroenter­ ology, Hepatology, and Nutrition, University of Texas Health Science Center at Houston Medical School; Chief of Service, Gastroenterology and Hepatology, Memorial Hermann Medical Center, Houston, Texas Hepatic Encephalopathy, Hepatorenal Syndrome, Hepatopulmonary Syndrome, and Systemic Complications of Liver Disease

Geoffrey C. Farrell, MD, FRACP

Professor of Hepatic Medicine, Australian National University Medical School; Gastroenterology and Hepatology Unit, Canberra Hospital, Canberra, Austra­ lian Capital Territory, Australia Liver Disease Caused by Drugs

James J. Farrell, MD

Associate Professor of Medicine, David Geffen School of Medicine at UCLA; Director, Pancreaticobiliary Endos­ copy, UCLA Medical Center, Los Angeles, California Digestion and Absorption of Nutrients and Vitamins

Richard J. Farrell, MD

Assistant Professor of Medicine, Harvard Medical School; Associate Physician, Gastroenterology Division, Beth Israel Deaconess Medical Center, Boston, Massachusetts Celiac Disease and Refractory Celiac Disease

Jordan J. Feld, MD, MPH

Assistant Professor of Medicine, University of Toronto Faculty of Medicine; Hepatologist, University Health Network, Toronto Western Hospital, Toronto, Ontario, Canada Hepatitis Caused by Other Viruses

Mark Feldman, MD

William O. Tschumy Jr., MD, Chair of Internal Medicine, Director of Internal Medicine Residency Program, and Medical Director of Research Services, Texas Health Presbyterian Hospital Dallas; Clinical Professor of Inter­ nal Medicine, University of Texas Southwestern Medical School, Dallas, Texas Gastritis and Gastropathies

ix

x

Contributors Carlos Fernández-del Castillo, MD

Associate Professor of Surgery, Harvard Medical School; Director, Pancreas and Biliary Surgery Program, Massachusetts General Hospital, Boston, Massachusetts Tumors of the Pancreas

Lincoln E. Ferreira, MD, PhD

Director, Digestive Endoscopy Unit, University Hospital, Universidade Federal de Juiz de Fora, Juiz de Fora, Brazil Endoscopic Treatment of Pancreatic Disease

Paul Feuerstadt, MD

Gastroenterology Fellow, Albert Einstein College of Medi­ cine; Montefiore Medical Center, Bronx, New York Intestinal Ischemia

Robert J. Fontana, MD

Associate Professor of Medicine and Medical Director of Liver Transplantation, University of Michigan Medical School, Ann Arbor, Michigan Acute Liver Failure

Chris E. Forsmark, MD

Professor of Medicine and Chief, Division of Gastroenter­ ology, Hepatology, and Nutrition, University of Florida College of Medicine, Gainesville, Florida Chronic Pancreatitis

Jeffrey M. Fox, MD, MPH

Assistant Clinical Professor of Medicine, Division of Gas­ troenterology, University of California, San Francisco, School of Medicine, San Francisco; Chief, Department of Gastroenterology, Kaiser Permanente, San Rafael, California Diverticular Disease of the Colon

Amy E. Foxx-Orenstein, DO

Associate Professor of Medicine, Mayo Clinic College of Medicine; Consultant, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota Ileus and Pseudo-obstruction

Frank K. Friedenberg, MD

Professor of Medicine, Temple University School of Medi­ cine; Attending Physician, Temple University Hospital, Philadelphia, Pennsylvania Gastroesophageal Reflux Disease

Lawrence S. Friedman, MD

Professor of Medicine, Harvard Medical School; Professor of Medicine, Tufts University School of Medicine, Boston; Chair, Department of Medicine, NewtonWellesley Hospital, Newton; Assistant Chief of Medicine, Massachusetts General Hospital, Boston, Massachusetts Chronic Abdominal Pain; Acalculous Biliary Pain, Acalculous Cho­ lecystitis, Cholesterolosis, Adenomyomatosis, and Polyps of the Gallbladder

Ralph A. Gianella, MD

Mark Brown Professor of Medicine, University of Cincin­ nati College of Medicine; Attending Physician, Univer­ sity Hospital, Cincinnati, Ohio Infectious Enteritis and Proctocolitis and Bacterial Food Poisoning

Gregory G. Ginsberg, MD

Professor of Medicine, University of Pennsylvania School of Medicine; Director of Endoscopic Services, Gastroen­ terology Division, University of Pennsylvania Health System, Philadelphia, Pennsylvania Foreign Bodies, Bezoars, and Caustic Ingestions

Robert E. Glasgow, MD

Associate Professor, Department of Surgery, University of Utah School of Medicine, Salt Lake City, Utah Treatment of Gallstone Disease

Gregory J. Gores, MD

Professor of Medicine, Mayo Clinic College of Medicine; Chair, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota Tumors of the Bile Ducts, Gallbladder, and Ampulla

David A. Greenwald, MD

Associate Professor of Clinical Medicine, Albert Einstein College of Medicine; Associate Division Director, Gastroenterology, and Fellowship Program Director, Montefiore Medical Center, Bronx, New York Protein-Losing Gastroenteropathy

Heinz F. Hammer, MD

Associate Professor of Internal Medicine and Gastroenter­ ology, Medical University Graz; Attending Gastroenter­ ologist, Division of Gastroenterology and Hepatology, Medical University Graz, Graz, Austria Maldigestion and Malabsorption

William V. Harford, Jr., MD

Professor of Internal Medicine, University of Texas South­ western Medical School; Director, Gastrointestinal Endoscopy, and Staff Physician, Veterans Affairs Medical Center, Dallas, Texas Diverticula of the Pharynx, Esophagus, Stomach, and Small Intestine; Abdominal Hernias and Gastric Volvulus

David J. Hass, MD

Clinical Instructor, Yale University School of Medicine; Clinical Faculty, Yale-New Haven Hospital, New Haven, Connecticut Complementary and Alternative Medicine

E. Jenny Heathcote, MB BS, MD, FRCP, FRCP(C)

Professor of Medicine, University of Toronto Faculty of Medicine; Hepatologist, University Health Network, Toronto Western Hospital, Toronto, Ontario, Canada Hepatitis Caused by Other Viruses

Maureen Heldmann, MD

Associate Professor of Radiology, Louisiana State University Health Sciences Center School of Medicine Shreveport; Director, Body CT and MRI, Louisiana State University Health Sciences Center, Shreveport, Louisiana Intestinal Obstruction

Christoph Högenauer, MD

Associate Professor of Internal Medicine, Medical University Graz; Attending Gastroenterologist, Division of Gastroenterology and Hepatology, Medical University Graz, Graz, Austria Maldigestion and Malabsorption

Christopher D. Huston, MD

Associate Professor, Department of Medicine and Depart­ ment of Microbiology and Molecular Genetics, Univer­ sity of Vermont College of Medicine; Attending Physician, Fletcher Allen Health Care, Burlington, Vermont Intestinal Protozoa

Steven H. Itzkowitz, MD

Professor of Medicine and Director, GI Fellowship Program, Mount Sinai School of Medicine, New York, New York Colonic Polyps and Polyposis Syndromes

Contributors Rajeev Jain, MD

Clinical Assistant Professor of Medicine, University of Texas Southwestern Medical School; Chief of Gastroen­ terology, Texas Health Presbyterian Hospital Dallas, Dallas, Texas Gastrointestinal and Hepatic Manifestations of Systemic Diseases

Dennis M. Jensen, MD

Professor of Medicine, David Geffen School of Medicine at UCLA and Veterans Affairs Greater Los Angeles Healthcare System; Staff Physician, UCLA Medical Center and West Los Angeles Veterans Affairs Medical Center, Los Angeles, California Gastrointestinal Bleeding

Robert T. Jensen, MD

Chief, Cell Biology Section, Digestive Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland Endocrine Tumors of the Pancreas and Gastrointestinal Tract

D. Rohan Jeyarajah, MD, FACS

Director of Surgical Oncology and Hepatobiliary Fellow­ ship Program, Methodist Dallas Medical Center, Dallas, Texas Diverticula of the Pharynx, Esophagus, Stomach, and Small Intestine; Abdominal Hernias and Gastric Volvulus

Ramon E. Jimenez, MD

Assistant Professor of Surgery, University of Connecticut Medical School, Farmington; Associate Attending Phy­ sician, Hartford Hospital, Hartford, Connecticut Tumors of the Pancreas

Ellen Kahn, MD

Professor of Pathology and Pediatrics, New York Univer­ sity School of Medicine, New York; Director of Labora­ tories, North Shore Gastroenterology Associates PC, Great Neck, New York Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine

Peter J. Kahrilas, MD

Gilbert H. Marqvardt Professor in Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois Esophageal Neuromuscular Function and Motility Disorders

Patrick S. Kamath, MD

Professor of Medicine, Mayo Clinic College of Medicine, Rochester, Minnesota Portal Hypertension and Gastrointestinal Bleeding

David A. Katzka, MD

Professor of Medicine, University of Pennsylvania School of Medicine; Division of Gastroenterology, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Esophageal Disorders Caused by Medications, Trauma, and Infection

Jonathan D. Kaunitz, MD

Professor of Medicine, David Geffen School of Medicine at UCLA; Staff Physician, West Los Angeles Veterans Affairs Medical Center, Los Angeles, California Gastric Secretion

Ciarán P. Kelly, MD

Professor of Medicine, Harvard Medical School; Chief, Blumgart Internal Medicine Firm, and Director, Gastro­ enterology Fellowship Training, Beth Israel Deaconess Medical Center, Boston, Massachusetts Celiac Disease and Refractory Celiac Disease; Antibiotic-Associated Diarrhea, Pseudomembranous Enterocolitis, and Clostridium difficile-Associated Diarrhea and Colitis

Seema Khan, MB BS

Associate Professor of Pediatrics, Jefferson Medical College of Thomas Jefferson University, Philadelphia, Pennsyl­ vania; Pediatric Gastroenterologist, Nemours/Alfred I. duPont Hospital for Children, Wilmington, Delaware Eosinophilic Disorders of the Gastrointestinal Tract

Arthur Y. Kim, MD

Assistant Professor of Medicine, Harvard Medical School; Assistant in Medicine, Infectious Diseases Unit, Massachusetts General Hospital, Boston, Massachusetts Bacterial, Parasitic, and Fungal Infections of the Liver, Including Liver Abscess

Michael B. Kimmey, MD

Clinical Professor of Medicine, University of Washington School of Medicine, Seattle; President, Tacoma Diges­ tive Disease Center, Tacoma, Washington Complications of Gastrointestinal Endoscopy

Kenneth L. Koch, MD

Professor of Medicine and Chief, Section on Gastro­ enterology, and Director, Digestive Health Center, Wake Forest University School of Medicine, Winston-Salem, North Carolina Gastric Neuromuscular Function and Neuromuscular Disorders

Kris V. Kowdley, MD

Clinical Professor of Medicine and Director, Center for Liver Disease, Virginia Mason Medical Center, Seattle, Washington Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis

Krzysztof Krawczynski, MD, PhD

Distinguished Consultant and Team Leader, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis E

Robert C. Kurtz, MD

Professor of Clinical Medicine, Weill Medical College of Cornell University; Chief, Gastroenterology-Nutrition Service, and Attending Physician, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York Tumors of the Small Intestine

J. Thomas Lamont, MD

Professor of Medicine, Harvard Medical School; Chief of Gastroenterology, Beth Israel Deaconess Medical Center, Boston, Massachusetts Antibiotic-Associated Diarrhea, Pseudomembranous Enterocolitis, and Clostridium difficile-Associated Diarrhea and Colitis

Charles S. Landis, MD, PhD

Fellow in Transplant Hepatology, Washington, Seattle, Washington Vascular Lesions of the Gastrointestinal Tract

University

of

xi

xii

Contributors Anne M. Larson, MD, FACP

Associate Professor of Internal Medicine and Medical Director of Liver Transplant Program, University of Texas Southwestern Medical Center, Dallas, Texas Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation

James Y. W. Lau, MD

Professor of Surgery, Chinese University of Hong Kong; Director of Endoscopy, Endoscopy Centre, Prince of Wales Hospital, Hong Kong, China Treatment of Peptic Ulcer Disease

Edward L. Lee, MD

Professor and Chair, Department of Pathology, Howard University College of Medicine; Howard University Hospital, Washington, DC Gastritis and Gastropathies

Anthony J. Lembo, MD

Associate Professor of Medicine, Harvard Medical School; Physician, Beth Israel Deaconess Medical Center, Boston, Massachusetts Constipation

Mike A. Leonis, MD, PhD

Assistant Professor of Pediatrics, University of Cincinnati College of Medicine; Staff Physician, Division of Gastroenterology, Hepatology, and Nutrition, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Other Inherited Metabolic Disorders of the Liver

Michael D. Levitt, MD

Professor of Medicine, University of Minnesota Medical School; Staff Physician, Minneapolis Veterans Affairs Medical Center, Minneapolis, Minnesota Intestinal Gas

James H. Lewis, MD, FACP, FACG

Professor of Medicine, Georgetown University School of Medicine; Director of Hepatology, Georgetown Univer­ sity Hospital, Washington, DC Liver Disease Caused by Anesthetics, Toxins, and Herbal Preparations

Hsiao C. Li, MD

Assistant Professor of Internal Medicine–Hepatology/ Oncology, University of Texas Southwestern Medical School, Dallas, Texas Gastrointestinal Lymphomas

Gary R. Lichtenstein, MD

Professor of Medicine, University of Pennsylvania School of Medicine; Director, Center for Inflammatory Bowel Diseases, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Ulcerative Colitis

Rodger A. Liddle, MD

Professor of Medicine, Duke University School of Medicine; Chief, Division of Gastroenterology, Duke University Hospital, Durham, North Carolina Gastrointestinal Hormones and Neurotransmitters

Steven D. Lidofsky, MD, PhD

Professor of Medicine and Pharmacology, University of Vermont College of Medicine; Director of Hepatology, Fletcher Allen Health Care, Burlington, Vermont Jaundice

Keith D. Lindor, MD

Professor of Medicine, Mayo Clinic College of Medicine, Rochester, Minnesota Primary Biliary Cirrhosis

Caroline Loeser, MD

Assistant Professor of Medicine, Section of Digestive Dis­ eases, Yale University School of Medicine; Attending Physician, Yale–New Haven Hospital, New Haven, Connecticut Ulcers of the Small and Large Intestine

John D. Long, MD

Associate Professor of Medicine, Wake Forest University School of Medicine; Director, GI Neuromuscular Disor­ ders Program, Wake Forest University Baptist Medical Center, Winston-Salem, North Carolina Anatomy, Histology, Embryology, and Developmental Anomalies of the Esophagus

Mark E. Lowe, MD, PhD

Professor of Pediatrics, University of Pittsburgh School of Medicine; Chief, Pediatric Gastroenterology, Hepatol­ ogy, and Nutrition, Children’s Hospital of Pittsburgh of UPMC, Pittsburgh, Pennsylvania Hereditary, Familial, and Genetic Disorders of the Pancreas and Pancreatic Disorders in Childhood

Emmy Ludwig, MD

Assistant Professor of Clinical Medicine, Weill Medical College of Cornell University; Assistant Attending Phy­ sician, Gastroenterology-Nutrition Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, New York Tumors of the Small Intestine

Matthias Maiwald, MD

Associate Professor of Medical Microbiology, Department of Microbiology and Infectious Diseases, Flinders Uni­ versity School of Medicine, Northern Territory Clinical School; Director of Microbiology, Department of Pathol­ ogy, Royal Darwin Hospital, Tiwi, Northern Territory, Australia Whipple’s Disease

Carolina Malagelada, MD

Research Associate, Autonomous University of Barcelona; Staff Physician, Digestive Diseases Department, Vall d’Hebron University Hospital, Barcelona, Spain Nausea and Vomiting

Juan-R. Malagelada, MD

Professor, Autonomous University of Barcelona; Chair, Digestive Diseases Department, Vall d’Hebron Univer­ sity Hospital, Barcelona, Spain Nausea and Vomiting

Peter W. Marcello, MD, FACS, FASCRS

Vice Chair, Department of Colon and Rectal Surgery, Lahey Clinic, Burlington, Massachusetts Diseases of the Anorectum

Lawrence A. Mark, MD, PhD

Assistant Professor of Dermatology and Charles W. Lewis Investigator, Indiana University School of Medicine; Wishard Dermatology Chief, Wishard Health Services, Indianapolis, Indiana Oral Diseases and Oral-Cutaneous Manifestations of Gastrointestinal and Liver Disease

Contributors Paul Martin, MD

Professor of Medicine and Chief, Division of Hepatology, University of Miami Leonard M. Miller School of Medi­ cine, Miami, Florida Liver Transplantation

Joel B. Mason, MD

Professor of Medicine and Nutrition, Tufts University School of Medicine; Staff Physician, Divisions of Gas­ troenterology and Clinical Nutrition, Tufts University Medical Center, Boston, Massachusetts Nutritional Assessment and Management of the Malnourished Patient

Jeffrey B. Matthews, MD

Christian R. Holmes Professor and Chair, Department of Surgery, University of Cincinnati College of Medicine, Cincinnati, Ohio Surgical Peritonitis and Other Diseases of the Peritoneum, Mesentery, Omentum, and Diaphragm

Lloyd Mayer, MD

Professor and Chair, Immunology Institute; Chief, Henry Janowitz Division of Gastroenterology; and Chief, Divi­ sion of Clinical Immunology, Mount Sinai School of Medicine, New York, New York Mucosal Immunity

Craig J. McClain, MD

Professor of Medicine, Professor of Pharmacology and Toxicology, and Associate Vice President for Transla­ tional Research, University of Louisville School of Med­ icine; Chief of Gastroenterology, Louisville Veterans Affairs Medical Center, Louisville, Kentucky Alcoholic Liver Disease

George B. McDonald, MD

Professor of Medicine, University of Washington School of Medicine; Member and Head, Gastroenterology/ Hepatology Section, Fred Hutchinson Cancer Research Center, Seattle, Washington Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation

Frederick H. Millham, MD

Associate Clinical Professor of Surgery, Harvard Medical School; Chair, Department of Surgery, NewtonWellesley Hospital, Newton, Massachusetts Acute Abdominal Pain

Joseph P. Minei, MD, MBA

Professor and Chair, Division of Burn, Trauma, and Critical Care, Department of Surgery, University of Texas Southwestern Medical Center; Surgeon-in-Chief, Parkland Health and Hospital System, Dallas, Texas Abdominal Abscesses and Gastrointestinal Fistulas

Ginat W. Mirowski, DMD, MD

Adjunct Associate Professor, Indiana University School of Dentistry, Indianapolis, Indiana Oral Disease and Oral-Cutaneous Manifestations of Gastrointestinal and Liver Disease

Joseph Misdraji, MD

Assistant Professor of Pathology, Harvard Medical School; Assistant Pathologist, Massachusetts General Hospital, Boston, Massachusetts Embryology, Anatomy, Histology, and Developmental Anomalies of the Liver

John Morton, MD, MPH

Associate Professor of Surgery, Director of Bariatric Surgery and Surgical Quality, and Section Chief of Mini­ mally Invasive Surgery, Stanford University School of Medicine, Stanford, California Bariatric Surgery

Sean J. Mulvihill, MD

Professor and Chair, Department of Surgery, and Ross Anderson Presidential Endowed Chair in Surgery, Uni­ versity of Utah School of Medicine; Senior Director for Clinical Affairs, Huntsman Cancer Institute, University of Utah, Salt Lake City, Utah Treatment of Gallstone Disease

Moises Ilan Nevah, MD

Assistant Professor of Medicine, University of Texas Health Science Center at Houston Medical School, Houston, Texas Hepatic Encephalopathy, Hepatorenal Syndrome, Hepatopulmonary Syndrome, and Systemic Complications of Liver Disease

Jeffrey A. Norton, MD

Professor of Surgery, Stanford University School of Medi­ cine; Chief, Surgical Oncology, Stanford University Medical Center, Stanford, California Endocrine Tumors of the Pancreas and Gastrointestinal Tract

Kjell Öberg, MD, PhD

Professor of Endocrine Oncology, Department of Medical Sciences, Uppsala University; Chair, Center of Excel­ lence Endocrine Tumors, Department of Endocrine Oncology, University Hospital, Uppsala, Sweden Gastrointestinal Carcinoid Tumors (Gastrointestinal Neuroendocrine Tumors) and the Carcinoid Syndrome

Jacqueline G. O’Leary, MD, MPH

Medical Director, Inpatient Liver and Transplant Unit, Baylor University Medical Center, Dallas, Texas Hepatitis C

Seamus O’Mahony, MD, FRCP

Senior Lecturer in Gastroenterology, University College Cork, National University of Ireland; Consultant Gastro­ enterologist, Cork University Hospital, Cork, Ireland Enteric Microbiota and Small Intestinal Bacterial Overgrowth

Susan R. Orenstein, MD

Professor Emerita, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania Eosinophilic Disorders of the Gastrointestinal Tract

Roy C. Orlando, MD

Mary Kay and Eugene Bozymski and Linda and William Heizer Distinguished Professor of Medicine and Adjunct Professor of Cell and Molecular Physiology, University of North Carolina School of Medicine, Chapel Hill, North Carolina Anatomy, Histology, Embryology, and Developmental Anomalies of the Esophagus

Mark T. Osterman, MD, MSCE

Assistant Professor of Medicine, University of Pennsylvania School of Medicine; Division of Gastro­ enterology, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Ulcerative Colitis

xiii

xiv

Contributors Stephen J. Pandol, MD

Professor of Medicine, David Geffen School of Medicine at UCLA; Staff Physician, Veterans Affairs Greater Los Angeles Healthcare System, Los Angeles, California Pancreatic Secretion

John E. Pandolfino, MD

Associate Professor of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois Esophageal Neuromuscular Function and Motility Disorders

Abhitabh Patil, MD

Assistant Professor of Medicine, Rush University Medical Center, Chicago, Illinois Vascular Diseases of the Liver

John H. Pemberton, MD

Professor of Surgery, Mayo Clinic College of Medicine; Consultant in Colon and Rectal Surgery, Rochester, Minnesota Ileostomy, Colostomy, and Pouches

V. S. Periyakoil, MD

Director, Stanford University Hospice and Palliative Med­ icine Fellowship Program, Stanford University School of Medicine; Associate Director of Palliative Care Ser­ vices, Veterans Affairs Palo Alto Health Care System, Palo Alto, California Palliative Care for Patients with Gastrointestinal and Hepatic Disease

Robert Perrillo, MD

Associate Director, Hepatology Division, and Director, Transplant Hepatology Fellowship, Baylor University Medical Center, Dallas, Texas

B. S. Ramakrishna, MD, DM, PhD

Professor of Gastroenterology, Christian Medical College, Vellore, Tamil Nadu, India Tropical Diarrhea and Malabsorption

Mrinalini C. Rao, PhD

Professor and Vice President of Faculty Affairs, Depart­ ment of Physiology and Biophysics, University of Illi­ nois College of Medicine at Chicago, Chicago, Illinois Intestinal Electrolyte Absorption and Secretion

Satish S. C. Rao, MD, PhD, FRCP(Lon)

Professor of Medicine, University of Iowa Carver College of Medicine; Director, Neurogastroenterology and Gastrointestinal Motility, University of Iowa Hospitals and Clinics, Iowa City, Iowa Fecal Incontinence

Andrea E. Reid, MD, MPH

GI/Hepatology/Nutrition Section, Veteran Affairs Medical Center, Washington, DC Nonalcoholic Fatty Liver Disease

John F. Reinus, MD

Professor of Clinical Medicine, Albert Einstein College of Medicine; Chief of Clinical Hepatology, Montefiore Medical Center, Bronx, New York Gastrointestinal and Hepatic Disorders in the Pregnant Patient

David A. Relman, MD

Professor of Medicine and Professor of Microbiology and Immunology, Stanford University School of Medicine; Chief, Infectious Diseases Section, Veterans Affairs Palo Alto Health Care System, Palo Alto, California

Hepatitis B and D

Whipple’s Disease

David A. Peura, MD

Joel E. Richter, MD

Emeritus Professor of Medicine, Division of Gastroenterol­ ogy and Hepatology, University of Virginia School of Medicine, Charlottesville, Virginia Helicobacter pylori

Patrick R. Pfau, MD

Associate Professor of Medicine, Section of Gastroenterol­ ogy and Hepatology, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin Foreign Bodies, Bezoars, and Caustic Ingestions

Daniel K. Podolsky, MD

Professor of Internal Medicine, Doris and Bryan Wilden­ thal Distinguished Chair in Medical Science, and Presi­ dent, University of Texas Southwestern Medical Center, Dallas, Texas Cellular Growth and Neoplasia

Jonathan Potak, MD

Assistant Professor of Medicine, Mount Sinai School of Medicine, New York, New York Colonic Polyps and Polyposis Syndromes

Daniel S. Pratt, MD

Assistant Professor of Medicine, Harvard Medical School; Liver-Biliary-Pancreas Center, Massachusetts General Hospital, Boston, Massachusetts Liver Chemistry and Function Tests

Deborah Denise Proctor, MD

Professor of Medicine and Director, Inflammatory Bowel Disease Program, Section of Digestive Diseases, Yale University School of Medicine; Attending Physician, Yale-New Haven Hospital, New Haven, Connecticut Ulcers of the Small and Large Intestine

Richard L. Evans Chair and Professor, Department of Medicine, Temple University School of Medicine, Philadelphia, Pennsylvania Gastroesophageal Reflux Disease

Eve A. Roberts, MD, FRCPC

Adjunct Professor of Pediatrics, Medicine, and Pharmacol­ ogy, University of Toronto Faculty of Medicine; Associ­ ate, Division of Gastroenterology, Hepatology, and Nutrition, Hospital for Sick Children, Toronto, Ontario, Canada Wilson Disease

Hugo R. Rosen, MD, FACP

Waterman Endowed Chair in Liver Research, Professor of Medicine and Immunology, Division Head of Gastroen­ terology and Hepatology, and Program Director, Hepati­ tis C Research Center, University of Colorado Denver Health Sciences Center, Denver, Colorado Liver Transplantation

Andrew S. Ross, MD

Digestive Disease Institute, Virginia Mason Medical Center, Seattle, Washington Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis

Jayanta Roy-Chowdhury, MD

Professor of Medicine and Genetics, Albert Einstein College of Medicine, Bronx, New York Liver Physiology and Energy Metabolism

Namita Roy-Chowdhury, MD

Professor of Medicine and Genetics, Albert Einstein College of Medicine, Bronx, New York Liver Physiology and Energy Metabolism

Contributors Bruce A. Runyon, MD

Professor of Medicine, Division of Gastroenterology and Hepatology, Loma Linda University School of Medicine; Chief of Liver Service, Loma Linda University Medical Center, Loma Linda, California Ascites and Spontaneous Bacterial Peritonitis

Michael A. Russo, MD

Assistant Professor of Pediatrics, Division of Pediatric Gastroenterology, Hepatology, and Nutrition, University of Texas Southwestern Medical School, Dallas; Attend­ ing Physician, Children’s Medical Center of Dallas at Legacy, Plano, Texas Anatomy, Histology, Embryology, and Developmental Anomalies of the Stomach and Duodenum

Hugh A. Sampson, MD

Professor of Pediatrics and Immunology, Mount Sinai School of Medicine; Staff Physician, Mount Sinai Hos­ pital, New York, New York Food Allergies

Bruce E. Sands, MD

Associate Professor of Medicine, Harvard Medical School; Interim Chief, Gastrointestinal Unit, and Medical Co-Director, MGH Crohn’s and Colitis Center, Massachusetts General Hospital, Boston, Massachusetts Crohn’s Disease

George A. Sarosi, Jr., MD

Associate Professor of Surgery, Robert H. Hux, MD, Pro­ fessor of Surgery, and Surgical Residency Program Director, University of Florida College of Medicine; Assistant Chief, Surgical Service, North Florida/South Georgia Veterans Affairs Medical Center, Gainesville, Florida Appendicitis

Thomas J. Savides, MD

Professor of Clinical Medicine, University of California, San Diego, School of Medicine; Clinical Service Chief, Gastroenterology, University of California, San Diego, Medical Center, La Jolla, California Gastrointestinal Bleeding

Lawrence R. Schiller, MD

Clinical Professor of Internal Medicine, University of Texas Southwestern Medical School; Director, Gastro­ enterology Fellowship Program, and Attending Physi­ cian, Digestive Health Associates of Texas, Baylor University Medical Center, Dallas, Texas Diarrhea

Mitchell L. Schubert, MD

Professor of Medicine and Physiology, Virginia Common­ wealth University Health System; Chief of Gastro­ enterology, McGuire Veterans Affairs Medical Center, Richmond, Virginia Gastric Secretion

Joseph H. Sellin, MD

Professor of Medicine and Director, GI Fellowship Program, Baylor College of Medicine; Chief, Division of Gastroenterology, Ben Taub General Hospital, Houston, Texas Diarrhea; Intestinal Electrolyte Absorption and Secretion

M. Gaith Semrin, MD, MBBS

Assistant Professor of Pediatric Gastroenterology, Hepato­ logy, and Nutrition, University of Texas Southwestern Medical School; Attending Physician, University of Texas Southwestern Medical Center, Dallas, Texas Anatomy, Histology, Embryology, and Developmental Anomalies of the Stomach and Duodenum

Vijay H. Shah, MD

Professor of Medicine and Physiology, Mayo Clinic College of Medicine, Rochester, Minnesota Portal Hypertension and Gastrointestinal Bleeding

Fergus Shanahan, MD, FRCP(UK), FRCPI, FACP, FRCP(C)

Professor and Chair, Department of Medicine, University College Cork, National University of Ireland; Consultant Gastroenterologist, Cork University Hospital, Cork, Ireland Enteric Microbiota and Small Intestinal Bacterial Overgrowth

Corey A. Siegel, MD

Assistant Professor, Dartmouth Medical School, Hanover; Director, IBD Center, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire Crohn’s Disease

Maria H. Sjogren, MD, MPH, FACP

Associate Professor of Medicine, Georgetown University School of Medicine, Washington, DC; Associate Profes­ sor of Preventive Medicine, Uniformed Services Univer­ sity of the Health Sciences F. Edward Hébert School of Medicine, Bethesda, Maryland; Staff Physician, Department of Gastroenterology/Hepatology, Walter Reed Army Medical Center, Washington, DC Hepatitis A

Rhonda F. Souza, MD

Associate Professor of Medicine, University of Texas Southwestern Medical School; Staff Physician, Veterans Affairs North Texas Health Care System, Dallas, Texas Barrett’s Esophagus

Stuart Jon Spechler, MD

Professor of Medicine and Berta M. and Cecil O. Patterson Chair in Gastroenterology, University of Texas South­ western Medical Center; Chief, Division of Gastroenter­ ology, Veterans Affairs North Texas Health Care System, Dallas, Texas Barrett’s Esophagus

William M. Steinberg, MD

Clinical Professor of Medicine, George Washington University School of Medicine and Health Sciences, Washington, DC Acute Pancreatitis

William E. Stevens, MD

Clinical Faculty, Department of Internal Medicine, Division of Gastroenterology, Presbyterian Hospital of Dallas, Dallas, Texas Vascular Diseases of the Liver

Andrew H. Stockland, MD

Assistant Professor of Radiology, Mayo Clinic College of Medicine; Consultant, Diagnostic Radiology, Mayo Clinic, Rochester, Minnesota Endoscopic and Radiologic Treatment of Biliary Disease

xv

xvi

Contributors Neil H. Stollman, MD

Associate Clinical Professor of Medicine, Division of Gas­ troenterology, University of California, San Francisco, School of Medicine, San Francisco; Chief, Division of Gastroenterology, Alta Bates Summit Medical Center, Oakland, California Diverticular Disease of the Colon

Frederick J. Suchy, MD

Professor and Chair, Department of Pediatrics, Mount Sinai School of Medicine; Pediatrician-in-Chief, Kravis Children’s Hospital, New York, New York Anatomy, Histology, Embryology, Developmental Anomalies, and Pediatric Disorders of the Biliary Tract

Jan Tack, MD, PhD

Professor of Medicine and Chair, Department of Patho­ physiology, University of Leuven; Clinic Head, Division of Gastroenterology, University Hospital Leuven, Leuven Belgium Dyspepsia

Nicholas J. Talley, MD, PhD

Professor of Medicine, Mayo Clinic College of Medicine; Division of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, Florida Irritable Bowel Syndrome

Scott Tenner, MD, MPH

Associate Professor of Medicine, State University of New York Health Science Center at Brooklyn; Director, Medical Education and Research, Maimonides Medical Center, Brooklyn, New York Acute Pancreatitis

Associate Professor of Medicine, Australian National University Medical School; Senior Staff Specialist in Gastroenterology and Hepatology, Canberra Hospital, Canberra, Australian Capital Territory, Australia Liver Disease Caused by Drugs

Professor and Vice Chair, Department of Internal Medi­ cine, University of Texas Southwestern Medical Center, Dallas, Texas Gastrointestinal and Hepatic Manifestations of Systemic Diseases

Richard H. Turnage, MD

Professor and Chair, Department of Surgery, University of Arkansas for Medical Sciences; UAMS Medical Center, Little Rock, Arkansas Intestinal Obstruction

Hospital–

Nimish Vakil, MD

Clinical Professor of Medicine, University of Wisconsin School of Medicine and Public Health, Madison; Gastroenterologist, Aurora Health Care, Waukesha, Wisconsin Peptic Ulcer Disease

Visiting Research Assistant Professor, Department of Physiology and Biophysics, University of Illinois College of Medicine at Chicago, Chicago, Illinois Intestinal Electrolyte Absorption and Secretion

Arnold Wald, MD

Professor of Medicine, Section of Gastroenterology and Hepatology, Department of Medicine, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin Other Diseases of the Colon and Rectum

David Q.-H. Wang, MD, PhD

Assistant Professor of Medicine, Harvard Medical School; Gastroenterologist, Beth Israel Deaconess Medical Center, Boston, Massachusetts Gallstone Disease

Timothy C. Wang, MD

Dorothy L. and Daniel Silberberg Professor of Medicine, Columbia University College of Physicians and Sur­ geons; Chief, Digestive and Liver Diseases, Columbia University Medical Center, New York, New York Adenocarcinoma and Other Tumors of the Stomach

David C. Whitcomb, MD, PhD

Professor of Medicine, Cell Biology and Physiology, and Human Genetics, University of Pittsburgh School of Medicine, Pittsburgh; Chief, Division of Gastroenter­ ology, Hepatology, and Nutrition, University of Pitts­ burgh Medical Center, Allison Park, Pennsylvania Hereditary, Familial, and Genetic Disorders of the Pancreas and Pancreatic Disorders in Childhood

Professor of Medicine, University of Alabama School of Medicine at Birmingham; Division of Gastroenterology and Hepatology, UAB Hospital, Birmingham, Alabama Gastrointestinal Consequences of Infection with Human Immunodeficiency Virus

Leonard Prosnitz Professor of Radiation Oncology, Duke University School of Medicine; Chief, Department of Radiation Oncology, Duke University Hospital, Durham, North Carolina Radiation Injury

Gavitt Woodard, BS

Research Coordinator, Stanford University School of Medicine, Stanford, California Bariatric Surgery

Stephan G. Wyers, MD

Constipation

Jayashree Venkatasubramanian, PhD

Whipple’s Disease

Christopher G. Willett, MD

Dwain L. Thiele, MD

Gastroenterologist, Beth Israel Deaconess Needham, Needham, Massachusetts

Professor and Consultant in Gastrointestinal Pathology, John Radcliffe Hospital, Oxford, United Kingdom

C. Mel Wilcox, MD

Narci C. Teoh, MBBS, PhD, FRACP

Sonal P. Ullman, MD

Axel von Herbay, MD

Assistant Professor of Surgery, University of Chicago Pritzker School of Medicine; General Surgery, Univer­ sity of Chicago Medical Center, Chicago, Illinois Surgical Peritonitis and Other Diseases of the Peritoneum, Mesentery, Omentum, and Diaphragm

Joseph C. Yarze, MD, FACP, FACG, FASGE, AGAF

Consultant, Gastroenterology Associates of Northern New York; Medical Director, GI Center, Glens Falls Hospital, Glens Falls, New York Chronic Abdominal Pain

Foreword It is a pleasure to write the foreword for the 9th edition of Sleisenger and Fordtran’s Gastrointestinal and Liver Disease: Pathophysiology/Diagnosis/Management, the classic textbook that has maintained its leadership position and progressively enhanced its reputation as the “go-to” book in its field. The extraordinary effort required to create a textbook that comprehensively and authoritatively presents the state of knowledge in an arena as broad and everexpanding as gastroenterology is a daunting task but one that has been achieved with distinction in this 9th edition. The excellence and reliability exhibited by this considerably revised and updated edition amply validates the lasting value of the series. Almost 40 years ago, Drs. Marvin Sleisenger and John Fordtran gathered a group of highly respected, thoughtful, and clinically excellent gastroenterologists to create what became the first edition of Gastrointestinal and Liver Disease. The book was a success from the onset. Subsequent editions have continued to successfully portray the accelerating evolution of gastroenterology. Gastroenterologists, internists, surgeons, pathologists, residents, and students from diverse backgrounds have relied on Gastrointestinal and Liver Disease as a thoughtful, extensively referenced, and user-accessible textbook. The images are well chosen and superbly produced. Important landmarks are noted, evaluated, and placed into proper perspective. The 9th edition is ably edited by Drs. Mark Feldman, Lawrence Friedman, and Lawrence Brandt—all three highly respected clinicians and educators. This monumental undertaking continues a grand tradition and succeeds in providing authoritative overviews of the status of gastro­ enterology in the early 21st century. The editors are to be applauded for their efforts. The creation of a textbook requires considerable judgment in defining the current state of the art for a diverse range of disorders and in selecting authors best qualified to evaluate and lead the field. The editors have chosen the authors well for the 9th edition, with an admixture of established leaders and younger colleagues who represent the next generation that will advance gastroenterology. A textbook differs in important ways from a journal. In a textbook, the state of knowledge is presented at a given, and set, time. A line is drawn in the sand. A successful chapter in a textbook defines the state of knowledge regarding a subject, provides a clear definition of what is known,

outlines how the knowledge evolved, suggests what challenges may be ahead, and offers a framework of advice as to the most efficient and effective means of establishing the diagnosis and developing a plan for treatment. The goal is to provide a foundation and a guide that will serve us well in the varied situations we encounter in clinical practice. There surely have been outstanding advances in gastroenterology to incorporate into the 9th edition. Notable examples are the further emergence of advanced diagnostic and therapeutic endoscopy, introduction of liver and intestinal transplantation, development of more precise diagnostic genetic and serologic tests, and game-changing results derived from the advent of minimally invasive surgery. New innovative and effective therapeutic agents have been developed. These days, drugs are increasingly designed to solve a specific identified problem. Increasingly, therapeutic agents provide the ability to modulate inflammation, inhibit fibrosis, and regulate the processes that lead to cell death. In the 9th edition of Gastrointestinal and Liver Disease, we are provided a sturdy platform to aid us now and direct us toward the future. What lies ahead is likely to be exciting beyond our most ambitious speculations. Knowledge will continue to evolve. Innovative, often life-changing therapies will be discovered. New challenges will emerge that must be faced. As time goes by, we all will undoubtedly rely even more on electronically transmitted information (including the online version of this textbook, in which the references will be updated regularly). The role for a comprehensive, respected medical textbook, however, is far from over. We salute those who wrote and those who edited Gastrointestinal and Liver Disease. Through their efforts we are provided with renewed confidence that tangible progress is being achieved on many fronts. An effective textbook reminds us of the past, firmly grounds us in the present, and guides our expectations and hopes for the future. Gastrointestinal and Liver Disease is such a book. We, the readers and our patients, will benefit from the guide provided. Willis C. Maddrey, MD University of Texas Southwestern Medical Center Dallas, Texas

xvii

Preface The 9th edition of Sleisenger and Fordtran’s Gastrointestinal and Liver Disease is the second edition for which the three of us have served together as editors and the first for which neither Dr. Fordtran nor Dr. Sleisenger has participated. We remain grateful to their inspiring examples as visionaries and superb editors and are pleased to present an edition that we believe is worthy of their high standards and commitment to excellence. This edition is the first for which a single purchase will provide all readers with both a copy of the book and access to the online version of the text; the latter will be updated regularly to keep readers abreast of new developments in the rapidly evolving field of gastroenterology and liver disease. Moreover, providing readers with online access has allowed us to reduce the bulk of the hard copy of the book by listing only 15 or so key references per chapter and referring the reader to the online version for the complete list of references cited in each chapter and linked to PubMed. We are confident that readers will appreciate the improved “portability” of the book. Like its predecessors, the 9th edition presents a critical overview of the state of gastrointestinal practice and its scientific basis by eminent authorities in their respective fields. With authors from at least 12 countries on 4 continents, the book is truly international in scope. We have been aided in our preparation of the 9th edition by the valuable feedback provided by our colleagues, trainees, reviewers, and readers; this has led to incremental refinements and enhancements since the highly acclaimed 8th edition, which received the first prize for a textbook in Gastroenterology from The British Medical Association in 2007. Attributes of the book, we believe, are its logical organization, minimized redundancies, polished writing, incorporation of color figures into the body of the text, inclusion of clear, visually appealing algorithms to summarize clinical decision making, and practical approaches to patient management. As always, we have tried to incorporate new, highquality scientific evidence as the basis of rational treatment, while alerting readers to emerging developments that hold promise for new approaches to patient management. To keep the book fresh and to ensure critical review of fields for which controversy exists, we have again rotated the authorship of many chapters, with approximately one third of the authors being new to this edition. We feel particularly fortunate to have enlisted contributors who are renowned authorities in their fields. Each has been chosen for his or her expertise and skills in communication, and it was a remarkable privilege for us to work with such accomplished colleagues. Reflecting recent trends in the practice of gastroenterology, three new chapters have been added to the 9th edition. First, a chapter by Drs. Woodard and Morton is devoted to Bariatric Surgery and complements the chapter on Obesity by Dr. Bray. This new chapter is an indication of the emerging role of the gastroenterologist not only in the selection of patients for bariatric surgery but also in the management of postsurgical complications and eventually in the endoscopic management of obesity. Second, a chapter on Barrett’s Esophagus by Drs. Spechler and Souza has been separated from the chapter on Gastroesophageal Reflux Disease by Drs. Richter and Freidenberg, again reflecting the

importance and controversies concerning the premalignant nature of Barrett’s esophagus and emerging approaches to its treatment. Third, a new chapter on Endoscopic Treatment of Pancreatic Disease complements the chapter on Endoscopic and Radiologic Treatment of Biliary Disease, both by Dr. Baron, with two different co-authors (Drs. Ferreira and Stockland, respectively), and reflects the expanding role of endoscopic interventions in the management of pancreatic disorders. Additionally, two chapters in the 8th edition have been merged into a single chapter in the 9th edition—Foreign Bodies, Bezoars, and Caustic Injury, by Drs. Pfau and Ginsberg. The organization of the 9th edition is otherwise similar to that of the 8th edition with the exception that the section on Nutrition in Gastroenterology (Section II) has now been placed, more logically we believe, before the section on Symptoms, Signs, and Biopsychosocial Issues (Section III). In Section I on Biology of the Gastrointestinal Tract, we are delighted to welcome Drs. Dotan and Mayer, who provide a comprehensive overview of Gastrointestinal Immunology and Inflammation. In Section III, we welcome Dr. Millham (Acute Abdominal Pain), Dr. Yarze (Chronic Abdominal Pain), Dr. Tack (Dyspepsia), Dr. Rao (Fecal Incontinence), and Drs. Savides and D. Jensen (Gastrointestinal Bleeding), all recognized experts in their respective fields. We are also happy to welcome back Dr. Drossman, who has provided a masterly discussion of Biopsychosocial Issues in Gastroenterology. Section IV deals with Topics Involving Multiple Organs and includes an expanded discussion of Eosinophilic Disorders of the Gastrointestinal Tract, with a section on eosinophilic esophagitis, by Drs. Khan and Orenstein, and an expanded chapter on Preparation for and Complications of Gastrointestinal Endoscopy by Dr. Kimmey. Among the many other noteworthy contributors in this section are Drs. Li and Collins (Gastrointestinal Lymphomas), Dr. Demetri (Gastrointestinal Stromal Tumors), Dr. Oberg (Carcinoid Tumors), Drs. R. Jensen and Norton (Pancreatic and Gastrointestinal Endocrine Tumors), Drs. Larson and McDonald (Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation), and Drs. Willett and Czito (Radiation Injury). The next six sections cover the Esophagus, Stomach and Duodenum, Pancreas, Biliary Tract, Liver, and Small and Large Intestine. Among the new authors are Drs. Kahrilas and Pandolfino (Esophageal Motor and Sensory Function and Motor Disorders), Drs. Kaunitz and Schubert (Gastric Secretion), Drs. Crowe and Peura (Helicobacter pylori), Dr. Vakil (Peptic Ulcer Disease), Drs. Wang and Afdhal (Gallstone Disease), Dr. Andersson (Acalculous Cholecystitis, Cholesterolosis, Adenomyomatosis, and Gallbladder Polyps), Drs. Blechacz and Gores (Tumors of the Bile Ducts, Gallbladder, and Ampulla), Dr. Misdraji (Anatomy, Histology, Embryology, and Developmental Anomalies of the Liver), Dr. Pratt (Liver Chemistry and Function Tests), Drs. O’Leary and Davis (Hepatitis C), Drs. Nevah and Fallon (Hepatic Encephalopathy, Hepatorenal Syndrome, Hepatopulmonary Syndrome, and Systemic Complications of Liver Disease), Drs. Belefer and Di Bisceglie (Hepatic Tumors and Cysts), Dr. Ramakrishna (Tropical Malabsorption and

xix

xx

Preface Tropical Diarrhea), Dr. Foxx-Orenstein (Ileus and Pseudoobstruction), Drs. Ludwig-Miller and Kurtz (Small Intestinal Neoplasms), and Dr. Marcello (Diseases of the Anorectum). The book concludes with an important section on Palliative, Complementary, and Alternative Medicine. The 9th edition of this book remains a living tribute to its founding editors, Drs. Marvin Sleisenger and John Fordtran, and the current editors are proud to strive to uphold the tradition of excellence established by our predecessors. We are particularly grateful to Druanne Martin, Anne Snyder, and Jeffrey Gunning and their colleagues at Elsevier for their

commitment to this classic book, which has been a labor of love for the editors. We believe that the 9th edition of Sleisenger and Fordtran’s Gastrointestinal and Liver Disease remains the ideal resource for all who seek excellence in the care of patients with gastrointestinal and liver disorders. Mark Feldman, MD Lawrence S. Friedman, MD Lawrence J. Brandt, MD

CHAPTE R

1

Gastrointestinal Hormones and Neurotransmitters Rodger A. Liddle

CHAPTER OUTLINE Cellular Communication  3 Neural Regulation of the GI Tract  5 Peptide Hormones of the GI Tract  6 Synthesis, Post-Translational Modification, and Secretion  6 Gastrin  6 Cholecystokinin  7 Secretin  7 Vasoactive Intestinal Polypeptide  8 Glucagon  8 Glucose-Dependent Insulinotropic Polypeptide  8 Pancreatic Polypeptide Family  9 Substance P and the Tachykinins  9 Somatostatin  9 Motilin  9 Leptin  10 Ghrelin  10 Other Chemical Messengers of the Gastrointestinal Tract  10 Acetylcholine  10 Catecholamines  11 Dopamine  11 Serotonin  11 Histamine  11

Cells throughout the gastrointestinal (GI) tract receive information in many forms, including chemical messengers that emanate from other cells. The initial stimulus for hormone secretion is the ingestion of food. Food provides central neural stimulation in the form of thought (anticipation) and sight, chemical stimulation in the form of odor and taste, nutrient stimulation of the epithelial cells lining the GI tract, and mechanical stimulation. These processes all stimulate the release of peptides and other transmitters from cells of the mucosa into the nearby space, where they act locally, or into the bloodstream, where they circulate to distant target tissues. Therefore, chemical messengers from the GI tract can have far-reaching effects throughout the body.

CELLULAR COMMUNICATION Chemical transmitters of the gut are produced by discrete cells of the GI mucosa and can be classified as endocrine, paracrine, synaptic (“neurocrine”), or autocrine (Fig. 1-1). Specialized signaling cells that secrete transmitters into the

Nitric Oxide  12 Adenosine  12 Cytokines  13 Signal Transduction  13 G Protein–Coupled Receptors  13 G Proteins  13 Receptors Not Coupled to G Proteins  14 Hormone and Transmitter Regulation of Gastrointestinal Growth  15 Growth Factor Receptors  15 Epidermal Growth Factor  16 Transforming Growth Factor-α  16 Transforming Growth Factor-β  16 Insulin-Like Growth Factors  16 Fibroblast Growth Factor and Platelet-Derived Growth Factor  16 Trefoil Factors  16 Other G Protein–Coupled Receptors  16 Taste Receptors  17 Intraluminal Releasing Factor Regulation of Gastrointestinal Hormones  17 Gastrointestinal Peptides That Regulate Satiety and Hunger  18 Enteroinsular Axis  18

blood are known as endocrine cells, and the transmitters they produce are known as hormones. Hormones bind to specific receptors on the surface of target cells at remote sites and regulate metabolic processes.1 In contrast with endocrine cells that act on distant target tissues, other signaling cells of the GI tract may produce transmitters that act on neighboring cells. This process is known as paracrine signaling and is typical of cells that produce somatostatin.2 Paracrine transmitters are secreted locally and cannot diffuse far. They bind to receptors on nearby cells to exert their biological actions. These actions are limited because they are taken up rapidly by their target cells, destroyed by extracellular enzymes, and adhere to extracellular matrix, all of which limit their ability to act at distant sites. Because paracrine signals act locally, their onset of action is generally rapid and can be terminated abruptly. By comparison, endocrine signaling takes much longer, and termination of signaling requires clearance of hormone from the circulation. A third form of signaling in the GI tract is neurotransmission. The enteric nervous system is a complex and sophisticated array of nerve cells and ganglia that is intimately involved in all aspects of GI function. When neurons of the

3

4

Section I  Biology of the Gastrointestinal Tract Table 1-1  Hormones and Transmitters of the Gastrointestinal Tract

Endocrine

Autocrine

Paracrine

Neurocrine

Figure 1-1.  Examples of cell-to-cell communication by chemical transmitters in the gastrointestinal tract. Hormones are secreted from endocrine cells into the blood, where they are carried to distant targets. Paracrine cells secrete transmitters into the paracellular space and act locally. Neurons secrete chemical transmitters or peptides into synapses or onto other cell types. Autocrine transmitters bind to receptors on the cell from which they originate.

GI tract are activated, signals in the form of neurotransmitters are released from the nerve terminals. These synapses deliver neurotransmitters to nerves, muscle cells, epithelial and secretory cells, and other specialized cells of the GI tract. Neurotransmitters are critical for the processes of digestion including the coordination of gut motility and secretion. Many of the same transmitters are produced by endocrine, paracrine, and neural cells. For example, cholecystokinin (CCK) is produced by typical endocrine cells of the upper small intestine and is secreted into the bloodstream on ingestion of a meal. However, CCK is also abundant in nerves of the GI tract and brain. In neural tissue, CCK functions as a neurotransmitter, although when secreted into the blood, CCK is a classic GI hormone. This conservation of transmitters allows the same messenger to have different physiologic actions at different locations and is made possible by the manner in which the transmitter is delivered to its target tissues. Endocrine cells secrete many different hormones into the blood, and their actions depend on the specificity of the receptor on the target tissues. In contrast, in synaptic transmission, the variety of neurotransmitters is more limited, and the specificity of action is dependent on the precise location at which the nerves synapse with the target cells. The concentration of signaling molecules can be adjusted quickly because the transmitter can be rapidly metabolized. In the synaptic cleft, transmitters are either rapidly destroyed or taken back up by the secretory neuron. Concentrations of these transmitters can be regulated rapidly by changes in their rate of synthesis, secretion, or catabolism. Many peptide transmitters have extremely short halflives (generally on the order of minutes), which allows the rapid initiation and termination of signaling. Endocrine transmitters of the GI tract consist predominantly of peptides (e.g., gastrin, secretin). Paracrine transmitters can be peptides, such as somatostatin, or nonpeptides, such as histamine, that act locally on neighboring cells. Neurotransmitters can be peptides, such as vasoactive intestinal polypeptide (VIP) and tachykinins, or small molecules, such as acetylcholine and norepinephrine, that are secreted, or nitric oxide (NO), which simply diffuses across the syn-

Peptides That Function Mainly as Hormones Gastrin Glucose-dependent insulinotropic peptide (GIP) Glucagon and related gene products (GLP-1, GLP-2, glicentin, oxyntomodulin) Insulin Motilin Pancreatic polypeptide Peptide tyrosine tyrosine (PYY) Secretin Peptides That May Function as Hormones, Neuropeptides, or Paracrine Agents Cholecystokinin (CCK) Corticotropin-releasing factor (CRF) Endothelin Neurotensin Somatostatin Peptides That Act Principally as Neuropeptides Calcitonin gene-related peptide (CGRP) Dynorphin and related gene products Enkephalin and related gene products Galanin Gastrin-releasing peptide (GRP) Neuromedin U Neuropeptide Y Peptide histidine isoleucine (PHI) or peptide histidine methionine (PHM) Pituitary adenylate cyclase–activating peptide (PACAP) Substance P and other tachykinins (neurokinin A, neurokinin B) Thyrotropin-releasing hormone (TRH) Vasoactive intestinal peptide (VIP) Peptides That Act as Growth Factors Epidermal growth factor Fibroblast growth factor Insulin-like factors Nerve growth factor Platelet-derived growth factor Transforming growth factor-β Vascular endothelial growth factor Peptides That Act as Inflammatory Mediators Interferons Interleukins Lymphokines Monokines Tumor necrosis factor-α Peptides That Act on Neurons Cholecystokinin Gastrin Motilin Nonpeptide Transmitters Produced in the Gut Acetylcholine Adenosine triphosphate (ATP) Dopamine γ-Aminobutyric acid (GABA) Histamine 5-Hydroxytryptamine (5-HT, serotonin) Nitric oxide Norepinephrine Prostaglandins and other eicosanoids Newly Recognized Hormones or Neuropeptides Amylin Ghrelin Guanylin and uroguanylin Leptin

aptic cleft. The major transmitters and hormones of the GI tract are listed in Table 1-1. Criteria for establishing whether a candidate transmitter functions as a true hormone requires the following: (1) that the peptide be released into the circulation in response to

Chapter 1  Gastrointestinal Hormones and Neurotransmitters a physiologic stimulus; and (2) that the target tissue response can be reproduced by infusing the transmitter into the blood, thereby producing the same blood levels that occur physiologically. If an identical target tissue response is elicited, the hormonal effect of the transmitter has been proved. These criteria have been satisfied for a limited number of GI hormones, including gastrin, CCK, secretin, motilin, and glucose-dependent insulinotropic peptide (GIP). Somatostatin is the prototype of a paracrine transmitter. However, depending on its location, somatostatin may also exert endocrine and neural actions. For example, intestinal somatostatin is released into the local circulation following ingestion of fat and acts on the stomach as an enterogastrone to inhibit gastric acid secretion. Some cells release messengers locally and possess cell surface receptors for the same messengers, thus enabling those cells to respond to their own secreted products. This mode of transmission, known as autocrine signaling, has been demonstrated for several growth factors and has been implicated in the growth of certain cancers, including colorectal cancer (see Chapter 3).3

NEURAL REGULATION OF THE GI TRACT The enteric nervous system plays an integral role in the regulation of gut mucosal and motor function.4 It is organized into two major plexuses (Fig. 1-2). The myenteric plexus lies between the external longitudinal and internal circular muscle layers. The submucosal plexus lies between the circular muscle layer and the mucosa. Although the enteric nervous system receives input from the central and autonomic nervous systems, it can function independently. Nerves of the myenteric plexus project fibers primarily to the smooth muscle of the gut, with only a few axons extending to the submucosal plexus. Most of the fibers of the submucosal plexus project into the mucosa and the submucosal and myenteric plexuses. Various peptide and nonpeptide neurotransmitters are found in the enteric nervous system. Studies using immunohistochemical staining have

Serosa Circular muscle

Longitudinal muscle

Submucosa

Myenteric plexus

Submucosal plexus

Muscularis mucosa Mucosal nerves Mucosa Figure 1-2.  Organization of the enteric nervous system. The enteric nervous system is composed of two major plexuses, one submucosal and one located between the circular and longitudinal smooth muscle layers. These neurons receive and coordinate neural transmission from the GI tract and central nervous system.

localized neurotransmitters to specific neurons in the GI tract. γ-Aminobutyric acid is found primarily in the myenteric plexus and is involved in regulating smooth muscle contraction. Serotonin is found within the plexus and functions as an interneuron transmitter. Adrenergic neurons originate in ganglia of the autonomic nervous system and synapse with enteric neurons. Peptides such as neuropeptide Y (NPY) are often secreted from the same adrenergic neurons and generally exert inhibitory effects, such as vasoconstriction.5 Other adrenergic neurons containing somatostatin project to the submucosal plexus, where they inhibit intestinal secretion. Coexistence of peptides and neurotransmitters in the same neurons is not unusual; in fact, the interplay among transmitters is critical for coordinated neural regulation.6 For example, the peptides VIP and peptide histidine isoleucine (PHI) are commonly found together, as are the tachykinins substance P and substance K, where they have complementary effects. Somatostatin is found in interneurons that project caudally. The inhibitory action of somatostatin is consistent with a role in causing muscle relaxation in advance of a peristaltic wave. The abundance of VIP in the myenteric plexus also suggests that its inhibitory actions are important for smooth muscle relaxation in gut motility. VIP neurons that project from the submucosal plexus to the mucosa most likely stimulate intestinal fluid secretion. Other neurons that innervate the mucosa contain acetylcholine. Mucosal cells of the intestine contain receptors for both VIP and acetylcholine, allowing these transmitters to exert synergistic effects, because VIP increases intracellular cyclic adenosine monophosphate (cAMP) levels and acetylcholine increases intracellular calcium in the target cell. Bipolar neurons that project to the mucosa and myenteric plexus act as sensory neurons and often contain substance P, calcitonin gene-related peptide (CGRP), and acetylcholine as neurotransmitters. These neurons participate in pain pathways and modulate inflammation. The ability of hormones to act on nerves locally within the submucosa of the intestine and affect more distant sites on nerves such as the vagus expands the potential organs that may be regulated by gut hormones.7 Chemical and mechanical stimuli cause the release of hormones from endocrine cells of the intestinal mucosa. These interactions initiate a wide variety of secretomotor responses, many of which are mediated by enteric neurons. Secretomotor circuits consist of intrinsic primary afferent neurons with nerve endings in the mucosa and extension through the myenteric and submucosal plexi. This circuitry allows nerves to stimulate mucosal cells to secrete fluid and electrolytes and at the same time stimulate muscle contraction. The same motor neurons also have axons that supply arterioles and can initiate vasodilator reflexes. Extrinsic primary afferent neurons can be of the vagus, with somal bodies in the nodose ganglia and axons that reach the gut through the vagus nerve, or of the spinal nerves of the thoracic and lumbar regions, whose cell bodies lie in the dorsal root ganglia. Information conducted by extrinsic primary afferent neurons includes pain, heat, and sensations of fullness or emptiness. These neurons are also targets for hormones. For example, the satiety effect of CCK in the bloodstream is mediated through the vagus nerve.8 Specific CCK receptors have been identified on the vagus, and blockade of these receptors abolishes the satiation induced by peripheral CCK. Endocrine, paracrine, and neural transmitters existing within the lamina propria modulate effects on the gut immune system.7 Lymphocytes, macrophages, mast cells, neutrophils, and eosinophils are potential targets for endo-

5

6

Section I  Biology of the Gastrointestinal Tract crine and neural transmitters and participate in the inflammatory cascade. Moreover, inflammatory mediators can act directly on enteric nerves. Serotonin released from endocrine cells is involved in intestinal anaphylaxis and stimulates vagal afferent fibers that possess the 5-hydroxytryptamine 3 (5-HT3) receptor.

Gene 5′

Exon 1

intron

Exon 2

intron

Exon 3

3′

transcription mRNA

poly A

Cap site translation

PEPTIDE HORMONES OF THE GI TRACT SYNTHESIS, POST-TRANSLATIONAL MODIFICATION, AND SECRETION

The expression of peptides is regulated at the level of the gene that resides on defined regions of specific chromosomes. The genes for most of the known GI peptides have now been identified. Specific gene regulatory elements determine if and when a protein is produced and the particular cell in which it will be expressed. Gut hormone gene expression is generally linked to peptide production and regulated according to the physiologic needs of the organism. For example, the production of a hormone may increase when gut endocrine cells are stimulated by food, changes in intraluminal pH, exposure to releasing factors, or other transmitters or hormones. These factors may simultaneously stimulate hormone secretion and increase gene expression. Ultimately, hormones are secreted into the circulation, where they can bind to receptors on target tissues. Once a biological response is elicited, signals may then be sent back to the endocrine cell to “turn off” hormone secretion. This negative feedback mechanism is common to many physiologic systems and avoids excess production and secretion of hormone. All GI peptides are synthesized via gene transcription of DNA into messenger RNA (mRNA) and subsequent translation of mRNA into precursor proteins known as preprohormones. Peptides that are to be secreted contain a signal sequence that directs the newly translated protein to the endoplasmic reticulum, where the signal sequence is cleaved and the prepropeptide product is prepared for structural modifications.9 These precursors undergo intracellular processing and are transported to the Golgi apparatus and packaged in secretory granules. Further modifications in peptide structure may occur within the Golgi apparatus (e.g., sulfation) that is important for the bioactivity of many peptide hormones, such as CCK. Secretory granules may be targeted for immediate release or stored in close proximity to the plasma membrane for release following appropriate cell stimulation. When GI endocrine cells are stimulated, mature hormone is secreted into the paracellular space and is taken up into the bloodstream. For many hormones, such as gastrin and CCK, multiple molecular forms exist in blood and tissues. Although there is only a single gene for these peptides, the different molecular forms result from differences in pretranslational or post-translational processing (Fig. 1-3). A common mechanism of pretranslational proc­ essing includes alternative splicing of mRNA, which generates unique peptides from the same gene. Post-translational changes include cleavage of precursor molecules. Enzymatic cleavage of the signal peptide produces a prohormone. Other post-translational features that result in mature GI peptides include peptide cleavage to smaller forms (e.g., somatostatin), amidation of the carboxyl terminus (e.g., gastrin), and sulfation of tyrosine residues (e.g., CCK). These processing steps are usually critical for biolo­ gical activity of the hormone. For example, sulfated CCK is 100-fold more potent than its unsulfated form. The vast biochemical complexity of gastroenteropancreatic hor-

Prepropeptide

Signal spacer peptide Post-translational processing

Propeptide

Peptide AB Peptide A Figure 1-3.  Schematic representation of the production of gastrointestinal peptides. The genetic information is transcribed into mRNA, which is translated to a prepropeptide. Subsequent enzymatic cleavage produces peptides of various lengths. mRNA, messenger RNA.

mones is evident in the different tissues that secrete these peptides. As GI peptides are secreted from endocrine as well as nervous tissue, the distinct tissue involved often determines the processing steps for production of the peptide. Many hormone genes are capable of manufacturing alternatively spliced mRNAs or proteins that undergo different post-translational processing and ultimately produce hormones of different sizes. These modifications are important for receptor binding, signal transduction, and consequent cellular responses.10 It has become possible to express human genes in other species. By introducing specific hormone-producing genes into pigs or sheep, human hormones have been produced for medicinal use.11 With the rapid sequencing of the human genome, it is likely that novel methods of gene expression will expand the therapeutic use of human proteins. Moreover, drugs are being developed that inhibit the transcription of DNA into mRNA or that block the gene elements responsible for turning on specific hormone production (e.g., antisense oligonucleotides).12 This technology is based on the principle that nucleotide sequences bind to critical DNA regions and prevent transcription into mRNA. Similarly, oligonucleotides can be made to interact with mRNA and alter (or inhibit) translation of a protein product. These principles may be applicable to the treatment of the growing list of diseases that result from aberrant protein processing.13,14

GASTRIN

As discussed in more detail in Chapter 49, gastrin is the major hormone that stimulates gastric acid secretion. Subsequently, gastrin was found to have growth-promoting effects on the gastric mucosa and possibly some cancers.15 Human gastrin is the product of a single gene located on chromosome 17. The active hormone is generated from a precursor peptide called preprogastrin. Human preprogastrin contains 101 amino acids (AAs), including a signal peptide (21 AAs), spacer sequence (37 AAs), gastrin component (34 AAs), and a 9-AA extension at the carboxyl terminus. The enzymatic processing of preprogastrin

Chapter 1  Gastrointestinal Hormones and Neurotransmitters produces all the known physiologically active forms of gastrin. Preprogastrin is processed into progastrin and gastrin peptide fragments of various sizes by sequential enzymatic cleavage. The two major forms of gastrin are G34 and G17, although smaller forms exist. The common feature of all gastrins is an amidated tetrapeptide (Try-Met-Asp-Phe-NH2) carboxyl terminus, which imparts full biological activity. Modification by sulfation at tyrosine residues produces alternative gastrin forms of equal biological potency. A nonamidated form of gastrin known as glycine-extended gastrin is produced by colonic mucosa. Glycine-extended gastrin has been shown in animal models to stimulate proliferation of normal colonic mucosa and enhance the development of colorectal cancer. It is not known whether local production of this form of gastrin contributes to human colon carcinogenesis, and the receptor for glycine-extended gastrin has not been identified.16 Most gastrin is produced in endocrine cells of the gastric antrum.17 Much smaller amounts of gastrin are produced in other regions of the GI tract, including the proximal stomach, duodenum, jejunum, ileum, and pancreas. Gastrin has also been found outside the GI tract, including in the brain, adrenal gland, respiratory tract, and reproductive organs, although its biological role in these sites is unknown. The receptors for gastrin and CCK are related and constitute the so-called gastrin-CCK receptor family. The CCK-1 and CCK-2 (previously known as CCK-A and -B) receptor complementary DNAs were cloned from the pancreas and brain, respectively, after which it was recognized that the CCK-2 receptor is identical to the gastrin receptor of the stomach.18 The CCK-1 receptor is present in the gallbladder and, in most species, in the pancreas. The CCK-1 receptor has a 1000-fold higher affinity for CCK than for gastrin. The CCK-1– and CCK-2–gastrin receptors have more than 50% sequence homology and respond differentially to various receptor antagonists and to gastrin. Gastrin is released from specialized endocrine cells (G cells) into the circulation in response to a meal. The specific components of a meal that stimulate gastrin release include protein, peptides, and amino acids. Gastrin release is profoundly influenced by the pH of the stomach. Fasting and increased gastric acidity inhibit gastrin release, whereas a high gastric pH is a strong stimulus for its secretion. Hypergastrinemia occurs in pathologic states associated with decreased acid production, such as atrophic gastritis. Serum gastrin levels can also become elevated in patients on prolonged acid-suppressive medications, such as his­ tamine receptor antagonists and proton pump inhibitors. Hypergastrinemia in these conditions is caused by stimulation of gastrin production by the alkaline pH environment. Another important but far less common cause of hypergastrinemia is a gastrin-producing tumor, also known as Zollinger-Ellison syndrome (see Chapter 32). The gastrin analog, pentagastrin, has been used clinically to stimulate histamine and gastric acid secretion in diagnostic tests of acid secretory capacity (see Chapter 49).

CHOLECYSTOKININ

CCK is a peptide transmitter produced by I cells of the small intestine and is secreted into the blood following ingestion of a meal. Circulating CCK binds to specific CCK-1 receptors on the gallbladder, pancreas, smooth muscle of the stomach, and peripheral nerves to stimulate gallbladder contraction and pancreatic secretion, regulate gastric emptying and bowel motility, and induce satiety.19 These effects serve to

coordinate the ingestion, digestion, and absorption of dietary nutrients. Ingested fat and protein are the major food components that stimulate CCK release. CCK was originally identified as a 33–amino acid peptide. However, since its discovery larger and smaller forms of CCK have been isolated from blood, intestine, and brain. All forms of CCK are produced from a single gene by posttranslational processing of a preprohormone. Forms of CCK ranging in size from CCK-58 to CCK-8 have similar biological activities.20 CCK is the major hormonal regulator of gallbladder contraction. It also plays an important role in regulating mealstimulated pancreatic secretion (see Chapter 56) In many species, this latter effect is mediated directly through receptors on pancreatic acinar cells but in humans, in whom pancreatic CCK-1 receptors are less abundant, CCK appears to stimulate pancreatic secretion indirectly through enteropancreatic neurons that possess CCK-1 receptors. In some species, CCK has trophic effects on the pancreas, although its potential role in human pancreatic neoplasia is speculative. CCK also has been shown to delay gastric emptying.21 This action may be important in coordinating the delivery of food from the stomach to the intestine. CCK has been proposed as a major mediator of satiety and food intake, an effect that is particularly noticeable when food is in the stomach or intestine. CCK inhibits gastric acid secretion by binding to CCK-1 receptors on somatostatin (D) cells in the antrum and oxyntic mucosa. Somatostatin acts locally to inhibit gastrin release from adjacent G cells and directly inhibits acid secretion from parietal cells.22 Clinically, CCK has been used together with secretin to stimulate pancreatic secretion for pancreatic function testing. It is also used radiographically or scintigraphically to evaluate gallbladder contractility. There are no known diseases of CCK excess. Low CCK levels have been reported in individuals with celiac disease who have reduced intestinal mucosal surface area and in those with bulimia nervosa.23,24 Elevated levels of CCK have been reported in some patients with chronic pancreatitis (see Chapter 59), presumably because of reduced pancreatic enzyme secretion and interruption of negative feedback regulation of CCK release.25

SECRETIN

The first hormone, secretin, was discovered when it was observed that intestinal extracts, when injected intravenously into dogs, caused pancreatic secretion.26 Secretin is released by acid in the duodenum and stimulates pancreatic fluid and bicarbonate secretion, leading to neutralization of acidic chyme in the intestine (see Chapter 56). Secretin also inhibits gastric acid secretion (see Chapter 49) and intestinal motility. Human secretin is a 27–amino acid peptide and, similar to many other GI peptides, is amidated at the carboxyl terminus. It is the founding member of the secretin-glucagonVIP family of structurally related GI hormones. Secretin is selectively expressed in specialized enteroendocrine cells of the small intestine called S cells.27 The secretin receptor is a member of a large family of G protein–coupled receptors (GPCRs) that is structurally similar to receptors for glucagon, calcitonin, parathyroid hormone, pituitary adenylate cyclase–activating peptide (PACAP), and vasoactive intestinal polypeptide (VIP). One of the major physiological actions of secretin is stimulation of pancreatic fluid and bicarbonate secretion (see Chapter 56). Pancreatic bicarbonate, on reaching the duodenum, neutralizes gastric acid and raises the duodenal pH, thereby “turning off” secretin release (negative feedback). It

7

8

Section I  Biology of the Gastrointestinal Tract has been suggested that acid-stimulated secretin release is regulated by an endogenous intestinal secretin-releasing factor.28 This peptide stimulates secretin release from S cells until the flow of pancreatic proteases is sufficient to degrade the releasing factor and terminate secretin release. Although the primary action of secretin is to produce pancreatic fluid and bicarbonate secretion, it is also an enterogastrone, a substance that is released when fat is present in the GI lumen and that inhibits gastric acid secretion. In physiologic concentrations, secretin inhibits gastrin release, gastric acid secretion, and gastric motility.29 The most common clinical application of secretin is in the diagnosis of gastrin-secreting tumors,30 as discussed in Chapter 32.

Proglucagon

Glucagon

Pancreas

Glucagon

GLP-1

GLP-2

Small intestine

GLP-1

GLP-2

Figure 1-4.  Different post-translational processing of glucagon in the pancreas and small intestine. The glucagon gene transcript is transcribed and translated into a prohormone (proglucagon) capable of producing glucagon and glucagon-like peptides (GLP-1 and GLP-2). However, only glucagon is produced in the pancreas because of specific processing. In the small intestine, GLP-1 and GLP-2 are the primary products.

VASOACTIVE INTESTINAL POLYPEPTIDE

VIP is a neuromodulator that has broad significance in intestinal physiology. VIP is a potent vasodilator that increases blood flow in the GI tract and causes smooth muscle relaxation and epithelial cell secretion.31,32 As a chemical messenger, VIP is released from nerve terminals and acts locally on cells bearing VIP receptors. VIP belongs to a family of GI peptides, including secretin and glucagon, that are structurally related. The VIP receptor is a G protein–coupled receptor that stimulates intracellular cAMP generation. Like other GI peptides, VIP is synthesized as a precursor molecule that is cleaved to an active peptide of 28 amino acids. VIP is expressed primarily in neurons of the peripheral-enteric and central nervous systems and is released along with other peptides, including primarily PHI and/or PHM (see Table 1-1).33 VIP is an important neurotransmitter throughout the central and peripheral nervous systems.34 Because of its wide distribution, VIP has effects on many organ systems; most notably, in the GI tract, VIP stimulates fluid and electrolyte secretion from intestinal epithelium and bile duct cholangiocytes.35,36 VIP, along with NO, is a primary component of nonadrenergic, noncholinergic nerve transmission in the gut.37 GI smooth muscle exhibits a basal tone, or sustained tension, caused by rhythmic depolarizations of the smooth muscle membrane potential. VIP serves as an inhibitory transmitter of this rhythmic activity, causing membrane hyperpolarization and subsequent relaxation of GI smooth muscle. Accordingly, VIP is an important neuromodulator of sphincters of the GI tract, including the lower esophageal sphincter and sphincter of Oddi. In certain pathologic conditions, such as achalasia and Hirschsprung’s disease, the lack of VIP innervation is believed to play a major role in defective esophageal relaxation and bowel dysmotility, respectively.38,39 Unlike GI endocrine cells that line the mucosa of the gut, VIP is produced and released from neurons and it is likely that most measurable VIP in serum is of neuronal origin. Normally, serum VIP levels are low and do not appreciably change with a meal. However, in pancreatic cholera, also known as Verner-Morrison syndrome and manifested by watery diarrhea, hypokalemia, and achlorhydria,40 VIP levels can be extraordinarily high.35 VIP-secreting tumors usually produce a voluminous diarrhea41 (see Chapter 32).

GLUCAGON

Glucagon is synthesized and released from pancreatic alpha cells and from intestinal L cells of the ileum and colon. Pancreatic glucagon is a 29–amino acid peptide that regulates glucose homeostasis via gluconeogenesis, glycogenolysis, and lipolysis and is counterregulatory to insulin. The gene for glucagon encodes not only preproglucagon but also

glucagon-like peptides (GLPs). This precursor peptide consists of a signal peptide, a glucagon-related polypeptide, glucagon, and GLP-1 and GLP-2. Tissue-specific peptide processing occurs through prohormone convertases that produce glucagon in the pancreas and GLP-1 and GLP-2 in the intestine (Fig. 1-4).42 Glucagon and GLP-1 regulate glucose homeostasis.43 Glucagon is released from the endocrine pancreas in response to a meal and binds to G protein–coupled receptors on skeletal muscle and the liver to exert its glucoregulatory effects. GLP-1 stimulates insulin secretion and augments the insulin-releasing effects of glucose on the pancreatic beta cell (see later, “Enteroinsular Axis”). GLP-1 analogs have been developed for the treatment of type II diabetes mellitus. A long-acting human GLP-1 analog improves beta cell function and can lower body weight in patients with type II diabetes.44,45 GLP-2 is an intestinal growth factor and may have therapeutic implications in the maintenance of the GI mucosal mass and the reversal of villus atrophy.

GLUCOSE-DEPENDENT INSULINOTROPIC POLYPEPTIDE

GIP was discovered based on its ability to inhibit gastric acid secretion (enterogastrone effect) and was originally termed gastric inhibitory polypeptide. It was subsequently shown that the effects on gastric acid secretion occur only at very high concentrations that are above the physiologic range. However, GIP has potent effects on insulin release that (like GLP-1) potentiates glucose-stimulated insulin secretion.46 Based on this action, GIP was redefined as glucose-dependent insulinotropic polypeptide. GIP is a 42–amino acid peptide produced by K cells in the mucosa of the small intestine. GIP is released into the blood in response to ingestion of glucose or fat. In the presence of elevated blood glucose levels, GIP binds to its receptor on pancreatic beta cells, activating adenylate cyclase and other pathways that increase intracellular calcium concentrations, leading to insulin secretion. Importantly, however, the effects on insulin secretion occur only if hyperglycemia exists; GIP does not stimulate insulin release under normoglycemic conditions. GIP receptors are also expressed on adipocytes through which GIP augments triglyceride storage, which may contribute to fat accumulation. Based on the insulinotropic properties of GIP, coupled with its effects on adipocytes, it has been proposed that GIP may play a role in obesity and development of insulin resistance associated with type II diabetes mellitus.47 Consistent with this proposal was the experimental finding that mice lacking the GIP receptor do not gain weight when placed on a high-fat diet.48 It remains

Chapter 1  Gastrointestinal Hormones and Neurotransmitters to be seen whether GIP antagonists can be used to treat obesity. In rare circumstances, receptors for GIP may be aberrantly expressed in the adrenal cortex, resulting in fooddependent Cushing’s syndrome.49,50

PANCREATIC POLYPEPTIDE FAMILY

Originally isolated during the preparation of insulin, pancreatic polypeptide (PP) is the founding member of the PP family.51 The PP family of peptides includes NPY and peptide tyrosine tyrosine (PYY), which were discovered because of the presence of a C-terminal tyrosine amide.52,53 PP is stored and secreted from specialized pancreatic endocrine cells (PP cells),54 whereas NPY is a principal neurotransmitter found in the central and peripheral nervous systems.55 PYY has been localized to enteroendocrine cells throughout the GI tract but is found in greatest concentrations in the ileum and colon.56 The PP-PYY-NPY family of peptides functions as endocrine, paracrine, and neurocrine transmitters in the regulation of a number of actions that result from binding to one of five receptor subtypes.57 PP inhibits pancreatic exocrine secretion, gallbladder contraction, and gut motility.58 PYY inhibits vagally stimulated gastric acid secretion and other motor and secretory functions.59 An abbreviated form of PYY lacking the first two amino acids of the normally produced 36 amino acid peptide, PYY3-36, has been shown to reduce food intake when administered to humans, indicating that intestinally released peptide may play a role in regulating meal size.60 NPY is one of the most abundant peptides in the central nervous system and, in contrast to PYY3-36, is a potent stimulant of food intake.61 Peripherally, NPY affects vascular and GI smooth muscle function.62

SUBSTANCE P AND THE TACHYKININS

Substance P belongs to the tachykinin family of peptides, which includes neurokinin A and neurokinin B. The tachykinins are found throughout the peripheral and central nervous systems, and are important mediators of neuropathic inflammation.63 Tachykinins, as a group, are encoded by two genes that produce preprotachykinin A and pre­ protachykinin B. Common to both is a well-conserved C-terminal pentapeptide. Transcriptional and translational processing produce substance P, neurokinin A, and/or neurokinin B, which are regulated in large part by alternative splicing. These peptides function primarily as neuropeptides. Substance P is a neurotransmitter of primary sensory afferent neurons and binds to specific receptors in lamina I of the spinal cord.64 Three receptors for this family of peptides have been identified—NK-1, NK-2, and NK-3.65 Substance P is the primary ligand for the NK-1 receptor, neurokinin A for the NK-2 receptor, and neurokinin B for the NK-3 receptor. However, all these peptides can bind and signal through all three receptor subtypes. Substance P has been implicated as a primary mediator of neurogenic inflammation. In the intestine, Clostridium difficile–initiated experimental colitis results from toxininduced release of substance P and consequent activation of the NK-1 receptor.66 These inflammatory sequelae can be blocked by substance P receptor antagonists. Substance P receptors are more abundant in the intestine of patients with ulcerative colitis and Crohn’s disease.67

SOMATOSTATIN

Somatostatin is a 14–amino acid cyclic peptide that was initially identified as an inhibitor of growth hormone secretion. Since its discovery, it has been found in almost every organ in the body and throughout the GI tract. In the gut, somatostatin is produced by D cells in the gastric and intes-

tinal mucosa and islets of the pancreas, as well as enteric neurons.68 Somatostatin has a number of pharmacologic effects that are mostly inhibitory. In the stomach, somatostatin plays an important role in regulating gastric acid secretion.69 In the antrum, D cells are open to the lumen, where they are directly exposed to acid. A low gastric pH stimulates D cells that lie in close proximity to gastrin-producing cells to secrete somatostatin and inhibit gastrin release (see Chapter 49). Reduced gastrin secretion decreases the stimulus for acid production and the pH of the stomach contents rises. Thus, some of the inhib­ itory effects of gastric acid on gastrin release (see earlier, “Gastrin”) are mediated by somatostatin. Somatostatin release is also influenced by mechanical stimulation, dietary components of a meal, including protein, fat, and glucose, and other hormones and neurotransmitters.70 Muscarinic stimulation appears to be the most important neural stimulus to somatostatin secretion. At least five somatostatin receptors have been identified that account for divergent pharmacologic properties.71 For example, receptor subtypes 2 and 3 couple to inhibitory G proteins but receptor subtype 1 does not. In addition, only somatostatin receptor subtype 3 inhibits adenylate cyclase. The inhibitory effects of somatostatin are mediated by a decrease in cAMP, Ca2+ channel inhibition, or K+ channel opening. In the gut, somatostatin has broad inhibitory actions. In addition to effects on gastric acid, somatostatin reduces pepsinogen secretion. Somatostatin profoundly inhibits pancreatic enzyme, fluid, and bicarbonate secretion and reduces bile flow.72 The effects of somatostatin on gut motility are largely inhibitory, with the exception that it stimulates the migrating motor complex, possibly through effects on motilin. Somatostatin also reduces intestinal transport of nutrients and fluid, reduces splanchnic blood flow, and has inhibitory effects on tissue growth and proliferation.73,74 Because of its varied physiologic effects, somatostatin has several clinically important pharmacologic uses. Many endocrine cells possess somatostatin receptors and are sensitive to inhibitory regulation. Therefore, somatostatin and more recently developed somatostatin analogs are used to treat conditions of hormone excess produced by endocrine tumors, such as acromegaly, carcinoid tumors, and islet cell tumors (including gastrinomas).75 Its ability to reduce splanchnic blood flow and portal venous pressure has led to somatostatin analogs being useful in treating esophageal variceal bleeding (see Chapter 90).76 The inhibitory effects on secretion have been exploited by using somatostatin analogs to treat some forms of diarrhea and reduce fluid output from pancreatic fistulas. Many endocrine tumors express abundant somatostatin receptors, making it possible to use radiolabeled somatostatin analogs, such as octreotide, to localize even small tumors throughout the body.

MOTILIN

Motilin is a 22–amino acid peptide produced by endocrine cells of the duodenal epithelium.77 Motilin is secreted into the blood in a periodic and recurrent pattern that is synchronized with the migrating motor complex (MMC) under fasting conditions. Elevations in blood motilin levels regulate the phase III contractions that initiate in the antroduodenal region and progress toward the distal gut. Motilin secretion is not stimulated by eating. Motilin binds to specific receptors on smooth muscle cells of the esophagus, stomach, and small and large intestines through which it exerts propulsive activity.78 Agonists to the motilin receptor such as erythromycin have pronounced

9

10

Section I  Biology of the Gastrointestinal Tract effects on GI motility, which occasionally produces undesired side effects of abdominal cramping and diarrhea.79 However, motilin agonists may be useful to treat conditions of impaired gastric and intestinal motility and are being investigated for the treatment of constipation-predominant irritable bowel syndrome.80

LEPTIN

Leptin is a 167–amino acid protein that is secreted primarily from adipocytes. Blood leptin levels reflect total body fat stores.81 Its primary action appears to be to reduce food intake. Leptin is a member of the cytokine family of signaling molecules. Five different forms of leptin receptors have been reported.82 A short form of the receptor appears to transport leptin from the blood across the blood-brain barrier, where it has access to the hypothalamus. A long form of the leptin receptor is located in hypothalamic nuclei, where leptin binds and activates the Janis kinase signal transduction and translation system (JAK STAT).83 Small amounts of leptin are produced by the chief cells of the stomach and by the placenta, and are present in breast milk. Peripheral administration of leptin reduces food intake. However, this effect is reduced as animals become obese. Interestingly, when injected into the central nervous system, obese animals respond normally to leptin and reduce food intake, suggesting that leptin “resistance” in obesity occurs at the level of the leptin receptor that transports leptin across the blood-brain barrier.84 Leptin’s ability to reduce food intake occurs within the brain by decreasing NPY (a potent stimulant of food intake) and by increasing α– melanocyte-stimulating hormone (α−MSH), an inhibitor of food intake.85 Peripherally, leptin acts synergistically with cholecystokinin to reduce meal size.86 In obese rats lacking the leptin receptor, the synergistic effects of leptin plus CCK to reduce meal size are lost, but could be restored with genetic reconstitution of the leptin receptor in the brain.87 One might expect loss of leptin-CCK synergy on meal size in those rare cases of human obesity caused by leptin receptor defects or even with leptin resistance. Blood levels of leptin increase as obesity develops and leptin appears to reflect total fat content.88 At the cellular level, large adipocytes produce more leptin than small adipocytes. Because of its effects on food intake, it was initially thought that exogenous leptin could be used therapeutically to treat obesity. However, only a very modest effect on weight loss has been demonstrated in clinical trials. Leptin deficiency has been reported as a cause of obesity in a few families, but this condition is extremely rare.89,90 Mutation of the leptin receptor has been described as a cause of obesity in at least one family.91

GHRELIN

Ghrelin is a 28–amino acid peptide produced by the stomach and is the natural ligand for the growth hormone secretagogue (GHS) receptor.92 When administered centrally or peripherally ghrelin stimulates growth hormone secretion, increases food intake, and produces weight gain.93,94 Circulating ghrelin levels increase during periods of fasting or under conditions associated with negative energy balance, such as starvation or anorexia. In contrast, ghrelin levels are low after eating and in obesity. Ghrelin appears to play a central role in the neurohormonal regulation of food intake and energy homeostasis. The gastric fundus is the most abundant source of ghrelin, although lower amounts of ghrelin are found in the intestine, pancreas, pituitary, kidney, and placenta. Ghrelin is produced by distinctive endocrine cells known as P/D1

cells95,96 that are of two types, open and closed. The open type is exposed to the lumen of the stomach, where it comes into contact with gastric contents, whereas the closed type lies in close proximity to the capillary network of the lamina propria.97 Both cell types secrete hormone into the bloodstream. Based on its structure, ghrelin is a member of the motilin family of peptides and, like motilin, ghrelin stimulates gastric contraction and enhances stomach emptying. The observations that circulating ghrelin levels increase sharply before a meal and fall abruptly after a meal suggest that it serves as a signal for initiation of feeding. The effects of food on plasma ghrelin levels can be reproduced by ingestion of glucose and appear to be unrelated to the physical effects of a meal on gastric distention. Circulating ghrelin levels are low in states of positive energy balance such as obesity and are inversely correlated with body mass index.98,99 Conversely, ghrelin levels are high in fasting, cachexia, and anorexia. Importantly, weight loss increases circulating ghrelin levels.100 Ghrelin released from the stomach acts on the vagus nerve to exert its effects on feeding. However, it is also active when delivered to the central nervous system and, in this location, ghrelin activates NPY and agouti-related protein-producing neurons in the arcuate nucleus of the hypothalamus, which is involved in the regulation of feeding.94,101 Gastric bypass patients do not demonstrate the premeal increase in plasma ghrelin that is seen in normal individuals.102 This lack of ghrelin release may be one of the mechanisms contributing to the overall effectiveness of gastric bypass surgery for inducing weight loss. Prader-Willi syndrome is a congenital obesity syndrome characterized by severe hyperphagia, growth hormone deficiency, and hypogonadism. Although obesity is ordinarily associated with low ghrelin levels, patients with PraderWilli syndrome have high circulating ghrelin levels that do not decline after a meal.103,104 The levels of ghrelin in this syndrome are similar to those that can stimulate appetite and increase food intake in individuals receiving infusions of exogenous ghrelin, suggesting that abnormal ghrelin secretion may be responsible for the hyperphagia in PraderWilli syndrome.105

OTHER CHEMICAL MESSENGERS OF THE GASTROINTESTINAL TRACT The enteric nervous system, through intrinsic and extrinsic neural circuits, controls GI function. This control is mediated by various chemical messengers, including motor and sensory pathways of the sympathetic and parasympathetic nervous systems. The parasympathetic preganglionic input is provided by cholinergic neurons and elicits excitatory effects on GI motility via nicotinic and muscarinic receptors. Sympathetic input occurs through postganglionic adrenergic neurons.

ACETYLCHOLINE

Acetylcholine is synthesized in cholinergic neurons and is the principal regulator of GI motility and pancreatic secretion. Acetylcholine is stored in nerve terminals and released by nerve depolarization. Released acetylcholine binds to postsynaptic muscarinic and/or nicotinic receptors. Nicotinic acetylcholine receptors belong to a family of ligand-gated ion channels and are homopentamers or heteropentamers composed of α, β, γ, δ, and ε subunits.106 The

Chapter 1  Gastrointestinal Hormones and Neurotransmitters α subunit is believed to be the mediator of postsynaptic membrane depolarization following acetylcholine receptor binding. Muscarinic receptors belong to the heptahelical GPCR family. There are five known muscarinic cholinergic receptors (M1 to M5). Muscarinic receptors can be further classified based on receptor signal transduction, with M1, M3, and M5 stimulating adenylate cyclase and M2 and M4 inhibiting this enzyme. Acetylcholine is degraded by the enzyme acetylcholinesterase, and the products may be recycled through high-affinity transporters on the nerve terminal.

CATECHOLAMINES

The primary catecholamine neurotransmitters of the enteric nervous system include norepinephrine and dopamine. Norepinephrine is synthesized from tyrosine and released from postganglionic sympathetic nerve terminals that innervate enteric ganglia and blood vessels. Tyrosine is converted to dopa by tyrosine hydroxylase. Dopa is initially converted into dopamine by dopa decarboxylase and packaged into secretory granules. Norepinephrine is formed from dopamine by the action of dopamine β-hydroxylase in the secretory granule. After an appropriate stimulus, norepinephrine-containing secretory granules are released from nerve terminals and bind to adrenergic receptors. Adrenergic receptors are G protein–coupled, have seven typical membrane-spanning domains, and are of two basic types, α and β. α-Adrenergic receptors are further classified into α1A, α1B, α2A, α2B, α2C, and α2D. Similarly, β receptors include β1, β2, and β3. Adrenergic receptors are known to signal through various G proteins, resulting in stimulation or inhibition of adenylate cyclase and other effector systems. Norepinephrine signaling is terminated by intracellular monoamine oxidase or by rapid reuptake by an amine transporter. The actions of adrenergic receptor stimulation regulate smooth muscle contraction, intestinal blood flow, and GI secretion.

DOPAMINE

Dopamine is an important mediator of GI secretion, absorption, and motility and is the predominant catecholamine neurotransmitter of the central and peripheral nervous systems. In the central nervous system, dopamine regulates food intake, emotions, and endocrine responses and, peri­ pherally, it controls hormone secretion, vascular tone, and GI motility. Characterization of dopamine in the GI tract has been challenging for several reasons. First, dopamine can produce inhibitory and excitatory effects on GI motility.107 Generally, the excitatory response, which is mediated by presynaptic receptors, occurs at a lower agonist concentration than the inhibitory effect, which is mediated by postsynaptic receptors. Second, localization of dopamine receptors has been hampered by identification of dopamine receptors in locations that appear to be species specific.108 Third, studies of dopamine in GI tract motility have often used pharmacologic amounts of this agonist. Therefore, the interpretation of results has been confounded by the ability of dopamine to activate adrenergic receptors at high doses. Classically, dopamine was thought to act via two distinct receptor subtypes, type 1 and type 2. Molecular cloning has now demonstrated five dopamine receptor subtypes, each with a unique molecular structure and gene locus.108 Dopamine receptors are integral membrane GPCRs, and each receptor subtype has a specific pharmacologic profile when exposed to agonists and antagonists. After release from the nerve terminal, dopamine is cleared from the synaptic cleft by a specific dopamine transporter.

SEROTONIN

Serotonin has long been known to play a role in GI neurotransmission.109 The GI tract contains more than 95% of the total body serotonin, and serotonin is important in various processes, including epithelial secretion, bowel motility, nausea and emesis.110 Serotonin is synthesized from tryptophan, an essential amino acid, and is converted to its active form in nerve terminals. Secreted serotonin is inactivated in the synaptic cleft by reuptake via a serotonin-specific transporter. Most plasma serotonin is derived from the gut, where it is found in mucosal enterochromaffin cells and the enteric nervous system. Serotonin mediates its effects by binding to a specific receptor. There are seven different serotonin receptor subtypes found on enteric neurons, enterochromaffin cells, and GI smooth muscle (5-HT1 to 5-HT7). The actions of serotonin are complex (Fig. 1-5).111 It can cause smooth muscle contraction through stimulation of cholinergic nerves or relaxation by stimulating inhibitory NO-containing neurons.110 Serotonin released from mucosal cells stimulates sensory neurons, initiating a peristaltic reflex and secretion (via 5-HT4 receptors) and modulates sensation through activation of 5-HT3 receptors.109 The myenteric plexus contains serotoninergic interneurons that project to the submucosal plexus and ganglia extrinsic to the bowel wall. Extrinsic neurons activated by serotonin participate in bowel sensation and may be responsible for abdominal pain, nausea, and symptoms associated with irritable bowel syndrome. Intrinsic neurons activated by serotonin are primary components of the peristaltic and secretory reflexes responsible for normal GI function. Serotonin may also activate vagal afferent pathways and, in the central nervous system, modulates appetite, mood, and sexual function. Because of these diverse effects, it is not surprising that selective serotonin reuptake inhibitor drugs (SSRIs), commonly used to treat depression and anxiety, have pro­minent GI side effects when compared with placebo treatment. Serotonin and its receptor have been implicated in the pathogenesis of motility disorders of the GI tract.112 Characterization of specific serotonin receptor subtypes has led to the development of selective agonists and antagonists for the treatment of irritable bowel syndrome and chronic constipation and diarrhea. For example, 5-HT3 receptor antagonists, which reduce intestinal secretion, are used to treat diarrhea-predominant irritable bowel syndrome. 5-HT4 receptor agonists elicit prokinetic effects and are used to treat constipation-predominant irritable bowel syndrome and other motility disorders.113,114 Serotonin can also be enzymatically converted to melatonin by serotonin N-acetyltransferase.115 Other than the pineal gland, the GI tract is the major source of the body’s melatonin. Melatonin is produced in enterochromaffin cells and released into the blood after ingestion of a meal. A number of actions on the GI tract have been described for melatonin, including reducing gastric acid and pepsin secretion, inducing smooth muscle relaxation, and preventing epithelial injury through an antioxidant effect.116 It has been proposed that melatonin released after a meal may contribute to postprandial somnolence.117

HISTAMINE

In the GI tract, histamine is best known for its central role in regulating gastric acid secretion (see Chapter 49) and intestinal motility. Histamine is produced by enterochromaffin-like cells of the stomach and intestine as well as enteric nerves. Histamine is synthesized from l-histidine by histidine decarboxylase and activates three

11

12

Section I  Biology of the Gastrointestinal Tract

CNS Longitudinal muscle 5-HT3 Excitatory motor neuron

5-HT4

5-HT4

Myenteric plexus

5-HT3

Inhibitory motor neuron

Circular muscle

5-HT3

Intrinsic primary

Figure 1-5.  Role of serotonin in the enteric nervous system. This model illustrates the location of 5hydroxytryptamine3 (5-HT3) and 5-HT4 receptor subtypes in the GI tract. CNS, central nervous system. (Modified from Talley NJ. Serotoninergic neuroenteric modu­ lators. Lancet 2001; 358:2061-8).

afferent neuron 5-HT3

5-HT4

Sensory neuron

Extrinsic Submucosal afferent plexus neuron 5-HT3 Mucosa

Enterochromaffin cells

NO bound to guanylyl cyclase NO synthase Arginine

Rapid diffusion of NO

GTP

cGMP

Relaxation

Nitric oxide Smooth muscle cell Activated neuron Figure 1-6.  Nitric oxide (NO) signals smooth muscle relaxation. NO, synthesized from arginine by nitric oxide synthase, diffuses across the plasma membrane into smooth muscle cells. NO binds to and activates guanylyl cyclase, which converts GTP to cGMP. cGMP causes smooth muscle relaxation. (Modified from Alberts B, Bray D, Lewis J, et al, editors. Molecular biology of the cell. 4th ed. New York: Garland Science; 2002. p 831.)

GPCR subtypes. H1 receptors are found on smooth muscle and vascular endothelial cells and are linked to phospholipase C (PLC) activation. As such, the H1 receptor mediates many of the allergic responses induced by histamine. H2 receptors are present on gastric parietal cells, smooth muscle, and cardiac myocytes. H2 receptor binding stimulates Gs (G proteins that stimulate adenylate cyclase) and activates adenylate cyclase. H3 receptors are present in the central nervous system and GI tract enterochromaffin cells. These receptors signal through Gi and inhibit adenylate cyclase.118 Histamine can also interact with the N-methyl-daspartate (NMDA) receptor and enhance activity of NMDAbearing neurons independently of the three known histamine receptor subtypes. Unlike other neurotransmitters, there is no known transporter responsible for termination of histamine’s action. However, histamine is metabolized to telemethylhistamine by histamine N-methyltransferase and is then degraded to telemethylimidazoleacetic acid by monoamine oxidase B and an aldehyde dehydrogenase.

known as endothelial NOS and neuronal NOS, respectively, and are constitutively active. Small changes in NOS activity can occur through elevations in intracellular calcium. The inducible form of NOS (type II) is apparent only when cells become activated by specific inflammatory cytokines. This form of NOS is capable of producing large amounts of NO and is calcium-independent. NOS is often colocalized with VIP and PACAP in neurons of the enteric nervous system.120 NO, being an unstable gas, has a relatively short half-life. Unlike most neurotransmitters and hormones, NO does not act via a membrane-bound receptor. Instead, NO readily diffuses into adjacent cells to activate guanylate cyclase directly (Fig. 1-6). NO activity is terminated by its oxidation to nitrate and nitrite. Many enteric nerves use NO to signal neighboring cells and induce epithelial secretion, vasodilation, or muscle relaxation. NO is also produced by macrophages and neutrophils to help kill invading organisms.121

NITRIC OXIDE

Adenosine is an endogenous nucleoside that acts through any of four GPCR subtypes.122 Adenosine causes relaxation of intestinal smooth muscle and stimulates intestinal secretion. Adenosine can also cause peripheral vasodilation and

NO is a unique chemical messenger produced from larginine by the enzyme nitric oxide synthase (NOS).119 Three types of NOS are known. Types I and III are also

ADENOSINE

Chapter 1  Gastrointestinal Hormones and Neurotransmitters activation of nociceptors that participate in neural pain pathways.

Extracellular

CYTOKINES

Cytokines are a group of polypeptides produced by various immunomodulatory cells and are involved in cell proliferation, immunity, and inflammation. Cytokines are induced by specific stimuli, such as toxins produced by pathogens, and often elicit a complex response involving other cellular mediators to eradicate the foreign substance. Cytokines may be categorized as interleukins (ILs), tumor necrosis factors (TNFs), lymphotoxins, interferons, colony-stimulating factors (CSFs), and others.123 Interleukins can be further subtyped into at least 35 separate substances, IL-1 to IL-35. There are two TNFs, TNF-α and TNF-β, which are also known as lymphotoxin-α. Interferons are produced during viral or bacterial infection and come in two varieties, interferon-α (also known as leukocyte-derived interferon or interferon-β) and interferon-γ. Interferon-α is produced by T lymphocytes and is used clinically for the treatment of viral hepatitis (see Chapters 78 and 79). The major CSFs are granulocyte mononuclear phagocyte CSF, mononuclear phagocyte CSF, and granulocyte CSF. These agents are used for chemotherapy-induced neutropenia and marrow support after bone marrow transplantation. Chemokines initiate and propagate inflammation and are of two groups, CXC (α chemokines) and CC (β chemokines). Other cytokines, such as transforming growth factor-β (TGF)-β and platelet-derived growth factor (PDGF), have proliferative effects.

Intracellular

Figure 1-7.  Molecular structure of a typical heptahelical G protein–coupled receptor. The amino terminus is extracellular and of variable length. It often contains N-linked glycosylation sites (Y) important in ligand binding. There are seven membrane-spanning domains and intracellular loops that contain sites for G protein binding and possible phosphorylation residues (orange circles).

Ligand

Extracellular

SIGNAL TRANSDUCTION Cells live in a constantly changing milieu. The structure and biochemical nature of this environment are dynamic and, for cells to function normally, they must be able to access this changing information. The biochemical mediators of this information are cell surface receptors and transmitters. Receptors transduce signals from the extracellular space to the intracellular compartment. Each step in the process from receptor activation to receptor desensitization, internalization, and resensitization represents a potential regu­ latory checkpoint and possible target for therapeutic intervention. Cell surface receptors include GPCRs, ion channels, and enzyme-coupled receptors.

α

β γ

Effector

Intracellular events Figure 1-8.  Hormones (ligands) bind to specific G protein–coupled receptors at a unique location within the receptor-binding pocket. On binding, the receptor conformation is altered so that a specific G protein α subunit is activated. G protein activation leads to dissociation of the α subunit from the βγ subunit and activation of effector pathways. These effectors include adenylate cyclase, ion channels, and an array of other systems. Intracellular

G PROTEIN–COUPLED RECEPTORS

GPCRs are seven membrane-spanning domain proteins associated with a heterotrimeric G protein (Fig. 1-7). The membrane regions consist of α-helical domains with a conserved structural motif.124 GPCRs contain an extracellular amino terminus and an intracellular carboxyl terminus (see Fig. 1-7). When stimulated by the appropriate chemical messenger, the GPCR undergoes a conformational change and couples to a specific G protein. The first crystal structure of a GPCR, for rhodopsin, was elucidated in 2000.125 The three-dimensional structure of the rhodopsin receptor reveals a highly organized heptahelical transmembrane component with a portion of the C-terminus perpendicular to the seventh and final membrane-spanning domains of the protein.

G PROTEINS

G proteins are molecular intermediaries that initiate the intracellular communication process on ligand binding to its GPCR (Fig. 1-8).126 G proteins are composed of three

subunits—α, β, and γ—and are classified according to their α subunit. They activate various effector systems, including adenylate cyclase, guanylate cyclase, phospholipases, and specific ion channels.127 G proteins that stimulate adenylate cyclase are classified as Gs; those that inhibit adenylate cyclase are called Gi.128 When an agonist binds to a Gs-coupled receptor, a conformational change occurs, allowing the receptor to associate with the Gαs subunit. Under basal (unstimulated) conditions, Gαs is bound to guanosine diphosphate (GDP); however, with agonist binding, GDP is released and replaced with guanosine triphosphate (GTP). The Gs-GTP complex then activates adenylate cyclase, resulting in the generation of cAMP from adenosine triphosphate (ATP) within the cell cytoplasm. cAMP phosphorylates effector proteins that ultimately lead to responses such as secretion, cell movement, and growth. Receptor activation also initiates

13

14

Section I  Biology of the Gastrointestinal Tract Table 1-2  Classification of G Protein α Subunits and Their Signaling Pathways Class

Signaling

Gαs Gαi and Gαo

Adenylate cyclase, calcium channels Adenylate cyclase, cyclic guanosine monophosphate, phosphodiesterase, c-Src, STAT 3 Phospholipase C-β Sodium-hydrogen exchange

Gαq Gα12 and Gα13

the dissociation of the α subunit from the βγ subunits. However, the βγ subunits remain tightly associated and themselves participate in a vast array of cellular signals. For example, not only can βγ subunits activate GPCR kinases, adenylate cyclase, and ion channels, they induce receptor desensitization and stimulate Ras-mediated mitogenactivated protein (MAP) kinase.129,130 The Gαs-GTP complex is gradually inactivated by guanosine triphosphatase (GTPase), which converts GTP to GDP. This enzymatic conversion occurs spontaneously by the G protein, which is itself a GTPase. The conversion of GTP to GDP terminates G protein stimulation of adenylate cyclase and is one way whereby the basal condition is restored. Certain GPCRs activate an inhibitory G protein (Gαi) that inhibits cAMP accumulation and antagonizes the effects of Gs-coupled events. In this manner, GPCRs can maintain fine control of the cellular cAMP concentration and subsequent intracellular signaling. Members of this GPCR family also activate phospholipases and phosphodiesterases, and are often involved with ion channel regulation. Other GPCRs couple with Gq and G12 (see Table 1-2). The Gq family of G protein subunits regulates the production of inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG).131 Following α subunit dissociation from βγ, when the α subunit reverts to the GDP-bound form, it reassociates with βγ. With reestablishment of the αβγ heterotrimer, along with other mechanisms of desensitization, receptor signaling via the separate subunits ceases.

Effector Systems

Following receptor occupation, G protein subunits cause activation of enzymes or other proteins, ultimately resulting in intracellular signaling events (Table 1-2). Enzymes such as adenylate cyclase or phospholipase C generate specific second messengers such as cAMP or IP3 and DAG. Some G proteins couple directly to specific ion channels, such as potassium or calcium channels, and initiate changes in ion permeability. The effector systems are not well understood for some receptors, such as those involved with cell growth and differentiation. Other G proteins such as Go may activate the phosphoinositide system. When bound to hormone, receptors that couple to Go activate PLC, which acts on inositol phospholipids found in the cell membrane. PLC can cause the hydrolysis of phosphatidylinositol 4,5-bisphosphate, generating 1,2-DAG and IP3. DAG and IP3 can regulate cell metabolism by increasing intracellular calcium levels.

Receptor Desensitization

To ensure the rapidity of hormone signaling, shortly after receptor stimulation, a series of events is initiated that ultimately acts to turn off signaling. The principal events in this process involve receptor desensitization and internalization, which reestablish cell responsiveness.

Phosphorylation of the receptor is one of the initial events involved in turning off the signal after agonist binding and occurs through binding of arrestin-like molecules, which uncouple the receptor from the G protein.132 This uncoupling and subsequent receptor internalization (sequestration) continue the process of signal termination and eventually lead to the reestablishment of cell responsiveness.

Receptor Resensitization

Internalization or sequestration of the receptor occurs within minutes of receptor occupancy. Agonist-activated receptors are phosphorylated by G protein–coupled receptor kinases at specific intracellular sites, which causes G protein uncoupling and initiates receptor endocytosis. GPCR endocytosis is followed by receptor dephosphorylation, recycling, and down-regulation. Chronic exposure of cells to high concentrations of hormones frequently leads to a decrease in cell surface–binding sites. This reduction in surface receptor expression is termed down-regulation and is the result of receptor internalization. The mechanisms used by the cell that distinguish receptor internalization and recycling from down-regulation are not clear. However, long-term agonist exposure to some receptors has been shown to activate signaling molecules that may be important in receptor down-regulation.

RECEPTORS NOT COUPLED TO G PROTEINS Enzyme-Coupled Receptors

Receptor Tyrosine Kinases Unlike GPCRs, where ligand-receptor interaction causes activation of a G protein intermediary, some ligand receptors possess intrinsic protein tyrosine kinase activity. These membrane-spanning cell surface receptors catalyze the transfer of phosphate from ATP to target proteins. Such receptors are structurally unique in that they contain glycosylated extracellular binding domains, a single transmembrane domain, and a cytoplasmic domain. The cytoplasmic domain contains a protein tyrosine kinase region and substrate region for agonist-activated receptor phosphorylation. With activation, these receptors may phosphorylate themselves or be phosphorylated by other protein kinases.133 In general, receptor tyrosine kinases exist in the cell membrane as monomers. However, with ligand binding, these receptors dimerize, autophosphorylate, and initiate other intracellular signal transduction pathways. Most receptor tyrosine kinases couple, via ligand binding, to Ras and subsequently activate MAP kinase. MAP kinase is then able to modulate the regulation of other cellular proteins, including transcription factors. Members of the receptor tyrosine kinase family include the insulin receptor, growth factor receptors (vascular endothelial growth factor, PDGF, epi­ dermal growth factor [EGF], fibroblast growth factor [FGF], insulin-like growth factor I [IGF] I, macrophage-CSF, nerve growth factor), and receptors involved in development.134 Receptor tyrosine kinases are discussed further in Chapter 3 in relation to cellular growth and neoplasia. Activated tyrosine kinase receptors participate in a number of intracellular signaling events that involve the phosphorylated cytoplasmic domain. Specific phosphorylated tyrosine residues serve as binding sites for Src homology regions 2 and 3 (SH2 and SH3 domains). The result of SH2 domain binding is activation or modulation of the signaling protein that contains this binding domain. In this manner, receptor tyrosine kinases activate diverse signaling pathways.135

Chapter 1  Gastrointestinal Hormones and Neurotransmitters Receptor Guanylate Cyclases Receptor guanylate cyclases use cyclic GMP (cGMP) as a direct intracellular mediator. These cell surface receptors contain an extracellular ligand-binding region, a single transmembrane domain, and a cytoplasmic guanylate cyclase catalytic domain.136 Ligand stimulation of a receptor guanylate cyclase results in activation of cGMP-dependent protein kinase, which is a serine-threonine protein kinase. The atrial natriuretic peptide (ANP) receptor is a representative receptor guanylate kinase, which mediates the potent smooth muscle relaxing activity of ANP. Nonreceptor Tyrosine Kinases Some cell surface receptors involved in inflammation and hematopoietic cell regulation work through tyrosine kinases but do not contain a cytoplasmic catalytic domain. The Src family of kinases is the primary component of this receptor signaling system.137 Receptor Tyrosine Phosphatases Leukocyte regulation is modulated by surface receptors whose function is to remove phosphate groups from specific phosphotyrosines. CD45 is a surface protein found in white blood cells that participates in T and B cell activation.138 CD45 contains a single membrane-spanning domain and a cytoplasmic region with tyrosine phosphatase activity. Depending on the substrate, dephosphorylation of signaling proteins may result in reduced or enhanced activity. Receptors in this family are important in inflammation and immune regulation and have been shown to participate in GI development, growth, and cancer. Receptor Serine-Threonine Kinases TGF-β (see “Growth Factor Receptors”) receptors are a unique group of surface proteins that are involved in various cell functions, including chemotaxis, inflammation, and proliferation. These receptors contain a single membrane domain and a cytoplasmic serine-threonine kinase region. Receptor stimulation initiates activation of the serinethreonine kinase and subsequent modulation of cellular protein function.139

are normally secreted in response to food ingestion and mediate many of the nutrient effects on the GI tract. They play a key role in cellular proliferation. Alterations in intestinal proliferation are manifested by atrophy, hyperplasia, dysplasia, or malignancy (see Chapter 3). Growth factors that have important effects on the GI tract include peptides of the EGF, TGF-β, IGF, FGF, and PDGF families, hepatocyte growth factors, trefoil factors, and many cytokines (including interleukins).140

GROWTH FACTOR RECEPTORS

Growth factors regulate cellular proliferation by interacting with specific cell surface receptors. These receptors are membrane proteins that possess specific binding sites for the growth factor ligand. An unusual form of signaling occurs when the ligand interacts with its receptor within the same cell. For example, PDGF receptors present on the intracellular surface of fibroblast cell lines are activated by intracellular ligand. This process is known as intracrine signaling. Most peptide growth factors, however, interact with receptors on different cells to regulate proliferation. Growth factor receptors can be single polypeptide chains containing one membrane-spanning region, such as the receptor for EGF, or they may be composed two subunit heterodimers, with one subunit containing a transmembrane domain and the other residing intracellularly but covalently bound to the transmembrane subunit (Fig. 1-9). Heterodimers may also dimerize to form a receptor composed of four subunits (e.g., IGF receptor). Binding of the ligand to its receptor usually causes aggregation of two or more receptors and activation of intrinsic tyrosine kinase activity. Growth factor receptors also have the ability to autophosphorylate when bound to ligand. In addition, receptor tyrosine kinase activity may phosphorylate other intracellular proteins important in signal transduction. Autophosphorylation attenuates the receptor’s kinase activity and often leads to down-regulation and internalization of the receptor. Mutation of the receptor at its autophos-

Proteoglycan

Ion Channel–Coupled Receptors

Ion channel–coupled receptors are involved in rapid signaling between cells. Ion channel signaling is particularly important in nerve cells and other electrically excitable tissues such as muscle. In nerve cells, a relatively small number of neurotransmitters are released that act directly on ion channel proteins, causing them to open or close. Ion channels are selective to specific anions or cations and, when open, allow the flow of those particular ions across the plasma membrane according to the concentration inside and outside the cell. This flow of ions regulates the excitability of the target cell, which can trigger cellular responses such as neurotransmission, muscle contraction, electrolyte and fluid secretion, and hormone release.

HORMONE AND TRANSMITTER REGULATION OF GASTROINTESTINAL GROWTH Growth of GI tissues is a balance between cellular proliferation and senescence. Many factors participate in maintenance of the GI mucosa. Nutrients and other luminal factors stimulate growth of the intestinal mucosa and are necessary to maintain normal digestive and absorptive functions. Hormones and transmitters serve as secondary messengers that

Plasma membrane

Cytoplasm

EGF receptor

Insulin receptor IGF-I receptor

PDGF TGFβ-I TGFβ-II TGFβ-III receptor receptor receptor receptor Tyrosine kinase Serine/Threonine kinase

Figure 1-9.  Growth factor receptors in the gastrointestinal tract. Schematic examples of growth factor receptor families are depicted in relation to the cell surface. Receptor regions that contain kinase activity are shown in boxes. On activation, these receptors have the ability to autophosphorylate or phosphorylate other proteins to propagate intracellular cell signaling. EGF, epidermal growth factor; IGF-I, insulin-like growth factor I; PDGF, platelet-derived growth factor; TGF, transforming growth factor (β-I, -II, -III). (Modified from Podolsky DK: Peptide growth factors in the gastro­ intestinal tract. In Johnson LR, editor. Physiology of the gastrointestinal tract. New York: Raven Press; 1994. p 129.)

15

16

Section I  Biology of the Gastrointestinal Tract phorylation site may lead to constitutive receptor activity and cellular transformation. Growth factor receptors may couple to various intracellular signaling pathways, including adenylate cyclase, phospholipase C, calciumcalmodulin protein kinases, MAP kinase, and nuclear transcription factors. Thus, growth factors play important and varied roles in most cells of the GI tract. It is not surprising, therefore, that mutations in growth factor receptors or downstream signaling proteins can lead to unregulated cell growth and neoplasia (see Chapter 3). An important action of growth factors is their ability to modulate the expression of transacting transcription factors that can regulate expression of many other genes.141 Early response genes such as jun and fos are activated rapidly after ligand binding and control the expression of many other genes involved in cellular proliferation. Other important transcriptional factors include c-myc and nuclear factor κB (NF-κB). The latter is found in the cytoplasm in an inactive form and, following ligand binding, translocates to the nucleus, where it activates other transcription factors. NF-κB is a key target for strategies to regulate cellular proliferation and inflammation. In its phosphorylated form Rb-1, originally identified in retinoblastoma, is an inhibitor of cellular proliferation that complexes with the tran­ scription factor p53. Dephosphorylation of Rb-1 releases p53, which activates other genes leading to cellular proliferation. Almost all growth factors of the GI tract exert paracrine effects. However, many growth factors also possess autocrine and even intracrine actions. It has become apparent that growth factors and other signaling molecules secreted into the lumen of the gut can have important local biological actions. Distant effects of growth factors found in the circulation may be important for growth of certain types of cancers, particularly lung and colon cancer.

EPIDERMAL GROWTH FACTOR

EGF was the first growth factor to be discovered. It is the prototype for a family of growth factors that are structurally related and have similarly related receptors. Other members of the family include TGF-α, amphiregulin, and heparinbinding EGF. EGF is identical to urogastrone (originally isolated from urine), which was shown to inhibit gastric acid secretion and promote healing of gastric ulcers. EGF is secreted from submaxillary glands and Brunner’s glands of the duodenum. It is likely that EGF interacts with luminal cells of the GI tract to regulate proliferation. EGF has important trophic effects on gastric mucosa, and the wide distribution of EGF receptors suggests that EGF has mitogenic actions on various cells throughout the gut. The EGF receptor has been reported to be responsible for gastric hyperplasia in patients with Ménétrier’s disease.142 Moreover, two patients were effectively treated with a mono­ clonal antibody that blocks ligand binding to the EGF receptor.143 EGF receptors are considered important targets for the experimental treatment of human cancer based on the evidence that they play a critical role in the growth and survival of certain tumors. Monoclonal antibodies as well as small tyrosine kinase inhibitors have been undergoing clinical evaluation for the treatment of human tumors.144

TRANSFORMING GROWTH FACTOR-α

TGF-α is produced by most epithelial cells of the GI tract and acts through the EGF receptor. Therefore, it shares trophic properties with EGF. It is believed to play a key role in gastric reconstitution after mucosal injury. Moreover, it

appears to be important in intestinal neoplasia because most gastric and colon cancers produce TGF-α (see Chapters 54 and 123).

TRANSFORMING GROWTH FACTOR-β

A family of TGF-β peptides exerts various biological actions, including stimulation of proliferation, differentiation, embryonic development, and formation of extracellular matrix.139 In contrast with the TGF-α receptor, there are three distinct TGF-β receptors (see Fig. 1-9).145 TGF-β modulates cell growth and proliferation in nearly all cell types and can enhance its own production from cells. It is likely that TGF-β plays a critical role in inflammation and tissue repair. TGF-β augments collagen production by recruitment of fibroblasts through its chemoattractant properties. This action can have beneficial or deleterious effects, depending on its site of deposition and abundance. For example, TGF-β may play a key role in the development of adhesions following surgery.146

INSULIN-LIKE GROWTH FACTORS

Alternative splicing of the insulin gene produces two structurally related peptides, IGF I and IGF II.147 IGFs signal through at least three different IGF receptors. The IGF I receptor is a tyrosine kinase, and the IGF II receptor is identical to the mannose 6-phosphate receptor. Although the exact function of IGFs in the GI tract is not clearly understood, they have potent mitogenic activity in intes­ tinal epithelium. IGF II appears to be critical for embryonic development.

FIBROBLAST GROWTH FACTOR AND PLATELET-DERIVED GROWTH FACTOR

At least seven related FGFs have been identified.148 These peptides have mitogenic effects on various cell types, including mesenchymal cells, and likely play an important role in organogenesis and neovascularization.149 Although not unique to the GI tract, PDGF is one of the most thoroughly studied growth factors. It is important for fibroblast growth, and its receptor is expressed in the liver and throughout the GI tract, where it appears to promote wound healing.

TREFOIL FACTORS

Trefoil factors (pS2, spasmolysin, and intestinal trefoil factor, also known as TTF1, 2, and 3, respectively) are a family of proteins expressed throughout the GI tract.150 They share a common structure, having six cysteine residues and three disulfide bonds, creating a cloverleaf appearance that stabilizes the peptide within the gut lumen. The pS2 peptide is produced in the gastric mucosa, spasmolysin is found in the antrum and pancreas, and intestinal trefoil factor is produced throughout the small and large intestines. These peptides are produced by mucous neck cells in the stomach or goblet cells in the intestine and are secreted onto the mucosal surface of the gut. It is likely that trefoil factors act on the apical surface of the epithelial cells, where they have growth-promoting properties on the GI mucosa.

OTHER G PROTEIN–COUPLED RECEPTORS

Other peptides signaling through GPCRs may also have growth-promoting effects. Three important examples include gastrin, CCK, and gastrin-releasing peptide (GRP). Gastrin stimulates the growth of enterochromaffin-like cells of the stomach and induces proliferation of the oxyntic mucosa containing parietal cells.151 Gastrin binds to CCK-2

Chapter 1  Gastrointestinal Hormones and Neurotransmitters receptors of the stomach and activates PLC and Ras pathways, which ultimately results in activation of protein kinase C and MAP kinase, respectively. MAP kinase, which can also be activated by tyrosine kinase receptors typical of growth factors, causes the phosphorylation of transcription factors that are involved in cellular proliferation. In some cells, cAMP and protein kinase A exert synergistic effects on cellular growth through activation of nuclear transcription factors such as cAMP-responsive element binding (protein) (CREB). However, in other cells, cAMP antagonizes proliferation. Therefore, depending on the cell type, the effects of growth factors such as EGF, IGF, and PDGF may be enhanced by hormones that stimulate cAMP production. Certain colon cancer cells possess CCK-2 receptors and respond to the proliferative effects of gastrin. Moreover, gastrin may be produced by some colon cancers, enabling it to exert an autocrine effect to promote cancer growth.152 Whether circulating gastrin initiates colon cancer development is unknown. CCK binds preferentially to the CCK-1 type receptor, which is abundant in gallbladder, the pancreas of many species, brain, and peripheral nerves of the gut. In the rodent, but not human, pancreas, CCK causes hypertrophy and hyperplasia of pancreatic acinar cells. Similar to the effects of gastrin, CCK activates phospholipase C and small GTP-binding proteins to activate MAP kinase. In animal models, CCK can promote pancreatic cancer growth.16 GRP (the mammalian analog of bombesin) was first recognized for its ability to stimulate gastrin secretion from the stomach. Neurons containing GRP are abundant in the oxyntic mucosal of the proximal stomach and, based on studies with a specific GRP antagonist, appear to play a major role in the cephalic phase of gastric acid secretion.153 It was later appreciated that GRP stimulates proliferation of G cells. GRP has received considerable attention for its growth-promoting effects on small cell lung cancer, pancreatic cancer, and certain colon cancers.154

TASTE RECEPTORS

The GI tract contains specialized taste receptor cells that detect chemicals and toxins. These cells are best characterized in the tongue, where they are concentrated in taste buds. Taste receptor cells can detect nutrients such as proteins, fats, sugars, and salt at submolar concentrations and other chemicals such as drugs and toxins at submicromolar concentrations.155 Detection of these chemical signals is important for several reasons. Distinguishing among chemicals can warn of dangerous chemical ingestants and induce a vomiting reflex, thus protecting the organism against poisoning. Alternatively, sensing various foods through taste can be pleasant and encourage food intake as well as facilitate digestion by stimulating salivary, gastric, and pancreatic secretions. Taste receptor cells contain a wide number of receptors and ion channels that serve as targets for different tastants.156 The membrane proteins can respond to molecules as simple as ions or as complex as fatty acids and proteins. Ions and amino acids activate taste receptor cells by directly interacting with ion channels. Such interactions often cause membrane depolarization and induce Ca2+ entry into the cell. Other types of molecules such as sweet and bitter tastants activate G protein–coupled receptors and stimulate production of second messengers, such as cAMP or inositol trisphosphate, which stimulate intracellular signaling processes. A large number of taste receptors have been identified. A family of receptors that detects sweet compounds, including certain l amino acids, are known as T1R1, 2, and 3. A larger

family, known as T2Rs, consisting of over 30 members, mediates bitter gustatory signals. The T1R and T2R families of receptors are G protein–coupled receptors that couple to α subunits, including α-gustducin and α-transducin. Interestingly, although taste receptors are abundant in the tongue, T2Rs, α-gustducin, and α-transducin have been identified in the gastric and intestinal mucosa and in the pancreas. In the intestine, T2Rs have been localized in some enteroendocrine and brush border cells. It is possible that chemical messengers stimulate T2Rs in these cells to secrete hormones or induce cellular responses, such as production of nitric oxide.157

INTRALUMINAL RELEASING FACTOR REGULATION OF GASTROINTESTINAL HORMONES Most GI hormones are secreted into the blood following the ingestion of a meal. However, the exact mechanism whereby luminal nutrients stimulate hormone secretion is unknown. Although the apical surface of most enteric endocrine cells is exposed to the intestinal lumen (“open cells”), it is unclear whether nutrients interact with specific receptors on the surface of endocrine cells or whether they are absorbed and then stimulate hormone secretion. It has been recognized that specific releasing factors for GI hormones are present in the lumen of the gut (Fig. 1-10). CCK was the first hormone shown to be regulated by an intraluminal releasing factor.158,159 Luminal CCK-releasing factor was purified from intestinal washings and shown to stimulate CCK release when instilled into the lumen of animals. Diazepam-binding inhibitor has also been shown to stimulate CCK release, as has a pancreatic peptide known as monitor peptide.160,161 Secretin may also be regulated by an intraluminal releasing factor.28 The existence of these releasing factors underscores the significance of bioactive peptides within the lumen of the gut.

CCK-RF

Trypsin

Food

Pancreas

Monitor peptide

CCK Figure 1-10.  Regulation of cholecystokinin (CCK) secretion by intraluminal releasing factors. Endocrine cells containing CCK are stimulated by trypsinsensitive releasing factors (CCK-RF) present in the lumen of the GI tract. Releasing factors secreted from the intestine are responsible for negative feedback regulation of pancreatic secretion. Under basal conditions, local trypsin inactivates CCK-RF; however, with ingestion of nutrients that compete as substrates for trypsin, CCK-RF is available to stimulate CCK secretion. Monitor peptide, a pancreatic releasing factor, may contribute to sustained CCK release and pancreatic secretion after a meal.

17

18

Section I  Biology of the Gastrointestinal Tract GASTROINTESTINAL PEPTIDES THAT REGULATE SATIETY AND HUNGER During a meal, ingested nutrients interact with cells of the mouth and GI tract. Endocrine cells of the stomach and small intestine possess receptors that are linked to the secretion of GI hormones. GI peptides (see Chapters 6 and 8) are then released into the surrounding space, where they exert paracrine actions or are taken up into the circulation, where they function as hormones.162 Each of these transmitters facilitates the ingestion, digestion, absorption, or distribution of nutrients that are essential for the organism. Some GI hormones control the size of an ingested meal and are known as satiety signals. Satiety hormones share several qualities.163 First, they decrease meal size. Second, blocking their endogenous activity leads to increased meal size. Third, reduction of food intake is not the result of an aversion to food. Fourth, secretion of the hormone is caused by ingestion of food that normally causes cessation of eating (Table 1-3). Most satiety signals interact with specific receptors on nerves leading from the GI tract to the hindbrain. CCK is one of the most extensively studied satiety hormones. In a time- and dose-dependent manner, CCK reduces food intake in animals and humans,164 an effect that is mediated by CCK-1 receptors residing on vagus nerve endings.165 The effect of CCK on food intake is a proven physiologic action because administration of a CCK receptor antagonist induces hunger and results in larger meal sizes. CCK also delays the rate at which food empties from the stomach, which may explain why the satiety actions of CCK are most apparent when the stomach is distended. Together, these findings indicate that CCK provides a signal for terminating a meal. GLP-1 is produced by L cells of the ileum and colon and is released in response to food in the intestine. Although the primary action of GLP-1 is to stimulate insulin secretion, it also delays gastric emptying. Moreover, infusion of GLP-1 increases satiety and produces feelings of fullness, thereby reducing food intake without causing aversion.166 GLP-1 receptors are found in the periventricular nucleus, dorsal medial hypothalamus, and arcuate nucleus of the hypothalamus, which are important areas in the regulation of hunger. Like CCK, central administration of GLP-1 suppresses food intake. PYY is also produced by L cells of the ileum and colon. Two forms of PYY are released into the circulation, PYY1-36 and PYY3-36. PYY1-36 binds to all subtypes of the neuropeptide Y family of receptors, whereas PYY3-36 has strong affinity for the Y2 receptor. When administered to animals, PYY3-36 causes a reduction in food intake, and mice lacking the Y2 receptor are resistant to the anorexigenic effects of PYY3-36, indicating that PYY3-36 signals satiety through this

Table 1-3  Gastrointestinal Peptides That Regulate Satiety and Food Intake Reduce food intake

Increase food intake

Cholecystokinin (CCK) Glucagon-like peptide-1 Peptide tyrosine tyrosine (PYY3-36) Gastrin-releasing peptide Amylin Apolipoprotein A-IV Somatostatin

Ghrelin

receptor.167 PYY3-36 has been shown in humans to decrease hunger scores and caloric intake.168 Interestingly, most of the GI peptide receptors involved in satiety are also found in the brain, where they mediate similar satiety effects. This may represent conservation of peptide signals that serve similar purposes. Leptin is referred to as an adiposity signal because it is released into the blood in proportion to the amount of body fat and is considered a long-term regulator of energy balance. Together with CCK, leptin reduces food intake and produces a greater reduction in body weight than either agent alone.86 Therefore, it appears that long-term regulators of energy balance can affect short-term regulators through a decrease in meal size, which may promote weight reduction. Hunger and initiation of a meal are intimately related. Ghrelin is intriguing because it is the only known circulating GI hormone that has orexigenic effects.102 Produced by the stomach, ghrelin levels increase abruptly before the onset of a meal and decrease rapidly after eating, suggesting that it signals initiation of a meal. Consistent with this role are studies demonstrating that administration of antighrelin antibodies or a ghrelin receptor antagonist suppresses food intake.169 It is not known if ghrelin is responsible for the hunger pains and audible bowel sounds that occur in people who are hungry. Bariatric surgery, in particular Roux-en-Y gastric bypass, is the most effective procedure for long-term weight loss in morbid obesity. Although it had been assumed that weight loss accompanying this procedure was the result of reduced gastric capacity and calorie malabsorption, recent evidence of reduced ghrelin release and exaggerated PYY release after a meal has suggested that hormonal factors may contribute to reduced calorie intake.170

ENTEROINSULAR AXIS GI hormones play an important role in the regulation of insulin secretion and glucose homeostasis. These hormones control processes that facilitate the digestion and absorption of nutrients, as well as disposal of nutrients that have reached the bloodstream. In particular, gut peptides control postprandial glucose levels through three different mechanisms: (1) stimulation of insulin secretion from pancreatic beta cells; (2) inhibition of hepatic gluconeogenesis by suppression of glucagon secretion; and (3) delaying the delivery of carbohydrates to the small intestine by inhibiting gastric emptying.171 Each of these actions reduces the blood glucose excursions that normally occur after eating. Approximately 50% of the insulin released after a meal is the result of GI hormones that potentiate insulin secretion.172 This interaction is known as the enteroinsular axis and the gut peptides that stimulate insulin release are known as incretins. The major incretins are GLP-1 and GIP. GLP-1 not only stimulates insulin secretion but also increases beta cell mass, inhibits glucagon secretion, and delays gastric emptying. GIP stimulates insulin secretion when glucose levels are elevated and decreases glucagonstimulated hepatic glucose production.173 Thus, on ingestion of a meal, glucose, as it is absorbed, stimulates GLP-1 and GIP secretion. Circulating glucose then stimulates beta cell production of insulin, and this effect is substantially augmented by incretins acting in conjunction with glucose to increase insulin levels. Postprandial hyperglycemia may also be controlled by delaying the delivery of food from the stomach to the small

Chapter 1  Gastrointestinal Hormones and Neurotransmitters Table 1-4  Gastrointestinal Peptides That Regulate Postprandial Blood Glucose Levels Stimulate Insulin Release Glucagon-like peptide-1 Glucose-dependent insulinotropic peptide Gastrin releasing peptide Cholecystokinin (potentiates amino acid–stimulated insulin release) Gastrin (in presence of amino acids) Vasoactive intestinal peptide (potentiates glucose-stimulated insulin release) Pituitary adenylate cyclase–activating peptide (potentiates glucosestimulated insulin release) Motilin Delay Gastric Emptying Cholecystokinin Amylin Secretin Inhibit Glucagon Release Amylin

intestine, allowing the rise in insulin to keep pace with the rate of glucose absorption. Several gut hormones that delay gastric emptying have been shown to reduce postprandial glucose excursions (Table 1-4).171 Amylin (islet amyloid polypeptide) is a 37–amino acid peptide synthesized primarily in the beta cells of the pancreatic islets together with insulin. Although it was originally recognized for its ability to form amyloid deposits in association with beta cell loss, it has more recently been found to suppress glucagon secretion, delay gastric emptying, and induce satiety.174 Insulin resistance in obese patients is associated with increased levels of both insulin and amylin. Type II diabetes mellitus is characterized by high circulating insulin levels and insulin resistance. In addition, insulin levels do not increase appropriately after a meal and significant hyperglycemia occurs, which is consistent with an impaired incretin effect. GIP secretion is preserved in type II diabetes; however, the insulinotropic effect of GIP is reduced.175 Although the precise cause is unknown, the defect in GIP-stimulated insulin release is most pronounced in the late phase of insulin secretion. In contrast to GIP, GLP-1 secretion has been shown to be reduced in insulinresistant type II diabetics. The lower GLP-1 levels are caused by impaired secretion rather than increased degradation of the hormone.176 Unlike GIP, the insulin response to infusion

of GLP-1 is preserved, indicating that the beta cell can respond normally to this incretin hormone. These observations suggest that GLP-1 administration could be a viable treatment for the hyperglycemia associated with diabetes.177 The growing evidence that beta cell failure may develop in type II diabetes supports the use of incretin hormones, such as GLP-1, or agents that delay GLP-1 degradation by the enzyme dipeptidyl peptidase-4 (DPP-4) to enhance beta cell function.178,179

KEY REFERENCES

Batterham RL, Cowley MA, Small CJ, et al. Gut hormone PYY(3-36) physiologically inhibits food intake. Nature. 2002; 418:650-4. (Ref 60.) Boehning D, Snyder SH. Novel neural modulators. Annu Rev Neurosci. 2003; 26:105-31. (Ref 119.) Burdick JS, Chung E, Tanner G, et al. Treatment of Ménétrier’s disease with a monoclonal antibody against the epidermal growth factor receptor. N Engl J Med 2000; 343:1697-701. (Ref 142.) Cummings DE, Weigle DS, Frayo RS, et al. Plasma ghrelin levels after diet-induced weight loss or gastric bypass surgery. N Engl J Med 2002; 346:1623-30. (Ref 100.) de Herder WW, Lamberts SW. Somatostatin and somatostatin analogues: Diagnostic and therapeutic uses. Curr Opin Oncol. 2002; 14:53-7. (Ref 75.) Drucker DJ. Enhancing incretin action for the treatment of type 2 diabetes. Diabetes Care 2003; 26:2929-40. (Ref 175.) Joseph IM, Zavros Y, Merchant JL, Kirschner D. A model for integrative study of human gastric acid secretion. J Appl Physiol. 2003; 94:160218. (Ref 70.) Miyawaki K, Yamada Y, Ban N, et al. Inhibition of gastric inhibitory polypeptide signaling prevents obesity. Nat Med. 2002; 8:738-42. (Ref 48.) Nakazato M, Murakami N, Date Y, et al. A role for ghrelin in the central regulation of feeding. Nature. 2001; 409:194-8. (Ref 94.) Pennefather JN, Lecci A, Candenas ML, et al. Tachykinins and tachykinin receptors: a growing family. Life Sci. 2004; 74:1445-63. (Ref 65.) Schlessinger J. Cell signaling by receptor tyrosine kinases. Cell. 2000; 103:211-25. (Ref 133.) Thomas RP, Hellmich MR, Townsend CM Jr, Evers BM. Role of GI hormones in the proliferation of normal and neoplastic tissues. Endocr Rev 2003; 24:571-99. (Ref 73.) Tracey KJ. The inflammatory reflex. Nature. 2002; 420:853-9. (Ref 123.) Vilsboll T, Zdravkovic M, Le-Thi T, et al. Liraglutide, a long-acting human glucagon-like peptide-1 analog, given as monotherapy significantly improves glycemic control and lowers body weight without risk of hypoglycemia in patients with type 2 diabetes. Diabetes Care 2007; 30:1608-10. (Ref 44.) Woods SC. GI satiety signals I. An overview of GI signals that influence food intake. Am J Physiol Gastrointest Liver Physiol 2004; 286:G7-13. (Ref 163.) Full references for this chapter can be found on www.expertconsult.com.

19

CHAPTE R

2

Mucosal Immunity Iris Dotan and Lloyd Mayer

CHAPTER OUTLINE Distinct Immune Responses in Gut-Associated Lymphoid Tissue  21 Controlled/Physiologic Inflammation  21 Oral Tolerance  22 Unusual Immunoglobulins of Gut-Associated Lymphoid Tissue  23 Immunoglobulin A  23 Immunoglobulin M  24 Other Immunoglobulins  24 Physiology of Gut-Associated Lymphoid Tissue: The Intestinal Barrier  24

Mucosal immunity refers to immune responses that occur at mucosal sites. The demands on the mucosal immune system are distinct from their systemic counterparts. At mucosal sites, the outside world is typically separated from the inner world by a single layer of epithelium. The mucosal immune system exists at a number of sites, including the respiratory tract (especially the upper respiratory tract), the urogenital tract, the mammary glands, and even the eye and ear. Regardless of the site, unique lymphoid and other cell populations are required to handle a wide array of environmental stimuli. Together, these sites are called mucosaassociated lymphoid tissue, or MALT.1-5 However, the site that is most often associated with mucosal immunity is the intestinal tract. The intestinal tract is unique in several aspects. In contrast to other mucosal sites, it is the least sterile, containing billions to trillions of microorganisms, mainly bacteria. These organisms, along with ingested food, represent an enormous antigenic load that must be tolerated to maintain the status quo in the intestine. This unusual environment and the demands associated with it have resulted in the development of a distinct immune system designated as gut-associated lymphoid tissue (GALT). The specific characteristics and peculiarities of GALT reflect the unique milieu in which it needs to function. To maintain homeostasis in the intestine, one of the most important tasks of GALT is to distinguish potentially harmful antigens, such as pathogenic bacteria or toxins, from ones that may benefit the body, such as those derived from food or commensal bacteria. To achieve homeostasis, unusual cell types, immunoglobulins (Igs), and secreted mediators need to function in a coordinated fashion. In contrast to the systemic immune system, whose focus is to act quickly within seconds of encountering a foreign antigen, the GALT is poised to respond but is predominantly tolerant, rejecting harmful antigens but allowing beneficial or harmless ones to persist without evoking immune responses such as allergic reactions or inflammation. The unique ways in which GALT performs in its demanding environment are the focus of this chapter, along with the consequences of

Functional Anatomy of Gut-Associated Lymphoid Tissue  25 Peyer’s Patches and Microfold Cells  25 Intestinal Epithelial Cells  25 Antigen Presentation in the Gut  26 Intestinal Mononuclear Cells  28 Intraepithelial Lymphocytes  28 Lamina Propria Mononuclear Cells  28 T Cell Differentiation  28 Dendritic Cells  29 Gut-Associated Lymphoid Tissue: Relevant Chemokines  29

abnormal GALT function that result in intestinal allergic or inflammatory diseases (discussed in other chapters).

DISTINCT IMMUNE RESPONSES IN GUT-ASSOCIATED LYMPHOID TISSUE As noted, the hallmark of mucosal, in contrast to systemic, immunity is suppression, exemplified by two linked phenomena, controlled or physiologic inflammation and oral tolerance. These processes are mediated by a unique anatomy, distinct resident cell populations, and selective antibody isotypes.

CONTROLLED/PHYSIOLOGIC INFLAMMATION

Within the lamina propria exist billions of activated plasma cells, memory T cells, memory B cells, macrophages, and dendritic cells.6,7 In fact, given the large surface area of the gastrointestinal (GI) tract and the resident cell populations that inhabit this space, the gut is the largest lymphoid organ in the body. Still, in contrast to activated lymphocytes in the peripheral immune system, significant inflammation is not present in the intestine. This phenomenon has been termed controlled/physiologic inflammation (Fig. 2-1). The entry and activation of the cells into the lamina propria is antigen-driven. Germ-free mice have few cells in their lamina propria. However, within hours to days following colonization with normal intestinal flora (without pathogens), there is a massive influx and activation of cells.8-11 Despite the persistence of an antigen drive (luminal bacteria), the cells fail to develop into aggressive inflammation-producing lymphocytes and macrophages. Bacteria or their products play a role in this persistent state of activation12 and likely contribute to the controlled inflammatory process as well. The failure to produce gastrointestinal pathology, despite the activation state of intestinal lymphocytes, is probably the consequence of regulatory mechanisms. The failure of lamina propria lymphocytes (LPLs) to generate “normal”

21

22

Section I  Biology of the Gastrointestinal Tract IELs

↓Cellular responses

Bacteria

DC Tight junctions Goblet cell

Treg

Defensins HBD-2, 3, 4 SlgA J Macrophage Plasma cell

Lymphocytes LPMC MAdCAM-1

α4β7 integrin Blood vessels

Figure 2-1.  Mechanisms for damping the mucosal immune response. The intestine uses a number of distinct mechanisms to dampen mucosal immune responses. The major source of antigen in the intestine is the commensal bacterial flora, but both innate and adaptive responses control local responses. Physical barriers such as mucins secreted by goblet cells and tight junctions between epithelial cells prevent invasion by luminal flora. Defensins such as HBD-2, -3, and -4 are thought to maintain the sterility of the crypt, whereas SIgA produced by local plasma cells prevents attachment and invasion by luminal bacteria, thereby reducing the antigenic load. Even in the face of antigenic challenge, the lymphocytes, macrophages, and dendritic cells in the intestine are programmed not to respond as a consequence of decreased expression of pattern recognition receptors (e.g., Toll-like receptors) and a decrease in the ability of lymphocytes to be activated through their antigen receptor. DC, dendritic cell; HBD, human β-defensin; IELs, intraepithelial lymphocytes; LPMC, lamina propria mononuclear cells; MAdCAM, mucosal addressin cell adhesion molecule; SIgA, secretory immunoglobulin A, a dimer with a connecting J chain; Treg, T regulatory cells (formerly known as suppressor T cells).

antigen-receptor mediated responses is an important factor in controlled inflammation (e.g., lack of expansion, despite a state of activation). LPLs respond poorly when activated via their T cell receptor (TCR), failing to proliferate although they still can produce cytokines.13,14 This is key to the maintenance of controlled inflammation.

ORAL TOLERANCE

The most recognized phenomenon equated with mucosal immunity and associated with suppression is oral tolerance.15-20 Oral tolerance can be defined as the active, antigen-specific nonresponse to antigens administered orally.18,21,22 Disruption of oral tolerance may result in food allergies (see Chapter 9) and food intolerances such as celiac disease (see Chapter 104). Part of the explanation for oral tolerance relates to the properties of digestion per se, where large potentially antigenic macromolecules are degraded so that potentially immunogenic substances are rendered nonimmunogenic. However, approximately 2% of dietary proteins enter the draining enteric vasculature intact.23 How does the body regulate the response to these potential antigens that have bypassed complete digestion? This is achieved by oral tolerance. Factors affecting the induction of oral tolerance are the age of the host, genetic factors, the nature of antigen, the tolerogen’s form, and dose. In addition, the state of the intestinal barrier affects oral tolerance and, when barrier function is reduced, oral tolerance decreases as well. Oral tolerance is difficult to achieve in the neonate, probably because of the rather permeable intestinal barrier that

exists in the newborn, as well as the immaturity of the mucosal immune system. Within 3 months of age (in the mouse), oral tolerance can be induced and many previous antibody responses to food antigens are suppressed. The limited diet in the newborn may further serve to protect the infant from generating a vigorous response to food antigens. Furthermore, the intestinal flora has been demonstrated to affect the development of oral tolerance. Probiotics such as Lactobacillus GG given to mothers before delivery and during lactation have provided protection against the development of atopic eczema in their offspring.24 The effects of probiotics on oral tolerance are probably mediated through modulation of cytokine responses,25 a positive effect on intestinal barrier function and restitution of tight junctions,26-27 suppression of intestinal inflammation via downregulation of Toll-like receptor (TLR) expression,28,29 and secretion of metabolites that may inhibit inflammatory cytokine production by mononuclear cells. The role of genetic factors in oral tolerance has been suggested in murine models in which certain strains develop tolerance more easily than others.30 The nature and form of the antigen also play a significant role in tolerance induction. Protein antigens are the most tolerogenic whereas carbohydrates and lipids are much less effective at inducing tolerance.31 The way the antigen is delivered is also critical. For example, a protein delivered in soluble form (e.g., ovalbumin) is tolerogenic, whereas aggregation of this protein reduces its potential to induce tolerance. This phenomenon may be associated with an alteration in the sites of antigen sampling.6 Exposure or

Chapter 2  Mucosal Immunity prior sensitization to an antigen through an extraintestinal route also affects the development of tolerance responses. Finally, the dose of antigen administered is critical to the form of oral tolerance generated. In mouse models, high doses of antigen are associated with clonal deletion or anergy.32 In this setting, tolerance is not transferable; transfer of T cells from tolerized animals does not lead to the transfer of tolerance. The mechanism underlying T cell deletion is possibly Fas-mediated apoptosis.33 On the other hand, low doses of antigen activate regulatory-suppressor T cells.34,35 Increasing numbers of such T cells occur, both in CD4 and CD8 lineages. Th3 cells were the initial regulatorysuppressor cells described as mediators of oral tolerance.35-37 These cells appear to be activated in Peyer’s patches and secrete transforming growth factor-β (TGF-β). This cytokine plays a dual role in mucosal immunity; it is the most potent suppressor of T and B cell responses while also promoting the production of IgA (TGF-β is the IgA switch factor).38-41 The production of TGF-β by Th3 cells elicited by low-dose antigen administration helps explain an associated phenomenon of oral tolerance, namely, bystander suppression. Whereas oral tolerance is antigen specific, the effector arm is antigen non-specific. If an irrelevant antigen is coadministered systemically with the tolerogen, suppression of T and B cell responses to that irrelevant antigen will also occur (hence, bystander suppression). Secreted TGF-β suppresses the response to the coadministered antigen. T regulatory 1 (Treg1, or Tr1) cells may also participate in bystander suppression and oral tolerance by producing interleukin-10 (IL-10), another potent immunosuppressive cytokine.42-44 Evidence for the activation of CD4+, CD25+ regulatory T cells during oral tolerance also exists, although their exact role in this process is still being investigated.45-49 Tolerance studies carried out in mice depleted of CD4+, CD25+ T cells coupled with neutralization of TGF-β have demonstrated that CD4+, CD25+ T cells and TGF-β together are involved in the induction of oral tolerance, partly through regulation of the expansion of antigen-specific CD4+ T cells.50 The ability to identify regulatory CD4+, CD25+ T cell subpopulations was enhanced by the recognition that these cells express the transcription factor FoxP3. Because not every cell within the CD4+, CD25+ population is a naturally occurring Treg cell, the ability to use FoxP3 as a marker of these Treg cells has been a major breakthrough in our ability to study these cells.51,52 Importantly, the absence of CD4+ T regulatory cell activity in mice results in inflammatory bowel disease (IBD), although this has not been demonstrated in humans.53-56 Preliminary data also support a role for antigen-specific CD8+ T cells in oral tolerance,57-61 as well as in the regulation of mucosal immune responses. Specifically, in vitro activation of human CD8+ peripheral blood T cells by normal intestinal epithelial cells (IECs) results in the expansion of CD8+, CD28− T cells with regulatory activity. Moreover, in the lamina propria of IBD patients, such CD8+, CD28− cells were significantly reduced, supporting a role for these epithelial-induced T regulatory (TrE) cells in the control of intestinal inflammation.62 Another factor affecting tolerance induction is the state of the intestinal barrier. In addition to the failure to generate tolerance in the neonatal period (when intestinal permeability is higher), several other states of barrier dysfunction are associated with aggressive inflammation and a lack of tolerance. During anaphylaxis, increased intestinal permeability caused by the disruption of tight junctions allows luminal antigens to pass through paracellular spaces.63-65 Treatment with interferon-γ (IFN-γ) can disrupt the mucosal barrier in mice and they fail to develop tolerance in response to

E E

E

L

L L

L

Figure 2-2.  M cell. This transmission electron micrograph from the noncolumnar region of the Peyer’s patch epithelium shows a cross-sectional view of a microfold (M) cell, as well as associated microvillus-covered epithelial cells and at least three lymphoid cells (L). Note the attenuated cytoplasm of the M cell (between arrows) that bridges the surface between microvillus-covered epithelial cells, forming tight junctions with them and producing a barrier between the lymphoid cells and the intestinal lumen (×9600). B, B cell; E, epithelial cell. (From Owen RL, Jones AL: Epithelial cell specialization within human Peyer’s patches: An ultrastructural study of intestinal lymphoid follicles. Gastroenterology 1974; 66:189-203.)

ovalbumin feeding. Even more interesting is the failure of N-cadherin–dominant negative mice to suppress mucosal inflammation (loss of controlled inflammation),66 possibly because of the enormous antigenic exposure resulting from the leaky barrier in these mice. Lastly, oral tolerance may also be associated with the cells serving as the antigen-presenting cell (APC; see later), as well as the site of antigen uptake. Orally administered reovirus type III is taken up in mice by microfold (M) cells expressing reovirus type III–specific receptors (Fig. 2-2).67 This epithelial uptake by M cells induces an active IgA response to the virus. Reovirus type I, on the other hand, infects intestinal epithelial cells (IECs) adjacent to M cells and this uptake induces tolerance to the virus. Thus, the route of entry (M cell versus IEC) of a specific antigen may dictate the type of immune response generated (IgA versus tolerance). Interestingly, poliovirus vaccine, one of the few oral vaccines effective in humans, binds to M cells, which may account for its ability to stimulate active immunity in the gut.68

UNUSUAL IMMUNOGLOBULINS OF GUT-ASSOCIATED LYMPHOID TISSUE IMMUNOGLOBULIN A

The unique antibody, secretory IgA (SIgA), is the hallmark of MALT-GALT immune responses (Fig. 2-3). Although IgG

23

24

Section I  Biology of the Gastrointestinal Tract Secretory component J chain Light chain

Heavy chain Figure 2-3.  Secretory immunoglobulin A (IgA) complex. Two IgA molecules are linked by a J chain and stabilized by secretory component (polymeric Ig receptor) to form dimeric secretory IgA.

Secretory IgA SC

Epithelial cell

mers are bound together by a J chain (also produced by plasma cells). Subsequently, the dimer binds to secretory component (SC), also known as the polymeric immunoglobulin receptor (pIgR), a highly specialized 55-kd glycoprotein produced by IECs. SC (pIgR) is expressed on the basolateral aspect of the IEC and binds only to dimeric IgA or to IgM (also polymerized with J chain; see later). Once bound to SC, SIgA is actively transported within vesicles to the apical membrane of the IEC. The vesicle fuses with the apical membrane and the SC-IgA complex is released into the lumen. Once in the lumen, SC serves its second function, protection of the SIgA dimer from degradation by luminal proteases and gastric acid. SIgA and SIgM are the only antibodies that can bind SC and therefore withstand the harsh environment of the GI tract. In addition to its unique form, SIgA is also unique as an immunoglobulin in that it is anti-inflammatory in nature. SIgA does not bind classic complement components but rather binds to luminal antigens, preventing their attachment to the epithelium or promoting agglutination and subsequent removal of the antigen in the mucus layer overlying the epithelium.69,72 This binding of luminal antigens by SIgA reflects immune exclusion, as opposed to the non­ specific mechanisms of exclusion exerted by the epithelium, the mucous barrier, proteolytic digestion, and other mechanisms.

IMMUNOGLOBULIN M

As noted, IgM is the other antibody capable of binding SC (pIgR). Like IgA, IgM uses J chains produced by plasma cells to form polymers—in the case of IgM, a pentamer. SC binds to the Fc portion of the antibody formed during polymerization. The ability of IgM to bind SC may be important in individuals with IgA deficiency in which secretory IgM (SIgM) may compensate for the absence of IgA in the lumen.

OTHER IMMUNOGLOBULINS

Polymeric Ig receptor

IgA plasma cell Polymeric IgA IgA

J chain

Figure 2-4.  Assembly and secretion of dimeric immunoglobulin A (IgA). IgA and J chain produced by IgA-committed plasma cells (bottom) dimerize to form polymeric IgA, which covalently binds to membrane-bound polymeric Ig receptor produced by epithelial cells (top). This complex is internalized, transported to the apical surface of the epithelial cell, and secreted into the lumen. SC, secretory component.

is the most abundant isotype in the systemic immune system, IgA is the most abundant antibody in mucosal secretions.69-71 In fact, given the numbers of IgA-positive plasma cells and the size of the MALT system, IgA turns out to be the most abundant antibody in the body. SIgA is a dimeric form of IgA produced by plasma cells in the lamina propria and transported into the lumen by a specialized pathway through the intestinal epithelium (Fig. 2-4). Two IgA mono-

Whereas SIgA is the major antibody isotype produced in GALT, IgG has been detected as well.73-74 The neonatal Fc receptor (FcRN), expressed by IECs, might serve as a bidirectional transporter of IgG75,76 and may be important in control of neonatal infections and IgG metabolism. In IBD, marked increases in IgG in the lamina propria and lumen have been observed.77 Even IgE production may play an important role in intestinal diseases in GALT. CD23 (low-affinity IgE Fc receptor) has been reported to be expressed by IEC. One model has suggested that CD23, or FcεRII, may play a role in facilitated antigen uptake and consequent mast cell degranulation in food allergy. In this setting, IgE transcytosis and mast cell degranulation may be associated with fluid and electrolyte loss into the lumen, an event that is intimately associated with allergic reactions in the gut and airways.78,79

PHYSIOLOGY OF GUT-ASSOCIATED LYMPHOID TISSUE: THE INTESTINAL BARRIER The cells, structures, and mediators separating the intestinal lumen from the lamina propria function as a physical barrier. However, this physical barrier is a biologically active structure that constantly interacts with its everchanging environment. The intestinal barrier changes not only on a daily basis but also over the years. Many barrier mechanisms are not fully developed at birth, and evidence exists to support less restricted antigen transport in neonates compared with adults, specifically in animals.

Chapter 2  Mucosal Immunity Physiologic factors operative in the upper GI tract influence the antigenic load that reaches the major sites of GALT in the small and large bowels. Detailed exploration of these factors are beyond the scope of this chapter and are discussed elsewhere, but include proteolysis, gastric acidity, and peristalsis. The mucous coat lining the intestinal tract is composed of a mixture of glycoproteins (mucins). The protein core of mucins is enriched in serine, threonine, and proline residues, and carbohydrate moieties are attached via Nacetylgalactosamine residues. At least six different mucin species have been identified,80 each with a distinct carbohydrate and amino acid composition. Mucus protects the intestinal wall by several mechanisms. Its stickiness and competitive binding to glycoprotein receptors decrease the ability of microorganisms to penetrate the intestine.81 It also generates a stream that moves luminal contents away from epithelial cells. Underneath the mucous layer, the physical barrier that prevents penetration of antigen across the intestinal epithelium consists of the actual epithelial cell (the transcellular route) and the tight intercellular spaces (the paracellular route) regulated by tight junction (TJ) complexes (e.g., zona occludens) and the subjunctional space.82 Of the two structures, TJs have the greater role in preventing macromolecular diffusion across the epithelium, because these junctions exclude almost all molecules present in the lumen (see Chapter 96). The barrier formed by the TJ is a dynamic structure, preserved even when epithelial cells themselves are damaged; this feature might be crucial for the prevention of intestinal inflammation (e.g., as seen in idiopathic IBD). The epithelial cells themselves serve as a physical barrier in several ways—their microvilli are at a distance of about 25 nm from each other and are negatively charged. Thus, a negatively charged luminal molecule would be repelled from passage even if its diameter were well below 25 nm. However, intact antigens may traverse the epithelium by fluid phase endocytosis and enter the circulation.83

FUNCTIONAL ANATOMY OF GUT-ASSOCIATED LYMPHOID TISSUE To accomplish the two major goals of the mucosal immune system in the intestine (maintenance of homeostasis and clearance of pathogens), several key features have been identified. Compartmentalization of cells into distinct regions and sites, despite being millimeters away from each other, is a hallmark of the GALT. Cell populations and the immune response in the epithelium, subepithelial region, lamina propria (LP), Peyer’s patches, and mesenteric lymph node (MLN) may differ substantially. The cells residing in these compartments differ not only topographically but also phenotypically and functionally, depending on the anatomic site in GALT. Cells with distinct phenotypes and functions are attracted to specific sites in GALT.

PEYER’S PATCHES AND MICROFOLD CELLS

The follicle-associated epithelium (FAE), which contains microfold (M) cells, is a specialized epithelium overlying the only organized lymphoid tissue of GALT, Peyer’s patches (PPs). The M cell, in contrast to the adjacent absorptive epithelial cell, has few microvilli, a limited mucin overlayer, a thin elongated cytoplasm, and a shape that forms a pocket surrounding subepithelial, T, B, macrophages, and dendritic cells (see Fig. 2-2). M cells are capable of taking up large particulate antigens from the lumen and transport-

ing them intact into the subepithelial space.84-86 M cells contain few lysosomes, so little or no processing of antigen occurs.87 M cells are exposed to the lumen, thus having a larger area for contact with luminal contents than adjacent epithelial cells. The M cell expresses several unique lectinlike molecules that help promote binding to specific pathogens—the prototype being poliovirus.88 Antigens that bind to the M cell and are transported to the underlying PP generally elicit a positive (SIgA) response. Successful oral vaccines bind to the M cell and not to the adjacent epithelium. Thus, M cells appear to be critical for the positive aspects of mucosal immunity. The M cell is a conduit to PPs. Antigens transcytosed across the M cell and into the subepithelial pocket are taken up by macrophages and DCs and carried into PPs. Once in the patch, TGF-β–secreting T cells promote B cell isotype switching to IgA.89 Importantly, there is a clear relationship between M cells and PPs. The induction of M cell differentiation has been shown to be dependent on direct contact between the epithelium and B lymphocytes in PPs.90 M cells do not develop in the absence of PPs. For example, M cells have not been identified in B cell–deficient animals in whom there are no PPs.91 Even though M cells and PPs may be involved in oral tolerance,92-94 PP-deficient mice are capable of developing tolerance after oral administration of soluble antigen.95 After activation in PPs, lymphocytes are induced to express specific integrins (α4β7), which provide a homing signal for mucosal sites (where the ligand is MadCAM-1).96-98 Cells then travel to MLNs and subsequently into the main intestinal lymphatic drainage system, the thoracic duct, and finally into the systemic circulation (Fig. 2-5). There, mucosally activated lymphocytes with their mucosal addressins circulate in the bloodstream to exit in high endothelial venules in various mucosal sites. Those bearing α4β7 molecules exit in the MALT-GALT lamina propria where they undergo terminal differentiation. Chemokines and their receptors (see later) as well as adhesion molecules and ligands may help direct this trafficking pattern.

INTESTINAL EPITHELIAL CELLS

The epithelium is composed of a single layer of columnar cells. These IECs are derived from the basal crypts and differentiate into absorptive villous or surface epithelium, secretory goblet cells, neuroendocrine cells, or Paneth cells (see Chapter 96). In addition to their function as a physical barrier in GALT discussed earlier, IECs contribute to innate and adaptive immunity in the gut and may play a key role in maintaining intestinal homeostasis.

Antigen Trafficking Across Intestinal Epithelial Cells

The ability of intact antigen to cross the lipid bilayer at the surface of the IEC (underneath the microvilli) is limited. However, invagination of apical membranes occurs regularly, allowing macromolecules to be carried into the cell within membrane-bound compartments. Binding to the surface of the cell depends on the structure of the antigen and the chemical composition of the microvillous membrane. For instance, bovine serum albumin binds less efficiently to the surface of the IEC than bovine milk protein and, as a consequence, is transported less efficiently.99 In addition, structural alterations in an antigen caused by proteolysis might also affect its binding, because this will change the physicochemical characteristics of the molecule.100 Several factors influence the transport of antigens from the apical to the basolateral surface of IECs. The rate of

25

26

Section I  Biology of the Gastrointestinal Tract Intestine Tonsil

Villi

Lumen

Follicle-associated epithelium

Mammary gland

Peyer’s patch Figure 2-5.  Mucosal lymphocyte migration. Following antigenic stimulation, T and B lymphocytes migrate from the intestine to the draining mesenteric lymph nodes, where they further differentiate and then reach the systemic circulation via the thoracic duct. Cells bearing the appropriate mucosal addressins then selectively home to mucosal surfaces which constitute the common mucosal immune system, including the intestine.

Circulation Mesenteric lymph node

vesicular passage to the basolateral membrane depends on the rate of endocytosis, the proportion of vesicles trafficking to the lysosome, and the speed of travel of membrane-bound compartments. Lysosomally derived enzymes determine the rate of breakdown of products contained in membrane compartments. These include proteases such as cathepsins B and D (found throughout the length of the intestine, particularly in the mid and distal thirds of the small intestine), as well as those enzymes that catalyze carbohydrate breakdown, such as acid phosphatase and mannosidase. It is the degree to which the organellar contents encounter such enzymes (in the lysosome or in endocytic vesicles) that determines the rate of intracellular destruction of macromolecules.101 Although cathepsins are capable of catalyzing antigens, they may not completely digest the protein, which may require further proteolysis by peptidases in the cytoplasm.

Recognition of Pathogen-Associated Molecular Patterns by Pattern Recognition Receptors

Classic APCs in the systemic immune system possess the innate capacity to recognize components of bacteria and viruses, called pathogen-associated molecular patterns (PAMPs). Receptors for these PAMPs are expressed on the cell surface (e.g., TLRs) and inside the cell (e.g., NOD2 [see later]). Despite the fact that IECs live adjacent to large numbers of luminal flora, IECs retain the ability to recognize components of these bacteria. In general, proinflammatory responses are normally down-regulated (i.e., expression of the lipopolysaccharide [LPS] receptor, TLR4, is absent) and expression of some of these pattern recognition receptors is maintained, such as TLR5, which recognizes bacterial flagellin. TLR5 is expressed basolaterally, so it is poised to identify organisms such as Salmonella species that have invaded the epithelial layer.102 Following invasion and engagement of TLR5, the IEC is induced to secrete a broad array of cytokines and chemokines that attract inflammatory cells to the local environment to control the spread of infection. In contrast to invading pathogens, some bacteria are probiotic and induce the IEC to produce anti-inflammatory cytokines (e.g., IL-10) and to increase the expression of peroxisome proliferator-activated receptor-γ (PPAR-γ).103 Furthermore, other bacterial products help promote the barrier and IEC differentiation (e.g., products of Bacteroides thetaiotaomicron).

Genitourinary tract Bronchus-associated lymphoid tissue

Thoracic duct

The significance of the ability of IECs to recognize PAMPs via surface TLRs, such as TLR5, or via intracellular nuclear oligomerization domain 1, 2 (NOD1, 2), has been increasingly recognized over the past decade. The latter ability has been shown to contribute to intestinal inflammation, because about 25% of patients with Crohn’s disease have mutations in the NOD2-CARD15 gene, interfering with their ability to mount an appropriate immune response to bacterial stimuli (discussed further in Chapter 111).104-108 In addition, TLRs that are weakly expressed by normal IECs are expressed at higher levels on IECs obtained from patients with IBD.109 Expression of different TLRs by IECs, as well as their contribution to innate and adaptive T and B cell responses in intestinal inflammation and homeostasis, has been demonstrated in several murine models.110,111 TLR expression by professional APCs is also down-regulated in the lamina propria. This finding, along with others described earlier, contribute to the immunologic nonresponsiveness of GALT. The importance of TLR and NOD2CARD15 expression and signaling in the intestine has been reviewed.112

ANTIGEN PRESENTATION IN THE GUT Effective immune responses to antigenic proteins depend on the antigen being presented to T lymphocytes by APCs. APCs internalize, digest, and then couple a small fragment of the antigen to a surface glycoprotein (major histocompatibility complex [MHC] class II, or HLA-D in humans) that eventually interacts with a TCR. Several cells in GALT can act as APCs, including B cells, macrophages, and dendritic cells. The ability of these cells to present antigen depends on the expression of class II MHC molecules on their surface. Class II MHC molecules are also present on the epithelium of the normal small intestine and, to a lesser extent, colonocytes in humans113 and rodents.114 In vitro studies have demonstrated that enterocytes isolated from rat and human small intestine can present antigens to appropriately primed T cells.115-117 This raises the possibility that IECs might present peptides to GALT T cells beneath the epithelium. Thus, IECs are capable of antigen trafficking and proc­ essing as well as antigen presentation to cells in the lamina propria in the appropriate context. Importantly, increased expression of MHC class II molecules by IECs has been

Chapter 2  Mucosal Immunity Luminal bacteria Food or bacterial antigen Inflammation ↑paracellular transport

Tight junction Stress response/ homeostasis? Autoregulatory or suppressor population?

γδ TCR

MICA/MICB

αβ TCR CD2

CD4+ or CD8+ IEL αβ TCR CD8 CD8+ CD28− IEL

β2m CD1d gp180

αβ TCR CD28 CD4+ CD25+ LPL

CD4+ IEL

MHC class I CD1d β2m gp180

αβ TCR CD8

MHC class II CD86 CD58

αβ TCR CD8

αβ TCR CD2

CD8+ CD28− LPL

CD4+ LPL

Homeostasis?

CD8+ IEL

Cytolytic or suppressor activity?

αβ TCR CD28+ CD4+ LPL

Tolerance? Autoregulatory or suppressor population?

Inflammation?

Figure 2-6.  Normal intestinal epithelial cell (IEC). The IEC is shown to express classic MHC molecules (classes I and II) that have the potential to present conventional antigen to local T cell populations and a broad array of nonclassic class I molecules (e.g., CD1d, MICA/MICB, and β2m [shown in the figure] and MR-1, ULBP, HLA-E, and FcRn [not shown]), which have the potential to present unconventional antigens to unique T cell populations. In addition, alternate pathways of activation appear to be functional in the intestine (e.g., activation via a CD58-CD2 interaction) and classic costimulatory molecules are not expressed on IECs, although CD86 may be induced in patients with ulcerative colitis. Other members of the B7 family are expressed (B7h and B7H-1) and may play a role in local T cell activation. β2m associates with MHC class I, CD1d, HLA-E, HLA-G, and FcRn. β2m, β2 microglobulin; gp180, membrane glycoprotein 180 (a CD8 ligand); IEL, intraepithelial lymphocyte; LPL, lamina propria lymphocyte, MHC, major histocompatibility complex; MICA/MICB, MHC class I–related chains A and B; TCR, T cell receptor.

reported in patients with IBD.118,119 Such overexpression would be expected to increase their potential to activate T lymphocytes and this has been reported.120,121 Drugs used to treat patients with IBD, such as 5-aminosalicylic acid (5-ASA) preparations, may reduce IEC MHC class II expression on IEC.122 In addition, IECs from normal individuals or IBD patients express a variety of costimulatory molecules required for T cell activation (Fig. 2-6). These include intercellular adhesion molecule 1 (ICAM-1), which binds to leukocyte function-associated antigen 1 (LFA-1) on the T cell and to B7 (CD80) on the APC. B7, which binds CD28 and cytotoxic T-lymphocyte antigen 4 (CTLA-4),123,124 has been shown to be expressed by IECs of patients with ulcerative colitis (UC). Interestingly, unique expression of these costimulatory molecules by IECs may be involved in the distinct regulation of mucosal responses. Failure of naive T cells to engage CD28 by B7 family members may result in T cell tolerance. This may be less of an issue in GALT, in which cells express the memory phenotype.125,126 Because small intestine IECs do not express B7-1 (CD80),127 activation of naive T cells by IECs is not probable, aiding in the down-regulation of T-cell

responses. On the other hand, increased expression of B7 during intestinal inflammation may serve to augment T cell stimulation.128 MHC class I and nonclassic class I molecules are also expressed by IECs. Thus, antigen presentation to unique T cell populations is possible and has been reported by several groups.116,129-135 Specifically, CD1d expressed on human IECs is able to present antigen, in a complex with CEACAM5, to CD8+ T cells.136-140 The role of other nonclassical class I molecules expressed by IECs (e.g., MR-1, TL, Hmt-1, MICA/B, HLA-E, HLA-G) is still unclear.141-144 In humans, IECs specifically activate CD8+ regulatory T cells.116 These regulatory cells may be involved in local tolerance and interaction with intraepithelial lymphocytes, which are CD8+ T cells (see later). The role of IECs in the regulation of mucosal immunity is best demonstrated in studies with tissues obtained from patients with IBD. IECs derived from IBD patients, in contrast to normal IECs, stimulate CD4+ T cells rather than regulatory CD8+ cells.120-121,145 Furthermore, oral antigen administration does not result in tolerance in IBD patients, but instead produces active immunity.146

27

28

Section I  Biology of the Gastrointestinal Tract INTESTINAL MONONUCLEAR CELLS Juxtaposed to the IEC are two unusual intestinal lymphocyte populations, each very different from the other—IELs and LPLs. The clear compartmentalization of these two distinct cell populations correlates with their ability to respond to distinct microenvironmental cues.

INTRAEPITHELIAL LYMPHOCYTES

IELs in the small intestine are largely (>98%) T cells and are characterized by being mostly CD8+ cells,147-154 including CD8+ αα T cells and CD4+, CD8+ double-positive and CD4−, CD8− double-negative cells. Greater numbers of these cells also express the γδ TCR, in contrast to the αβ TCR expressed by T cells in the systemic immune system.155 Approximately half of murine small bowel IELs express the γδ TCR,156 whereas the murine and human large intestines contain primarily αβ CD4+ or CD8+ T cells similar to those found in the systemic immune system. Based on their phenotype, IELs have been classified into two subsets, a and b. Subpopulation a includes TCR αβ T cells selected in the thymus with conventional MHC class I and II expression. Subpopulation b includes TCR αβ CD8+ αα, TCR γδ DP, and TCR γδ DN cells. Both subpopulations have been shown to be cytolytic, killing via granzyme or by engagement of Fas. They both also secrete Th1 cytokines such as IL-2, TNF-α, and IFN-γ. However, antigen-specific type a IELs can transfer protection against a variety of pathogenic organisms, whereas type b IELs are not able to transfer immunologic protection and do not possess immunologic memory. This is possibly caused by the activation of type b IELs by IECs in situ by nonclassical MHC molecules, rather than by the polymorphic MHC-expressed molecules on professional APCs that activate type a IELs.156 IELs express a variety of activation markers and are CD45 RO+ (memory cells). IELs also express GALT-specific integrin αEβ7.157-158 Expression of this integrin is induced by TGF-β, and its ligand on IECs is E-cadherin, which is involved in cell signaling and cytoskeletal rearrangement.158 When isolated, IELs are difficult to activate through their TCR and they barely proliferate, even in response to potent stimuli.153 On the other hand, they may be activated by alternative pathways, such as via CD2. Type a IELs secrete cytokines such as IL-7 that are dif­ ferent from the ones secreted by their peripheral blood counterparts.152,159-161 Functionally it has been suggested that IELs potentially kill epithelial cells that have undergone some form of stress, such as infection, transformation, or invasion by other cells.154-156,162 Alternatively, it has been proposed that IELs are active in suppressing local immunity, although the evidence that they actually function in luminal antigen recognition is weak. IELs do not travel in and out of the epithelium. Rather, the epithelial cells grow over the IELs as they move from the crypt to the surface. Thus, IELs likely serve as sentinels for epithelial integrity.

LAMINA PROPRIA MONONUCLEAR CELLS

The LP is the major effector site in GALT. Recently, it has been suggested that the LP may also be an inductive site, because antigen presentation by professional and nonprofessional APCs may occur in the LP itself. The LP is also considered a graveyard for activated lymphocytes. Lamina propria lymphocytes (LPLs) are more prone to undergo apoptosis compared with their peripheral counterparts.

This may be a regulatory mechanism limiting the potentially inflammatory effects of activated lymphocytes. Indeed, a major defect reported in Crohn’s disease (CD) is the resistance of LPL to undergo apoptosis when activated inappropriately (see later). Clearly, GALT operates under a distinct set of rules compared with the systemic immune system. This is reflected not only in its functional anatomy (no organized structure) but also in its responses and regulation. As noted, highly specialized cells mediate these effects, some detected only in GALT. Lamina propria mononuclear cells (LPMNCs) are a heterogeneous group of cells.163,164 The most populous cell type is the IgA-positive plasma cell, but there are more than 50% T and B cells (together comprising the LPL population), macrophages, and dendritic cells. In contrast to IELs, which express αEβ7, LPLs express the mucosal addressin α4β7. Similar to IEL, they express an activated memory phenotype and do not proliferate in response to engagement of the TCR. Alternate pathways of LPL activation are mainly via CD2 and CD28.159,165,166 Down-regulating the ability of LPLs to respond to stimulation via the TCR (i.e., antigen) may be another mechanism involved in dampening immune responses to normal luminal contents, along with the increased tendency for LPLs to undergo apoptosis if activated inappropriately. The mechanism underlying this latter apoptotic phenomenon possibly relates to engagement of the death receptor Fas and its ligand on activated LPLs, and by the imbalance between the intracellular anti- and pro-apoptotic factors, Bcl2 and Bax. Defects in this proapoptotic balance have been reported in patients with Crohn’s disease.167,168 These observations all contribute to the normal scenario in the lamina propria, called controlled physiologic inflammation (see earlier). This state of inflammation is the norm in the gut, whereas it would be considered to be indicative of disease in any other organ. When regulatory mechanisms go awry—an increase in cell recruitment coupled with a decrease in apoptosis—the result is uncontrolled inflammation, such as that observed in IBDs.

T CELL DIFFERENTIATION

As noted, there is an organized lymphoid structure in the LP, the Peyer’s patch (see Fig. 2-5). There, B and T lymphocytes interact with antigen sampled via M cells in the FAE. Activation and maturation of T lymphocytes from naive Th0 cells to distinct biased subpopulations is strongly influenced by the microenvironment. Specifically, contact with dendritic cells (DCs), professional APCs within GALT, and their secreted mediators will skew T lymphocytes to one of several effector cells. IL-2, IFN-γ and TNF-α–secreting T helper 1(Th1) cells develop from Th0 cells when DCs secrete the IL-12/p35-40 heterodimer.169 This heterodimer induces activation and phosphorylation of the transcription factor STAT-4 (signal transducer and activator of transcription factor 4).170 STAT-4 in turn induces IFN-γ expression and production. IFN-γ then induces activation of STAT-1 and consequently of T-bet; this is the master transcription factor that induces Th1 cytokine production and IL-12 receptor β2 expression while simultaneously suppressing Th2 cytokine production. Thus, a cycle promoting Th1 and suppressing Th2 responses is created. Overactivation of T-bet is possibly an essential step for Th1-mediated mucosal diseases, such as those seen in some patients with Crohn’s disease.170 Another Th1-promoting cytokine is IL-18, mediating its effects by augmenting IL-12 receptor β2 expression on T cells and by AP-1-(c-fos/c-jun)–dependent transactivation

Chapter 2  Mucosal Immunity of the IFN-γ promoter. It also activates nuclear factor κB (NF-κB) in T cells.169 In contrast, when IL-4 is secreted, STAT-6 is activated, followed by activation of the transcription factor GATA-3, which is capable of promoting the expression of several Th2 cytokines, including IL-4, IL-5, and IL-13.171 In addition to IL-4, IL-13 also plays an important role in Th2 development and IgE synthesis in an IL-4–independent fashion. These Th2 cytokines, which also include IL-6, -9, and -10, appear to play a role in the development of food allergies (see Chapter 9). IL-5 induces B cells expressing surface IgA to differentiate into IgA-producing plasma cells. IL-6 causes a marked increase in IgA secretion, with little effect on IgM or IgG synthesis.172 Thus, in the normal state in GALT, a Th2 bias might exist. However, despite the presence of classic Th1 and Th2 cells, the dominant tone in the intestine is one of regulation. Regulatory T cells, such as Th3 and Tr1 cells (see earlier, “Oral Tolerance”), develop in an environment in which IL-10 is predominant. Th3 and Tr1 cells secrete the regulatory cytokines TGF-β and IL-10, respectively. In addition to these regulatory cells, other cells such as CD4+, CD25+, Foxp3+, and even CD8+ T regulatory cells have been identified. All these cell populations may be involved in the induction of oral tolerance and controlled inflammation in GALT (see earlier).

DENDRITIC CELLS

DCs play an important role in tolerance and immunity in the gut. DCs continuously migrate within lymphoid tissues and present self antigens (likely from dying apoptotic cells to maintain self-tolerance) as well as non– self-antigens.173 Within the lamina propria of the distal small intestine, DCs express the chemokine receptor CX3CR1 and form transepithelial dendrites that allow direct sampling of luminal antigen.174 It has been suggested that IECs expressing chemokine (C-C motif) ligand 25 (CCL25), the ligand for chemokine (C-C motif) receptor 9 (CCR9) and for CCR10, attract DCs to the small bowel, whereas CCL28, the ligand for CCR3 and CCR10, attracts DCs to the colon.175-177 DCs process internalized antigens more slowly than macrophages,178 which probably contributes to local tolerance.179,180 Tolerance induction by DCs is associated with their degree of maturation at the time of antigen presentation to T cells (e.g., immature DCs activate Tregs), downregulation of costimulatory molecules CD80 and CD86, production of the suppressive cytokines IL-10, TGF-β and IFN-α, and interaction with the costimulatory molecule CD200.181-183

GUT-ASSOCIATED LYMPHOID TISSUE: RELEVANT CHEMOKINES Many of the chemokines secreted in GALT are produced by IECs, one more piece of evidence for their active participation in the regulation of intestinal immune responses. This is especially true in IBDs, in which the secretion of IECderived chemokines and cytokines is increased, mainly because of enhanced bacterial translocation and IFN-γ production, contributing to the augmentation of mucosal inflammation. Of the chemokines secreted, those secreted by IECs have the capacity to attract inflammatory cells, such as lymphocytes, macrophages, and DCs. Chemokine (C-C motif) ligand 5 (CCL5, formerly called RANTES), secreted predominantly by macrophages, can be

produced by human IECs as well.184 CCL5 may have a role in innate as well as adaptive mucosal immunity185 and, interestingly, increased CCL5 expression has been demonstrated in the mucosa of patients with IBD.186-189 Several CXC cytokines are constitutively expressed by lymphocytes, endothelial cells, and human colonic IECs.190,191 These include CXCL9, also known as monokine induced by interferon-γ (MIG), CXCL10, also known as interferon-γ inducible protein 10 (IP-10), which is a chemokine that appears to promote Th1 responses and therefore may be relevant in Crohn’s disease, and CXCL11, also known as IFN-γ–inducible T cell α chemoattractant (ITAC). Expression of CXCL9, 10, and 11, and their polarized basolateral secretion, increases after IFN-γ stimulation. CXC chemokines attract Th1 cells expressing high levels of CXCR3 (CXC receptor 3).192 They also contribute to natural killer (NK) T cell chemotaxis and increased cytolytic responses193 and activate subsets of DCs.194 By attracting CD4+ Th1 cells that produce IFN-γ, up-regulation of expression and secretion of CXC chemokines occurs, because IECs express IFN-γ receptors. This appears to contribute to a positive feedback loop that may be relevant in inflammatory states, specifically in IBD and celiac disease. Importantly, blockade of the CXCR3-CXCL10 axis has been shown to be beneficial in ameliorating murine colitis195 and is currently being investigated in human IBD. In contrast to the inflammation-related CXCR3 receptor, a tissue-specific chemokine receptor, CCR9, is constitutively expressed on small bowel IELs and LPLs.196-198 CCL25, also known as thymus-expressed cytokine (TECK), is the ligand for CCR9 and is differentially expressed in the jejunal and ileal epithelium, where levels of expression decrease from the crypt up to the villous epithelium.199 In murine models, it was shown that CCL25-CCR9 is associated with selective localization of MLN-activated CD8 αβ+ lymphocytes coexpressing αEβ7 to the small intestine.200 CCL25 expression by IECs has been shown to be increased in inflamed small bowel in Crohn’s disease patients, with increased CCR9 expression by peripheral blood lymphocytes (PBLs) and decreased expression by LPLs,197 thus supporting its role in the specific attraction of peripheral lymphocytes to the small bowel in CD. This key chemokine/chemokine receptor pair (CCL25-CCR9) has also been used as a target for therapeutic intervention in CD, with preliminary positive results.201 Fractalkine is a unique chemokine expressed by IECs. It combines the properties of chemokines and adhesion molecules. Fractalkine attracts NK cells, monocytes, CD8+ T lymphocytes and, to a lesser extent, CD4+ T lymphocytes that express CX3CR1.202 Expression of fractalkine is increased in CD, specifically in the basolateral aspect of IECs.203 Mucosa-associated epithelial chemokine (MEC) may also have a role in intestinal immunity. This chemokine and its receptors, CCR3 and CCR10, are expressed by colonic IECs. CD4+ memory lymphocytes and eosinophils are attracted by MEC in vitro, although its function in vivo has not yet been demonstrated.204 MDC-CCL2 is a chemokine that is constitutively expressed and secreted by colonic IECs. MDC-CCL2 is unique in that it attracts CCR4+ Th2 cytokine-producing lymphocytes. Polarized basolateral secretion of MDC-CCL2 from stimulated colonic IEC lines has been reported.205 The specific recruitment of lymphocytes that preferentially secrete anti-inflammatory (Th2) cytokines supports a role for the IEC in orchestrating normal mucosal homeo­stasis, and adds to the accumulating evidence that these cells possess the ability to regulate mucosal immune responses.

29

30

Section I  Biology of the Gastrointestinal Tract The chemokine CCL20, also known as macrophage inflammatory protein-3α (MIP-3α), is unique in its ability to attract immature DCs specifically, as well as memory CD4+ T lymphocytes.206-208 CCL20 is also expressed and produced in the human small intestine, mainly in the FAE, and by colonic IECs, and has been suggested to be the mediator of lymphocyte adhesion to the α4β7 ligand MAdCAM1.206 CCL20 expression and secretion are increased in colonic IECs derived from IBD patients.209 Its stimulated secretion is polarized to the basolateral compartment, supporting its ability to attract immune cells into the LP. Mucosal memory T cells and IECs express CCR6, the cognate receptor for CCL20. CCR6 and CCR9 are coexpressed in T cells expressing the α4β7 integrin, characteristic of mucosal lymphocytes. This may suggest that in inflammatory states, and to some extent in the normal state, CCL20 and CCL25 expression by IECs attracts CCR6- or CCR9-positive lymphocytes that are activated in mesenteric lymph nodes, enter the peripheral blood, and then are recruited to the intestinal mucosa, where they can undergo activation-induced apoptosis, if they are aberrantly activated, or terminal differentiation.

KEY REFERENCES

Groux H, O’Garra A, Bigler M, et al. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 1997; 389:737-42. (Ref 42.) Hugot JP, Chamaillard M, Zouali H, Lesage S, et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature 2001; 411:599-603. (Ref 104.) Kerneis S, Bogdanova A, Kraehenbuhl JP, Pringault E. Conversion by Peyer’s patch lymphocytes of human enterocytes into M cells that transport bacteria. Science 1997; 277:949-52. (Ref 90.)

Kraus TA, Toy L, Chan L, et al. Failure to induce oral tolerance to a soluble protein in patients with inflammatory bowel disease. Gastroenterology 2004; 26:1771-8. (Ref 146.) MacDonald TT. T cell immunity to oral allergens. Curr Opin Immunol 1998; 10:620-7. (Ref 20.) Mowat AM, Viney JL. The anatomical basis of intestinal immunity. Immunol Rev 1997; 156:145-66. (Ref 2.) Neurath MF, Finotto S, Glimcher LH. The role of Th1/Th2 polarization in mucosal immunity. Nat Med 2002; 8:567-73. (Ref 169.) Neurath MF, Weigmann B, Finotto S, et al. The transcription factor T-bet regulates mucosal T cell activation in experimental colitis and Crohn’s disease. J Exp Med 2002; 195:1129-43. (Ref 170.) Neutra MR. Current concepts in mucosal immunity. V. Role of M cells in transepithelial transport of antigens and pathogens to the mucosal immune system. Am J Physiol 1998; 274:G785-91. (Ref 85.) Niess JH, Reinecker HC. Lamina propria dendritic cells in the physiology and pathology of the gastrointestinal tract. Curr Opin Gastroenterol 2005; 21:687-91. (Ref 182.) Ogura Y, Bonen DK, Inohara N, et al. A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature 2001; 411:603-6. (Ref 105.) Perera L, Shao L, Patel A, et al. Expression of nonclassical class I molecules by intestinal epithelial cells. Inflam Bowel Dis 2007; 13: 298-307. (Ref 141.) Sakaguchi S, Toda M, Asano M, et al. T cell-mediated maintenance of natural self-tolerance: Its breakdown as a possible cause of various autoimmune diseases. J Autoimmun 1996; 9:211-20. (Ref 45.) Singh UP, Venkataraman C, Singh R, et al. CXCR3 axis: Role in inflammatory bowel disease and its therapeutic implication. Endocr Metab Immune Dis Drug Targets 2007; 7:111-23. (Ref 195.) Steinman RM, Hawiger D, Nussenzweig MC. Tolerogenic dendritic cells. Annu Rev Immunol 2003; 21:685-711. (Ref 173.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

3

Cellular Growth and Neoplasia Daniel C. Chung and Daniel K. Podolsky

CHAPTER OUTLINE Mechanisms of Normal Cell Homeostasis  31 Cellular Proliferation  31 Programmed Cell Death and Senescence  32 Signaling Pathways That Regulate Cellular Growth  32 Intestinal Tumor Development: Multistep Formation and Clonal Expansion  34 Neoplasia-Associated Genes  35 Oncogenes  35 Tumor Suppressor Genes  37 DNA Repair Genes  40 Oncogenic Signaling Pathways  40

Neoplasia in the gastrointestinal (GI) tract remains one of the most common types of diseases that gastroenterologists confront. Advances in our understanding of the cellular and molecular basis of GI neoplasia have provided a foundation for the development of novel diagnostic and therapeutic approaches. Although some features are tissue site–specific, many mechanisms of tumorigenesis are common to all sites throughout the GI tract. This chapter reviews mechanisms of normal cell growth and the fundamental cellular and molecular alterations that result in malignant transformation. The common principles discussed in this chapter provide the framework for consideration of specific GI neoplasms in later chapters.

MECHANISMS OF NORMAL CELL HOMEOSTASIS CELLULAR PROLIFERATION

Neoplasia is the ultimate result of the disruption of exquisite mechanisms regulating normal cell growth. Growth is determined by the balance of cellular proliferation, differentiation, senescence, and programmed cell death. Proliferation occurs as cells traverse the cell cycle (Fig. 3-1). In preparation for cell division, there is a period of deoxyribonucleic acid (DNA) synthesis, designated the S phase. After an intervening gap period, designated the G2 phase, actual mitosis occurs in the M phase. After another intervening gap period, the G1 phase, DNA replication can begin again. The commitment to proceed through DNA replication and cell division occurs during the G1 phase at the so-called start or restriction (R) point. Cells may exit this cycle of active proliferation before reaching the R point and enter a quiescent phase, G0. Cells can subsequently re-enter the cell cycle from the G0 state (see Fig. 3-1). The duration of each

Environmental Mutagenesis  41 Chemical Carcinogenesis  41 Dietary Factors  41 Biological Features of Tumor Metastasis  42 Epithelial-Mesenchymal Transition  42 Angiogenesis and Lymphangiogenesis  42 Metastasis Genes  43 Molecular Medicine: Current and Future Approaches in Gastrointestinal Oncology  43 DNA-Based Approaches  43 Oncofetal Proteins  44

cell cycle phase as well as the overall length of the cycle vary among cell types. Regulation of cell cycle progression appears to be achieved principally by cyclins and cyclin-dependent kinase activity at the G1/S and G2/M phase transitions. Cyclin proteins are classified on the basis of their structural features and temporal expression patterns during the cell cycle (see Fig. 3-1). Cyclins A and B are expressed predominantly during the S and G2 phases. In contrast, cyclins D and E proteins are most active during the G1 phase.1 Overexpression of cyclin D1 in fibroblasts results in more rapid entry of cells into the S phase. Cyclin D1 is frequently overexpressed in a number of GI and non-GI malignancies, including those originating from the oral cavity, esophagus, breast, and bladder.2 Each cyclin forms a complex with a cyclin-dependent kinase (cdk) in a cell cycle–dependent fashion. Cyclins function as catalysts for cdk activity (see Fig. 3-1). Cdks physically associate with cyclins through their catalytic domains. The cyclin-cdk complexes regulate cell cycle progression through phosphorylation of key target proteins, including the retinoblastoma gene product (pRb) as well as the Rb family members p130 and p107.3 The final result is progression out of G1 into the S phase of the cell cycle. The cell cycle is also regulated by multiple cdk inhibitors; p21CIP1/WAF1 and p27KIP1 are inhibitors of cyclin E/cdk2. Originally discovered to be part of the complex containing cyclin D1 and cdk4/6, p21CIP1/WAF1 is transcriptionally activated by the TP53 tumor suppressor gene product (see Fig. 3-1).4 p16INK4A is another cdk inhibitor that specifically inhibits cdk4 and cdk65 and is part of a larger family of related inhibitors that includes p14, p15, and p18. p16INK4A is frequently inactivated in esophageal squamous cell cancers and pancreatic ductal adenocarcinomas, a finding that is consistent with its function as a tumor suppressor gene.6,7 p16INK4A disrupts the complex of cyclin D1

31

32

Section I  Biology of the Gastrointestinal Tract CIP1/WAF1

p21

INK4A

p16

–

KIP1

or p27

–

+ cdk4/6

cdk2

cyc D1

cyc E

pRb pRb E2F

P P P

G1

GO

E2F

S

Cyclin A

M

G2 Cyclin B Figure 3-1.  Regulation of the cell cycle by cyclins, cdks, and cdk inhibitors. In the normal cell cycle, DNA synthesis (in which chromosomal DNA is duplicated) occurs in the S phase, whereas mitosis (in which nuclei first divide to form a pair of new nuclei, followed by actual cellular division to form a pair of daughter cells) takes place in the M phase. The S and M phases are separated by two gap phases, the G1 phase after mitosis and before DNA synthesis, and the G2 phase following the S phase. During these gap phases, the cell is synthesizing proteins and metabolites, increasing its mass, and preparing for the S phase and M phase. Cell cycle progression is regulated primarily at two points, the G2/M and G1/S boundaries, through the coordinated activities of cyclins and cyclindependent kinases (cdks), which in turn are negatively regulated by cdk inhibitors (Ink4 and Cip/kip families). The mid-G1 phase is characterized by the interaction between cyclin D1 and cdk4/6. This complex hyperphosphorylates the retinoblastoma protein (pRb) and its family members (e.g., p130). Another important complex at the G1/S boundary is that of cdk2 and cyclin E (cyc E). The result is to release transcription factors such as E2F that are complexed with pRb. In turn, E2F binds to and activates the promoters of genes important in DNA synthesis.

and cdk 4/6, thereby freeing p21CIP1/WAF1 and p27KIP1 to inhibit the activity of cyclin E/cdk2.8

PROGRAMMED CELL DEATH AND SENESCENCE

Apoptosis (or programmed cell death) is an important mechanism that counterbalances cell proliferation, and escape from normal apoptotic mechanisms plays a critical role in oncogenesis. Apoptosis is characterized by distinctive features that include chromatin compaction, condensation of cytoplasm, and mild convolution of the nucleus and cytoplasm. These changes are followed by nuclear fragmentation and marked convolution of the cell surface. Eventually, membrane-bound apoptotic bodies that represent the cellular residue are produced and phagocytosed. Apoptosis is distinguished biochemically by cleavage of doublestranded DNA, which results in fragmented DNA.

Studies of the roundworm Caenorhabditis elegans have led to the initial identification of the gene ced-3, a protease that is the major effector of apoptosis. Two key regulators of ced-3, designated ced-9 and ced-4, were found to prevent or induce apoptosis, respectively.9 The mammalian oncogene bcl-2 shares homology with ced-9 and protects lymphocytes and neurons from apoptosis10; bcl-2 complexes with bax, a protein that by itself contributes to apoptosis.11 Of note, both bcl-2 and bax are part of larger gene families, and the stoichiometric relationships among different combinations of the encoded proteins can determine the balance between cell survival and cell death.12 Two well-defined pathways that trigger apoptosis have been described in detail. One pathway is mediated through membrane-bound death receptors, which include tumor necrosis factor (TNF) receptors, Fas, and DR5, whereas the other pathway involves activation of TP53 expression by environmental insults such as ionizing radiation, hypoxia, or growth factor withdrawal, with a subsequent increase in the bax-to-bcl-2 ratio. Both pathways converge to disrupt mitochondrial integrity and release of cytochrome c (Fig. 3-2). The so-called apoptosome complex (cytochrome c, caspase 9, and Apaf1) then activates downstream caspases, such as caspase 3, eventuating in cell death. Activation of caspases, intracellular cysteine proteases that cleave their substrates at aspartate residues, is a key step in programmed cell death in mammalian cells. Replicative senescence also plays a role in determining overall growth in cell populations. Most primary cells when grown in vitro have a limited replicative potential and eventually undergo senescence.13 In contrast, malignant cells can replicate indefinitely. Up-regulation of the telomerase enzyme is essential to escape from replicative senescence. Telomeres are repetitive DNA sequences at the ends of all chromosomes that regulate chromosomal stability. Telomeres shorten with each cell division and, when they have been reduced to a certain critical length, senescence occurs. Cancer cells are able to maintain their telomere length despite multiple cell divisions through the reactivation of telomerase enzyme activity.14

SIGNALING PATHWAYS THAT REGULATE CELLULAR GROWTH

Cellular proliferation is achieved through transition of cells from G0 arrest into the active cell cycle (see Fig. 3-1). Although progression through the cell cycle is controlled by the regulatory proteins just described, overall proliferation is modulated by external stimuli. Growth factors that bind to specific transmembrane receptors on the cell surface may be especially important. The cytoplasmic tails of these transmembrane receptor proteins produce an intracellular signal after ligand binding. In addition to peptide growth factors, extracellular matrix and cell-cell adhesion molecules have a significant impact on cell proliferation. Although the full spectrum of molecules that play a role in cell-matrix and cell-cell adhesion is still not defined, it is known to include integrins, cadherins, selectins, and proteoglycans. Interactions with these adhesion molecules lead to changes in the cell cytoskeleton, indirectly modulating external growth stimuli. Alterations in cell-matrix or cell-cell interactions are particularly important in contributing to the invasive phenotype characteristic of malignant cells. Interaction of ligands with their receptors at the cell surface induces intracellular signals that ultimately result in alterations in gene transcription. Three important receptor subtypes appear to initiate cellular signaling through

Chapter 3  Cellular Growth and Neoplasia Death receptors (TNF-R1, Fas, DR5)

Apoptotic signal (e.g., ionizing radiation) DD FLIP TP53

Caspase 8 bax, bak

bcl-2 bcl-xl

BID

Mitochondria

Apoptosome

Apaf1, caspase 9, cytochrome c

Caspase 3

Cell death Figure 3-2.  Apoptosis (programmed cell death) counterbalances cellular proliferation to regulate overall tissue growth. A complex interplay of proapoptotic and antiapoptotic molecules results in the downstream activation of caspases that mediate cell death. Some of these signals are initiated through environmental insults that activate the TP53 tumor suppressor gene, and some are initiated through death receptors, including TNF-R1, Fas, and DR5. Death receptors activate caspase 8, which in turn activates BID. In addition, there is an interplay between proapoptotic (bax, bak) and antiapoptotic (bcl-2, bcl-xl) molecules. Both pathways converge on the mitochondria, resulting in the release of cytochrome c and formation of the apoptosome (Apaf1, caspase 9, and cytochrome c). This leads to activation of multiple caspases, in particular caspase 3, and ultimately to cell death. BID, bcl-2 interacting domain; DD, death domain; FLICE, FADD-like IL-1β-converting enzyme; FLIP, FLICE (also known as caspase 8) inhibitory protein; TNF-R1, tumor necrosis factor receptor 1.

ligand-receptor interaction at the cell surface: (1) tyrosine kinases; (2) serine and threonine kinases; and (3) G protein– coupled receptors. The receptors for many peptide growth factors contain intrinsic tyrosine kinase activity within their intracellular tail. After ligand binding, tyrosine kinase activity is stimulated, leading to phosphorylation of tyrosine residues in target proteins within the cell. The full spectrum of proteins phosphorylated by each tyrosine kinase remains to be determined. Most receptors also autophosphorylate tyrosine residues present in the receptors themselves to initiate signaling and, in some cases, this also causes attenuation of their own activity to effect an intramolecular feedback regulatory mechanism. The receptors for many peptide growth factors, including epidermal growth factor (EGF), belong to this receptor class. Other receptors on the cell surface possess kinase activity directed toward serine or threonine residues rather than tyrosine. These receptors also phosphorylate a variety of cellular proteins, leading to a cascade of biological responses.

Multiple sites of serine and threonine phosphorylation are present on many growth factor receptors, including the tyrosine kinase receptors, suggesting the existence of significant interactions among various receptors present on a single cell.15 The transforming growth factor-β (TGF-β) receptor complex is one important example of a serine-threonine kinase–containing transmembrane receptor. Many receptors are members of the so-called sevenmembrane–spanning receptor family. These receptors are coupled to guanine nucleotide binding proteins, and are designated G proteins. G proteins undergo a conformational change that is dependent on the presence of guanosine phosphates.16 Activation of G proteins can trigger a variety of intracellular signals, including stimulation of phospho­ lipase C and the generation of phosphoinositides (most importantly, inositol 1,4,5-triphosphate) and diacylglycerol through hydrolysis of membrane phospholipids, as well as modulation of the second messengers cyclic AMP and GMP.17 Somatostatin receptors exemplify a G protein– coupled receptor prevalent in the GI tract. Binding of growth factors and cytokines to cell surface receptors typically produces alterations in a variety of cellular functions that influence growth. These functions include ion transport, nutrient uptake, and protein synthesis. However, the ligand-receptor interaction must ultimately modify gene expression within the nucleus to affect cell proliferation. The regulation of the content and activity of transcriptional factors within the nucleus is the final step in pathways that translate an external stimulus to a change in cell proliferation. These transcriptional factors modulate the expression of genes that control cell proliferation and phenotype. The Wnt pathway is one important example of a signaling pathway that regulates the cell cycle machinery to control the proliferation of intestinal epithelial cells (Fig. 3-3). Although the details of the specific interactions between the Wnt ligand and its receptor Frz, a member of the seven-membrane receptor family, in the GI tract are not fully clarified, an active Wnt signal ultimately results in the accumulation of β-catenin in the nucleus, where it binds with the transcription factor TCF-4 to activate a set of target genes.18 Inhibition of the Wnt signal in mice can be achieved by deletion of TCF-4 or overexpression of a Wnt inhibitor designated Dickkopf1, which results in dramatic hypoproliferation of the intestinal epithelium.19,20 This hypopro­ liferation appears to be mediated by decreased expression of the TCF-4 target gene c-MYC, which directly represses p21CIP1/WAF1.21 Thus, a Wnt signal stimulates proliferation of intestinal epithelial cells by repressing the cell cycle inhibitor p21CIP1/WAF1. Cyclin D1 has an extremely short half-life (60

(60.9 × W) − 54 (22.7 × W) − 495 (17.5 × W) + 651 (15.3 × W) + 679 (11.2 × W) + 879 (13.5 × W) + 987

(60.1 × W) − 51 (22.5 × W) + 499 (12.2 × W) + 746 (14.7 × W) + 996 (8.7 × W) + 829 (10.5 × W) + 596

A, age in years; H, height in centimeters; W, weight in kilograms.

Table 4-4  Relative Thermic Effect of Various Levels of Physical Activity Activity Level Resting Very light Light Moderate Heavy

Examples Standing, driving, typing Walking 2-3 miles/hr shopping, light housekeeping Walking 3-4 miles/hr, biking, gardening, scrubbing floors Running, swimming, climbing, basketball

Activity Factor 1.0 1.1-2.0 2.1-4.0 4.1-6.0 6.1-10.0

Adapted from Alpers DA, Stenson WF, Bier DM. Manual of nutritional therapeutics. Boston: Little, Brown; 1995.

Table 4-5  Metabolic Stress Factors for Estimating Total Energy Expenditure in Hospitalized Patients Injury or Illness Second- or third-degree burns, >40% BSA Multiple trauma Second- or third-degree burns, 20%-40% BSA Severe infections Acute pancreatitis Second- or third-degree burns, 10%-20% BSA Long bone fracture Peritonitis Uncomplicated postoperative state

overestimation of energy requirements and is calculated as follows:

relative Stress Factor* 1.6-1.8 1.5-1.7 1.4-1.5 1.3-1.4 1.2-1.4 1.2-1.4 1.2 1.2 1.1

*A stress factor of 1.0 is assumed for healthy controls. BSA, body surface area.

Also, in patients who have large artifactual increases in weight because of extracellular fluid retention, such as the patient with ascites, the IBW should be used to estimate energy requirements, rather than the ABW. Method Without a Stress Factor Table 4-7 outlines a simple method for estimating total daily energy requirements in hospitalized patients based on body mass index (BMI).6 With this method, energy expressed per kilogram is inversely related to BMI. Common sense needs to be applied when using any means to estimate energy expenditure in hospitalized individuals because illness commonly interjects artifacts into these calculations (e.g., ascites). Over the past two decades, the trend generally has been toward a more conservative approach to caloric delivery in acutely ill patients. One reason for this conservatism is that acute illness and its management often exacerbate preexisting diabetes or produces de novo glucose intolerance. As a result, hyperglycemia is a frequent consequence of enteral, and especially parenteral, nutrition. The issue seems to be particularly germane for intensive care unit (ICU) patients, in whom even modest hyperglycemia results in worse clinical outcomes, usually of an infectious nature. Clinical trials of high quality in surgical ICU (SICU)7 and medical ICU (MICU)8 patients have found that morbidity is substantially and significantly reduced in those randomized to intensive insulin therapy who maintained serum glucose levels below 111 mg/dL compared with those whose glucose values were maintained below 215  mg/dL. Mortality also was significantly lower in those in the SICU randomized to receive tight glucose control, although in the MICU study such reductions in mortality caused by tight glucose control only were realized in those residing in the MICU longer than three days. These clinical observations substantiate years of animal studies showing that even modest hyperglycemia impairs immune function in a variety of tissues.9 The clinical benefits of tight glucose control in the ICU, however, have not always been reproducible,10 and come at the cost of more frequent hypoglycemic episodes,7,8,10 so the issue of how tight glucose control should be remains controversial. Results of a recent meta-analysis of 29 trials in critically ill patients recapitulate the previously observed discrepancies between SICU and MICU patients.11 Overall, the relative risk of septicemia was reduced approximately 25% in those randomized to tight glucose control, but this salutary effect largely was attributable to the SICU patients, in whom the reduction in septicemia was almost 50%. In contrast, no benefit was observed in MICU patients. Also in this meta-analysis, no demonstrable benefit in overall mortality was evident in any of the categories of critically ill patients.

PROTEIN In using this formula, adjustments are necessary when the ABW is a misleading reflection of lean body mass. An adjusted ideal body weight should be substituted for ABW in obese individuals who are more than 30% heavier than their ideal body weight (desirable body weight, more commonly referred to as ideal body weight [IBW], appears in Table 4-6). The use of an adjusted IBW helps prevent an

Twenty different amino acids are found commonly in human proteins. Some amino acids (histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan, valine, and possibly arginine) are considered essential because their carbon skeletons cannot be synthesized by the body. Other amino acids (glycine, alanine, serine, cysteine, cystine, tyrosine, glutamine, glutamic acid, asparagine, and aspartic acid) are nonessential in most circumstances because they can be made from endogenous

49

50

Section II  Nutrition in Gastroenterology Table 4-6  Desirable Weight in Relation to Height for Men and Women 25 Years or Older* Men, Medium Frame

Women, Medium Frame

Weight (lb) Height (ft/inches)

5′1″ 5′2″ 5′3″ 5′4″ 5′5″ 5′6″ 5′7″ 5′8″ 5′9″ 5′10″ 5′11″ 6′0″ 6′1″ 6′2″ 6′3″

Range

Midpoint

113-124 116-128 119-131 122-134 125-138 129-142 133-147 137-151 141-155 145-160 149-165 153-170 157-175 162-180 167-185

118.5 122 125 128 131.5 135.5 140 144 148 153 157 161.5 166 171 176

Weight (lb) Height (ft/inches)

Range

Midpoint

4′8″ 4′9″ 4′10″ 4′11″ 5′0″ 5′1″ 5′2″ 5′3″ 5′4″ 5′5″ 5′6″ 5′7″ 5′8″ 5′9″ 5′10″

93-104 95-107 98-110 101-113 104-116 107-119 110-123 113-127 117-132 121-136 125-140 129-144 133-148 137-152 141-156

98.5 101 104 107 110 113 116.5 120 124.5 128.5 132.5 136.5 140.5 144.5 148.5

*Corrected to nude weights and heights by assuming 1-inch heel for men, 2-inch heel for women, and indoor clothing weight of 5 and 3 lb for men and women, respectively. Data from Metropolitan Life Insurance Company. New height standards for men and women. Statistical Bulletin 1959;40:1-4.

Table 4-7  Estimated Energy Requirements for Hospitalized Patients Based on Body Mass Index (BMI)*

Table 4-8  Recommended Daily Protein Intake*

Body Mass Index (BMI; kg/m2)

Clinical Condition

20%) from other causes and who are not TPN-dependent also frequently display a biochemical profile of EFAD,23 although whether such a biochemical state carries adverse clinical consequences with it is unclear. Moreover, TPN lacking any source of fat may lead to EFAD in adults if no exogenous source of EFAs is available; the plasma pattern of EFAD may be observed as early as 10 days after glucose-based TPN is started and before the onset of any clinical features.24 In this situation, EFAD is probably attributable to the increase in plasma insulin concentrations caused by TPN, because insulin inhibits lipolysis and therefore the release of endogenous essential FAs. The biochemical diagnosis of EFAD is defined as an absolute and relative deficiency in the two EFAs in the plasma FA profile. The full clinical EFAD syndrome includes alopecia, scaly dermatitis, capillary fragility, poor wound healing, increased susceptibility to infection, fatty liver, and growth retardation in infants and children.

MAJOR MINERALS

Major minerals are inorganic nutrients that are required in large (>100 mg/day) quantities, and are important for ionic equilibrium, water balance, and normal cell function. Malnutrition and nutritional repletion can have dramatic effects on major mineral balance. The evaluation of macromineral deficiency and recommended daily allowance (RDA) of minerals for healthy adults are shown in Table 4-9.

MICRONUTRIENTS Micronutrients (the vitamins and trace minerals) are a diverse array of dietary components that are necessary to sustain health. The physiologic roles of micronutrients are as varied as their composition. Some are used in enzymes as coenzymes or prosthetic groups, others as biochemical substrates or hormones and, in some cases, their functions are not well defined. The average daily dietary intake for each micronutrient required to sustain normal physiologic operations is measured in milligrams or smaller quantities.

May not reflect body stores

In this way, micronutrients are distinguished from macronutrients (carbohydrates, fats, and proteins) and macro­ minerals (calcium, magnesium, and phosphorus). An individual’s dietary requirement for any given micronutrient is determined by many factors, including its bioavailability, the amount needed to sustain its normal physiologic functions, a person’s gender and age, any diseases or drugs that affect the nutrient’s metabolism, and certain lifestyle habits, such as smoking and alcohol use. The U.S. National Academy of Sciences Food and Nutrition Board regularly updates dietary guidelines that define the quantity of each micronutrient that is “adequate to meet the known nutrient needs of practically all healthy persons.” These RDAs underwent revision between 1998 and 2001, and the values for adults appear in Tables 4-10 and 4-11. The formulation of an RDA takes into account the biological variability in the population, and, therefore, RDAs are set 2 SDs above the mean requirement; this allows the requirements of 97% of the population to be met and ingestion of quantities that are somewhat less than the RDA usually are sufficient to meet the needs of a particular individual. A “tolerable upper limit (TUL),” which is “the maximal daily level of oral intake that is likely to pose no adverse health risks,” also has been established for most of the micronu­ trients (see Tables 4-10 and 4-11). Present recommendations for how much of each micronutrient is needed in individ­ uals on TPN are based on far less data than what were available for the development of the RDAs. Nevertheless, it is important to have guidelines, and Table 4-12 provides such recommendations.

VITAMINS

Vitamins are categorized as fat-soluble (A, D, E, K) or watersoluble (all others) (see Table 4-10). This categorization remains physiologically meaningful; none of the fat-soluble vitamins appear to serve as coenzymes, whereas almost all of the water-soluble vitamins appear to function in that role. Also, the absorption of fat-soluble vitamins is primarily through a micellar route, whereas the water-soluble vitamins are not absorbed in a lipophilic phase in the intestine (see Chapter 100).

TRACE MINERALS

Compelling evidence exists for the essential nature of 10 trace elements in humans—iron, zinc, copper, chromium, Text continued on p. 58

Chapter 4  Nutritional Assessment and Management of the Malnourished Patient Table 4-10  Salient Features of Vitamins Vitamin

Deficiency (RDA)*

Toxicity (TUL)†

Assessment of Status

A

Follicular hyperkeratosis and night blindness are early indicators. Conjunctival xerosis, degeneration of the cornea (keratomalacia) and dedifferentiation of rapidly proliferating epithelia are later indications of deficiency. Bitot spots (focal areas of the conjunctiva or cornea with foamy appearance) are an indication of xerosis. Blindness, caused by corneal destruction and retinal dysfunction, may ensue. Increased susceptibility to infection also a consequence (1 µg of retinol equivalent to 3.33 IU of vitamin A; F, 700 µg; M, 900 µg)

Retinol concentration in the plasma, as well as vitamin A concentrations in milk and tears, are reasonably accurate measures of status. Toxicity is best assessed by elevated levels of retinyl esters in plasma. A quantitative measure of dark adaptation for night vision and electroretinography are useful functional tests.

D

Deficiency results in disordered bone modeling called rickets in childhood and osteomalacia in adults. Expansion of the epiphyseal growth plates and replacement of normal bone with unmineralized bone matrix are the cardinal features of rickets; the latter feature also characterizes osteomalacia. Deformity of bone and pathologic fractures occur. Decreased serum concentrations of calcium and phosphate may occur (1 µg equivalent to 40 IU; 5 µg, ages 19-50; 10 µg, age 51-70; 15 µg, age > 70 yr)

In adults, >150,000 µg may cause acute toxicity—fatal intracranial hypertension, skin exfoliation, and hepatocellular injury. Chronic toxicity may occur with habitual daily intake of >10,000 µg—alopecia, ataxia, bone and muscle pain, dermatitis, cheilitis, conjunctivitis, pseudotumor cerebri, hepatic fibrosis, hyperlipidemia, and hyperostosis are common. Single large doses of vitamin A (30,000 µg), or habitual intake of >4500 µg/day during early pregnancy can be teratogenic. Excessive intake of carotenoids causes a benign condition characterized by yellowish discoloration of the skin. Habitually large doses of canthaxanthin, a carotenoid, have the additional capability of inducing a retinopathy (3000 µg) Excess amounts result in abnormally high concentrations of calcium and phosphate in the serum; metastatic calcifications, renal damage, and altered mentation may occur (50 µg)

E

Deficiency caused by dietary inadequacy is rare in developed countries. Usually seen in (1) premature infants, (2) individuals with fat malabsorption, and (3) individuals with abetalipoproteinemia. Red blood cell fragility occurs and can produce hemolytic anemia. Neuronal degeneration produces peripheral neuropathies, ophthalmoplegia, and destruction of the posterior columns of the spinal cord. Neurologic disease is frequently irreversible if deficiency is not corrected early enough. May contribute to hemolytic anemia and retrolental fibroplasia in premature infants. Has been reported to suppress cell-mediated immunity (15 mg) Deficiency syndrome uncommon except in (1) breast-fed newborns, in whom it may cause “hemorrhagic disease of the newborn,” (2) adults who have fat malabsorption or are taking drugs that interfere with vitamin K metabolism (e.g., warfarin, phenytoin, broad-spectrum antibiotics), and (3) individuals taking large doses of vitamin E and anticoagulant drugs. Excessive hemorrhage is the usual manifestation (F, 90 µg; M, 120 µg)

K

Depressed levels of vitamin K–dependent procoagulants and potentiation of oral anticoagulants have been reported, as has impaired leukocyte function. Doses of 800 mg/day have been reported to increase slightly the incidence of hemorrhagic stroke (1000 mg)

Rapid intravenous infusion of vitamin K1 has been associated with dyspnea, flushing, and cardiovascular collapse; this is likely related to the dispersing agents in the dissolution solvent. Supplementation may interfere with warfarin-based anticoagulation. Pregnant women taking large amounts of the provitamin menadione may deliver infants with hemolytic anemia, hyperbilirubinemia, and kernicterus (TUL not established)

The serum concentration of the major circulating metabolite, 25-hydroxyvitamin D, is an excellent indicator of systemic status except in chronic kidney disease, in which the impairment of renal 1-hydroxylation results in dissociation of the monoand dihydroxy vitamin concentrations; measuring the serum concentration of 1,25-dihydroxyvitamin D is then necessary. Plasma or serum concentration of alpha-tocopherol is used most commonly. Additional accuracy is obtained by expressing this value per mg of total plasma lipid. Red blood cell peroxide hemolysis test is not entirely specific but is a useful functional measure of the antioxidant susceptibility of cell membranes.

The prothrombin time is typically used as a measure of functional vitamin K status; it is neither sensitive nor specific for vitamin K deficiency. Determination of undercarboxylated prothrombin in the plasma is more accurate but is less widely available.

53

Table 4-10  Salient Features of Vitamins—cont’d Vitamin

Deficiency (RDA)*

Toxicity (TUL)†

Assessment of Status

Thiamine (vitamin B1)

Classic deficiency syndrome (beriberi) is described in Asian populations consuming polished rice diet. Alcoholism and chronic renal dialysis are also common precipitants. High carbohydrate intake increases need for B1. Mild deficiency commonly produces irritability, fatigue, and headaches. More pronounced deficiency produces various combinations of peripheral neuropathy and cardiovascular and cerebral dysfunction. Cardiovascular involvement (wet beriberi) includes congestive heart failure and low peripheral vascular resistance. Cerebral disease includes nystagmus, ophthalmoplegia, and ataxia (Wernicke’s encephalopathy), as well as hallucinations, impaired short-term memory, and confabulation (Korsakoff’s psychosis). Deficiency syndrome responds within 24 hours to parenteral thiamine but is partially or wholly irreversible after a certain stage (F, 1.1 mg; M, 1.2 mg) Deficiency is usually seen in conjunction with deficiencies of other B vitamins. Isolated deficiency of riboflavin produces hyperemia and edema of nasopharyngeal mucosa, cheilosis, angular stomatitis, glossitis, seborrheic dermatitis, and normochromic, normocytic anemia (F, 1.1 mg; M, 1.3 mg)

Excess intake is largely excreted in the urine, although parenteral doses of >400 mg/day are reported to cause lethargy, ataxia, and reduced tone of the gastrointestinal tract (TUL not established)

The most effective measure of vitamin B1 status is the erythrocyte transketolase activity coefficient, which measures enzyme activity before and after addition of exogenous TPP; red blood cells from a deficient individual express a substantial increase in enzyme activity with addition of TPP. Thiamine concentrations in the blood or urine are also measured.

Toxicity not reported in humans (TUL not established)

The most common method of assessment is determining the activity coefficient of glutathione reductase in red blood cells (the test is invalid for individuals with glucose-6phosphate dehydrogenase deficiency). Measurements of blood and urine concentrations are less desirable methods. Assessment of status is problematic; blood levels of the vitamin are not reliable. Measurement of urinary excretion of the niacin metabolites N-methylnicotinamide and 2-pyridone are thought to be the most effective means of assessment.

Riboflavin (vitamin B2)

Niacin (vitamin B3)

Pyridoxine (vitamin B6)

B12

Pellagra is the classic deficiency syndrome and is often seen in populations in which corn is the major source of energy. Still endemic in parts of China, Africa, and India. Diarrhea, dementia (or associated symptoms of anxiety or insomnia), and a pigmented dermatitis that develops in sun-exposed areas are typical features. Glossitis, stomatitis, vaginitis, vertigo, and burning dysesthesias are early signs. Occasionally occurs in carcinoid syndrome because tryptophan is diverted to other synthetic pathways (F, 14 mg; M, 16 mg) Deficiency usually seen in conjunction with other water-soluble vitamin deficiencies. Stomatitis, angular cheilosis, glossitis, irritability, depression, and confusion occur in moderate to severe depletion; normochromic, normocytic anemia has been reported in severe deficiency. Abnormal EEGs and, in infants, convulsions also have been reported. Sideroblastic anemias are responsive to B6 administration. Isoniazid, cycloserine, penicillamine, ethanol, and theophylline are drugs that can inhibit B6 metabolism (Ages 19-50, 1.3 mg; >50 yr, 1.5 mg for women, 1.7 mg for men) Dietary inadequacy is a rare cause of deficiency, except in strict vegetarians. The vast majority of cases of deficiency arise from loss of intestinal absorption: this may be a result of pernicious anemia, pancreatic insufficiency, atrophic gastritis, small intestinal bacterial overgrowth, or ileal disease. Megaloblastic anemia and megaloblastic changes in other epithelia (see “Folate”) are the result of sustained depletion. Demyelination of peripheral nerves, the posterior and lateral columns of the spinal cord and nerves within the brain may occur. Altered mentation, depression, and psychoses occur. Hematologic and neurologic complications may occur independently. Folate supplementation, in doses exceeding 1000 µg/day, may partly correct the anemia, thereby masking (or perhaps exacerbating) the neuropathic complications (2.4 µg)

Human toxicity is known largely through studies examining hypolipidemic effects; includes vasomotor phenomenon (flushing), hyperglycemia, hepatocellular injury, and hyperuricemia (35 mg)

Chronic use with doses exceeding 200 mg/day (in adults) may cause peripheral neuropathies and photosensitivity (100 mg)

Many useful laboratory methods of assessment exist. The plasma or erythrocyte PLP levels are most common. Urinary excretion of xanthurenic acid after an oral tryptophan load or activity indices of RBC aminotransferases (ALT and AST) all are functional measures of B6-dependent enzyme activity.

A few allergic reactions have been reported from crystalline B12 preparations and are probably caused by impurities, not the vitamin (TUL not established)

Serum or plasma concentrations are generally accurate. Subtle deficiency with neurologic complications, as described in the “Deficiency” column, can best be established by concurrently measuring the concentration of plasma B12 and serum methylmalonic acid because the latter is a sensitive indicator of cellular deficiency.

Chapter 4  Nutritional Assessment and Management of the Malnourished Patient Table 4-10  Salient Features of Vitamins—cont’d Vitamin

Deficiency (RDA)*

Toxicity (TUL)†

Assessment of Status

Folate

Women of childbearing age are the most likely to develop deficiency. The classic deficiency syndrome is megaloblastic anemia. The hemopoietic cells in the bone marrow become enlarged and have immature nuclei, which reflect ineffective DNA synthesis. The peripheral blood smear demonstrates macro-ovalocytes and polymorphonuclear leukocytes with an average of more than 3.5 nuclear lobes. Megaloblastic changes also occur in other epithelia that proliferate rapidly, such as the oral mucosa and gastrointestinal tract, producing glossitis and diarrhea, respectively. Sulfasalazine and diphenytoin inhibit absorption and predispose to deficiency (400 µg of dietary folate equivalent (DFE); 1 µg folic acid = 1 µg DFE; 1 µg food folate = 0.6 µg DFE) Overt deficiency is uncommonly observed in developed countries. The classic deficiency syndrome is scurvy, characterized by fatigue, depression, and widespread abnormalities in connective tissues (e.g., inflamed gingivae, petechiae, perifollicular hemorrhages, impaired wound healing, coiled hairs, hyperkeratosis, and bleeding into body cavities). In infants, defects in ossification and bone growth may occur. Tobacco smoking lowers plasma and leukocyte vitamin C levels (F, 75 mg; M, 90 mg; requirement for cigarette smokers increased by 35 mg/day)

Daily dosage >1000 µg may partially correct the anemia of B12 deficiency and may, therefore, mask (and perhaps exacerbate) the associated neuropathy. Large doses also are reported to lower seizure threshold in individuals prone to seizures. Parenteral administration is rarely reported to cause allergic phenomena from dispersion agents (1000 µg)

Serum folate levels reflect short-term folate balance, whereas RBC folate is a better reflection of tissue status. Serum homocysteine levels rise early in deficiency but are nonspecific because B12 or B6 deficiency, renal insufficiency, and older age may also cause elevations.

Quantities exceeding 500 mg/day (in adults) sometimes cause nausea and diarrhea. Acidification of the urine with vitamin C supplementation, and the potential for enhanced oxalate synthesis, have raised concerns regarding nephrolithiasis, but this has yet to be demonstrated. Supplementation with vitamin C may interfere with laboratory tests based on redox potential (e.g., fecal occult blood testing, serum cholesterol, and serum glucose). Withdrawal from chronic ingestion of high doses of vitamin C supplements should occur gradually over one month because accommodation does seem to occur, raising a concern for rebound scurvy (2000 mg) Toxicity has not been reported in humans, with doses as high as 60 mg/day in children (TUL not established)

Plasma ascorbic acid concentration reflects recent dietary intake, whereas leukocyte levels more closely reflect tissue stores. Plasma levels in women are approximately 20% higher than in men for any given dietary intake.

Diarrhea is reported to occur with doses exceeding 10 g/day (TUL not established)

Whole blood and urine concentrations of pantothenic acid are indicators of status; serum levels are not thought to be accurate.

C (ascorbic and dehydroascorbic acid)

Biotin

Pantothenic acid

Isolated deficiency is rare. Deficiency in humans has been produced experimentally (by dietary inadequacy), by prolonged TPN lacking the vitamin, and by ingestion of large quantities of raw egg white, which contains avidin, a protein that binds biotin with such high affinity that it renders it bio-unavailable. Alterations in mental status, myalgias, hyperesthesias, and anorexia occur. Later, seborrheic dermatitis and alopecia develop. Biotin deficiency is usually accompanied by lactic acidosis and organic aciduria (30 µg) Deficiency rare: reported only as a result of feeding semisynthetic diets or as an antagonist to the vitamin. Experimental isolated deficiency in humans produces fatigue, abdominal pain and vomiting, insomnia, and paresthesias of the extremities (5 mg)

Plasma and urine concentrations of biotin are diminished in the deficient state. Elevated urine concentrations of methyl citrate, 3-methylcrotonylglycine, and 3-hydroxyisovalerate are also observed in deficiency.

*RDA, recommended daily allowance; established for female (F) and male (M) adults by the U.S. Food and Nutrition Board, 1999-2001. In some cases, data are insufficient to establish an RDA, in which case the adequate intake (AI) established by the board is listed. † TUL, tolerable upper level; established for adults by the U.S. Food and Nutrition Board, 1999-2001. ALT, alanine aminotransferase; AST, aspartate aminotransferase; EEG, electroencephalogram; PLP, pyridoxyl 5-phosphate; RBC, red blood cell; TPN, total parenteral nutrition; TPP, thiamine pyrophosphate. Adapted from Goldman L, Ausiello D, Arend W, et al, editors. Cecil Textbook of Medicine. 22nd ed. Philadelphia: WB Saunders; 2004. With permission.

55

56

Section II  Nutrition in Gastroenterology Table 4-11  Salient Features of Trace Minerals Mineral

Deficiency (RDA)*

Toxicity (TUL)†

Assessment of Status

Chromium

Deficiency in humans is only described for patients on long-term TPN whose TPN contained inadequate chromium. Hyperglycemia, or impaired glucose intolerance, is uniformly observed. Elevated plasma free fatty acid concentrations, neuropathy, encephalopathy, and abnormalities in nitrogen metabolism are also reported. Whether supplemental chromium may improve glucose tolerance in mildly glucose intolerant, but otherwise healthy, individuals remains controversial (F, 25 µg; M, 35 µg) Dietary deficiency is rare; it has been observed in premature and low birth weight infants fed exclusively a cow’s milk diet and in individuals on long-term TPN without copper. The clinical manifestations include depigmentation of skin and hair, neurologic disturbances, leukopenia and hypochromic, microcytic anemia, and skeletal abnormalities. The anemia arises from impaired uptake of iron, and is therefore a secondary form of iron deficiency anemia. The deficiency syndrome, except the anemia and leukopenia, is also observed in Menkes disease, a rare inherited condition associated with impaired copper utptake (900 µg) Intake of 30 mg/kg body weight of fluoride is likely to cause death. Excessive chronic intake (0.1 mg/kg/day) leads to mottling of the teeth (dental fluorosis), calcification of tendons and ligaments, and exostoses and may increase the brittleness of bones (10 mg) Large doses (>2 mg/day in adults) may induce hypothyroidism by blocking thyroid hormone synthesis. Supplementation with >100 µg/day to an individual who was formerly deficient occasionally induces hyperthyroidism (1.1 mg)

Practical methods for detecting marginal deficiency are not available. Marked deficiency is reliably detected by diminished serum copper and ceruloplasmin concentrations, as well as low erythrocyte superoxide dismutase activity.

Copper

Fluoride

Iodine

In the absence of supplementation, populations relying primarily on food from soils with low iodine content have endemic iodine deficiency. Maternal iodine deficiency leads to fetal deficiency, which produces spontaneous abortions, stillbirths, hypothyroidism, cretinism, and dwarfism. Rapid brain development continues through the second year, and permanent cognitive deficits may be induced by iodine deficiency during that period. In adults, compensatory hypertrophy of the thyroid (goiter) occurs, along with varying degrees of hypothyroidism (150 µg)

Estimates of intake or clinical assessment are used because no reliable laboratory test exists.

Urinary excretion of iodine is an effective laboratory means of assessment. The thyroidstimulating hormone (TSH) level in the blood is an indirect, not entirely specific, means of assessment. Iodine status of a population can be estimated by the prevalence of goiter.

Chapter 4  Nutritional Assessment and Management of the Malnourished Patient Table 4-11  Salient Features of Trace Minerals—cont’d Mineral

Deficiency (RDA)*

Toxicity (TUL)†

Assessment of Status

Iron

The most common micronutrient deficiency in the world. Women of childbearing age constitute the highest risk group because of menstrual blood losses, pregnancy, and lactation. Hookworm infection is the most common cause, worldwide. The classic deficiency syndrome is hypochromic microcytic anemia. Glossitis and koilonychia (spoon nails) are also observed. Easy fatigability often develops as an early symptom, before the appearance of anemia. In children, mild deficiency of insufficient severity to cause anemia is associated with behavioral disturbances and poor school performance (postmenopausal F, 8 mg; M, 8 mg; premenopausal F, 18 mg)

Negative iron balance initially leads to depletion of iron stores in the bone marrow: a bone marrow biopsy and the concentration of serum ferritin are accurate and early indicators of such depletion. As the severity of deficiency proceeds, serum iron (SI) decreases and total iron binding capacity (TIBC) increases; an iron saturation (= SI/TIBC) of 200 mg of zinc in a single day (in adults). It is manifested by epigastric pain, nausea, vomiting, and diarrhea. Hyperpnea, diaphoresis, and weakness may follow inhalation of zinc fumes. Copper and zinc compete for intestinal absorption: chronic ingestion of >25 mg zinc/day may lead to copper deficiency. Chronic ingestion of >150 mg/day has been reported to cause gastric erosions, low high-density lipoprotein cholesterol levels, and impaired cellular immunity (40 mg)

Until the deficiency syndrome is better defined, an appropriate measure of status will be difficult to develop.

No effective clinically available assessment exists. Rare cases of deficiency are associated with hypouricemia, hypermethionemia, and low levels of urinary sulfate with elevated excretion of sulfite, xanthine and hypoxanthine. Erythrocyte glutathione peroxidase activity and plasma, or whole blood, selenium concentrations are the most commonly used methods of assessment. They are moderately accurate indicators of status.

There are no accurate indicators of zinc status available for routine clinical use. Plasma, erythrocyte, and hair zinc concentrations are frequently misleading. Acute illness, in particular, is known to diminish plasma zinc levels, in part by inducing a shift of zinc out of the plasma compartment and into the liver. Functional tests that determine dark adaptation, taste acuity, and rate of wound healing lack specificity.

*Recommended Daily Allowance (RDA) established for female (F) and male (M) adults by the U.S. Food and Nutrition Board, 1999-2001. In some cases, insufficient data exist to establish an RDA, in which case the adequate intake (AI) established by the Board is listed. † Tolerable upper level (TUL) established for adults by the U.S. Food and Nutrition Board, 1999-2001. TPN, total parenteral nutrition. Adapted from Goldman L, Ausiello D, Arend W, et al, editors. Cecil Textbook of Medicine. 22nd ed. Philadelphia: WB Saunders; 2004. With permission.

57

58

Section II  Nutrition in Gastroenterology Table 4-12  Guidelines for Daily Delivery of Parenteral Micronutrients in Adults and Children Micronutrient Vitamin A D E K C Folate Niacin Riboflavin Thiamine B6 B12 Pantothenic acid Biotin Trace Elements Copper Chromium Manganese Zinc Molybdenum Iodine* Selenium Iron

Adults

Children

1000 µg (= 3300 IU) 5 µg (= 200 IU) 10 mg (= 10 IU) 1 mg 100 mg 400 µg 40 mg 3.6 mg 3 mg 4 mg 5 µg 15 mg 60 µg

700 µg 10 µg 7 mg 200 µg 80 mg 140 µg 17 mg 1.4 mg 1.2 mg 1.0 mg 1.0 µg 5 mg 20 µg

0.5-1.5 mg 10-15 µg 0.1 mg 2.5-4.0 mg 15 µg — 100 µg 1-2 mg

20 µg/kg/day 0.2 µg/kg/day 1.0 µg/kg/day 50 µg/kg/day 0.25 µg/kg/day — 2.0 µg/kg/day 1 mg/day

*

Naturally occurring contamination of parenteral nutrition appears to provide sufficient quantities of iodine. Adult vitamin guidelines adapted from American Society of Parenteral and Enteral Nutrition (ASPEN). Board of Directors and the Clinical Guidelines Task Force. Guidelines for the use of parenteral and enteral nutrition in adult and pediatric patients. JPEN J Parenter Enteral Nutr 2002;26:144. Children’s values adapted from Greene HL, Hambidge KM, Schanler R, Tsang RC. Guidelines for the use of vitamins, trace elements, calcium, magnesium, and phosphorus in infants and children receiving total parenteral nutrition: Report of the Subcommittee on Pediatric Parenteral Nutrient Requirements from the Committee on Clinical Practice Issues of the American Society for Clinical Nutrition. Am J Clin Nutr 1988; 48:1324; Am J Clin Nutr 1989; 49:1332; and Am J Clin Nutr 1989; 50:560.

selenium, iodine, fluorine, manganese, molybdenum, and cobalt (see Table 4-11). The biochemical functions of trace elements have not been as well characterized as those of the vitamins, but most of their functions appear to be as components of prosthetic groups or as cofactors for a number of enzymes. Aside from iron, the trace mineral depletion that clinicians are most likely to encounter is zinc deficiency. Zinc depletion is a particularly germane issue to the gastroenterologist because the gastrointestinal tract is a major site for zinc excretion. Chronically excessive losses of gastrointe­ stinal secretions, such as chronic diarrhea in inflammatory bowel disease, is a known precipitant for zinc deficiency, and in this setting zinc requirements often increase several-fold.25 Nevertheless, a biochemical diagnosis of zinc deficiency is problematic, as is true for many of the other essential trace minerals. Accurate laboratory assessment of zinc status is complicated by the very low concentrations of zinc in bodily fluids and tissues, a lack of correlation between serum and red blood cell levels of zinc with levels in the target tissues, and the reality that functional tests have yet to be devised. Furthermore, it is well recognized that in acute illness a shift in zinc occurs from the serum compartment into the liver, further obscuring the diagnostic value of serum zinc levels.26,27 Thus, it is often best simply to proceed with empirical zinc supplementation in patients whose clinical scenario puts them at high risk of zinc deficiency. Some reports have indicated that TPN solutions that deliver several-fold more manganese than what is recommended in Table 4-12 may lead to deposition of the mineral

in the basal ganglia, with extrapyramidal symptoms, seizures, or both.28 Because the content of manganese varies widely in the different trace element mixtures available for TPN compounding, the health professional needs to be mindful of this issue as protocols for TPN admixtures are developed.

PHYSIOLOGIC AND PATHOPHYSIOLOGIC FACTORS AFFECTING MICRONUTRIENT REQUIREMENTS Age

An evolution of physiology continues throughout the life cycle, with an impact on the requirements of certain micronutrients with aging; specific RDAs for older adults now have been developed. The mean vitamin B12 status of most populations, for example, declines significantly with older age, in large part because of the high prevalence of atrophic gastritis and its resultant impairment of protein-bound vitamin B12 absorption.29 Approximately 10% to 15% of the older ambulatory population is thought to have significant vitamin B12 depletion because of this phenomenon, and neuropathic degeneration may occur in older individuals whose plasma vitamin B12 levels are in the low-normal range (150 to 300  pg/mL), even in the absence of hematologic manifestations. For this reason, the use of sensitive indicators of cellular depletion of vitamin B12, such as serum methylmalonic acid levels in conjunction with serum levels of vitamin B12, now are recommended for diagnosis.30 Some experts also suggest that older adults should consume a portion of their vitamin B12 requirement in the crystalline form (i.e., as a supplement) rather than relying only on the naturally occurring protein-bound forms found in food.31 Older adults also require greater quantities of vitamins B6 and D and calcium to maintain health compared with younger adults, and these requirements are reflected in the new RDAs (see Tables 4-10 and 4-11).

Malabsorption and Maldigestion

Both fat- and water-soluble micronutrients are absorbed predominantly in the proximal small intestine, with the only exception being vitamin B12. Diffuse mucosal diseases, which affect the proximal portion of the gastrointestinal tract, are therefore likely to result in multiple deficiencies. Even in the absence of proximal small intestinal disease, however, extensive ileal disease, small bowel bacterial overgrowth, and chronic cholestasis may interfere with the maintenance of adequate intraluminal conjugated bile acid concentrations and thereby may impair absorption of fatsoluble vitamins. Conditions that produce fat malabsorption frequently are associated with selective deficiencies of the fat-soluble vitamins. The early stages of many vitamin deficiencies are not apparent clinically and therefore may go undetected until progression of the deficiency has resulted in significant morbidity. This can be disastrous in conditions such as spinocerebellar degeneration from vitamin E deficiency, which often is irreversible.32 Fat-soluble vitamin deficiencies are well-recognized complications of cystic fibrosis and congenital biliary atresia in which fat malabsorption often is overt, but monitoring also is necessary in conditions associated with more subtle fat malabsorption, such as the latter stages of chronic cholestatic liver disease.33,34 Restitution of vitamin deficiencies sometimes can be difficult when severe fat malabsorption is present and initial correction may require parenteral administration. In severe fat malabsorption, chemically modified forms of vitamins D and E that largely bypass the need for the lipophilic phase

Chapter 4  Nutritional Assessment and Management of the Malnourished Patient of intestinal absorption are commercially available for oral use and can be helpful. The polyethylene glycol succinate form of vitamin E (Nutr-E-Sol) is very effective in patients with severe fat malabsorption who cannot absorb conventional alpha-tocopherol.35 Similarly, hydroxylated forms of vitamin D (1-hydroxyvitamin D [Hectorol] and 1,25dihydroxyvitamin D [Rocaltrol]) can be used in patients resistant to the more conventional forms of vitamin D. Intermittent monitoring of serum calcium levels is indicated in the first few weeks of therapy when hydroxylated forms of vitamin D are administered because they are considerably more potent than vitamin D2 or D3 and risk of vitamin D toxicity exists. In contrast, water-miscible preparations of fat-soluble vitamins, in which a conventional form of vitamin A or E is dissolved in polysorbate 80 (e.g., AquasolE, Aquasol-A), have not been proven to improve overall absorption. At the time of this writing, Aquasol-A is no longer available as an oral supplement. Maldigestion usually results from chronic pancreatic insufficiency, which if untreated frequently causes fat malabsorption and deficiencies of fat-soluble vitamins. Vitamin B12 malabsorption also can be demonstrated in this setting, but clinical vitamin B12 deficiency is rare unless other conditions known to diminish its absorption also are present, such as atrophic gastritis29 or chronic administration of proton pump inhibitors (PPIs).36 Whether the long-term administration of PPIs alone warrants occasional checks of vitamin B12 status is a matter of debate. Regardless, malabsorption of vitamin B12 from atrophic gastritis or with PPIs is confined to dietary sources of vitamin B12. Small supplemental doses of crystalline vitamin B12 are absorbed readily in both cases. Histamine H2 receptor antagonists also inhibit protein-bound vitamin B12 absorption, although the effect generally is believed to be less potent than with the PPIs.37 Many medications may adversely affect micronutrient status. The manner in which drug-nutrient interaction occurs varies; some of the more common mechanisms are described in Table 4-13. A comprehensive discussion of drug-nutrient interactions is beyond the scope of this

Table 4-13  Interactions of Drugs on Micronutrient Status Drug(s)

Nutrient

Mechanism(S)

Dextroamphetamine, fenfluramine, levodopa Cholestyramine

Potentially all micronutrients

Induces anorexia

Vitamin D, folate

PPIs

Vitamin B12

Sulfasalazine

Folate

Isoniazid

Pyridoxine

NSAIDs

Iron

Penicillamine

Zinc

Adsorbs nutrient, decreases absorption Modest bacterial overgrowth, decreases gastric acid, impairs absorption Impairs absorption and inhibits folatedependent enzymes Impairs uptake of vitamin B6 Gastrointestinal blood loss Increases renal excretion

NSAIDs, nonsteroidal anti-inflammatory drugs; PPIs, proton pump inhibitors. From Goldman L, Ausiello D, Arend W, et al, editors. Cecil Textbook of Medicine. 22nd ed. Philadelphia: WB Saunders; 2004. With permission.

chapter and the reader is referred to other references for a detailed discourse on this topic.38

STARVATION During periods of energy and/or protein deficit, an array of compensatory mechanisms serves to lessen the pathophy­s­ iologic impact of these deficits. These responses decrease the metabolic rate, maintain glucose homeostasis, conserve body nitrogen, and increase the uptake of adipose tissue TGs to meet energy needs. To appreciate how acute illness disrupts this compensatory scheme, it is first necessary to understand how the body adapts to starvation in the absence of underlying disease. During the first 24 hours of fasting, the most readily available energy substrates (i.e., circulating glucose, FAs, and TGs, and liver and muscle glycogen) are used as fuel sources. The sum of energy provided by these stores in a 70-kg man, however, is only about 5000 kJ (1200 kcal) and therefore is less than a full day’s requirements. Hepatic glucose production and oxidation decrease, whereas whole-body lipolysis increases, and the latter provides additional FAs and ketone bodies.39 Oxidation of the FAs released from adipose tissue TGs accounts for approximately 65% of energy consumed during the first 24 hours of fasting. During the first several days of starvation, obligate glucose-requiring tissues such as the brain and blood cells, which collectively account for about 20% of total energy consumption, can use only glycolytic pathways to obtain energy. Because FAs cannot be converted to carbohydrate, these glycolytic tissues must use glucose or substrates that can be converted to glucose. Glucogenic amino acids derived from skeletal muscle (chiefly alanine and glutamine) are a major source of substrate for this purpose. Approximately 15% of the REE is provided by oxidation of protein.40 The relative contribution of gluconeogenesis to hepatic glucose production increases as the rate of hepatic glycogenolysis declines because the latter process becomes redundant; after 24 hours of fasting, only 15% of liver glycogen stores remain. During short-term starvation (1 to 14 days), several adaptive responses appear that lessen the loss of lean mass. A decline in levels of plasma insulin, an increase in plasma epinephrine levels, and an increase in lipolytic sensitivity to catecholamines stimulate adipose tissue lipolysis.41,42 The increase in FA delivery to the liver, in conjunction with an increase in the ratio of plasma glucagon-to-insulin concentrations, enhances the production of ketone bodies by the liver. A maximal rate of ketogenesis is reached by three days of starvation, and plasma ketone body concentration is increased 75-fold by seven days. In contrast to FAs, ketone bodies can cross the blood-brain barrier and provide most of the brain’s energy needs by seven days of starvation.43 The use of ketone bodies by the brain greatly diminishes glucose requirements and thus spares the need for muscle protein degradation to provide glucose precursors. If early protein breakdown rates were to continue throughout starvation, a potentially lethal amount of muscle protein would be catabolized in less than three weeks. Similarly, the heart, kidney, and skeletal muscle change their primary fuel substrate to FAs and ketone bodies. Other tissues such as bone marrow, renal medulla, and peripheral nerves switch from full oxidation of glucose to anaerobic glycolysis, resulting in increased production of pyruvate and lactate. The latter two compounds can be converted back to glucose in the liver using energy derived from fat oxidation via the Cori

59

60

Section II  Nutrition in Gastroenterology cycle, and the resulting glucose is available for systemic consumption. This enables energy stored as fat to be used for glucose synthesis. Whole-body glucose production decreases by more than 50% during the first few days of fasting because of a marked reduction in hepatic glucose output. As fasting continues, the conversion of glutamine to glucose in the kidney represents almost 50% of total glucose production. Energy is conserved by a decrease in physical activity because of fatigue and a reduction in REE of approximately 10%, resulting from increased conversion of active thyroid hormone to its inactive form and suppressed sympathetic nervous system activity. During long-term starvation (14 to 60 days), maximal adaptation is reflected by a plateau in lipid, carbohydrate, and protein metabolism. The body relies almost entirely on adipose tissue for its fuel, providing more than 90% of daily energy requirements.44 Muscle protein breakdown decreases to less than 30  g/day, causing a marked decrease in urea nitrogen production and excretion. The decrease in osmotic load diminishes urine volume to 200  mL/day, thereby reducing fluid requirements. Total glucose production decreases to approximately 75 g/day, providing fuel for glycolytic tissues (40  g/day) and the brain (35  g/day) while maintaining a constant plasma glucose concentration. Energy expenditure decreases by 20% to 25% at 30 days of fasting and remains relatively constant thereafter, despite continued starvation. The metabolic response to short- and long-term starvation differs somewhat between lean and obese persons. Obesity is associated with a blunted increase in lipolysis and decrease in glucose production compared with that in lean persons.45,46 In addition, protein breakdown and nitrogen losses are less in obese persons, thereby helping conserve muscle protein.47 The events that mark the terminal phase of starvation have been studied chiefly in laboratory animals. Body fat mass, muscle protein, and the sizes of most organs are markedly decreased. The weight and protein content of the brain, however, remain relatively stable. During this final phase of starvation, body fat stores reach a critical level, energy derived from body fat decreases, and muscle protein catabolism is accelerated. Death commonly occurs when there is a 30% to 50% loss of skeletal muscle protein.48 In humans, it has been proposed that there are certain thresholds beyond which lethality is inevitable—depletion of total body protein between 30% and 50% and of fat stores between 70% and 95%, or reduction of BMI below 13 kg/m2 for men and 11 kg/m2 for women.49,50

MALNUTRITION In the broadest sense, malnutrition implies a sustained imbalance between nutrient availability and nutrient requirements. This imbalance results in a pathophysiologic state in which intermediary metabolism, organ function, and body composition are variously altered. Sustained is an important element of this definition because homeostatic mechanisms and nutrient reserves usually are adequate to compensate for any short-term imbalance. Customarily, the term malnutrition is used to describe a state of inadequacy in protein, calories, or both, and is more precisely called protein-energy malnutrition (or proteincalorie malnutrition). Occasionally, it is used to describe a state of excessive availability, such as a sustained excess of calories (e.g., obesity) or a vitamin (e.g., vitamin toxicity).

PROTEIN-ENERGY MALNUTRITION

There are different pathways whereby protein-energy malnutrition (PEM) may evolve. Primary PEM is caused by inadequate intake of protein and/or calories or, less commonly, when the protein ingested is of such poor quality that one or more essential amino acids becomes a limiting factor in the maintenance of normal metabolism. Secondary PEM is caused by illness or injury. Acute illnesses and injuries increase bodily requirements for protein and energy substrate and they impair the digestion, absorption, and uptake of these nutrients in various ways. Consequently, secondary PEM usually arises from multiple factors. Illness and injury also commonly induce anorexia (see later for mechanisms), and so primary and secondary factors often act in concert to create PEM in the setting of illness. Illness or injury may directly interfere with nutrient assimilation; for example, extensive ileal disease or resection may directly produce fat malabsorption and a caloric deficit. The most common causes of secondary PEM, however, are the remarkable increases in protein catabolism and energy expenditure that occur as a result of a systemic inflammatory response. REE may increase as much as 80% above basal levels in a manner roughly proportional to the magnitude of the inflammatory response, which in turn is roughly proportional to the severity and acuity of the illness. Thus, for example, REE in patients with extensive second- and third-degree burns (the prototype for maximal physiologic stress) may approach twice normal; with sepsis, REE is about 1.5 times normal; and with a localized infection or fracture of a long bone, REE is 25% above normal.5 Such stress factors can be used to construct a formula for predicting the caloric needs of ill individuals (see Table 4-5). Protein catabolism during illness or injury also increases in proportion to the severity and acuity of the insult, and therefore parallels the increase in energy consumption. The magnitude of increase in protein catabolism, however, is proportionately greater than that observed with energy consumption, such that urinary urea N losses, which reflect the degree of protein catabolism in acute illness, are about 2.5 times the basal level with maximal stress.5 This increase in catabolism results in a net loss of protein because the rate of synthesis usually does not rise in concert with the rise in catabolism.51 No known storage form of protein exists in the body and, therefore, any net loss of protein represents a loss of functionally active tissue. A healthy adult typically loses about 12  g  N in the urine/day, and excretion may increase to as much as 30  g/day during critical illness. Because 1  g of urinary N represents the catabolism of approximately 30  g of lean mass, it follows that severe illness may produce a daily loss of up to ~0.5  kg of lean mass as a result of excess protein catabolism. Most of this loss comes from the skeletal muscle, where the efflux of amino acids increases two- to sixfold in critically ill patients.52 Mobilization of amino acids from skeletal muscle appears to be an adaptive response. Once liberated, these amino acids, in part, are deaminated and used for gluconeogenesis; they also are taken up by the liver and other visceral organs. The proteolysis of muscle under stress thus enables the body to shift amino acids from the skeletal muscle (the somatic protein compartment) to the visceral organs (the visceral protein compartment), the functions of which are more critical for immediate survival during illness. Nevertheless, with sustained stress, the limitations of this adaptive response become evident, and even the visceral protein compartment sustains a contraction in mass.44

Chapter 4  Nutritional Assessment and Management of the Malnourished Patient Primary versus Secondary Protein-Energy Malnutrition: A Body Compartment Perspective

The type of tissue lost as malnutrition evolves is critical in determining the pathologic ramifications of weight loss. Over 95% of energy expenditure resides in the lean body mass, which therefore contains the bulk of metabolism that sustains homeostasis. It is the maintenance of this body compartment that is most critical for health. Lean body mass can be subdivided further into somatic and visceral protein compartments, blood and bone cells, and extracellular lean mass, such as plasma and bone matrix (Fig. 4-1). In total or semistarvation in otherwise healthy individuals, adipose tissue predominates as a primary energy source; thus, fat mass contracts to a much greater degree proportional to the loss of lean mass.44 Alterations in metabolism from injury or illness, however, produce a proportionately greater loss of muscle mass so that it matches or exceeds the loss in fat mass.53,54 Although the lean mass that is lost in illness preferentially is from the somatic protein compartment, with sustained stress there also will be a significant contraction of the visceral protein compartment (Table 4-14). The metabolic forces associated

Table 4-14  Body Compartment Losses in Simple Starvation versus Metabolic Stress

Parameter Starvation Metabolic stress

Skeletal Muscle Wasting

Visceral Wasting

Loss of Fat Mass

+ +++

+/−* ++/−*

+++ +++

*

Relatively spared early in the process; can become pronounced with extended starvation or metabolic stress.

Blood cells, bone cells, etc. 7%

Extracellular lean mass (plasma, bone mineral, etc.) 36%

Visceral mass 7%

with acute illness and injury are potent, and restoration of muscle mass is unlikely with nutritional support unless the underlying inflammatory condition is corrected. There is increasing interest in attenuating or reversing net catabolism with the use of exogenous anabolic agents in conjunction with nutrition, although to date it remains unclear whether the clinical benefits of using exogenous growth hormone and other anabolic agents in acute illness outweigh their potential side effects.55,56 Another important ramification of the potency of the catabolic state associated with acute illness is that most of the weight that is gained with the provision of nutritional support is the result of increases in fat mass and body water; only minor increases in lean mass are observed until the inflammatory focus is resolved.57 Cytokines are the most important mediators of the alterations in energy and protein metabolism that accompany illness and injury. In a wide spectrum of systemic illnesses, increased secretion of interleukin-1 (IL-1), tumor necrosis factor-α (TNF-α), IL-6, and interferon-γ (IFN-γ) has been observed to be associated with increased energy expenditure and protein catabolism, as well as the shift of amino acids into the visceral compartments.58-60 Such observations concur with in vitro studies in human cells and animal models that have shown remarkably potent effects of these cytokines in this regard (Table 4-15). In the wasting syndrome associated with cancer, proteolysis-inducing factor and lipid-mobilizing factor are two humoral mediators that appear to be unique to cancer cachexia, contrib­ uting to protein catabolism and loss of adipose tissue, respectively.61

Protein-Energy Malnutrition in Children

Undernutrition in children differs from that in adults because it affects growth and development. Much of our understanding of undernutrition in children comes from observations made in underdeveloped nations in which poverty, inadequate food supply, and unsanitary conditions lead to a high prevalence of PEM. The Waterlow classification of malnutrition takes into account a child’s weight for height (wasting) and height for age (stunting) (Table 4-16).62 The characteristics of the three major clinical PEM syndromes in children—kwashiorkor, marasmus, and nutritional dwarfism—are outlined in Table 4-17.63 Although these three syndromes are classified separately, overlap syndromes often coexist in the same patient. Marasmus Weight loss and marked depletion of subcutaneous fat and muscle mass are characteristic features of children with marasmus. Ribs, joints, and facial bones are prominent and the skin is thin, loose, and lies in folds. In contrast, the visceral protein compartment is relatively spared, a fact that often is reflected by a normal serum albumin level, which in turn sustains normal oncotic pressure in the vascular compartment, thus minimizing edema.

Muscle mass 22% Fat mass 28% Figure 4-1.  Body composition analysis by weight in a healthy adult. The speckled segments and gray segment collectively represent lean body mass. The speckled segments alone represent body cell mass. (From Mason JB: Gastrointestinal cancer; nutritional support. In Kelsen D, Daly J, Kern S, et al, editors. Principles and Practice of Gastrointestinal Oncology. Philadelphia, Pa: Lippincott Williams & Wilkins; 2002. p 127. With permission.)

Kwashiorkor The word kwashiorkor, from the Ga language of West Africa, means “disease of the displaced child” because it commonly was seen after weaning. The presence of peripheral edema distinguishes children with kwashiorkor from those with marasmus and nutritional dwarfism. Children with kwashiorkor also have characteristic skin and hair changes (see sections on hair and skin changes, later). The abdomen is protuberant because of weakened abdominal muscles, intestinal distention, and hepatomegaly, but ascites is rare.

61

62

Section II  Nutrition in Gastroenterology Table 4-15  Major Cytokines That Mediate Hypercatabolism and Hypermetabolism Associated with Metabolic Stress Cytokine

Cell Source

Metabolic Effects

Tumor necrosis factor-α

Monocytes/macrophages, lymphocytes, Kupffer cells, glial cells, endothelial cells, natural killer cells, mast cells

IL-1

Monocytes/macrophages, neutrophils, lymphocytes, keratinocytes, Kupffer cells

IL-6

Monocytes/macrophages, keratinocytes, endothelial cells, fibroblasts, T cells, epithelial cells Lymphocytes, pulmonary macrophages

Decreased FFA synthesis Increased lipolysis Increased peripheral amino acid loss Increased hepatic amino acid uptake Fever Increased ACTH levels Increased hepatic acute-phase protein synthesis Fever Increased acute-phase protein synthesis Fever

IFN-γ

Increased monocyte respiratory burst

ACTH, adrenocorticotropic hormone; FFA, free fatty acid; IL, interleukin; IFN, interferon. Adapted from Smith M, Lowry S. The hypercatabolic state. In: Shils M, Olson J, Shike M, Ross AC, editors. Modern Nutrition in Health and Disease. Baltimore: Williams & Wilkins; 1999. p 1555.

Table 4-16  Waterlow Classification of Protein-Energy Malnutrition in Children Parameter Weight for Height (Wasting) Percentage of median NCHS standard Standard deviation from the NCHS median Height for Age (Stunting) Percentage of median NCHS standard Standard deviation from NCHS median

Normal

Mild

Moderate

Severe

90 to 100

80 to 89

70 to 79

30  kg/m2. The percentage of obese adults older than 20 years has been estimated since 1991, and these maps show the changing pattern. There has been a rapid and impressive increase in the percentage of obese people, both men and women, in the United States. Data of the Behavioral Risk Factor Surveillance System. (From the Centers for Disease Control and Prevention. U.S. obesity trends 1985-2007. Available at http://www.cdc.gov/nccdphp/dnpa/ obesity/trend/maps/index.htm.)

No data

1991

1996

2003

2006

35 inches (>88 cm) >40 inches (>102 cm) 40  kg/m2) was associated with disability in individuals who reported arthritis and those who did not report arthritis. Even lighter weight obese individuals had an increased likelihood of disability compared with normal-weight respondents.83,84 Rheumatoid arthritis and BMI have a paradoxical relationship. In a study of rheumatoid arthritis that accrued 123 deaths in 3460 patient-years of observation, the BMI was found to be inversely related to mortality, although the study period was relatively short and the number of subjects relatively small.85 Several skin changes are associated with excess weight.86 Stretch marks, or striae, are common and reflect the pressure on the skin from expanding lobular deposits of fat. Acanthosis nigricans refers to a deepening of the pigmentation in the folds of the neck, knuckles, and extensor surfaces that occurs in many overweight individuals. In normal-weight

105

106

Section II  Nutrition in Gastroenterology individuals, this can be a sign of increased risk of malignancy or insulin resistance but, when associated with obesity, such risks are unusual. Hirsutism in obese women may reflect altered increased androgen production, which can impair menstrual cycles and ovulation.4 Psychosocial Dysfunction Overweight is stigmatized in both children and adults87-89; that is, overweight individuals are exposed to the consequences of public disapproval of their fatness. The disapproval of obesity has, if anything, worsened over the past 40 years.90 Overweight children have a negative self-image and also a significant decrease in physical and social functioning compared with normal-weight children. A study using the Medical Outcomes Study Short-form Health Survey (SF-36) found that obese people had profound abnormalities in health-related quality of life.91 Overweight women appear to be at greater risk of psychological dysfunction, compared with overweight men, possibly because of increased societal pressures on women to be thin.92 A systematic review has shown that in four of eight studies that met criteria for inclusion in the investigation, obese subjects had an increased risk of dementia.93

BENEFITS OF WEIGHT LOSS

Weight loss improves a person’s health outlook in many ways.4 Weight loss reduces the risk of death in obese patients treated with bariatric surgery. In one study from Sweden94 with more than 4000 patients, half of whom received one of three bariatric operations (see later), there was a reduction in mortality of 24% after 10.9 years. Another study from Utah95 compared over 7000 patients who received a gastric bypass (see later) with those matched for weight and age and showed a reduction in mortality of 56% after 7.1 years.95 Weight reduction also reduces the risk from diseases that result from obesity. The Diabetes Prevention Program (DPP) is a clear example of a reduction in the risk of developing diabetes with weight loss. After an average of 2.8 years of follow-up in 3234 individuals with impaired glucose tolerance, those who were randomized to the intensive lifestyle treatment lost 7% of their body weight and showed a 58% reduction in the risk of developing diabetes.96 Blood pressure also benefits from weight loss. In the Framingham study, a modest weight loss of at least 6.8 kg or more led to a 28% reduction in the risk of hypertension in middle-aged adults and a 37% reduction in older adults.97 In a clinical trial using lifestyle interventions to lower blood pressure (TOHP II), the risk of being hypertensive was reduced 42% at 6 months and 22% at 18 months. In those who maintained a weight loss of 4.5 kg for 30 months, the risk of hypertension was reduced 65%.98 Apneic episodes also are influenced by changes in weight. Relative to stable weight, a 10% weight loss predicted a 26% reduction in the apnea-hyperpnea index.99 In a systematic review of long-term weight loss studies in obese adults, dietary and lifestyle approaches and pharmacologic interventions improved markers of cardiovascular disease, particularly in patients with cardiovascular risk factors at the beginning of the study.100 Quality of life also improves following weight loss.101

EVALUATION The hazards of excess weight and the benefits of weight loss point to the need for careful evaluation of the overweight patient. The National Heart, Lung, and Blood Institute

has provided an algorithm for evaluating the overweight patient. It is a useful framework for viewing the information that is collected during the evaluation of individual patients (Fig. 6-4). The basic components in the evaluation of any overweight or obese patient are a record of the historical events associated with the patient’s weight problem, a physical examination, and an appropriate laboratory assessment. Here, I have used the criteria recommended by the U.S. Preventive Services Task Force102 and also taken into account reports from the National Heart, Lung, and Blood Institute2 and the World Health Organization.1 The importance of evaluating overweight individuals has increased as the epidemic of overweight has worsened, and the number of patients potentially needing treatment has increased.

CLINICAL HISTORY

It is important to identify specific events associated with the increase in the patient’s body weight. Has there been a sudden increase in weight, or has body weight been rising steadily over a long period of time? Three categories of weight gain are identified: 22 pounds). In addition to total weight gain, the rate of weight gain after age 20 years needs to be considered when deciding the degree of risk for a given patient. Weight gain is associated with an increased risk to health and the more rapidly the patient is gaining weight, the more concerned the health care provider should be. Successful and unsuccessful weight loss programs undertaken by the patient also should be identified. A sedentary lifestyle increases the risk of early death. It is important to determine whether the patient comes from a family in which overweight is common, the usual setting, or whether she or he has become overweight in a family in which few people are overweight. The latter setting suggests a need to search for environmental factors that may be contributing to weight gain. Studies have shown that alteration in the melanocortin-4 receptor occurs in 2.5% to 5.5% of children and adolescents with a BMI > 30 kg/m2.28 This genetic defect is among the most common of those associated with any chronic disease, and evaluation for this defect may become important in the treatment of overweight people.

PHYSICAL EXAMINATION

The first step in the clinical examination of the overweight patient is to determine vital signs, which include BMI and waist circumference, as well as pulse and blood pressure.103 Accurate measurement of height and weight is the initial step in the clinical assessment,104 because these are needed to determine the BMI (see earlier). The BMI has a reasonable correlation with body fat and a curvilinear relationship to risk. Risk arbitrarily has been subdivided by cut points derived from data collected on whites. It is now clear, however, that different ethnic groups have different percentages of body fat for the same BMI105 and BMI thus needs to be interpreted in an ethnically specific context. An Asian Conference selected lower levels of BMI to define overweight (BMI < 23 kg/m2) and obesity (BMI > 25 kg/m2); the same BMI presumably carries a different level of risk in various populations. These differences need to be taken into consideration when making clinical judgments about the degree of risk for the individual patient. During weight loss, body weight is more useful than the BMI, because height doesn’t change during this period, and the need to use the height squared makes it more difficult for the physician and patient to calculate. After BMI, waist circumference is the second vital sign in the evaluation of the overweight individual. Waist

Chapter 6  Obesity Patient encounter Hx of BMI ≥25? No BMI measured in past 2 years?

Measure Wt, Ht, and waist circumference Calculate BMI

Yes

BMI ≥25 or waist circumference >88 cm (F) >102 cm (M)

Yes

BMI ≥30 or waist circumference Yes >88 cm (F) >102 cm (M) and >1 risk factor

Assess risk factors

No Hx of BMI ≥25?

No

Yes

Does patient want to lose wt?

No Reinforcement/ educate on weight management

Devise goals and strategy for weight loss and risk factor control

Advise to maintain weight/address other risk factors

Yes

No

Yes

Progress made/goal achieved? No

Examination Treatment

Periodic weight check

Maintenance counseling: Diet Behavior therapy Exercise

Assess reasons for failure to lose weight

Figure 6-4.  Algorithm for the diagnosis and treatment of obesity developed by the National Heart, Lung and Blood Institute (NHLBI). BMI, body mass index; Ht, height; Hx, history; Wt, weight. (From Carpenter KM, Hasin DS, Allison DB, Faith MS. Relationships between obesity and DSM-IV major depressive disorder, suicide ideation, and suicide attempts: Results from a general population study. Am J Public Health 2000; 90:251-7.)

circumference is the easiest measurement to evaluate central adiposity. It is determined using a metal or nonstretchable plastic tape. Measurements at the level of the umbilicus or at the midpoint between the lower rib and suprailiac crest are the two most common locations. Waist circumference is a good strategy for following the clinical progress of weight loss and is particularly valuable if patients become more physically active. Physical activity may slow loss of muscle mass and thus slow weight loss, whereas fat continues to be mobilized. Waist circumference can help in making these distinctions. The relationship of central fat to risk factors for health varies among populations as well as within them. Japanese Americans and Indians from South Asia have relatively more visceral fat and are thus at higher risk for a given BMI or total body fat than whites. Even though the BMI is below 25 kg/ m2, central fat may be increased, particularly in Asian populations, and may increase the risk of disease.106 Central adiposity is important, particularly with a BMI between 22 and 29  kg/m2. Blood pressure should be measured carefully. Hyper­ tension is amenable to improvement with diet107 and is an important criterion for diagnosis of the metabolic syndrome. The patient should sit quietly for 5 minutes before measuring the blood pressure with a calibrated instrument to increase the accuracy of measurement. The blood pressure criteria from the Seventh Joint National Commission recommendations should be followed. A normal blood pres-

sure is less than 120/80  mm Hg. Prehypertension was defined by this group as a systolic blood pressure (SBP) of 120 to 139 mm Hg and diastolic BP (DBP) of 80 to 89 mm Hg. Hypertension is then a SBP/DBP of 140/90 mm Hg and clearly needs treatment if such blood pressure values are confirmed. Individuals with prehypertension need to be carefully observed. Acanthosis nigricans (see earlier) in normal-weight individuals may signify increased insulin resistance or malignancy, but this is not usually the case in obesity. If this is suspected, however, further evaluation is necessary.

LABORATORY STUDIES

Serum lipids, glucose, C-reactive protein (now measured as high-sensitivity CRP [hs-CRP]), and other values indicated from the history and physical examination should be measured. An increased fasting glucose, low HDL cholesterol, and high triacylglycerol levels are atherogenic components of the metabolic syndrome. Along with elevated blood pressure, it is possible to categorize the patient as having the metabolic syndrome by using criteria proposed by the National Cholesterol Education Program (see Table 6-2). An individual has the metabolic syndrome if three of the five criteria are abnormal. Measurement of LDL cholesterol also is important because it may need treatment independently of obesity or central adiposity. A positive hs-CRP assay along with an elevated serum LDL cholesterol level is a clear risk factor for heart disease.

107

108

Section II  Nutrition in Gastroenterology PREVENTION Studies designed to prevent obesity have been conducted in children and in adults.4 For children and adolescents, many school-based programs have been tried and, although there have been some promising results, the long-term impact of such programs has been small. In one successful study in children, a reduction in television watching slowed their gain in BMI.108 Another study has shown that decreasing children’s consumption of carbonated beverages, primarily soft drinks, was associated with slower weight gain than that of children who were not given this advice.109 In studies involving adults, however, there are few successful preventive programs.

TREATMENT Realism is one important aspect of treatment for obesity. For most forms of treatment, including behavior therapy, diet, and exercise, weight loss levels off at less than 10% below baseline. For many patients this is a frustrating experience, because their dream weight would require a loss of almost 30% of their body weight.110 A weight loss of less than 17% would be a disappointment with participation in a weight loss program. Yet, other than surgery, a weight loss of 10% is the expected outcome. It is important for the patient and physician to realize that an initial weight loss of 10% of body weight should be considered a success and that this amount of weight loss lessens the health risks of obesity.100 Because obesity left to itself will lead to a number of associated diseases, there are two therapeutic strategies: (1) wait until associated diseases develop (e.g., diabetes, hypertension, or dyslipidemia) and treat them individually; or (2) treat the obesity itself, thus reducing the risk of developing diabetes, hypertension, and other associated diseases. The second approach is preferable.

DIETS

To lose weight, a person must consume fewer calories (eat less food) than the body needs for daily activities. Many diet plans are available for overweight individuals (Table 6-3).111 These can be categorized as those that are low in fat (Ornish, Jenny Craig), those that are low in carbohydrate (Atkins, South Beach), those that restrict most nutrients, the so-called balanced deficit diets (Weight Watchers, Volumetrics, Slim Fast, eDiets), those that highlight one type of food or another (e.g., the low glycemic

index diet), or diets that highlight special foods (e.g., the grapefruit diet). The efficacy of dietary counseling versus control therapy has been examined in a meta-analysis.112 A random effects model of 46 studies of dietary counseling showed a maximum net treatment effect of −1.9 BMI units (95% confidence interval [CI]: −2.3, −1.5) or approximately a 6% weight loss over 12 months. There was a loss of about 0.1 BMI unit/month for the 12 months of active treatment and a regain of about 0.02 to 0.03 BMI unit/month during subsequent phases of the program. There were many different strategies used in the studies reviewed in this meta-analysis, but there was no clear basis for selecting one dietary approach over another.

Popular Diets

Low-Fat Diets Low-fat diets are a standard strategy to help patients lose weight. One benefit from a very low level of fat intake is the slowing or reversal of coronary artery disease.113 A metaanalysis of five randomized controlled trials of low-fat diets, however, has shown that these diets produce significant weight loss, but no more so than the control diets.114 Subsequently, 48,835 women were randomly assigned in a large clinical trial to low-fat or control diets.115 Weight loss was 2.2 kg below baseline at year one and 0.6 kg at an average of 7.5 years of follow-up. At both time points, weight loss was significantly less in the women on the low-fat diet compared with those on the normal-fat diet, and there was a clear relationship between the decrease in percentage of fat ingested and weight loss (P < 0.001 for trend). A metaanalysis of weight loss studies has found that over the first six months, low-fat diets produce weight loss and that heavier individuals lose more weight than lighter weight individuals.116 Low Energy Density Diets The theory behind the use of low energy density diets117 is that filling the stomach with low-fat, high-fiber foods (low energy density) reduces hunger and produces satiety. Conversely, in experimental settings, people eat more food when it is more energy dense—that is, has more calories per unit weight. Low-Carbohydrate Diets The most popular diets are the low-carbohydrate, highprotein, high-fat diets. Daily carbohydrate intake in some of these diets is as low as 13 g; when carbohydrate intake is less than 50 g/day, ketosis develops. In short-term metabolic ward studies, patients did not increase the intake

Table 6-3  Nutritional Content of Several Popular Diets

Diet Atkins eDiets Jenny Craig Ornish Slim-Fast South Beach Volumetrics Weight Watchers

Protein (%)

Fat (%)

Saturated Fat (%)

29 24 20 16 21 22 22 20

60 23 18 6 22 39 23 24

20 5 7 1 6 9 7 7

Adapted from Rating the Diets from Atkins to Dr. Sears Zone. Consum Rep 2005; 70:18-22.

Carbohydrate (%)

Fiber (g/1000 kcal)

Daily Servings of Fruits and Vegetables

11 53 62 77 57 38 55 56

12 19 16 31 21 19 20 20

6 12 6 17 12 3 14 11

Chapter 6  Obesity of other foods to compensate for the lower calories in a very low carbohydrate diet.118 Several randomized clinical trials have reported greater weight loss in the low-carbohydrate diet group up to six months, but not at one year.119 Four clinical trials have compared the effect of macronutrient composition on weight loss in one-year studies120,121 and two-year studies.122,123 The two one-year studies compared the Atkins, Zone, and Ornish diets, with the Weight Watchers diet in one study120 or the LEARN manual in another study.121 One two-year study compared the Atkins diet, a Mediterranean-style diet, and the American Heart Association low-fat diet122 and the other compared four diets: 20% or 40% fat with 15% or 25% protein.123 In the first one-year study,120 160 participants were randomly assigned to each diet. After one year, there was no significant difference in the weight loss of patients using any of the four diets. The weight loss was as follows: −3.9 ± 6 kg (−8.58 ± 13.2 pounds) with the Atkins diet; −4.9 ± 6.9  kg with the Zone diet (CI, 10.78 ± 15.18 pounds); −4.6 ± 5.4 kg (−10.12 + 11.88 pounds) with the Weight Watchers diet; and −6.6 ± 9.3 kg (−14.52 ± 20.46 pounds) with the Ornish diet— thus, no differences among the diets. The principal determinant of weight loss was the degree of adherence to the diet, not the diet itself. In the second one-year trial, 311 premenopausal women were randomly assigned to one of four diets.121 In this study, the Atkins diet produced more weight loss at 12 months (−0.7 kg) compared with the other three diets (Zone, −1.6 kg; LEARN, −2.6 kg; Ornish, −2.2 kg), a difference that was not statistically different. Also in this study, adherence to the diet was the principal determinant of success. The first two-year study compared a Mediterranean diet, Atkins diet, and low-fat diet in a group composed of 90% men living in a town in Israel. At the end of two years, the weight loss was −4.7 kg for the low-carbohydrate group, −2.9  kg for the low-fat group, and −4.4  kg for the Mediterranean diet group. After reaching a plateau at six to seven months, the group on the Mediterranean diet had another drop in weight equal to that of the low-carbohydrate group.122 The second two-year study was the most complex, using a 2 × 2 design that had four carbohydrate levels (35%, 45%, 55%, and 65%) resulting from the two fat levels (20% or 40%) and the two protein levels (15% or 25%).123 In this study, 811 individuals were randomized to one of the four diets. The weight loss at one year averaged 7% across diets, with no significant differences. Thereafter, there was a small weight regain. Attendance at support, education, and diet groups strongly predicted success. The authors concluded that “the content of dietary fat, carbohydrate, and protein had little influence on body weight loss over two years in obese people.”

Very Low-Calorie Diets

Very low-calorie diets (i.e., diets with an energy level below 800  kcal/day), can be used for rapid weight loss prior to major surgery. In other settings, the weight rebound that usually occurs at the end of a program with very lowcalorie diets may not make them worth the effort for some people, and may deter them from using similar diets in the future. A systematic review of 29 studies of weight loss programs using a very low-calorie diet that lasted more than two years124 found that participants in the very lowcalorie diet program lost significantly more weight than those eating hypoenergetic balanced diets. The very lowcalorie diets, however, have been replaced largely by portion-controlled diets, in which the calories from bever-

ages, bars, or frozen meals provided at breakfast or lunch are fixed by the manufacturer. In a four-year study, this approach resulted in early initial weight loss, which then was maintained.125

Commercial Programs

A number of commercial and self-help programs, including Overeaters Anonymous, Take Off Pounds Sensibly (TOPS), Weight Watchers, Jenny Craig, Herbalife, OPTIFAST, LA Health, and eDiets, are available to the consumer. Tsai and Wadden126 have examined the effectiveness of a number of these programs. The Weight Watchers program is done in groups, in contrast to Jenny Craig and LA Weight Loss, in which clients are seen individually. Jenny Craig uses prepackaged food and Weight Watchers and LA Weight loss use diet plans. All three programs encourage physical activity. In one trial lasting two years and including 423 subjects, the participants in the intervention group attended the Weight Watchers meetings and experienced a mean weight loss of 5.3% at one year and 3.2% at two years, compared with 1.5% at one year and 0% weight loss for the control group that received the self-help intervention with two visits to a dietitian.127

LIFESTYLE MODIFICATION

A basic strategy in helping obese patients lose weight is through lifestyle changes. The first step in this process is to determine whether the individual is really ready to make lifestyle changes. Patients often have a dream weight that involves a weight loss of nearly 30% of their initial body weight.110 An initial loss of 5% to 10% of body weight is a more realistic goal, because it will significantly reduce many of the health hazards described above, if they are present.100 Behavioral strategies include helping patients learn to monitor their eating behavior by recording what is eaten, the setting in which it is eaten, and the situations that trigger eating. With this information, the health care provider can help a patient change his or her eating habits. Patients should be encouraged to use a defined eating plan. People who are successful in losing weight and maintaining weight loss tend to monitor their behavior, eat low-fat diets, increase their physical activity, and practice positive self-thinking and techniques for stress reduction, as documented by the National Weight Loss Registry.128 Use of the Internet is a promising new tool.129

EXERCISE

Exercise is one strategy for balancing energy intake and expenditure, whether as a primary treatment for weight loss or for prevention of weight regain. Walking expends approximately 100 kcal/mile. A deficit of 3500 kcal (500 kcal/day) maintained for one week should result in a loss of 0.45 kg (1 pound). To obtain this effect from exercise alone, an individual would need to walk five miles/day, seven days/ week. For this reason, exercise alone has not been very effective as a primary weight loss technique.4 Moderate to vigorous exercise for 60 min/day, six days/week, produced more weight loss (−1.4 kg in women; −1.8 kg in men) than that of a nonexercise group in over 12 months.130 A metaanalysis of weight loss trials lasting at least one year has found that exercise-alone groups had minimal weight loss. Use of resistance training, as opposed to aerobic exercise, may help retain lean body mass and reduce the associated fall in resting energy expenditure.131 For individuals wanting to monitor their exercise, inexpensive pedometers can be worn on the belt. A mile is about 2000 steps, and increasing the number of monitored

109

110

Section II  Nutrition in Gastroenterology Table 6-4  Drugs Approved by the U.S. Food and Drug Administration for the Treatment of Obesity Drug

Trade Name

Dosage

Pancreatic Lipase Inhibitor* Orlistat Xenical Norepinephrine-Serotonin Reuptake Inhibitor* Sibutramine Meridia (United States); Reductil (Europe and other countries) Noradrenergic Drugs† Diethylpropion Tenuate, Tepanil Tenuate Dospan Phentermine Adipex-P, Fastin, Phentercot Ionamin Benzphetamine Didrex Phendimetrazine Bontril XR, Plegine Plegine, Prelu-2, X-Trozine

DEA Schedule

120 mg three times daily before meals

—

5-15 mg daily

IV

25 mg twice daily 75 mg every morning 15 to 37.5 mg daily 15-30 mg daily 25-50 mg daily to three times daily 35 mg twice to three times daily 105 mg daily

IV IV III III

DEA, Drug Enforcement Administration. *Approved for long-term use; †Approved for short-term use.

steps walked each day is a good way to encourage walking.132

PHARMACOTHERAPY

Because all medications inherently have more risks than diet and exercise, medications should only be used when the benefits justify the risk. Current medications for the treatment of obesity can be divided into two broad categories: (1) those that act primarily on the central nervous system to reduce food intake; and (2) those that act primarily outside the brain. Wherever the primary site of action may be, however, the net effect must be a reduction in food intake, an increase in energy expenditure, or both. There currently are several drugs available in the United States to treat obesity133-135 (Table 6-4).

Mechanisms of Drug Action

The brain plays a central role in regulating food intake by receiving and processing information from the environment and internal milieu.34 A number of neurotransmitter systems, including monoamines, amino acids, and neuropeptides, are involved in modulating food intake. The monoamines include norepinephrine, serotonin, dopamine, and histamine, as well as certain amino acids. The serotonin system has been one of the most extensively studied of the monoamine pathways. Its receptors modulate both the quantity of food eaten and macronutrient selection. Stimulation of the serotonin receptors in the paraventricular nucleus reduces fat intake, with little or no effect on the intake of protein or carbohydrate. This reduction in fat intake is pro­bably mediated through 5-HT2C receptors, because its effect is attenuated in mice that cannot express the 5-HT2C receptor. Sibutramine blocks reuptake of serotonin and norepinephrine. Lorcaserin is a drug in clinical trials that acts directly on serotonin receptors in the brain. Stimulation of α1-adrenergic receptors also reduces food intake,134 as shown by the α1 agonist phenylpropanolamine. Some of the α1 receptor antagonists used to treat hypertension produce weight gain, further indicating a role for this receptor in weight control. In contrast, stimulation of α2adrenergic receptors increases food intake in experimental animals, and a polymorphism in the α2a adrenoceptor has been associated with reduced metabolic rate in humans. Activation of β2 receptors in the brain reduces food intake. These receptors can be activated by β agonists, which release norepinephrine in the vicinity of these receptors or block the reuptake of norepinephrine. Sibutramine is both a serotonin and norepinephrine reuptake inhibitor and also uses this mechanism.

Histamine receptors also can modulate feeding. Stimulation of the H1 receptor in the central nervous system reduces feeding. Experimentally, this has been addressed by modulating the H3 autoreceptor, which controls histamine release. When the autoreceptor is stimulated, histamine secretion is reduced and food intake increases. Blockade of this H3 autoreceptor decreases food intake. The histamine system is important in control of feeding because some psycho­ active drugs bind to histamine receptors and produce weight gain.19 The opioid receptors were the first group of peptide receptors shown to modulate feeding. They also modulate fat intake. Both mu and kappa opioid receptors can stimulate feeding. Stimulation of the mu opioid receptors increases the intake of dietary fat in experimental animals. Two other peptides, corticotropin-releasing hormone (CRH) and the closely related urocortin, reduce food intake and body weight in experimental animals. The endocannabinoid system is the most recent addition to the central controllers of feeding.36 Tetrahydrocanna­ binol, isolated from the marijuana plant, stimulates food intake. Isolation of the cannabinoid receptor was followed by the identification of two fatty acids, anandamide and 2-arachidonoylglycerol, which are endogenous ligands for this receptor; infusion of either ligand into the brain stimulates food intake. The CB1 receptor is a preganglionic receptor, meaning that its activation inhibits synaptic transmission. Antagonists to this receptor have been shown to reduce food intake and lead to weight loss.36 In addition to the drugs that act on the central nervous system, there are also drugs that act peripherally.33 Thus, for example, blockade of intestinal lipase by orlistat will produce weight loss. A second drug in this class, cetilistat, is in clinical trials. Pancreatic and intestinal peptides modulate food intake and also are candidates for treatment targets. Intestinal glucagon-like peptide-1 (GLP-1) acts on the pancreas, intestine, and brain to reduce food intake and slow gastric emptying. GLP-1 and exenatide, a drug that works by this GLP-1 mechanism, increase insulin secretion from the pancreas, reduce glucagon release from the pancreas, and reduce food intake by acting on GLP-1 receptors in the brain. Amylin is secreted from the pancreatic beta cell and can reduce food intake. Pramlintide is an example of a drug that works by mimicking the effects of amylin, which reduce food intake by acting on receptors in the brain.

FDA-Approved Medications

The FDA has approved several drugs for the treatment of obesity, shown in Table 6-4. Two of them, sibutramine and

Chapter 6  Obesity orlistat, are approved for long-term use (12 months); the others are approved for up to a few weeks, which is usually interpreted as 12 weeks. Sibutramine.  Sibutramine (Meridia in the United States and Reductil in Europe and other countries) has been marketed in the United States since March 1998. It is a selective reuptake inhibitor of serotonin and norepinephrine into neurons, but does not act on any known receptors. Sibutramine promotes satiety, but may also increase energy expenditure by blocking the reduction in metabolic rate that accompanies weight loss. In a randomized, placebocontrolled, six-month dose-ranging study of 1047 patients, there was a clear dose-response effect in dosages of 1 to 30  mg/day. In this study, 67% of subjects treated with sibutramine achieved a 5% weight loss from baseline, and 35% lost 10% or more.136 In a one-year trial of 456 patients who received sibutramine (10 or 15  mg/day) or placebo, 56% of those who stayed in the trial for 12 months lost at least 5% of their initial body weight, and 30% of the patients lost 10% of their initial body weight while taking the 10-mg dose.137 In a trial of patients who initially lost weight eating a very low-calorie diet before being randomized to sibutramine (10  mg/day) or placebo, sibutramine produced additional weight loss, whereas the placebo-treated patients regained weight.134,138 Sibutramine also is effective for weight maintenance. The Sibutramine Trial of Obesity Reduction and Maintenance (STORM) Trial began with a six-month open-label phase to induce weight loss using 10 mg/day of sibutramine. Patients who lost more than 8 kg then were randomized to sibutramine or placebo.139 During the 18-month double-blind phase of this trial, the placebo-treated patients steadily regained weight, maintaining only 20% of their weight loss at the end of the trial. In contrast, the subjects treated with sibutramine maintained 80% of their weight loss after two years.139 The blood pressure levels of the sibutramine-treated patients were still higher than in patients treated with placebo (see later). Clinical trials with sibutramine have shown that about 75% of patients treated with 15  mg/day of sibutramine achieved more than 5% weight loss, and 80% of those maintained that loss for 2 years if they stayed on the drug. In a meta-analysis of clinical trials with sibutramine,135 the drug produced a weighted mean weight loss of 6.35 ± 6.47 kg (−13.9 pounds) compared with 2.18 ± 5.23 kg (−4.8 pounds) for the placebo group, giving a net effect, or what is often called the placebo-subtracted weight loss, of −4.16 kg (95% CI: −4.73 to −3.59). About 5% of patients do not tolerate sibutramine because of adverse effects on blood pressure and pulse. Some patients (∼25%) are nonresponders. This drug, like other sympathomimetic drugs, produces a small increase in mean heart rate and mean blood pressure, as observed in clinical trials; however, the blood pressure response is variable. Sibutramine should be used with caution in patients with cardiovascular disease and in individuals taking selective serotonin reuptake inhibitors. It should not be used within two weeks of taking monoamine oxidase inhibitors or with other noradrenergic agents. A subset (∼5%) of patients appears to be sensitive to the blood pressure effects and cannot tolerate the drug. Other side effects, including dry mouth, insomnia, and asthenia, are similar to those of other noradrenergic drugs. Sibutramine is not associated with valvular heart disease, primary pulmonary hypertension, or substance abuse. Orlistat.  Orlistat (Xenical) is available by prescription in a dosage of 120 mg three times daily and Alli is offered over-

the-counter (OTC) at a lower dosage, 60  mg three times daily; both drugs should be taken before meals. Orlistat inhibits the enzymatic action of pancreatic lipase. In a twoyear trial with orlistat,140 patients received a hypocaloric diet that was 500 kcal/day less than their calculated requirements for the first year, and a diet that was calculated to maintain body weight in the second year. By the end of the first year, the placebo-treated patients lost 6.1% of their initial body weight and the drug-treated patients lost 10.2%. At the end of the second year, the patients who were switched from orlistat to placebo after one year gained weight, from 10% to 6% below baseline, a gain of 4%. Patients switched from placebo to orlistat lost weight, from 6% to 8.1% below baseline (a loss of 2.1%), an amount essentially identical to the 7.9% weight loss in the patients treated with orlistat for the full 2 years. In a three-year study,141 a very low-energy diet was used for eight weeks, and subjects who lost a minimum of 5% of their body weight were randomized to lifestyle or lifestyle plus orlistat. Weight loss continued to decline for three months and remained below randomization levels at 12 months in the orlistat group, but had risen above randomization level by six months in the lifestyle controls. At the end of three years, those on orlistat were still 2.4 kg lighter than the controls. Clinical trials show that ∼70% of patients will achieve more than 5% weight loss and at 2 years, 70% of them will have maintained that loss. There are clinical trials documenting orlistat use for up to four years.142 One advantage to orlistat’s use is its beneficial effect on LDL cholesterol. Because orlistat blocks fat absorption, the LDL reduction is about twice that seen with weight loss alone. A meta-analysis has shown that the orlistat-treated patients had a weighted mean weight loss of −5.70 ± 7.28 kg (−12.6 pounds) compared with −2.40 ± 6.99  kg (−5.3 pounds), giving a net, or placebo-subtracted weighted mean, weight loss of −2.87 (95% CI −3.21 to −2.53) (−6.4 pounds).135 In regard to safety, orlistat is poorly absorbed; all its side effects are those expected from inhibition of lipase in the intestine. It can produce fecal incontinence, anal leakage, bloating, and borborygmi, but these tend to occur early in treatment and deter very few patients. It also can lower levels of fat-soluble vitamins. A multivitamin taken when orlistat is not taken can prevent the reduction in fat-soluble vitamin levels. Sympathomimetic Amines.  Four sympathomimetic drugs have been approved by the FDA.134 Two (phentermine, diethylpropion) are schedule IV drugs and the other two (benzphetamine and phendimetrazine) are schedule III drugs (see Table 6-4). These drugs are only approved for a few weeks of use, which usually is interpreted as up to 12 weeks. Phentermine is not available in Europe. Obtaining written informed consent if phentermine is prescribed for longer than 12 weeks is good medical practice, because of the paucity of published reports on the long-term use of phentermine. Cannabinoid CB1 Antagonists.  Rimonabant is a cannabinoid receptor antagonist to the CB-1 receptor and initially was approved by European regulatory authorities. Marketing approval in Europe was withdrawn on October 23, 2008 by the European Medicines Agency (EMEA), and Sanofi-Aventis stopped development of this drug on October 5, 2008.

Non–FDA-Approved Medications

Fluoxetine.  Fluoxetine is a selective serotonin reuptake inhibitor (SSRI) that blocks serotonin transporters, thus pro-

111

112

Section II  Nutrition in Gastroenterology longing the action of serotonin. Fluoxetine is approved by the FDA for the treatment of depression. Fluoxetine at a dosage of 60  mg/day (three times the usual dose for treatment of depression) was effective in reducing food intake4 and body weight in overweight patients. A meta-analysis of six studies using fluoxetine has shown a wide range of results with a mean weight loss in one study of 14.5 kg and a weight gain of 0.40 kg in another.133 In the meta-analysis by Avenell and colleagues,143 the weight loss at 12 months was 0.33 kg (95% CI, −1.49 to 0.82 kg). Goldstein and colleagues reviewed the trials with fluoxetine that included one 36-week trial in type II diabetic subjects, one 52-week trial in subjects with uncomplicated overweight, and two 60-week trials in subjects with dyslipidemia, diabetes, or both.144 A total of 719 subjects were randomized to fluoxetine and 722 to placebo. Six months of treatment were completed by 522 subjects on fluoxetine and 504 subjects on placebo. Weight losses in the placebo and fluoxetine groups at six months were 2.2 and 4.8  kg and at one year were 1.8 and 2.4 kg, respectively. The regain of 50% of the lost weight during the second six months of treatment with fluoxetine makes this drug inappropriate for the long-term treatment of obesity. Fluoxetine, however, although not a good drug for long-term treatment of obesity, may be preferred for the treatment of depressed obese patients over some of the tricyclic antidepressants, which are associated with significant weight gain. Bupropion.  Bupropion is a norepinephrine and dopamine reuptake inhibitor approved for the treatment of depression and for help in smoking cessation. Two multicenter clinical trials, one in obese subjects with depressive symptoms and one in uncomplicated overweight patients, have tested this drug. In the study of overweight patients with depressive symptom ratings of 10 to 30 on a Beck Depression Inventory, 213 patients were randomized to 400 mg/day of bupropion and 209 subjects were assigned to placebo over a 24-week period. In the bupropion group, 121 subjects completed the trial and lost 6.0% ± 0.5% of initial body weight; the 108 subjects in the placebo group who completed the trial lost 2.8% ± 0.5% (P < 0.0001).145 A study in uncomplicated overweight subjects randomized 327 subjects to bupropion, 300  mg/day, bupropion 400  mg/day, or placebo in equal proportions.146 At 24 weeks, 69% of those randomized remained in the study and the percent losses of initial body weight were 5% ± 1%, 7.2% ± 1%, and 10.1% ± 1% for the placebo, bupropion 300-mg, and bupropion 400-mg groups, respectively (P < 0.0001). The placebo group was randomized to the 300- or 400-mg group at 24 weeks and the trial was extended to week 48. By the end of the trial, the dropout rate was 41%, and the weight losses in the bupropion 300and 400-mg groups were 6.2% ± 1.25% and 7.2% ± 1.5% of initial body weight, respectively.146 Thus, it appears that nondepressed subjects may respond to bupropion with more weight loss than those with depressive symptoms. Topiramate.  Topiramate is an antiepileptic drug that was discovered to be associated with weight loss in its clinical trials for epilepsy.147 Weight losses of 3.9% of initial weight were seen at three months and losses of 7.3% of initial weight were seen at one year. Bray and colleagues reported a six-month, placebo-controlled, dose-ranging study of topiramate in which 385 obese subjects were randomized to placebo or topiramate at 64, 96, 192, or 384 mg/day.148 These doses were gradually reached by a tapering increase and were reduced in a similar manner at the end of the trial. Weight loss from baseline to 24 weeks was 2.6%, 5%, 4.8%, 6.3%, and 6.3% in the placebo, 64-, 96-, 192-, and 384-mg

groups, respectively. The most frequent adverse events were paresthesias, somnolence, and difficulty with concentration, memory, and attention. This trial was followed by two other multicenter trials.149,150 The first trial randomized 1289 obese subjects to placebo or topiramate 89, 192, or 256 mg/day. This trial was terminated early because of the sponsor’s decision to pursue a time-release form of the drug. The 854 subjects who completed one year of the trial before it was terminated lost 1.7%, 7%, 9.1%, and 9.7% of their initial body weight in the placebo, 89-, 192-, and 256-mg groups, respectively. Subjects in the topiramate groups also had significant improvement in blood pressure and glucose tolerance.149 The second trial enrolled 701 subjects who were treated with a very low-calorie diet to induce an 8% loss of initial body weight.150 The 560 subjects who achieved an 8% weight loss were randomized to topiramate 96 or 192  mg/day, or placebo. This study also was terminated early. At the time of termination, 293 subjects had completed 44 weeks of the trial. The topiramate groups lost 15.4% and 16.5% of their baseline weight and the placebo group lost 8.9%.150 Topiramate also is effective in producing weight loss in diabetic patients.151 Although topiramate still is available as an antiepileptic drug, the development program to obtain an indication for overweight was terminated by the sponsor because of the associated adverse events. Zonisamide.  Zonisamide is an antiepileptic drug that has serotonergic and dopaminergic activity in addition to inhibiting sodium and calcium channels. Weight loss was noted in the clinical trials for the treatment of epilepsy, again suggesting the drug as a potential agent for weight loss. Gadde and colleagues tested this possibility by performing a 16-week randomized controlled trial in 60 obese subjects.152 Subjects were placed on a calorie-restricted diet and randomized to zonisamide or placebo. Zonisamide was started at 100 mg/day and increased to 400 mg/day. At 12 weeks, the dosage in those subjects who had not lost 5% of initial body weight was increased to 600  mg/day. The zonisamide group lost 6.6% of initial body weight at 16 weeks compared with 1% in the placebo group. Thirtyseven subjects completing the 16-week trial elected to continue the trial for 32 weeks, 20 in the zonisamide group and 17 in the placebo group. At the end of 32 weeks, the 19 subjects in the zonisamide group lost 9.6% of their initial body weight compared with 1.6% for the 17 subjects in the placebo group.152 Metformin.  Metformin is a biguanide approved for the treatment of diabetes mellitus that reduces hepatic glucose production, decreases glucose absorption from the gastrointestinal tract, and enhances insulin sensitivity. In clinical trials in which metformin was compared with sulfonylureas, it produced weight loss.134 In one French trial, called BIGPRO, metformin was compared with placebo in a oneyear multicenter study of 324 middle-aged subjects with upper body adiposity and the metabolic syndrome (insulin resistance syndrome). The subjects on metformin lost significantly more weight (1 to 2 kg) than those in the placebo group, and the study concluded that metformin may have a role in the primary prevention of type 2 diabetes.153 In a meta-analysis of three of these studies, Avenell and colleagues143 reported a weighted mean weight loss at 12 months of 1.09 kg (95% CI, −2.29 to 0.11 kg). The best trial of metformin for obesity, however, is the Diabetes Prevention Program (DPP) study of individuals with impaired glucose tolerance. This study included a double-blind comparison of metformin 850 mg twice daily

Chapter 6  Obesity with placebo. During the 2.8 years of this trial, the 1073 patients treated with metformin lost 2.5% of their body weight (P < 0.001) compared with the 1082 patients treated with placebo, and the conversion from impaired glucose tolerance to diabetes was reduced by 31% compared with placebo. In the DPP trial, metformin was more effective in reducing the development of diabetes in persons in the subgroup who were most overweight and in the younger members of the cohort.96 Although metformin does not produce enough weight loss to qualify as a weight loss drug (FDA criteria require 5% weight loss), it would appear to be a very useful choice for overweight individuals who have diabetes or are at high risk for diabetes. One area in which metformin has found use is in treating overweight women with the polycystic ovary syndrome, in whom the modest weight loss may contribute to increased fertility and reduced insulin resistance.154 Pramlintide.  Amylin is a peptide found in the beta cells of the pancreas that is cosecreted along with insulin to circulate in the blood. Amylin and insulin are deficient in type 1 diabetics, in whom beta cells are destroyed immunologically. Pramlintide, a synthetic amylin analog, has a prolonged biological half-life.155 Pramlintide is approved by the FDA for the treatment of diabetes. Unlike insulin and many other diabetic medications, pramlintide use is associated with weight loss. In a study in which 651 subjects with type 1 diabetes were randomized to placebo or subcutaneous pramlintide 60 µg three or four times daily, along with an insulin injection, the hemoglobin A1c level decreased 0.29% to 0.34% and weight decreased −1.2 kg relative to placebo.156 Maggs and colleagues analyzed the data from two one-year studies in insulin-treated type 2 diabetic subjects randomized to pramlintide 120 µg twice daily or 150 µg three times daily.157 Weight decreased by 2.6  kg and hemoglobin A1c decreased 0.5%. When weight loss was then analyzed by ethnic group, African Americans lost 4  kg, whites lost 2.4 kg, and Hispanics lost 2.3 kg; the improvement in diabetes correlated with the weight loss, suggesting that pramlintide is effective in ethnic groups with the greatest burden from overweight. The most common adverse event was nausea, which usually was mild and confined to the first four weeks of therapy. Exenatide.  GLP-1 is derived from the processing of the preproglucagon peptide, which is secreted by L cells in the terminal ileum in response to a meal. Increased GLP-1 inhibits glucagon secretion, stimulates insulin secretion, stimulates gluconeogenesis, and delays gastric emptying.155 It has been postulated to be responsible for the superior weight loss and improvement in diabetes seen after gastric bypass surgery for overweight.158 GLP-1 is rapidly degraded by dipeptidyl peptidase-4 (DPP-4), an enzyme that is elevated in obese individuals. Bypass operations for overweight increase GLP-1, but do not change the levels of DPP-4.155 Exenatide (exendin-4) is a 39–amino acid peptide produced in the salivary gland of the Gila monster lizard. It has 53% homology with GLP-1, but has a much longer half-life. Exenatide is approved by the FDA for treatment of type 2 diabetics who are inadequately controlled while being treated with metformin or sulfonylureas. In humans, exenatide reduces fasting and postprandial glucose levels, slows gastric emptying, and decreases food intake by 19%.155 The side effects of exenatide in humans are headache, nausea, and vomiting that are lessened by gradual dose escalation. Several clinical trials of 30 weeks’ duration have been reported using exenatide at a daily

dosage of 5 µg twice daily subcutaneously.159-161 In one trial with 377 type 2 diabetic subjects who were failing maximal sulfonylurea therapy, exenatide produced a fall of 0.74% more in hemoglobin A1c than placebo. Fasting glucose levels also decreased and there was a progressive weight loss of 1.6 kg.161 The interesting feature of this weight loss is that it occurred without change in lifestyle, diet, or exercise. In a 26-week randomized controlled trial, exenatide produced a 2.3-kg weight loss compared with a gain of 1.8 kg in the group receiving insulin glargine.162 The FDA has issued a notice about potential risks of pancreatitis.

Drug Combinations

The first important clinical trial combining drugs that acted by separate mechanisms used phentermine and fenfluramine.163 This trial showed that by using combination therapy, a highly significant weight loss was achieved of almost 15% below baseline and with fewer side effects than those seen with the individual agents. This combination became very popular, but because of reports of aortic valvular regurgitation associated with its use, fenfluramine was withdrawn from the market worldwide on September 15, 1997.164 Several other combinations of existing drugs are now under development, including phentermine with topiramate (Qnexa), phentermine with zonisamide, and naltrexone with bupropion. Initial data have been published on all three combinations, but longer term studies are needed to evaluate the potential drug-drug interactions and side effects produced.

SURGERY

Surgical intervention for obesity has become ever more popular (see Chapter 7).165 The Swedish Obese Subjects Study has offered surgical intervention for obese Swedish patients aimed at reducing their obesity through a gastrointestinal operation. The control group included obese Swedish patients who did not get surgical treatment but were treated with the best alternatives in the Swedish health care system. The effect of weight change on dyslipidemia, blood pressure, and serum insulin levels in the surgically treated group two and 10 years after surgery was compared with these parameters in the control group.166 There was a graded effect of weight change on HDL cholesterol, triglycerides, systolic and diastolic blood pressure, insulin, and glucose. This surgery has now been shown to reduce mortality.94,95 A comparison of surgically and nonsurgically treated patients has shown that weight loss improves long-term health outcomes, but at a cost of significant short-term health problems.167

KEY REFERENCES

Avenell A, Brown TJ, McGee MA, et al. What interventions should we add to weight reducing diets in adults with obesity? A systematic review of randomized controlled trials of adding drug therapy, exercise, behaviour therapy or combinations of these interventions. J Hum Nutr Diet 2004; 7:293-316. (Ref 143.) Bray GA. The metabolic syndrome and obesity. Totowa, NJ: Humana Press; 2007. (Ref 4.) Bray G, Greenway F. Pharmacological treatment of the overweight patient. Pharmacol Rev 2007: 59:151-84. (Ref 134.) Christakis NA, Fowler JH. The spread of obesity in a large social network over 32 years. N Engl J Med 2007; 357:370-9. (Ref 15.) Dansinger ML, Tatsioni A, Wong JB, et al. Meta-analysis: The effect of dietary counseling for weight loss. Ann Intern Med 2007; 147:41-50. (Ref 112.) Farooqi IS, O’Rahilly S. Genetic factors in human obesity. Obes Rev 2007; 8(Suppl 1):37-40. (Ref 27.) Freedman DM, Ron E, Ballard-Barbash R, et al. Body mass index and all-cause mortality in a nationwide US cohort. Int J Obes (Lond) 2006; 30:822-9. (Ref 45.)

113

114

Section II  Nutrition in Gastroenterology Klein S, Burke LE, Bray GA, et al; American Heart Association Council on Nutrition, Physical Activity, and Metabolism. Clinical implications of obesity with specific focus on cardiovascular disease: A statement for professionals from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism: Endorsed by the American College of Cardiology Foundation. Circulation 2004; 110:2952-67. (Ref 46.) Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002; 346:393-403. (Ref 96.) National Institutes of Health. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults— The Evidence Report. National Institutes of Health. Obes Res 1998; 6(Suppl 2):51S-209S. (Ref 2.)

Poirier P, Giles TD, Bray GA, et al; American Heart Association; Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism. Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss: An update of the 1997 American Heart Association Scientific Statement on Obesity and Heart Disease from the Obesity Committee of the Council on Nutrition, Physical Activity, and Metabolism. Circulation 2006; 113:898-918. (Ref 50.) Rucker D, Padwal R, Li SK, et al. Long term pharmacotherapy for obesity and overweight: updated meta-analysis. BMJ 2007; 335:1194-9. (Ref 135.) Sjöström L, Narbro K, Sjöström CD, et al. Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med 2007; 357:741-52. (Ref 94.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

7

Bariatric Surgery Gavitt Woodard and John Morton

CHAPTER OUTLINE Epidemiology of Morbid Obesity  115 Bariatric Surgery as Treatment for Morbid Obesity  115 Preoperative Evaluation  115 Efficacy  116 Effects on Mortality and Morbidity  116 Comorbidity Resolution  117

EPIDEMIOLOGY OF MORBID OBESITY Morbid obesity is the leading public health crisis of the industrialized world (see Chapter 6).1,2 The prevalence of obesity in the United States continues to rise at an alarming rate, with two thirds of adults currently considered overweight, half of whom are obese.3 Being overweight is defined by the body mass index (BMI): normal BMI = 25 kg/m2; BMI for obesity > 30 kg/m2; BMI for morbid obesity > 40 kg/m2; and BMI for supermorbid obesity > 50 kg/m2. Rising rates of obesity are seen across the United States in men and women and in all major racial, ethnic, and socioeconomic groups.4 Morbid obesity reduces life expectancy by five to 20 years and, for the first time in history, it is predicted that the current generation may have a shorter life expectancy than the last.5,6

BARIATRIC SURGERY AS TREATMENT FOR MORBID OBESITY

Bariatric surgery remains the only effective and enduring treatment for morbid obesity. Since 1997, the number of bariatric surgical procedures in the United States has grown sevenfold as evidence has proven their safety and efficacy.7 Weight loss operations can be classified as malabsorptive or restrictive (Fig. 7-1). Roux-en-Y gastric bypass (RNYGB), which accounts for 88% of bariatric procedures in the United States, is restrictive and malabsorptive. Biliopancreatic diversion–duodenal switch (BPD-DS), the other malabsorptive procedure, is not as commonly performed in the United States because of its higher risk profile. Purely restrictive procedures include the laparoscopic adjustable gastric banding (LAGB), vertical banded gastroplasty (VBG), gastrectomy, and sleeve gastrectomy procedures, all of which reduce the size of the stomach so it is unable to accommodate more than a few ounces of food. VBG is no longer performed commonly because of its potential for staple line dehiscence and subsequent weight gain. Advantages, disadvantages and complications of the major weight loss operations are shown in Tables 7-1 and 7-2.

Surgical Complications  118 Volume Effect and Center of Excellence Movement  119

PREOPERATIVE EVALUATION To qualify for bariatric surgery, patients must meet the 1991 NIH consensus criteria, which include having a BMI > 40 kg/m2 or a BMI > 35 kg/m2 with obesity-related comorbidities, and at least six months of documented, medically supervised weight loss attempts.8 Some bariatric surgeons require patients to lose additional weight through diet and exercise between the time of initial bariatric surgery con­ sultation and the date of operation. This additional required preoperative weight loss is not correlated with comorbidity resolution or complication rates, but is associated with shorter operative times and greater weight loss at one year after the surgery; therefore, it should be encouraged in all patients.9,10 Prior to surgery, patients must complete an extensive screening process, including consultation with a surgeon, psychologic evaluation, nutrition consultation, chest roentgenography, electrocardiography, pulmonary function testing, a sleep study, and an esophagogastroduodenoscopy (EGD). The EGD was recommended by the European Association for Endoscopic Surgery to detect and treat any upper gastrointestinal lesions that may cause postoperative complications or influence the decision of which type of bariatric surgery should be performed.11 In a study of 272 gastric bypass patients who underwent preoperative EGD, 12% had clinically significant preoperative findings that included erosive esophagitis (3.7%), Barrett’s esophagus (3.7%), gastric ulcer (2.9%), erosive gastritis (1.8%), duodenal ulcer (0.7%), and gastric carcinoid (0.3%); 1.1% had more than one lesion. Given that 12% of patients who eventually underwent RNYGB had clinically significant preoperative findings, and only 67% of them had upper gastrointestinal symptoms, it is important to perform EGD preoperatively, because the excluded distal stomach cannot be evaluated easily after a RNYGB procedure.12 Patients undergoing LAGB surgery also should undergo a preoperative EGD, especially to evaluate gastroesophageal reflux disease (GERD). Gastric banding leads to satiety and weight loss by creating a small restrictive stomach with a slow

115

116

Section II  Nutrition in Gastroenterology gastric emptying time. Patients who overfill their small stomach pouch post-LAGB can force food and stomach acid back up into their lower esophagus, thereby worsening any preexisting GERD.13,14 In addition, overzealous banding adjustment may lead to pseudoachalasia with an increased pressure zone below the lower esophageal sphincter, furthering any incompetence.

EFFICACY

Roux-en-Y Gastric Bypass (RNYGB)

Vertical Banded Gastroplasty (VBG)

The steep rise in the use of bariatric surgery can be attributed to its proven efficacy as a treatment for morbid obesity. Two meta-analyses have provided strong validation that bariatric surgery leads to successful weight loss and mortality reduction.15,16 A meta-analysis by Buchwald and colleagues that included 22,094 patients found the mean percentage of excess weight loss (EWL) for all patients to be 61.2%.15 EWL is highest for VBG (68.2%), lower for RNYGB (61.6%), and lowest for LAGB (47.5%). A meta-analysis by Maggard and associates found similar weight loss trends at three or more years postoperatively, with the greatest weight loss achieved after the malabsorptive procedures of BPD-DS (53 kg) and RNYGB (42 kg), and less weight loss after the restrictive LAGB (35 kg) and gastroplasty (32 kg).16

EFFECTS ON MORTALITY AND MORBIDITY

Laparoscopic Adjustable Gastric Band (LAGB)

Biliopancreatic Diversion (BPD) with Duodenal Switch

Figure 7-1.  Types of weight loss operations. (From the American Society for Bariatric Surgery. The story of surgery for obesity, 2005. Available at www.asbs.org.)

Such substantial weight loss is associated with a clear reduction in long-term mortality. A retrospective cohort study of 9,949 RNYGB patients matched to 9,628 severely obese controls found that having RNYGB surgery reduces the adjusted long-term mortality from any cause of death by 40%.17 Among RNYGB patients, mortality was decreased 56% from coronary artery disease, 92% from diabetes, and 60% from cancer. In another study, a 14% decrease in cancer incidence was shown in patients who underwent RNYGB, the biggest reductions in which were seen among types of cancers that are considered obesity-related: eso­ phageal adenocarcinoma (2% reduction), colorectal cancer (30% reduction), breast cancer (9%), uterine cancer (78%), non-Hodgkin’s lymphoma (46%), and multiple myeloma (54%).18 The lower cancer risk of patients after RNYGB

Table 7-1  Advantages and Disadvantages of Bariatric Surgery Procedures for Weight Loss

Procedure RNYGB

Weight Loss Mechanism

Mean Weight Loss (kg) Advantages

Disadvantages

Reversible?

Good weight loss

Vitamin deficiencies; internal hernias Less weight loss; may require adjustments; foreign body

LAGB

Restriction and malabsorption Restriction

VGB

Restriction

Relatively low risk of vitamin deficiencies; low rate of complications Ease of construction

BPD

Malabsorption

Greatest weight loss

No longer performed Staple dehiscence High risk of malabsorption and vitamin deficiencies

After 12 mo

After 36 mo

No

43.5

41.5

Band may be removed

30.2

34.8

Yes

32.1

32.0

No

51.9

53.1

BPD, biliopancreatic diversion; LAGB, laparoscopic adjustable gastric banding; NR, not reported; RNYGB, Roux-en-Y gastric bypass; VGB, vertical banded gastroplasty. Data from Maggard MA, Shugarman LR, Suttorp M, et al. Meta-analysis: Surgical treatment of obesity. Ann Intern Med 2005; 142:547-59.

Chapter 7  Bariatric Surgery Table 7-2  Complication Rates (%) for Bariatric Surgery Procedures Procedure RNYGB LAGB VGB BPD

Mortality Rate 1.0 0.4 0.2 NR, but at least ≥RNYGB

GI Symptoms

Reflux

Vomiting

16.9 7.0 17.5 37.7

10.9 4.7 2.2 NR

15.7 2.5 18.4 5.9

Nutritional Abnormalities 16.9 NR 2.5 NR, but high potential

Reoperation

Leak Rate

1.6 7.7 11.3 4.2

2.2 NR 1.0 1.8

Bleeding 2.0 0.3 0.7 0.2

BPD, biliopancreatic diversion; GI, gastrointestinal; LAGB, laparoscopic adjustable gastric banding; NR, not reported; RNYGB, Roux-en-Y gastric bypass; VGB, vertical banded gastroplasty. Data from Maggard MA, Shugarman LR, Suttorp M, et al. Meta-analysis: Surgical treatment of obesity. Ann Intern Med 2005; 142:547-59.

presumably was caused by weight loss, which has been shown in many studies to reduce cancer incidence. Furthermore, once obese patients lose weight, they may have better access to needed health surveillance, such as Pap smears and colonoscopy. Finally, given that increased BMI leads to worse surgical oncologic outcomes, it may be surmised that with weight loss, a better surgical outcome may be anticipated. Overall, bariatric surgery dramatically improves survival and decreases mortality from all disease-related causes of death. Only rate of deaths not caused by disease, including deaths resulting from accidents and suicide, increased after bariatric surgery and were 58% higher in the RNYGB patients17; it has been speculated that alcohol abuse may explain why accidents and suicides were higher in the surgical group. One study demonstrated altered alcohol metabolism after gastric bypass surgery, with the gastric bypass patients having a greater peak alcohol level and a longer time for the alcohol level to reach a zero blood level than in controls.30 Another study of bariatric surgery candidates found that 9% reported having attempted suicide and 19% reported having abused alcohol preoperatively.19 There is concern that this vulnerable patient population has additional difficulty with the psychological adjustments to weight loss, further supporting the need for psychological counseling before and after surgery.20,21 In addition to benefiting from a decreased mortality, bariatric patients benefit from decreased morbidity. Morbidly obese patients suffer from more intense gastrointestinal symptoms (e.g., abdominal pain, heartburn) and sleep disturbances than normal-weight patients. By six months after RNYGB, however, the frequency and severity of gastrointestinal symptoms of many morbidly obese patients have decreased to levels seen in normal weight patients. Dysphagia is common in morbidly obese patients, all of whom experience increased intra-abdominal pressure. Dysphagia is the only gastrointestinal symptom that worsens after RNYGB, probably from further increase in esophageal pressure because of overeating and overfilling of the restrictive small gastric pouch; this observation again underscores the need for preoperative and postoperative education regarding diet.27 Quality of life, as measured by the validated SF36 survey, improves greatly after RNYGB surgery. Preoperatively, morbidly obese patients score significantly lower than U.S. population norms in the categories of general health, vitality, physical functioning, bodily pain, emotional, and social functioning. As soon as three months following RYNGB, these same patients scored no differently than U.S. norms in these categories.28

COMORBIDITY RESOLUTION

Weight loss surgery is a singular medical intervention that can reverse or improve the numerous medical conditions associated with obesity. RNYGB results in a substantial reduction in cardiac risk factors with the following resolution rates: diabetes (82%), hypertension (70%), and hyperlipidemia (63%).29 Gastric bypass has assembled the most evidence of comorbidity resolution; however, all weight loss operations result in some degree of improvement. The meta-analysis by Buchwald and coworkers15 found that bariatric surgery reverses, ameliorates, or eliminates major cardiovascular risk factors: Hypertension was resolved by banding (38%), gastroplasty (73%), gastric bypass (75%), and BPD-DS (81%). Diabetes was resolved by banding (48%), gastroplasty (68%), gastric bypass (84%), and BPD-DS (98%). Hyperlipidemia was improved by banding (71%), gastroplasty (81%), gastric bypass (94%), and BPD-DS (99%). One study found resolution of all conventional abnormal risk factors, including serum levels of total cholesterol, lowdensity lipoprotein, high-density lipoprotein (HDL), triglyceride, high-sensitivity C-reactive protein, homocysteine, and lipoprotein A at one year after RNYGB.30 The Swedish Obese Subjects (SOS) Study has provided further demonstration of the ameliorative effect of bariatric surgery. At 10 years of follow-up, surgically treated obese patients had 25% reduction in hypertension, 43% improvement in HDL, and 75% reduction in diabetes compared with the medically treated group.31 Beyond the significant improvement in cardiac risk factors, weight loss surgery also provides substantial relief of the many medical problems that obesity causes. One study found that a leading digestive health complaint, GERD, is cured or improved at a 96% rate.29 Other studies also have documented a highly significant reduction in GERD symptoms after bariatric surgery.27,32 One study of patients with severe GERD prior to RNYGB found signi­ ficant declines in use of proton pump inhibitors (44% to 9%) and H2 blockers (60% to 10%) at postoperative times ranging from six to 36 months.33 GERD resolution rates following RNYGB are so robust that RNYGB is a suggested treatment for recalcitrant GERD in morbidly obese patients, especially given the lack of efficacy of antireflux surgery in the obese.34,35 Prior to surgery, morbidly obese patients report significantly more symptoms of abdominal distress, including pain, gnawing sensations, nausea, vomiting, and abdominal distention than normal controls. After RNYGB, these symptoms are reduced to levels comparable with those of normal controls.27,32 Irritable bowel syndrome (IBS) is a constella-

117

118

Section II  Nutrition in Gastroenterology tion of symptoms (see Chapter 118), including abdominal pain or discomfort, altered bowel habits (diarrhea, consti­ pation), increased flatus, and bloating or distention. Preoperative morbidly obese patients rate their severity of IBS symptoms significantly higher than controls. Following RNYGB, these same patients rated the severity of their symptoms significantly lower than they did preoperatively and at levels equivalent to those of control IBS patients. Nonalcoholic fatty liver disease (NAFLD) comprises a histologic spectrum of fatty liver ranging from simple steatosis to portal fibrosis, nonalcoholic steatohepatitis (NASH), and cirrhosis.36 The most advanced forms of NAFLD are strongly associated with the metabolic syndrome—that is, obesity and hypertension, hypertriglyceridemia, and diabetes mellitus—and NAFLD is the most prevalent liver disease in the United States.37,38 In a postmortem series of obese nondrinkers, hepatic steatosis was present in 76% and NASH was present in 18.5%.39 Mounting evidence suggests that current bariatric surgical procedures actually may be beneficial for NAFLD.40 One study documented that histologic improvements as measured by steatosis (89.7% improvement), hepatocellular ballooning (58.9% improvement), and centrilobular-perisinusoidal fibrosis (50% improvement) occur within a mean period of 18 months after RNYGB. The diagnosis of NASH was made in 58.9% of patients preoperatively and none of them were found to have NASH in postoperative liver biopsies.41 Weight loss surgery has been shown to improve sleep apnea with the following rates: banding (95%), gastroplasty (77%), gastric bypass (87%), and BPD-DS (95%).15 Another study has demonstrated that after RNYGB, patients have significantly fewer sleep disturbances with regard to falling asleep, insomnia, and feeling rested on awakening.27 Morbidly obese patients undergoing RNYGB have demonstrated an 88% resolution or improvement of joint disease.29 Furthermore, weight loss surgery has been demonstrated to eliminate or improve obesity-associated venous stasis disease, gout, asthma, pseudotumor cerebri, urinary incontinence, and infertility.29

SURGICAL COMPLICATIONS The risk of operative mortality and complications may temper some enthusiasm for bariatric surgery. Based on a meta-analysis by Maggard and colleagues,16 mortality rates were shown to depend on the procedure performed. The average mortality rates for the different procedures are banding (0.4%, 0.01% to 2.1%), gastroplasty (0.2%, 0% to 16.8%), gastric bypass (1%, 0.2% to 2.5%), and BPD-DS (0.9%, 0.01% to 1.3%). Complications of bariatric procedures include anastomotic leak or stenosis, pulmonary embolus, gastrointestinal bleeding, nutritional deficiencies, wound complications, bowel obstructions, ulcers, hernias, respiratory, cardiac, and implant device-related complications. Among the different surgical procedures, the rate of complications is proportional to the amount of weight loss produced by each operation: banding (7%), gastroplasty (18%), gastric bypass (17%), and BPD-DS (38%).16,42 Marginal ulcers are estimated to occur in 1% to 16% of gastric bypass patients.43,44 Perforated marginal ulcers occur in 1% of RNYGB patients. Ulcer perforation is linked to smoking and use of nonsteroidal anti-inflammatory drugs (NSAIDs) or glucocorticoids.45 The use of nonabsorbable sutures, as opposed to absorbable sutures, for the inner layer of the gastrojejunal anastomosis is associated with increased

ulcer incidence.46 The presence of H. pylori also increases risk for marginal ulcers.47 Postoperatively, it is common practice for bariatric surgeons to begin a six-month ulcer prophylaxis program with proton pump inhibitors. If a marginal ulcer is recalcitrant to medical therapy, the possibility of a gastric-gastric fistula must be entertained. If a gastricgastric fistula is present with a marginal ulcer, then surgical correction is mandated. Gastrointestinal bleeding occurs postoperatively in 2.0% of RNYGB, 0.7% of VBG, 0.3% of LAGB, and 0.2% of BPD-DS procedures. Another potential complication is an anastomotic gastric pouch or duodenal leak, which occurs after 2.2% of RNYGB, 1.0% of VBG, and 1.8% of BPD-DS procedures.16 Nutritional and vitamin deficiencies and electrolyte abnormalities occur in 16.9% of RNYGB patients and 2.5% of patients having VBG.16 Patients who do not take daily vitamins postoperatively or patients who experience frequent vomiting are at increased risk of developing such deficiencies, most common of which are protein, iron, vitamin B12, folate, calcium, and the fat-soluble vitamins A, D, E, and K.48 The parietal cells of the stomach produce intrinsic factor, which is necessary for vitamin B12 absorption in the terminal ileum. Patients who undergo RNYGB may develop vitamin B12 deficiency because RNYGB separates the parietal cells in the fundus and body of the stomach from the smaller gastric pouch, which receives ingested food. There is, therefore, no contact between ingested food and intrinsic factor until the intersection of the Roux limb in the jejunum.49,50 In addition, following RNYGB, the parietal cells of the stomach often cease to produce intrinsic factor, presumably because the fundus no longer has any contact with food.51 It has been shown that restrictive bariatric surgery does not cause vitamin B12 deficiency because the parietal cells remain in contact with the ingested food.52 Fat-soluble vitamin deficiencies are most often seen following BPD-DS operations because food has very little exposure to the biliary and pancreatic secretions necessary for fat digestion, and there is little exposure of food to the ileum, where fat is normally absorbed. Calcium and folate deficiency can occur because they are absorbed in the duodenum and proximal jejunum. These segments of the digestive tract are commonly bypassed in gastric bypass and duodenal switch surgery. The fat-soluble vitamin D is necessary for calcium absorption, and so vitamin D deficiency will further contribute to any calcium deficiency.48 Thiamine deficiency may lead to Wernicke’s encephalopathy, a syndrome of confusion, ataxia, ophthalmoplegia, and impaired short-term memory. If thiamine deficiency is suspected, the patient should be given intravenous or intramuscular thiamine immediately in order to increase the chances of symptom resolution.53 Meta-analysis has revealed that gastroesophageal reflux occurs postoperatively in 10.9% of RNYGB, 2.2% of VBG, and 4.7% of LAGB patients.16 Approximately half of LAGB patients will experience some degree of heartburn and acid regurgitation.54 Pseudoachalasia and esophageal dysmotility are late complications of LAGB and usually reverse on removal of the gastric band.55 Among RNYGB patients, postoperative dysphagia is significantly worse than normal weight controls, but not significantly worse than the patient’s matched preoperative symptoms.27 A high incidence of gallstone formation has been well documented when morbidly obese patients undergo rapid surgically induced weight loss.56 It was shown in a doubleblind, randomized, placebo-controlled trial that a daily dosage of 600 mg ursodiol for the first six months after surgery reduces the incidence of gallstones to 2%.57 It is

Chapter 7  Bariatric Surgery therefore recommended that all bariatric patients take ursodiol for six months postoperatively to reduce the risk of this largely preventable complication. Beyond the type of procedure, there are identified risk factors for complications after bariatric surgery, including older age, male gender, greater BMI, comorbidities, and Medicare insurance status.58-62 The increased risk for Medicare patients is beyond age, given that eligibility for Medicare is disability, which may affect outcomes. Although patients with the most risk factors carry the highest risk for surgery, they also may derive the most benefit from bariatric surgery, given the disease burden they carry.63 Of note, complications may not affect long-term weight loss, which is the outcome that best predicts long-term mortality risk.9

VOLUME EFFECT AND CENTER OF EXCELLENCE MOVEMENT Many of the reported studies regarding morbidity and mortality had been completed prior to the many improvements in current surgical technique.28 In addition, surgeon and hospital experience can mitigate the risks associated with weight loss surgery. The best demonstrated and most protective effect against complications is an experienced surgeon and hospital.63-66 Clearly, there is benefit in having this complex and demanding surgery performed by experienced and committed surgeons operating in a dedicated health care facility. None of the previously mentioned perioperative risk factors can be modified, with the exception of the volume status of the surgeon and hospital. For gastric bypass surgery, it has been demonstrated that a high-volume surgeon and high-volume hospital lead to decreased morbidity and mortality.63-66 In the United States, this volume outcome effect has been recognized by the Centers for Medicare and Medicaid Services, which now require Medicare patients to undergo surgery only at a Bariatric Surgery Center of Excellence.67 Numerous criteria enable a center to become a Bariatric Surgery Center of Excellence, but the primary criteria are a surgeon volume of more than 50 cases and hospital volume exceeding 125 cases annually. Although a referral

to a Bariatric Surgery Center of Excellence may lead to decreased morbidity and mortality, this referral pattern must be balanced with appropriate and sufficient access to care for a vulnerable population without other therapeutic options.

KEY REFERENCES

Adams TD, Gress RE, Smith SC, et al. Long-term mortality after gastric bypass surgery. N Engl J Med 2007; 357:753-61. (Ref 17.) Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a syste­ matic review and meta-analysis. JAMA 2004; 292:1724-37. (Ref 15.) Fontaine KR, Redden DT, Wang C, et al. Years of life lost due to obesity. JAMA 2003; 289:187-93. (Ref 5.) Hagedorn JC, Encarnacion B, Brat GA, Morton JM. Does gastric bypass alter alcohol metabolism? Surg Obes Relat Dis 2007; 3:543-8. (Ref 30.) Klein S, Mittendorfer B, Eagon JC, et al. Gastric bypass surgery improves metabolic and hepatic abnormalities associated with nonalcoholic fatty liver disease. Gastroenterology 2006; 130:1564-72. (Ref 40.) Maggard MA, Shugarman LR, Suttorp M, et al. Meta-analysis: surgical treatment of obesity. Ann Intern Med 2005; 142:547-59. (Ref 16.) Nguyen NT, Goldman C, Rosenquist CJ, et al. Laparoscopic versus open gastric bypass: A randomized study of outcomes, quality of life, and costs. Ann Surg 2001; 234:279-89; discussion 289-91. (Ref 28.) Nguyen NT, Paya M, Stevens CM, et al. The relationship between hospital volume and outcome in bariatric surgery at academic medical centers. Ann Surg 2004; 240:586-93. (Ref 64.) Ogden CL, Carroll MD, Curtin LR, et al. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA 2006; 295:1549-55. (Ref 1.) Olshansky SJ, Passaro DJ, Hershow RC, et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med 2005; 352:1138-45. (Ref 6.) Santry HP, Gillen DL, Lauderdale DS. Trends in bariatric surgical procedures. JAMA 2005; 294:1909-17. (Ref 7.) Schauer PR, Ikramuddin S, Gourash W, et al. Outcomes after laparoscopic Roux-en-Y gastric bypass for morbid obesity. Ann Surg 2000; 232:515-29. (Ref 29.) Sjöström L, Lindroos AK, Peltonen M, et al. Lifestyle, diabetes, and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 2004; 351:2683-93. (Ref 31.) Sugerman HJ, Brewer WH, Shiffman ML, et al. A multicenter, placebocontrolled, randomized, double-blind, prospective trial of prophy­ lactic ursodiol for the prevention of gallstone formation following gastric-bypass-induced rapid weight loss. Am J Surg 1995; 169:91-6. (Ref 57.) Williams DB, Hagedorn JC, Lawson EH, et al. Gastric bypass reduces biochemical cardiac risk factors. Surg Obes Relat Dis 2007; 3:8-13. (Ref 58.) Full references for this chapter can be found on www.expertconsult.com.

119

CHAPTE R

8

Eating Disorders Anne E. Becker and Christina Wood Baker

CHAPTER OUTLINE Epidemiology  121 Causative Factors  121 Satiety  122 Appetite  123 Energy Storage  123 Onset and Course  123 Diagnosis and Evaluation  123 Anorexia Nervosa  124 Bulimia Nervosa  124 Eating Disorder Not Otherwise Specified  125 Binge Eating Disorder  125 Night Eating Syndrome and Nocturnal Sleep-Related Eating Disorder  125

Eating disorders are mental illnesses characterized by disturbances in weight control, body image, and/or dietary patterns. Diagnostic categories include (1) anorexia nervosa (AN); (2) bulimia nervosa (BN); and (3) eating disorder not otherwise specified (EDNOS; Fig. 8-1; Table 8-1). Several variants of EDNOS, such as binge-eating disorder and night eating syndrome, are well-described in the literature. The focus of this chapter is eating disorders in adults; other disturbances in eating that typically have onset in infancy and early childhood, such as pica, rumination syndrome, and feeding disorder of early childhood, are not discussed here. Although eating disorders are classified as mental illnesses, their associated behaviors commonly result in and present with medical sequelae, many of which are gastrointestinal. Because associated chronic undernutrition, overweight, and/or purging behaviors often result in serious medical complications that can be chronic, individuals with eating disorders benefit from the ongoing care of a multidisciplinary treatment team. Indeed, eating disorders (AN and BN) are among the mental disorders with the highest mortality risk.1

EPIDEMIOLOGY Eating disorders have been described across diverse global settings, although epidemiologic data are best established for populations in North America and Europe. The incidence rate for AN is approximately eight cases/100,000 population/year, with a point prevalence of AN estimated at 0.3% in young women of the United States and Western European general populations. BN is more common than AN, with an incidence of 12 cases/100,000 population/year in the United States and Western Europe2 and a 12-month prevalence of 0.5% among adult women in the United States. Lifetime prevalence estimates for U.S. women based on the National Comorbidity Study Replication (NCS-R) are

Purging Disorder  125 Differential Diagnosis  125 Nutritional and Medical Evaluation  126 Nutritional Evaluation  126 Gastrointestinal Abnormalities Associated with Eating Disorders  128 Management  131 Psychiatric Treatment  131 Weight Management  134 Medical Management of Gastrointestinal Symptoms of Patients with Eating Disorders  135

0.9% for AN and 1.5% for BN; U.S. men have a lifetime prevalence of 0.3% for AN and 0.5% for BN.3 The most common presentation of an eating disorder in outpatient settings is EDNOS, although fewer prevalence data are available for this broad and heterogeneous category. One large study from Portugal reported a prevalence of 2.37% for EDNOS in female students in grades 9 to 12.4 Relatively high prevalence rates also are reported for specific symptoms associated with disordered eating. For example, in 2007, 6.4% of school-going female adolescents in the United States reported vomiting or laxative use, and 16.3% reported fasting within the previous month to lose weight.5 Within the diagnostic category of EDNOS, there is great interest in a clinical variant termed binge-eating disorder (BED). The lifetime prevalence of BED in the United States is 3.5% for adult women and 2.0% for adult men. Additional variants of disordered eating (that are not classified as distinct disorders in the Diagnostic and Statistical Manual of Mental Disorders, DSM) include the night eating syndrome (NES) and nocturnal sleep-related eating disorder (NSRED). The prevalence of NES in young adult women has been reported as 1.6%6 and of NSRED as 0.5%.7 Eating disorders occur across ethnically and socioeconomically diverse populations, but each of the eating disorders is more common in women than in men. Males account for less than 10% of individuals with AN, 10% of those with BN, 34% of those with NES, and 40% of those with BED.8,9

CAUSATIVE FACTORS Although incompletely understood, the cause of eating disorders is almost certainly multifactorial, with psychodevelopmental,10 sociocultural,11 and genetic12 contributions to risk. For example, exposure to risk factors for dieting appears to elevate risk for BN,13 just as childhood exposure

121

122

Section II  Nutrition in Gastroenterology Eating disorders

Eating disorder, not otherwise specified

Anorexia nervosa

Bulimia nervosa

Unwilling to maintain minimal healthy weight Excessive fear of weight gain Distorted perception of weight or body and/or related medical danger Amenorrhea

Recurrent binge-eating Recurrent behavior to purge or neutralize excessive intake or control weight Excessive concern with weight or body shape

Binge/purge subtype

Restricting subtype

Recurrent binge-eating and/or purging

Restrictive pattern of eating without binge-eating or purging

Purging subtype

Non-purging subtype

Recurrent self-induced vomiting, and/or abuse of laxatives, enemas, diuretics, or stimulants

Behaviors to neutralize excessive intake limited to food restriction and exercise

Atypical or subthreshold symptoms Examples: Normal weight, recurrent purging, but no bingeeating Restrictive eating, with normal weight Binge-eating and purging once weekly Chewing and spitting out food

Binge-eating disorder Recurrent binge-eating without recurrent purging or other behaviors to neutralize the excess energy intake Binge-eating that is associated with distress and often with secrecy and shame Binge-eating that is often rapid and irrespective of hunger or satiety

Night-eating syndrome Evening hyperphagia Morning anorexia Nighttime wakings, snacking during nighttime wakings

Figure 8-1.  Key diagnostic features of the eating disorders.

Table 8-1  Behaviors Used to Neutralize Excessive Food Intake or to Prevent Weight Gain Purging behaviors Self-induced vomiting (including syrup of ipecac use) Laxative and/or enema abuse Diuretic abuse Stimulant abuse (e.g., caffeine, ephedra, methylphenidate, cocaine) Non-purging behaviors Excessive physical activity Fasting, skipping meals, restrictive eating pattern Inappropriate withholding or underdosing of insulin (among individuals with diabetes mellitus)

to negative comments about weight and shape elevate risk for BED.14 Body dissatisfaction in a social context in which thinness,15,16 self-efficacy, and control are valued may be an important means whereby dieting is initiated and disordered eating attitudes and behaviors ensue. Dietary restraint may precipitate a cycle of hunger, binge eating, and purging.17 Among numerous risk correlates, childhood gastrointestinal (GI) complaints have been found associated with earlier age of onset and greater severity of BN in a retrospective study,18 and picky eating and digestive prob­ lems were found prospectively associated with AN in adolescence.10 It has been suggested that physiologic vulnerabilities potentially may increase risk for an eating disorder. Neurobiologic targets have been identified as possibly playing a role in the pathogenesis of AN, BN, and BED. Researchers have studied the psychobiology of eating disorders and the

neurophysiologic correlates and determinants of energy intake, hunger, and satiety for decades. Findings highlight the multifactorial and phenotypically diverse nature of eating behavior. For example, energy intake is influenced by complex interactions among signaling molecules from peripheral systems (e.g., GI peptides, vagal stimulation) and central nervous system (CNS) neuropeptides and neuroamines. As is true in the search for treatments for obesity, it is unlikely that single mechanisms will become the basis for therapeutic interventions for eating disorders; however, the greater our understanding of the physiology of ingestive behavior, the more likely we are to establish integrated therapy models in the future. There is a vast literature on this topic, and what follows is simply a brief overview of the more commonly investigated mechanisms.

SATIETY

Serotonin has long been a focus of attention for its possible role in disrupted satiety. There is substantial evidence that altered 5-hydroxytryptamine (5-HT; serotonin) functioning contributes to dysregulated appetite, mood, and impulse control in eating disorders and persists after recovery from AN and BN, possibly reflecting premorbid vulnerability.19,20 There also is evidence that cholecystokinin (CCK) levels are altered in eating disorder populations. Findings for AN are inconsistent: There is some evidence that young women with AN have high levels of pre- and postprandial CCK, which may impede treatment progress by contributing to postmeal nausea and vomiting,21,22 whereas other reports have shown decreased CCK compared with that of controls.23 In those with BN, there is consistent evidence for an impaired satiety response, characterized by a blunted postprandial CCK response and satiety, as well as delayed

Chapter 8  Eating Disorders gastric emptying.24,25 In contrast, individuals with BED and obesity do not differ in postmeal CCK responses from those with obesity but no BED.26 The relationships between CCK, binge eating, and BMI need further clarification. Lastly, protein tyrosine (PYY) functioning also appears to be dysregulated in BN and AN, but not in BED. Young women with AN have higher levels of PYY, the intestinally derived anorexigen that elicits satiety, compared with controls, perhaps con­tributing to reduced food intake.27 In individuals with BN, expected elevations in PYY after meals are blunted,28,29 possibly playing a role in impaired satiety. A recent report found no differences between BED and non-BED groups in fasting levels and postmeal changes in PYY.30

APPETITE

The orexigenic peptide ghrelin is of interest for its role in eating disorders because it influences secretion of growth hormone (GH), stimulates appetite and intake, induces adiposity, and is implicated in signaling to the hypothalamic nuclei involved in energy homeostasis. There are two consistent findings in the literature examining ghrelin and AN: (1) circulating ghrelin levels are elevated, likely a consequence of prolonged starvation31,32; and (2) GH and appetite responses to ghrelin are blunted, suggesting altered ghrelin sensitivity.33,34 In BN, plasma levels of ghrelin are normal or elevated28,29; of most interest is the postprandial blunted response (i.e., reduced suppression of ghrelin35). Investigations of ghrelin functioning in individuals with BED have reported lower circulating levels of pre- and postmeal ghrelin, possibly reflecting down-regulation in response to chronic overeating and smaller decreases in ghrelin after eating.30

ENERGY STORAGE

Leptin and adiponectin are hormonal signals associated with longer term regulation of body fat stores. Leptin is also directly implicated in satiety through its binding to the ventral medial nucleus of the hypothalamus, an area termed the satiety center. Leptin and adiponectin are both altered in patients with eating disorders. A number of studies have found evidence for hyperadiponectinemia and hypoleptinemia in populations of underweight AN with reversal following restoration of weight36,37; increased adiponectin levels may act protectively to support energy homeostasis during food deprivation. Individuals with BN also exhibit decreased plasma levels of leptin, which are inversely correlated with length of illness and severity of symptoms.38 The mechanism of altered leptin functioning in BN is unclear because blunted postmeal leptin levels are not observed in individuals with BED.26 There are other mechanisms of interest, including but not limited to neuropeptide Y, GLP-1 and GLP-2, orexins A and B, the endocannabinoids, resistin (adipose tissue–specific secretory factor), pancreatic polypeptide, and brain-derived neurotrophic factor, but more research is necessary to clarify their roles in the pathophysiology of eating disorders. One high priority for research is clarifying whether observed psychobiological abnormalities are antecedents or consequences of disturbed eating behavior that return to normal after recovery, because this information would shed light on cause and possible treatment targets. Patients often report intense discomfort after eating as a reason that they continue to restrict intake. The discomfort may be dismissed as perceptual or psychological in the absence of any positive medical findings to support the symptoms, but there may in fact be disruptions in CNS or peripheral signals contributing to reported symptoms.

ONSET AND COURSE AN and BN most commonly have their onset in adolescence,39 and BED usually manifests in the early 20s,40 but eating disorders can occur throughout most of the lifespan and appear to be increasing in frequency in middle-aged and older women.41,42 Diagnostic migration from one eating disorder category to another is common.43 Lifetime comorbidity of AN, BN, and BED with other psychiatric disorders is high at 56.2%, 94.5%, and 63.6%, respectively.3 Mortality associated with AN and BN combined is five times higher than expected and is one of the highest mortality rates among mental disorders.1 Some data support the chronicity of AN, reporting that slightly under half of survivors with AN make a full recovery, with 60% attaining a normal weight and 47% regaining normal eating behavior; 34% improve but only achieve partial recovery, whereas 21% follow a chronic course.44 Other data suggest that recovery rates for AN may be more favorable than previously believed,3 with one large twin cohort study reporting a five-year clinical recovery rate of 66.8%.45 In contrast, after a five-year follow-up of 216 patients with BN and EDNOS, 74% and 78% of patients, respectively, were still in recovery.46 In a six-year longitudinal study of patients with BED, 43% of individuals continued to be symptomatic.47 In summary, despite well-established treatments available for the eating disorders, up to 50% of treated individuals continue to be symptomatic.48 Some prevention strategies developed for eating disorders show promise, including programs that induce cognitive dissonance about the “thin ideal.”49,50

DIAGNOSIS AND EVALUATION A substantial percentage of individuals with eating disorders in the United States do not receive treatment for this problem.3 Despite clear diagnostic criteria for the eating disorders, clinical detection often is problematic and up to 50% of cases may go unrecognized in clinical settings. Moreover, individuals with eating disorders are often reluctant to disclose their symptoms,51 and those with BN and BED can have a normal physical examination. Although individuals with AN are underweight by definition, this is easily missed in clinical settings. Even when noted on evaluation, the medical seriousness of low weight is frequently unappreciated.52 Finally, when an eating disorder is suspected or confirmed, patients may decline or avoid mental health care. Indeed, a feature of AN can be denial of the medical seriousness of symptoms.8 Given that many individuals with eating disorders initially present in primary care or medical subspecialty settings, recognition of clinical signs and symptoms across diverse health care settings will facilitate appropriate referrals and make diagnostic evaluation and treatment plans more efficient. One study has reported that individuals with BN are more likely to seek help for their GI complaints prior to seeking treatment for their eating disorder.53 Thus, familiarity with the diagnostic features and gastrointestinal complications of eating dis­orders will help to identify the most appropriate interventions, including the full spectrum of treatment resources available, for a comprehensive treatment plan. Formal screening for eating disorders can be timeconsuming, and although shorter measures are being developed,54 these have many limitations in clinical settings.55

123

124

Section II  Nutrition in Gastroenterology When an eating disorder is suspected, however, a directed clinical interview about restrictive or binge eating and inappropriate compensatory measures to control weight (see Table 8-1) is essential in determining the scope and severity of symptoms that underlie specific GI complaints and pose medical risk. Accurate and timely diagnosis of eating disorders is challenging for several reasons. First, patients may be unreliable reporters of their history and BN and BED may be present without any abnormal physical findings. Next, some dieting and exercise behaviors are appropriate, and discerning pathologic behavior that is consistent with a clinically significant eating disorder can be difficult. There is con­ siderable overlap in symptoms among the eating disorders; diagnostic specificity, however, is critical to effective management. Given the frequent reluctance of patients to disclose symptoms of an eating disorder, targeted history taking may be essential to making a prompt diagnosis. In some cases, an eating disorder may not be suspected nor the diagnosis confirmed until physical findings suggestive of purging are detected, a suggestive pattern is noticed in weight changes, and/or there is difficulty gaining weight, notwithstanding appropriate nutritional treatment and the exclusion of other potential causes for low weight.

ANOREXIA NERVOSA

Anorexia nervosa is characterized by an unwillingness or incapacity to maintain a minimally normal weight (commonly described as at least 85% of expected weight or a body mass index [BMI] of 17.5 kg/m2), fear of gaining weight (despite being thin), a disturbance in the way weight is experienced (e.g., a denial of the medical seriousness of being underweight or feeling fat despite emaciation), and amenorrhea (in postmenarcheal females). Individuals with AN typically restrict their food selections and caloric intake, but approximately half of those with AN also routinely binge-eat and/or engage in inappropriate compensatory behaviors, such as induced vomiting or laxative use to prevent weight gain (see Fig. 8-1). AN is subdivided further into restricting type (i.e., those who primarily control their weight through dieting, fasting, or exercising) and bingeeating–purging type (i.e., those who routinely purge calories to control weight and may or may not routinely binge-eat).8 In middle-aged and older women, new-onset AN may present in conjunction with difficulty making life transitions and fear of aging.42 The diagnosis of AN may be delayed when patients present to a GI specialty practice without disclosing their concerns and behaviors relating to weight. Presentation with GI complaints, even if related to real symptoms or disease, can sometimes prove to be a red herring, drawing attention away from and delaying diagnosis of an eating disorder. One study evaluated 20 consecutive patients presenting to a GI practice who were ultimately diagnosed with an eating disorder, and found that patients did not receive a diagnosis of an eating dis­ order for an average of 13 months after presentation. Notably, all patients stated a desire to gain weight and denied attempts to lose weight via exercise, purging, or dietary restriction.56 Individuals with AN are not always able or willing to frame their difficulty maintaining a healthy weight as intentional; thus, the diagnosis initially may be unsuspected and delayed.

BULIMIA NERVOSA

The clinical hallmark of BN is recurrent binge eating accompanied by inappropriate compensatory behaviors to control weight or to purge calories consumed during a binge. On

average, these behaviors must occur twice weekly for at least three months to meet diagnostic criteria.8 Also intrinsic to the diagnosis of BN is the excessive influence of weight and/or shape on self-image. By definition, binge eating is consumption of an unusually large amount of food during a “discrete period of time” (i.e., not overeating or grazing all day), accompanied by the feeling that the eating cannot be controlled.8 Many patients describe an emotional numbing during the period of eating. For some, this state appears to motivate the bingeing. Most clinicians are familiar with self-induced vomiting as the primary purging behavior, but individuals with BN often use alternative or additional means of preventing weight gain, including abuse of laxatives and/or enemas, diuretics (especially among health care workers), stimulants (including methylphenidate, cocaine, ephedra, and caffeine), underdosing of insulin (for those with diabetes mellitus), fasting or restrictive eating, and excessive exercise (see Table 8-1). Intentional consumption of gluten to promote weight loss in adolescents with celiac disease also has been reported.57 Whereas most compensatory behaviors to prevent weight gain fall within the diagnostic subtype of purging BN, excessive exercise and fasting are behaviors categorized within the subtype of nonpurging BN.8 Because of the absence of the more classic purging behaviors and because of their frequent indistinct nature, this variant of BN often is challenging to identify. As with overeating and dieting, it frequently is difficult to determine the line between culturally normative and pathologic behavior with excessive exercise. Generally, clinical suspicion should be raised when an individual continues to exercise despite an injury or illness, or if he or she is exercising routinely in excess of what a coach is recommending for the team. It is recommended that clinicians explore the presence of purging behaviors if an eating disorder is suspected. Although it is not certain that a patient will respond candidly, individuals probably are more likely than not eventually to disclose information about symptoms when asked.51 Some patients report feeling relieved when clinicians pose such questions if they previously had not been able to discuss their symptoms. On occasion, however, patients report learning about techniques from clinicians’ questions, so it is advisable to provide a psychoeducational context for the questions (e.g., by conveying serious physical consequences associated with the behavior, such as ipecac use) and to avoid introducing information about a dangerous behavior (e.g., underdosing insulin), depending on the clinical context. Patients also benefit from learning that treatment is available and that their clinicians understand the illness. Whereas all these purging and other behaviors aimed at neutralizing calorie intake and controlling weight can pose medical risks when chronic, some of them pose more immediate, and potentially lethal, consequences. Patients should be informed of these acute life-threatening risks and steps should be taken to eradicate such behaviors immediately. For example, because of the serious neurotoxicity, cardiotoxicity, and risk of death associated with repeated syrup of ipecac ingestion,58 its ongoing use is a clinical emergency and may require immediate hospitalization. Many patients are unaware of the serious risk associated with syrup of ipecac use. Similarly, ephedra, now banned in the United States, poses risk of stroke or adverse cardiac events, even in young adults.59 Some ephedra-free supplements marketed as weight loss agents also may be proarrhythmic and pose medical risks.60 Although patients find it difficult to abstain from purging behaviors, they may be willing to

Chapter 8  Eating Disorders substitute less immediately harmful behaviors while treatment is initiated. Individuals with BN are excessively concerned with weight and body shape. For example, they may be preoccupied with a running mental tally of calories and plans to neutralize them to prevent weight gain. Their self-image frequently is poor and anchored to their weight. It is not unusual for individuals with AN or BN to weigh themselves daily, even several times each day, and experience fluctuations in self-esteem and mood based on the result.

EATING DISORDER NOT OTHERWISE SPECIFIED

EDNOS covers a broad range of clinical manifestations, including atypical symptoms, symptoms of BED, symptoms consistent with NES, and subthreshold, yet clinically significant, eating disorders. Although EDNOS is a residual category, it nonetheless is the most commonly diagnosed eating disorder in outpatient settings, and therefore there is much interest in refining diagnostic categories of eating disorders.61 EDNOS currently includes eating disorders that do not meet threshold criteria for duration or frequency for AN or BN as well as BED, NES, purging disorder and other atypical variants. Diagnostic classification of eating disorders will be updated in the DSM-V, with an expected publication date of 2012.

BINGE EATING DISORDER

BED is a variant of EDNOS, although there is a substantial literature to support its prevalence and consistent response to specific therapeutic strategies. Like BN, BED is characterized by recurrent binge-eating. To meet provisional criteria for BED, the binge eating episodes must occur two days per week, on average, for at least six months. Unlike BN, however, BED is not associated with recurrent inappropriate compensatory behaviors to prevent weight gain. BED is distinguished from nonpathologic overeating by several possible associated symptoms, including rapid eating, eating irrespective of hunger or satiety, eating alone because of shame, and negative feelings after a binge.8 Apart from overweight or obesity, BED patients frequently present without any specifically associated physical findings. Although in some cases binge eating associated with BED may cause or perpetuate weight gain, many with BED develop symptoms only after they have become overweight. Individuals with BED frequently are distressed enough about their symptoms to seek medical help, although they may present seeking a solution to their weight gain rather than their binge eating. A substantial percentage of patients seeking weight loss treatment will have comorbid BED or NES. Therefore, medical subspecialists are likely to encounter these patients before they have been diagnosed with BED.

NIGHT EATING SYNDROME AND NOCTURNAL SLEEP-RELATED EATING DISORDER

NES is a pathologic eating pattern that may be considered a variant of EDNOS. First described in 1955,62 it, too, is characterized by recurrent bouts of overeating—but not necessarily bingeing—without associated inappropriate compensatory behaviors to prevent weight gain. As such, some individuals may appear to meet criteria for NES and BED, but these are distinct syndromes with relatively little overlap,7,63 and proposed criteria for the syndrome exclude a concomitant diagnosis of BN or BED.64 There is no clear consensus on core criteria for NES, although most investigators propose morning anorexia, evening hyperphagia, and sleep disturbance (operationalized in various ways); some

propose the additional criterion of eating in relation to sleep disturbance, such as during a nighttime awakening.65 In one study, NES in obese subjects was associated with an average of 3.6 awakenings/night compared with just 0.3 awakenings/night for matched controls. Subjects with NES ate during 52% of their awakenings, taking in a mean of 1134 kJ/ episode, considerably less than the usual intake of a binge associated with BN or BED.64 NES also can occur in nonobese individuals66 but is more common in the obese and may contribute to poor outcome in weight loss treatment programs.64,67 NSRED is also characterized by nighttime snacking, but individuals typically are totally or partially unconscious (e.g., they are in stage 3 or 4 sleep) during the snacking and frequently do not remember it.7

PURGING DISORDER

Emerging evidence raises the possibility of an additional distinctive eating disorder variant characterized by recurrent purging symptoms in the absence of clinically significant binge pattern eating. The proposed name for this diagnostic category is purging disorder. Crossover between this variant and BN appears to be rare, lending support to the hypothesis that this represents a distinctive clinical phenomenon68; data also support comparable severity to BN. Lifetime prevalence of purging disorder has been estimated as 1.1% to 5.3% of young adult women. Course, outcome, and treatment strategies for purging disorder require further research.69

DIFFERENTIAL DIAGNOSIS

The differential diagnosis of the eating disorders includes evaluation and exclusion of medical causes of weight loss, weight gain, anorexia, hyperphagia, vomiting, and other associated symptoms. These considerations are especially germane in cases of atypical or early- or late-onset eating disorders.42 Medical causes of appetite and/or weight loss include hyperthyroidism, diabetes, malignancy, and infectious diseases, among the systemic disorders, and substance abuse, depression, dementia, delirium, and psychosis. Illnesses associated with weight gain include hypothyroidism, Cushing’s disease, and organic brain disease. The differential diagnosis of hyperphagia is broad and includes PraderWilli syndrome, dementia (including Alzheimer’s disease), and intracranial lesions. Hyperphagia also has been associated with the use of certain medications, particularly many of the psychotropic agents (e.g., lithium, valproate, tricyclic antidepressants, mirtazapine, and conventional and atypical antipsychotic agents), pregnancy,70 and poststarvation refeeding.71 Psychiatric illnesses associated with loss of appetite and weight loss include major depression, anxiety, and substance-use disorders. Moreover, comorbid psychiatric illness is common among those with eating disorders,3 and frequently complicates their diagnosis and treatment. Thus, identi­fication of excessive concern with weight and food intake, unrealistic or inappropriate weight goals, or resistance to attempts to restore normal weight and/or limit excessive exercise can be helpful in distinguishing an eating disorder from another psychiatric illness or in revealing the presence of an underlying comorbid eating disorder. Because individuals with BN and EDNOS can have an unremarkable physical examination on presentation, the diagnosis may remain obscure until the patient discloses his or her symptoms, or until the clinician suspects an eating disorder based on other elements of the clinical history (e.g., weight fluctuations or menstrual irregularities). Although there potentially is much phenomenologic overlap among the eating disorders and individuals do cross

125

126

Section II  Nutrition in Gastroenterology Table 8-2  Distinguishing Features of Eating Disorders

Eating Disorder

Physical Signs Included in Diagnostic Criteria

Restrictive Pattern Eating

Binge Pattern Eating

AN

Amenorrhea, underweight

Typically

May occur in AN

BN

None (patients generally are of normal weight or overweight) None (patients frequently are overweight or obese) Nighttime awakenings (patients range from normal weight to obese)

May occur as behavior to control weight No

None (patients range from underweight to obese)

Frequently

Must occur twice weekly for at least 3 mo Must occur at least 2 days/wk for at least 6 mo Not intrinsic to the disorder; snacks of high-calorie foods typically contain fewer calories than a binge Frequently

BED NES

EDNOS

No

Purging and Other Behaviors to Control Weight or Neutralize Effects of Calorie Intake

Excess Concern with Body Image or Weight

Purging may occur in up to half of those with AN Must occur to meet diagnostic criteria

Yes

No

Not intrinsic to current research diagnostic criteria8 Not intrinsic to the syndrome

No

Frequently

Yes

Frequently occurs but not necessary to meet diagnostic criteria

AN, anorexia nervosa; BED, binge-eating disorder; BN, bulimia nervosa; EDNOS, eating disorder, not otherwise specified; NES, night-eating syndrome.

over from one diagnostic category to another (Table 8-2),43 categories are mutually exclusive, according to DSM-IV* diagnostic criteria.8 Even though a transdiagnostic approach to eating disorders classification and treatment has been proposed,72 differing responses to treatment make it desirable to establish a clear diagnosis to optimize care. A weight criterion distinguishes anorexia nervosa from bulimia nervosa in some cases. Individuals who are substantially underweight (e.g., 85% or less of expected body weight) and who otherwise meet criteria for AN most likely should be classified as having AN, even if bingeing, purging, or both are present. Individuals with BED or BN also can have symptom overlap. BN is distinguished by recurrent purging and other behaviors directed at neutralizing excessive calorie intake so as to prevent weight gain, as well as an excessive concern with weight.

NUTRITIONAL AND MEDICAL EVALUATION

In addition to excluding medical causes of weight and appetite changes, medical evaluation for confirmed or suspected eating disorders includes obtaining a full history of the patient’s eating behaviors, with attention to daily number of calories ingested, purging behavior (e.g., vomiting, use of ipecac, or laxative use), and exercise patterns. Often, medical evaluation will be guided by an assessment of nutritional status, which includes determination of the appropriateness of weight for height, age, and gender.

NUTRITIONAL EVALUATION

There are several established means for evaluating nutritional status in the office, central to which is measuring weight and height (see Chapter 5). Assessment of the appropriateness of weight for height is one of the key factors intrinsic to determining the urgency of medical and psychiatric care. For patients with AN, it is important not to rely on self-reported weight, given the strong possibility of an inaccurate report. Individuals with AN often go to great effort to conceal their low weights. For example, some patients “water load” prior to a clinical encounter, some attach weights to themselves, and others layer loose and bulky clothing to create the illusion of being of normal weight. Assessment of weight, therefore, should factor in the possibility that a patient may wish to conceal a low weight or weight loss. Some clinicians will find it helpful to have a scale in a private area (i.e., not in a hallway) and a clear and consistent protocol for weighing patients with AN. This might include asking them to void prior to being weighed, to change into a hospital gown, and to remove heavy jewelry. When patients have a history of consuming water prior to an appointment to increase their measured weight, it may be helpful to check a urine specific gravity in order to adjust interpretation of the office weight. Standard means of evaluating the appropriateness of weight for height include use of the BMI. This is calculated as follows: 2 BMI = weight ( in kg ) height ( in m )

*The multiaxial diagnostic system inherent in the Diagnostic and Statistical Manual (of Mental Disorders (DSM) reflects several dimensions relevant to health, mental health, and functioning.8 Axis I refers to psychiatric illnesses exclusive of personality disorders and mental retardation. Axis II refers to personality disorders and mental retardation. Axis III refers to medical conditions that are potentially relevant to the mental disorder(s). Axis IV refers to acute and chronic psychosocial stressors. Axis V refers to the level of psychological, social, and occupational functioning.

Although BMI may not be an appropriate standard for evaluating a healthy weight status in professional athletes, with relatively high lean muscle mass, and in some ethnic groups (e.g., Polynesians may have a different cut point for obesity73), BMI generally is appropriate for men and women aged 18 years or older. A BMI within the range of 18.5 to

Chapter 8  Eating Disorders 24.9 kg/m2 for men and women is considered normal. A BMI of 17.5 kg/m2 or less is the threshold for meeting the underweight criterion for AN in the ICD-10 (International Classification of Diseases 10).8,74 A BMI in the range of 25 to 29.9 kg/m2 is consistent with overweight, and a BMI higher than 30 kg/m2 reflects obesity.75 An alternative means for evaluating weight for height that is especially useful to assess nutritional compromise and weight recovery in those with eating disorders expresses the patient’s weight as a percentage of expected body weight. The formula is as follows: Expected body weight (%) = (patient weight expected weight for height and gender ) × 100% Expected weight for height = 100 pounds + 5 pounds per inch above 5 feet ± 10% for women, 106 pounds + 6 pounds per inch above 5 feet ± 10% for men and, if the patient is shorter than 5 feet, then the same number of pounds per inch is subtracted for each inch below 5 feet.76 Although this is a linear equation (compared with the quadratic equation for BMI) and may be less useful at extreme heights, it is straightforward to calculate. Moreover, conceptually, this formula may be easier for patients and families to understand, especially in setting weight goals or limits. A 90% to 110% range of expected body weight is considered within the normal range and is a good place to begin for setting weight gain goals for patients with AN. Within this range, the goal will be refined by clinical history (including the patient’s history of baseline, minimal and maximal weights), whether and when menses return, and medical parameters, such as reversal of bone loss. Patients below 85% of expected body weight likely meet the weight criterion for AN, and those below 75% of expected body weight are seriously nutritionally compromised and generally require inpatient care.77 For patients who are overweight or obese (>110% or >120% expected body weight, respectively), it may not be realistic or desirable to set weight goals within the normal range. Considerations in weight management are discussed subsequently. Seriously nutritionally compromised patients require inpatient care for both efficacy and safety of weight management. For underweight patients without this degree of compromise, the primary goals of nutritional management are increasing caloric requirements to regain weight, ensuring adequate intake and balance of macro- and micronu­ trients, and reestablishing a dietary pattern of three meals daily. Patients’ diets are supplemented routinely with calcium (if dietary intake is inadequate) and multivitamins containing vitamin D. Some may require additional dietary guidance and adjustments because many patients restrict not only calories, but specific foods or food groups as well. For patients with BN, BED, and EDNOS, dietary intervention includes moderating excessive calorie intake and establishing a pattern of eating that is less vulnerable to emotional cues and excessive hunger. Many patients with eating disorders are quite knowledgeable about nutrition and commonly wish to avoid meeting with a nutritionist. Conversely, information from a nutritional assessment is invaluable to the treatment team and even well-informed patients are likely to benefit from reinforcement of more healthful food choices, meal patterns, and appropriate intake.

Medical Evaluation

Medical evaluation includes a clinical history with special attention to weight fluctuations and any purging or other

Figure 8-2.  Dental erosions resulting from chronic vomiting. (Adapted with permission from the Department of Psychiatry, Massachusetts General Hospital, Boston.)

inappropriate behaviors to neutralize calorie intake to control weight (see Table 8-1). Ascertainment of syrup of ipecac use (as an emetic) and nonadherence to insulin protocols in patients with diabetes mellitus is essential, given the potentially lethal sequelae of these behaviors.78 Symptoms of medical complications of undernutrition, overnutrition, excessive exercise, or purging should be assessed and a menstrual history should be clarified. Physical examination includes a comprehensive assessment of potential complications of nutritional deficiencies, underweight, overweight, excessive exercise, and purging behaviors. If an eating disorder is suspected, physical examination may reveal signs to confirm nutritional compromise (e.g., bradycardia, hypotension, hypothermia, lanugo, breast tissue atrophy, muscle wasting, peripheral neuropathy) or to suggest chronic purging (e.g., Russell’s sign, an excoriation on the dorsum of the hand from chronic scraping against the incisors); hypoactive or hyperactive bowel sounds; an attenuated gag reflex79; dental erosion (perimolysis; Fig. 8-2)80; or parotid hypertrophy (Fig. 8-3).81 Medical complications of behaviors associated with AN, BN, BED, and EDNOS are potentially serious and are too numerous to review in detail here; selected complications are listed in Table 8-3. Complications that are common and/ or associated with serious morbidity should be actively sought on physical examination and laboratory studies, so that appropriate interventions can be initiated. Examples of such important and common findings include abnormal vital signs (e.g., hypotension, orthostatic hypotension, bradycardia, hypothermia), low weight or overweight, osteopenia or osteoporosis,82 and dental pathology (e.g., perimolysis [erosion of the tooth enamel], caries, or both).81,83,84 Cardiac complications can be lethal and include prolonged QT interval, QT dispersion, ventricular arrhythmias, and cardiac syncope.85,86 Neurologic findings in AN include cortical atrophy and increased cerebral ventricular size.87 Endocrinologic abnormalities include menstrual abnormalities, low serum estradiol levels, low serum testosterone levels, hypercortisolism, and euthyroid sick syndrome, with resultant hypotension and cold intolerance.88 Reported complications of eating disorders during pregnancy include miscarriage, inadequate weight gain of the mother, intrauterine growth retardation, premature delivery, infants of low birth weight and low Apgar scores, and perinatal death.89-92

127

128

Section II  Nutrition in Gastroenterology b-human chorionic gonadotropin and possibly a serum prolactin level therefore are recommended. Additional studies such as follicle-stimulating hormone (FSH) to evaluate ovarian function or neuroimaging studies to exclude a pituitary lesion may be indicated in some clinical scenarios. Bone densitometry using dual-energy x-ray absorptio­ metry (DEXA) scans of the hip and spine are useful in identifying bone loss and can be repeated after a year to assess further bone loss if disease continues. Osteopenia and osteoporosis may be present in as many as 90% and 40%, respectively, of women with AN, and are associated with risk of fractures and kyphosis.103,104 An electrocardiogram is recommended to evaluate patients with eating disorders because idiopathic QT prolongation can occur with AN,105 and QT intervals may be prolonged in those with BN and EDNOS, even in the absence of hypokalemia.106 Use of pharmacologic agents that can prolong the QT interval (e.g., olanzapine or desipramine), as well as purging that leads to hypokalemia, may further increase the risk of cardiac arrhythmia in this patient population. Abuse of ipecac may result in potentially fatal cardiotoxicity and arrhythmias.78 Figure 8-3.  Patient with parotid hypertrophy resulting from chronic vomiting. (Adapted with permission from the Department of Psychiatry, Massachusetts General Hospital, Boston.)

Laboratory Evaluation

Whereas choice of laboratory studies to evaluate medical complications of eating disorders will depend on the clinical history and presentation, it is useful to obtain serum electrolyte levels among individuals in whom AN or BN is suspected or confirmed. For example, hypokalemia occurred in 4.6% of a large sample of outpatients with eating disorders in one study93 and in 6.8% of individuals with BN in another moderately sized sample.94 In the latter study, hypokalemia was significantly more common in patients with BN than in those without BN. Although assessment for hypokalemia is not efficient for identifying occult cases of BN, it will assist in the identification and monitoring of individuals at risk for cardiac arrhythmias secondary to their eating disorder. Hypochloremia, hypomagnesemia, hyponatremia, hypernatremia, and hyperphosphatemia also are seen in patients with eating disorders.88,94-96 In addition, for patients with AN, a serum glucose determination is recommended to identify hypoglycemia, which can be severe in this population.97 Although hyperamylasemia reportedly is common in BN (i.e., in 25% to 60% of cases), laboratory analysis of serum amylase generally is not clinically useful for detecting BN or gauging the severity of bingeing and purging symptoms.98 An elevated serum amylase level in a patient with AN or BN often reflects increased salivary isoamylase activity98,99; however, pancreatitis should be considered, when clinically appropriate, given its occurrence in this patient population. A complete blood count is recommended to assess for anemia, neutropenia, leukopenia, and thrombocytopenia among patients with AN. A retrospective study of 67 patients with AN found that 27% had anemia, 17% had neutropenia, 36% had leukopenia, and 10% had thrombocytopenia.100 Evaluation of the cause of amenorrhea is suggested, even if it is presumed to be related to decreased pulsatility of gonadotropin-releasing hormone secondary to weight loss.88 Menstrual irregularities are common among women with eating disorders, but women with symptomatic eating disorders still may be menstruating at presentation101 and women with AN can become pregnant102; a quantitative

GASTROINTESTINAL ABNORMALITIES ASSOCIATED WITH EATING DISORDERS GI signs and symptoms are common in those with eating disorders (Tables 8-3, 8-4). It has been asserted that the most dramatic changes in bodily function caused by AN are in the GI tract.107 There is also evidence that many individuals with eating disorders may present with a GI complaint prior to seeking treatment for an eating disorder. In one small retrospective study, 8 of 13 inpatients with eating disorders had sought care for a GI complaint, and 6 of them had sought such GI care before tending to their eating disorder.108 Several cross-sectional studies of hospital inpatients with eating disorders have suggested that 78% to 98% have GI symptoms.109-113 For example, constipation is a frequently reported symptom in AN and BN; in a study of 28 inpatients with an eating disorder, 100% of patients with AN and 67% of patients with BN had constipation.112 Nausea, vomiting, gastric fullness, bloating, diarrhea, and decreased appetite also are seen commonly in AN and bloating, flatulence, decreased appetite, abdominal pain, borborygmi, and nausea commonly are reported in BN. In one study of 43 inpatients with severe bulimia nervosa, 74% reported bloating, 63% reported constipation, and 47% reported nausea; borborygmi and abdominal pain also were more frequent than in the comparison group of healthy controls.110 Moreover, certain GI symptoms have been shown to be more common in dieters (specifically, abdominal pain, bloating, and diarrhea)114 and in those with binge eating (nausea, vomiting, and bloating) than in normal controls.115 A large study of obese individuals with GI symptoms found a strong association between BED and abdominal pain and bloating, after adjusting for BMI.116 Finally, a study of 101 consecutive women admitted to an inpatient eating disorders program found the vast majority of study participants (98%) to have functional gastrointestinal disorders (FGIDs). Among these, 52% had irritable bowel syndrome, 51% had functional heartburn, 31% had functional abdominal bloating, 24% had functional constipation, 23% had functional dysphagia, and 22% had functional anorectal pain; 52% of respondents met criteria for three or more FGIDs. Whereas the authors found psychological predictors for several of the FGIDs, they were not associated with functional

Chapter 8  Eating Disorders Table 8-3  Selected Clinical Features and Complications of Behaviors in Patients with Eating Disorders Clinical Feature or Complication Associated with Weight Loss and Food Restriction or binge-eating in anorexia nervosa

Associated with Purging or refeeding Behaviors in anorexia nervosa, bulimia nervosa, or EDNOS

Cardiovascular

Arrhythmia Bradycardia Congestive heart failure (in refeeding syndrome) Decreased cardiac size Diminished exercise capacity Dyspnea Hypotension Mitral valve prolapse Orthostasis Prolonged QT interval QT dispersion Syncope

Ventricular arrhythmia Cardiomyopathy (with ipecac use) Prolonged QT interval Orthostasis Syncope

Dermatologic

Brittle hair Dry skin Hair loss Hypercarotenemia Lanugo

Russell’s sign (knuckle lesions from repeated scraping against the incisors)

Oral, pharyngeal

Cheilosis

Dental erosion and caries Sialadenosis Pharyngeal and soft palatal trauma Angular cheilitis Perimolysis Vocal fold pathology

Gastrointestinal*

Anorectal dysfunction Delayed gastric emptying Elevated liver enzyme levels Elevated serum amylase levels Gastroesophageal reflux Hepatic injury Pancreatitis Prolonged whole-gut transit time Rectal prolapse Slow colonic transit Superior mesenteric artery syndrome

Abdominal pain Acute gastric dilatation Barrett’s esophagus Bloating Constipation Delayed gastric emptying Diarrhea Dysphagia Elevated liver enzyme levels Elevated serum amylase levels Esophageal bleeding Esophageal ulcers, erosions, stricture Gastroesophageal reflux Mallory-Weiss tear

System Affected

During refeeding: Acute gastric dilatation, necrosis, and perforation Elevated liver enzyme levels Hepatomegaly Pancreatitis

Endocrine and metabolic

Amenorrhea Euthyroid sick syndrome Hypercholesterolemia Hypocalcemia Hypoglycemia Hyponatremia Hypothermia Low serum estradiol, low serum testosterone levels Osteopenia, osteoporosis Pubertal delay, arrested growth As part of the refeeding syndrome: Hypomagnesemia Hypophosphatemia

Gastroesophageal reflux Gastric necrosis and perforation Hematemesis Pancreatitis Prolonged intestinal transit time Rectal bleeding Rectal prolapse Hypercholesterolemia Hyperphosphatemia Hypochloremia Hypoglycemia Hypokalemia Hypomagnesemia Hyponatremia Hypophosphatemia Metabolic acidosis Metabolic alkalosis Secondary hyperaldosteronism

129

130

Section II  Nutrition in Gastroenterology Table 8-3  Selected Clinical Features and Complications of Behaviors in Patients with Eating Disorders—cont’d Clinical Feature or Complication Associated with Weight Loss and Food Restriction or binge-eating in anorexia nervosa

Associated with Purging or refeeding Behaviors in anorexia nervosa, bulimia nervosa, or EDNOS

Acute kidney injury Amenorrhea Atrophic vaginitis Breast atrophy Infertility Pregnancy complications (including low birth weight, premature birth, and perinatal death)

Abnormal menses Azotemia Pregnancy complications (including low birth weight infant)

Neurologic

Cognitive changes Cortical atrophy Delirium (in refeeding syndrome) Peripheral neuropathy Ventricular enlargement

Stroke (associated with ephedra use) Neuropathy (with ipecac use) Reduced or absent gag reflex

Hematologic

Anemia Leukopenia Neutropenia Thrombocytopenia

System Affected Genitourinary and reproductive

EDNOS, eating disorder, not otherwise specified. *Gastrointestinal complications associated with binge pattern eating in any of the eating disorders, are not all listed, and include weight gain, acute gastric dilatation, gastric rupture, gastroesophageal reflux, increased gastric capacity, and increased stool volume.

Table 8-4  Common Gastrointestinal Symptoms in Patients with Eating Disorders Abdominal pain Belching Bloating Borborygmi Changes in appetite Constipation Diarrhea Dyschezia Flatulence Nausea Vomiting

abdominal bloating or functional dysphagia in this study population.109 Specific GI findings are commonly associated with eating disorders. Delayed whole-gut transit time98 and delayed gastric emptying appear to be common among inpatients with AN or BN.24,117-120 Delayed colonic transit has been reported in AN.121 Mild esophagitis is common (e.g., 22% of a case series of 37 consecutive patients) in patients with chronic BN, but more serious esophageal disease is rare.122,123 Abnormal esophageal motor activity has been reported in AN and BN.124,125 Barrett’s esophagus, Mallory-Weiss tears, and gastroesophageal reflux have been reported in association with chronic vomiting associated with BN.126 Unusual GI manifestations and catastrophic complications have been described in case reports of patients with eating disorders, including acute gastric dilation, gastric necrosis and perforation, and occult gastrointestinal bleeding (attributed to transient intestinal ischemia in the setting of endurance running).127-129 Rectal bleeding and rectal prolapse have been reported in patients with AN and BN.130,131

Other studies have found abnormal gastric function in patients with eating disorders, including diminished gastric relaxation (in patients with BN),132 bradygastria (in patients with AN, BN, and EDNOS),133 and higher gastric capacity (in patients with BN).134 Some evidence suggests that the GI abnormalities associated with eating disorders may be related to the duration or presence of active eating disorder symptoms. Physiologic sequelae of disordered eating, such as contracted or expanded gastric capacity, altered gastric motility, delayed large bowel transit (through reflex pathways),135 and possibly blunted postprandial cholecystokinin release, may perpetuate symptoms that exacerbate the excessive body image concern driving abnormal eating patterns.24 There is evidence that subjective reports of GI symptoms do not correlate well with physiologic data in patients with eating disorders.136 GI findings associated with eating disorders are listed in Table 8-3. In one study,137 elevated liver biochemical test results were documented in 4.1% of 879 patients presenting for treatment of an eating disorder. A probable cause distinct from the eating disorder was identified for 47% of the study participants but, for the remaining 53% of subjects, the abnormal results could not be attributed to a condition other than their eating disorder. Elevated enzymes were seen in underweight and normal-weight study participants. The results of this study suggest that abnormal liver biochemical tests are neither a specific nor a common marker for an eating disorder, and other possible causes should be excluded before attributing such abnormality to the eating disorder.137 In contrast, a study of 163 adolescent and young adult women outpatients with AN or EDNOS and low weight (excluding those with acute illness, alcohol abuse, hepatitis from viral or other known causes, and on medications associated with elevated liver enzyme levels) found elevated aminotransferase levels (alanine aminotransferase and/or aspartate aminotrans­ferase) in 19.6% of AN patients with a BMI lower than 16 kg/m2; 8.7% of AN patients with BMI higher than 16 kg/m2, and 15.2% of low-weight EDNOS

Chapter 8  Eating Disorders patients.138 Elevated liver biochemical test results and hepatomegaly also are observed on the initiation of refeeding in AN.137,139 There also are several case reports of severe liver dysfunction or damage in patients with AN attributed to malnutrition and associated hypoperfusion.140-142 Although many of the common GI complications of eating disorders are relatively benign, others, such as acute gastric dilation, gastric necrosis, and gastric rupture,143-147 although uncommon, are serious or even catastrophic. Esophageal rupture is another potentially catastrophic risk with chronic vomiting.139 Acute pancreatitis has been reported in patients with AN and BN,145,148,149 and also can be associated with refeeding in AN.138 In addition, there is a case report of severe steatosis resulting in fatal hepatic failure in a patient with severe AN150 and one of death resulting from duodenal obstruction secondary to a binge in a patient with BN.151 Both help-seeking and diagnosis may be delayed or com­ plicated by an undisclosed or unrecognized eating dis­ order.127,145,152 Conversely, esophageal dysfunction can be obscured by bulimic symptoms79,124 and can be misdiagnosed as AN.125 Superior mesenteric artery (SMA) syndrome can complicate AN and occurs when the support of the SMA is lost with weight loss and the duodenum is compressed between the aorta and the SMA. Because it manifests with vomiting, a concurrent diagnosis can be missed if this symptom is attributed to the eating disorder.153

MANAGEMENT Optimally, management of patients with eating disorders includes integration of mental health, nutrition, and primary care (Fig. 8-4). Occasionally, medical subspecialty consultation and care are helpful. Multidisciplinary management is desirable for several reasons. First, patients are at risk of medical, psychological, and nutritional complications of their disease. Second, patients commonly selectively avoid care essential to their ultimate recovery. For example, a patient may wish to avoid the detection of an injury so that she or he can continue to participate in a team sport; another may find it difficult to undergo the psychological work necessary to address antecedents of her illness; or another may wish to bypass active weight management. Conversely, a patient may attempt to pursue relief for specific medical complications to the exclusion of appropriate psychological or nutritional therapies. It often is helpful, if not essential, to establish a treatment agreement at the outset of care for a patient with an eating disorder. This is particularly relevant for patients for whom the severity of their symptoms may compromise medical and psychological health to a degree that hospital level care is likely during the course of treatment. A treatment agreement allows the caregivers to establish initial treatment goals and criteria for which they may wish to adjust the level of intensity of care. This will allow transparency of expectations for the patient, facilitate a rapid response to emerging crises, and help avoid split opinions among the team during the course of care. A treatment agreement also clarifies for patients the contingencies for nonadherence or poor health. As part of the initiation of care, patients should be asked to give permission for open communication among the members of the clinical team. If a patient cannot agree to this, it signals potential difficulties in providing coordinated care, and the lack of agreement should be reconciled. Depending on the patient’s age and circumstances, a plan for how and what information will be shared with parents also should be established.

Evaluation

Education and involvement of patient Identification of most appropriate level of care

Outpatient

Partial hospitalization

Identification of interdisciplinary team and referrals

Primary care clinician(s)

Medical specialty care

Nutritionist

Inpatient

Residential care

Referrals

Mental health clinician(s)

Psychotherapist

Psychopharmacologist

Adjunctive psychotherapist(s) (e.g., group or family) Division of tasks (e.g., weighing, medical parameter surveillance, dietary plan, monitoring symptom frequency and severity) Consensus on whether there will be any weight, symptom severity, or symptom frequency parameters that necessitate a more intensive level of care, adjunctive treatment, or restriction of activities Plan for how often and how to communicate

Obtain permission from the patient for communication Draft treatment agreement if warranted

Initiate treatment Reevaluate and adjust care as necessary Figure 8-4.  Key components of team management of patients with an eating disorder.

PSYCHIATRIC TREATMENT

Psychiatric treatment generally begins with psychotherapy. In many cases, pharmacotherapy is useful as an adjunctive treatment for BN and BED. Active weight management is indicated for AN, and there is a role for weight loss treatment in some patients with BED. Usually, psychotherapy can be used to support weight management goals, although optimally it should be coordinated with the efforts of the nutritionist and primary care clinician on the team. Regardless of the mode of psychotherapy chosen, specific behavioral strategies directed at establishing normal eating patterns and drawing the patient’s attention to triggers for abnormal patterns can augment treatment. Among these, patients are encouraged to identify and avoid emotion-, schedule-, and food-related triggers to episodes of bingeing

131

132

Section II  Nutrition in Gastroenterology

Time

Location

Food consumed

Feelings about Eating Before

During

After

Triggers

7 a.m.

At home

Coffee, 1/2 bagel

Hungry

OK

Good—only ate 1/2 of the bagel

None; it was breakfast time

9 a.m.

At desk

Diet cola

Avoiding a snack

OK

Glad to avoid food

None

11 a.m.

At desk

Diet cola

Hungry—trying to avoid eating

OK

Glad to avoid food

None

1:30 p.m.

Cafeteria

Salad with fat-free dressing on the side and a diet cola

Very hungry— trying not to eat too much

OK since dressing was low-cal

Glad to avoid a bigger lunch than I had

None

2:00 p.m.

At desk

3 mini candy bars, and a cookie

Starving

Gross

Gluttonous

Candy looked tempting and I was hungry

4:00 p.m.

In meeting

2 cups of coffee

Still hungry

OK

Good; glad to avoid food

Hungry again

7:00 p.m.

Walking home

Pizza (3 slices)

Starving: upset about interaction at work; want comfort food

Temporarily distracted

Terrible—way too many calories

Boss reprimanding me

8:45 p.m.

At home in front of TV

Most of a pint of ice cream, a box of cookies, 3 donuts, and a spoonful of peanut butter

Still upset; feeling that I’ve already blown it for the day and may as well eat more and then purge

Numb

Disgusted with myself for bingeing but relieved after purging

Still upset about work; knowing there was unfinished cookie dough ice cream in freezer; roommate left donuts; alone

Figure 8-5.  Food journal page from a hypothetical patient with bulimia nervosa.

and to plan three regular meals and two between-meal snacks to prevent excessive hunger. Finally, a food journal (Fig. 8-5) kept for a few days and reviewed in a treatment session will help many patients identify relationships among psychosocial stressors, hunger, and symptoms and may provide a concrete framework from which to relate symptoms to other psychological concerns.

Psychotherapy

A variety of psychotherapies have established efficacy for the eating disorders. Recent guidelines and reviews have summarized findings from empirical studies and highlighted the paucity of recommendations for treatment of AN and EDNOS.77,154-157 Cognitive behavioral therapy (CBT) and interpersonal therapy (IPT) have received a great deal of research attention for the treatment of eating disorders. CBT is a structured, manual-based approach that addresses the relationships among thoughts, feelings, and behaviors; IPT is another short-term therapy focused on present-day interpersonal events and roles in relationships. The choice of psychotherapeutic modality will be guided by the diagnosis, medical and psychiatric comorbidities, desirability of targeting the eating disorder symptoms versus bro­adening the therapeutic goals, treatment history, patient strengths and preferences, and the availability of care. Initial recommendations should be evidence-based when possible; however, clinical judgment is important for identifying individual needs and situations in which alternative treatment choices are appropriate.154 In practice, patients

with AN or EDNOS can benefit from a flexible approach to treatment that uses appropriate components of the various therapeutic modalities because there is a dearth of empirical data to enable evidence-based recommendations and because some patients do not respond to evidencebased treatments.158 There is limited empirical evidence for AN treatments. The one consistent finding is that family therapy focused on parental control of nutrition emerges as the treatment of choice for adolescents, particularly those who are younger and who have a shorter duration of illness.155,159 Although evidence-based recommendations are limited, guidelines do suggest therapies to be considered for the psychological treatment of AN: cognitive analytic therapy (CAT), CBT, IPT, focal psychodynamic therapy and family interventions focused explicitly on eating disorders.154 A study comparing CBT, IPT and nonspecific supportive clinical management in the treatment of underweight AN outpatients found that the supportive treatment produced better global outcomes than IPT and was superior to CBT, over 20 weeks, in its impact on global functioning.160 The efficacy of CBT for underweight individuals remains unclear, but it appears useful as a posthospitalization treatment for AN, contributing to improved outcomes and relapse prevention in adults after weight restoration.161 Factors consistently predicting treatment outcome have not been identified.155 A number of treatments for BN have strong empirical support. CBT and IPT have been found effective, with CBT superior at reducing behavioral symptoms.156 CBT leads to

Chapter 8  Eating Disorders faster improvement in symptoms, with better outcomes at the end of treatment, but at follow-up assessment there are no differences between CBT and IPT.162 All guidelines recommend CBT (16 to 20 sessions over four to five months) as the first-line treatment of choice for BN,77,154,156 but not all patients respond to CBT, and IPT is an effective alternative. CBT and IPT can be delivered in a group format as well as individually.163,164 Other promising treatment options with preliminary empirical support include dialectical behavior therapy (DBT, an approach developed for borderline personality disorder that focuses on assisting patients in developing skills to regulate affect165) and a manual-based guided self-change approach.166 For a subset of patients, self-help or guided self-help with an evidence-based CBT manual167 is an appropriate starting point for treatment in a steppedcare approach154 or if other treatments are not available.168 A variety of factors have been shown to be associated with treatment outcome in BN, but two emerge consistently— severity (higher frequency of binge eating) and duration of illness are associated with poorer outcomes.156 There are limited data to guide treatment decisions for the large proportion of individuals with eating disorders who are diagnosed with EDNOS. The main exception is the subgroup of those with BED. As with BN, some individuals will benefit from an evidence-based self-help program as a first step in treatment or if other treatments are not available.77,154,168 Studies have found that self-help intervention, delivered in a variety of ways (with varying levels of professional or peer support), leads to better outcomes when compared with control groups, with reductions in binge eating, binge days, and psychological features associated with BED (for a review, see Brownley and colleagues157 and Sysko and Walsh168). After consideration of self-help, American Psychiatric Association (APA)77 and National Institute for Clinical Excellence (NICE)154 guidelines recommend CBT adapted for BED as an initial treatment choice. Group CBT has been found effective for treating binge eating in overweight individuals.169,170 There is some support for individually based CBT, although methodologic limitations preclude firm conclusions. Group IPT and adapted DBT are options to consider if CBT is not a good match for the individual or is unavailable. In one study, IPT was found to lead to similar abstinence rates as CBT at one-year follow-up.170 DBT has shown promising results, with a recovery rate of 56% at six months after treatment in one randomized controlled trial (RCT).171 It is important to note that treatments for BED usually do not result in weight loss, but they may still be of benefit with regard to weight by preventing further weight gain.157 This issue of dual treatment goals—weight loss and reducing binge eating—is explored in depth later in this chapter (see “Weight Management”). Further research is needed to establish and replicate factors associated with treatment outcome. Across all diagnoses and treatments, there has been little attention to differential outcomes by socioeconomic factors. Future studies are needed to explore whether treatment efficacy differs by gender, age, race, ethnicity, socioeconomic status, or cultural group.155,156 Given the frequent psychiatric comorbidity associated with eating disorders, as well as psychosocial risk correlates, some patients with an eating disorder will benefit from psychodynamic psychotherapy and a flexible and eclectic approach depending on patient capabilities, goals, treatment history, and other psychosocial considerations.

Pharmacotherapy

Pharmacologic management has an adjunctive role for the treatment of BN and BED. Of numerous agents that have

been studied, only one, fluoxetine, has U.S. Food and Drug Administration (FDA) approval for an eating disorder (bulimia nervosa). There is insufficient empirical support for efficacy of any agent in treating the primary symptoms of AN. Similarly, there are no clinical trial data to support recommendations for the pharmacologic management of EDNOS (with the exception of trials addressing BED and NES). Finally, there are not adequate available clinical trial data to support recommendations for pharmacologic management of eating disorders in children and adolescents.172 Among a variety of agents evaluated for treatment of the primary symptoms of AN, several have been studied because of their association with weight gain; of these, none is in routine clinical use. Although some data have suggested that olanzapine may be beneficial in promoting clinical improvement in AN,173,174 a recent RCT combining olanzapine with CBT versus placebo with CBT did not demonstrate significant between-group differences in improvement in BMI.175 Given the lack of data supporting efficacy and safety in patients with AN, no pharmacologic agents currently can be generally recommended to promote weight gain in this patient population. Pharmacologic agents associated with weight gain for other indications should be used judiciously and with a candid discussion with the patient about the anticipated risks and benefits of appetite and weight changes. Finally, if such an agent is selected, symptoms should be monitored carefully to look for onset, recurrence, or increase in bingeing or purging behaviors. Other agents may have a limited role in the management of AN but do not have FDA approval for this indication. Fluoxetine has not been found to be effective for treating the primary symptoms of AN in underweight patients176 and has unclear benefit in stabilizing weight-recovered patients with AN.177,178 Sertraline (50 to 100 mg/day) was associated with significant clinical improvements in a small, open, controlled trial of patients with AN.179 Comorbid psychiatric illness is common among patients with AN and may improve with pharmacologic management, but depressive symptoms in severely underweight patients may not respond as well to antidepressant medication as in normal-weight patients. Notwithstanding the very limited role for psychotropic medication in the management of AN, patients likely will need calcium and vitamin D supplementation if dietary sources are inadequate.88 Although oral contraceptive agents may mitigate some of the symptoms of hypoestrogenemia associated with AN, they do not protect against bone loss in this population.150,180 It is useful for clinicians to bear in mind that weight restoration is the treatment of choice for underweight individuals with AN for medical stabilization, and probably also as a prerequisite to developing the psychological insight necessary for recovery. In contrast to the limitations of medication management for AN, a number of medications have established shortterm modest efficacy for the treatment of BN, although remission rates are low.181 CBT has better efficacy than medication to reduce the symptoms associated with BN, but there is some support for augmenting psychotherapy with medication, and this is fairly routine clinical practice. It is optimal to use pharmacotherapy as an adjunct to, rather than a substitute for, psychotherapy; psychotherapy, however, may not be available or beneficial to all patients. Some evidence supports treatment with fluoxetine (60 mg/ day) alone in a primary care setting.182 Fluoxetine (60 mg/ day) also has been found superior to placebo for treating bulimic symptoms in patients who have not responded adequately to CBT or IPT.183

133

134

Section II  Nutrition in Gastroenterology Of medications with established efficacy in treating BN, only fluoxetine has FDA approval for this indication. Fluoxetine (60 mg/day) generally is well tolerated in this patient population and has been shown to be effective for symptom reduction and for maintenance therapy for up to 12 months.184,185 Desipramine and imipramine (both at conventional antidepressant dosages, as tolerated) also have efficacy in symptom reduction, but are not as well tolerated in this patient population.186 Topiramate has shown efficacy in reduction of binge and purge symptoms in two short-term RCTs in individuals with BN.187-189 Other agents that have demonstrated at least some efficacy (but with less data available) are trazodone,190 ondansetron (in patients with severe BN),191 and sertraline.192 Flutamide has shown some efficacy in reducing binge (but not purge) frequency in one small RCT, but was associated with hepatotoxicity and teratogenicity and cannot be recommended for the treatment of BN.193 A number of studies has investigated the efficacy of naltrexone in treating bulimic symptoms,186 but only at higher doses was it superior to placebo and in reducing symptoms in patients who had previously not responded to alternative pharmacotherapy.194 Monitoring of liver biochemical test results is essential when this drug is used. Other medications with efficacy are relatively contraindicated for those with BN given their potential adverse effects. For example, bupropion was associated with a seizure risk of 5.8% during a clinical trial195 and there have been case reports of spontaneous hypertensive crises in patients with BN who were taking monoamine oxidase (MAO) inhibitors.196 Finally, although fluvoxamine has shown some efficacy for BN relapse prevention in one RCT,197 another RCT combining fluvoxamine with stepped-care psych­ otherapy not only did not show efficacy of this agent, but also reported grand mal seizures in participants on the active drug.198 Several trials have investigated the efficacy of phar­ macologic treatment of BED. Of the selective serotonin reuptake inhibitors (SSRIs), sertraline,199 fluvoxamine,200 citalopram,201 and fluoxetine202 have shown some efficacy in reducing symptoms associated with BED in RCTs. In addition, atomoxetine,203 orlistat in combination with CBT,204 and zonisamide205 have shown efficacy in an RCT, although zonisamide was not well tolerated in this BED study population and significantly greater binge remission rates in the orlistat group were not maintained at the three-month post-treatment follow-up.206 Two agents, sibutramine207 and topiramate,208 have shown efficacy in reducing symptoms of BED (the latter in BED co-morbid with obesity) in mul­ tisite placebo-controlled trials. None of these medications has FDA approval for the treatment of BED. Notwithstanding some efficacy of medication, studies have suggested that CBT is a superior treatment and that augmentation of CBT with medication may not enhance treatment response.209

WEIGHT MANAGEMENT

Active weight management is a cornerstone of treatment for AN. As essential as weight gain is to reduce or reverse the medical and cognitive sequelae of severe undernutrition, it is one of the great challenges in the successful treatment of this illness. By definition, individuals with AN are unreasonably fearful of gaining weight and many of them remain unconvinced of the serious medical impact of their selfstarvation. Ideally, patients can be engaged in the process of weight recovery by identifying some clear benefit (e.g., permission to remain on an athletic team or participate in a performance, or to avoid a compulsory medical leave from school or work). Such behavioral reinforcement can be an essential adjunct to a nutritional plan that provides bal-

anced nutrition and calories adequate for weight gain and reestablishes routine meals. Outpatient weight recovery is best addressed with the collaboration of a nutritionist experienced in the treatment of AN. As calories are added and foods are reintroduced into the diet, patients may initiate or increase compensatory behaviors (e.g., exercise, purging) to control weight gain. If possible, behavioral restrictions on exercise can be implemented if patients are not meeting weight gain goals. Caloric supplements often are added as snacks to help patients meet nutritional and weight gain goals. Patients with early satiety and delayed gastric emptying have a particularly difficult time adding calories because gastrointestinal discomfort and bloating enhance their concerns about feeling and being “fat.”210,211 If supportive psychotherapy, nutritional guidance, behavioral reinforcements and limits, nutritional supplements, and restricted exercise do not result in adequate weight gain, a higher intensity of care may be indicated. Supervised meals, partial hospitalization (e.g., a structured day treatment program, often including a 12-hour day of various treatment modalities, during which the patient returns to his or her own home in the evenings), or even increasing the frequency of outpatient therapy appointments, may be sufficient to promote weight gain. However, if bingeing and purging symptoms are emerging or increasing or if the patient is losing weight, inpatient care may be required for weight restoration. Even in this setting, behavioral methods to promote weight gain are preferred to nasogastric feeding or total parenteral nutrition. The latter options are avoided if possible, but in the setting of severe malnutrition they may be necessary. Severely malnourished patients—especially those below 70% to 75% of expected body weight—require inpatient care for refeeding. Patients with AN are at particularly high risk of refeeding syndrome, which can occur with any means of refeeding (see Chapters 4 and 5).212 Refeeding syndrome, typically associated with hypophosphatemia in the setting of depletion and cellular shifts in the early weeks following refeeding, can result in delirium, congestive heart failure, and death.213 Risk for refeeding syndrome can be reduced for at-risk patients by using an initially low-calorie prescription that is advanced slowly. During at least the first two weeks of refeeding, serum electrolyte, phosphorus, and magnesium levels should be monitored closely (e.g., six to eight hours after feeding begins, then daily for a week, then at least every other day until the patient is stabilized214). Heart rate, respiratory rate, lower extremity edema, and signs of congestive heart failure also should be evaluated daily for at least a week and then gradually at longer intervals as the patient stabilizes, and cardiac telemetry should be used to monitor heart rhythm during the first two weeks so that supplementation and other appropriate measures can be instituted if hypophosphatemia or other signs of refeeding syndrome develop. Delirium may occur in the second week of refeeding, or later, and may last for several weeks.215-218 Some experimental data have suggested that a healthful dieting intervention may be beneficial in reduction of bulimic symptoms219 but, conventionally, weight loss treatment has been discouraged in patients with BN because dieting can stimulate bingeing and purging. Weight loss is often a primary or secondary treatment goal for individuals with BED because of comorbid obesity. Models of binge eating have proposed that dietary restriction is an antece­ dent to binge eating; thus, there has been debate about the optimal means and order of addressing co-occurring binge eating and obesity. Most data, however, have shown that a variety of weight loss approaches do not exacerbate binge

Chapter 8  Eating Disorders eating and may help reduce symptoms; one prospective study found no evidence that a reduced calorie diet precipitated binge eating in women with obesity.220 Behavioral weight loss treatment (BWLT)221 and very low-calorie diets (VLCDs)222,223 have been found effective for reducing symptoms of BED. Available evidence also supports that CBT and BWLT are equally effective in terms of binge eating outcomes, although the rates of change differ for binge eating and weight loss. Binge eating decreases faster with CBT, whereas weight decreases more rapidly with BWLT, so treatment priorities may inform recommendations. The addition of exercise to treatment for BED is associated with greater decreases in binge eating and BMI.224 Although a number of studies have found that treating binge eating does not translate to weight loss, some studies have found that reductions in binge eating can assist in modest weight loss among those with BED, especially when complete remission is achieved.225 BED is common in individuals presenting for obesity surgery, so there is much interest in clarifying how BED affects the outcome of bariatric surgery and, conversely, how surgery might influence binge eating behavior. The prevalence of BED in preoperative gastric bypass patients has been found to range from 2% to 49%, and up to 64% of bariatric surgery candidates have binge eating behaviors.226 The most current studies examining presurgery binge eating, postsurgery binge eating, and long-term weight outcomes suggest that a presurgery history of binge eating does impart risk for poorer long-term weight outcome.226,227 Postsurgical binge eating usually is seen in patients who reported binge eating prior to surgery and is rare among those without presurgery binge eating.226 Patients who continue to experience binge eating after surgery have poorer weight outcomes. One study has shown that a history of binge eating before Roux-en-Y gastric bypass is associated with significantly less weight loss at one- and two-year follow-ups.227 Many patients with binge eating will have positive outcomes after bariatric surgery, but adjunctive treatment is likely to be important for optimizing outcomes and preventing relapse. There are few data on bulimia or self-induced vomiting and bariatric surgery. Cases of AN developing after bariatric surgery have been reported.228,229

MEDICAL MANAGEMENT OF GASTROINTESTINAL SYMPTOMS OF PATIENTS WITH EATING DISORDERS

Individuals with eating disorders are likely to have cooccurring GI symptoms for which consultation may be sought. GI complaints are the most common somatic complaint among adolescents with partial eating disorders.230 Similarly, childhood GI complaints may influence later risk or timing or onset and severity for an eating disorder. In some cases, behaviors associated with eating disorders result in serious GI complications. In other cases, GI symptoms may be mild and not correlate with underlying pathology, but may compromise efforts to nutritionally rehabilitate the patient. Given the evidence that restrictive eating, binge pattern eating, and purging behaviors may underlie or exacerbate some of the GI symptoms, concurrent management of the eating disorder is integral to prevent worsening of the GI manifestations of illness. Careful differential diagnosis also is necessary to avoid misattribution of symptoms to an eating disorder and to detect primary GI pathology that may be obscured by an eating disorder. Available data suggest that individuals with an eating disorder are significantly more likely to seek GI specialty care than healthy controls.53

Moreover, presentation to a GI practice rather than to an eating disorder specialty practice results in delayed diagnosis and a greater number of clinical tests than controls with slow transit constipation.56 Because the GI consul­tation may precede help-seeking related to the primary symptoms of the eating disorder, the patient’s care will benefit from identification of an associated eating disorder, evaluation of its severity, and appropriate counsel about the necessity of team management and referrals to mental health, nutritional, and primary care clinicians. A case series of individuals with coexisting comorbid eating dis­orders and celiac disease has illustrated how synchronous GI and eating disorders reciprocally influence management. For example, celiac disease can mimic, exacerbate, or promote recovery from an eating disorder, whereas an eating disorder can reduce adherence to treatment for celiac disease.57 Subjective reports of Gl symptoms may not reliably indicate pathology122,231; moreover, they may be mediated by affect103 or body image concerns.93 Thus, when patients complain of bloating and constipation, it is useful to determine to what extent these complaints stem from fear of gaining weight or reflect decreased GI motility. A number of studies have evaluated improvement in GI function after nutritional rehabilitation (Table 8-5). These studies have yielded mixed results, and conclusions have been limited by small sample sizes and nonrandomized design. For example, in one study,122 gastric emptying improved in patients with restricting type AN, but did not improve in patients with binge eating–purging type AN after a 22-week treatment period of increasing dietary intake up to 4000 cal/day and CBT. Self-reported GI symptom scores improved after treatment in this same study, but remained abnormal and did not correlate with gastric emptying as evaluated with ultrasound.122 Another study of a mixed sample of adolescents and adults with AN did not demonstrate significant improvement of gastric emptying after weight gain (N = 6), despite normalization of heart rate and blood pressure.232 Other studies have suggested that nutritional rehabilitation is associated with improved gastric emptying in inpatients with AN, but it is unclear whether such improvement is related to refeeding per se, or to weight gain.97,122,106,233 Constipation is a frequent complaint of patients with AN and BN and may have multiple causes. Colonic transit appears to be delayed in patients with constipation and AN, but colonic transit has been shown to return to normal within three to four weeks of refeeding in hospitalized patients with AN.107,121 In one study, however, anorectal dysfunction in anorexic patients with severe constipation did not significantly improve with refeeding. The investigators suggested that abnormal defecatory perception thresholds and expulsion dynamics in AN may have contributed to the patients’ unremitting constipation.121 Laxative abuse occurs in AN234 and BN.235 Some patients use laxatives as their chief method of purging and may gradually escalate their daily dose to very large amounts. Although the relationship of laxative abuse to colonic dysfunction remains controversial,236-238 it has been observed that patients with chronic laxative abuse complain of consti­ pation while tapering off their laxatives. Rectal prolapse has been described with AN and BN and is thought to be linked to constipation, laxative use, excessive exercise, and increased intra-abdominal pressure secondary to selfinduced vomiting.116,117 Delayed intestinal transit and its associated clinical symptoms present a particularly interesting clinical challenge in patients with eating disorders. Studies to date, however, have been small, short term, and not randomized, so only limited conclusions regarding management can be

135

136

Section II  Nutrition in Gastroenterology Table 8-5  Selected Studies of Effects of Nutritional Rehabilitation on Gastrointestinal Symptoms Associated with Eating Disorders in Adults condition and Method of Assessment

Study Population

Results

109

Gastrointestinal symptoms using a survey (GISS)

16 with AN and 12 healthy volunteers

110

Response of GI symptoms in BN to nutritional and psychotherapeutic inpatient treatment using GISS Eating Disorders Inventory; Zung Depression Inventory

43 inpatients with severe bulimia; 32 healthy volunteers as untreated comparison group

111

GI symptoms and gastric emptying using a double isotope technique to measure gastric emptying and self-reporting of GI symptoms

14 adult inpatients with AN (13 women, 1 man) and 14 normal male controls

120

Gastric emptying in BN and AN using gamma cameraassessed gastric emptying

121

Colonic transit and motility using radiopaque markers and anorectal manometry

22 patients with AN (12 patients on self-selected diet; 10 patients on refeeding diet); 10 with BN (untreated comparison group); 10 controls (second untreated comparison group) 13 adult inpatients with AN and chronic constipation and 20 age-matched healthy female controls

135

Constipation using radiopaque markers and anorectal manometry

12 adult inpatient women with AN and constipation and 12 healthy female controls

136

Gastric dysmotility using ultrasonographic gastric emptying test, bowel symptom questionnaire, and psychological assessments

23 inpatients with AN (12 with binge/purge AN and 11 with restrictive AN) and 24 age-matched healthy controls

80% of patients reported one or more serious GI complaints; significant improvement noted in most symptoms after nutritional rehabilitation, but patients remained more symptomatic on GISS than comparison group 95% of patients had two or more GI complaints; patients had more GI symptoms (nausea, dysphagia, heartburn, borborygmi, belching, bloating, flatulence, abdominal pain, constipation, diarrhea) on admission compared with controls; patients remained more symptomatic than comparison group, even after nutritional rehabilitation, despite some improvement Upper GI symptoms were present in 78% of AN patients; gastric emptying was significantly slower in AN patients than in comparison group pretreatment; gastric emptying of liquids and solids was significantly faster after treatment completion (n = 11) (returned to normal in six, remained slow in two, and faster than normal in three); no change in GI symptoms and gastric emptying for three patients who did not gain weight Gastric emptying time of solid meal in AN subjects on self-selecting diet was significantly longer than that of controls; gastric emptying times of AN subjects on refeeding diet and BN subjects not significantly different from controls; weight gain among refed AN subjects had no significant effect on gastric emptying of solid meal Colonic transit significantly slower for patients within three wk of admission compared with those with >three wk of treatment; no significant differences between patients in hospital >three wk and controls; two of four patients with slow colonic transit who were restudied after six weeks in hospital achieved normal transit time; no significant differences between AN patients and controls with respect to anorectal manometry 8 of 12 AN patients had slow colonic transit times that normalized after four wk of refeeding; 5 of 12 patients had anorectal dysfunction that did not normalize with refeeding Significantly delayed gastric emptying and higher self-reported symptom scores in patients compared with controls pretreatment; no correlation found between gastric symptoms and emptying or psychological tests; self-reported GI symptoms improved significantly following treatment but remained mostly in pathologic range; significantly improved gastric emptying found in treatment completers with restricting but not binge/purge AN; gastric emptying time in treatment completers still longer than controls at baseline

reference

AN, anorexia nervosa; BN, bulimia nervosa; GI, gastrointestinal.

Chapter 8  Eating Disorders drawn. Existing data suggest that reestablishing regular food intake or weight gain will improve delayed gastric emptying and slowed colonic transit, although this may not be sufficient to restore normal GI function. Patients may resist active weight management or cessation of their disordered pattern of eating, despite their having a serious eating disorder and associated GI complications. This resistance may be exacerbated by early satiety, abdominal pain, bloating, or constipation, all of which may reinforce the patient’s excessive concern with weight or conviction that his or her diet needs to be further restricted. Management of symptoms is complicated further because subjective symptom reports correlate imperfectly with pathology and some of the complaints may be mediated by psychiatric symptoms or illness, including depression, anxiety, or distorted body image. Because refeeding and establishing normal and healthful dietary patterns are both treatment goals and likely to improve symptoms, careful nutritional rehabilitation is a reasonable and conservative initial step in management of suspected delayed gastric emptying and slow colonic transit for inpatients with AN or BN. Patients are likely to benefit from the support and reassurance that many of the GI symptoms commonly associated with eating disorders (e.g., bloating, constipation, nausea, vomiting, and diarrhea) will improve as eating and weight return to normal. Additional management strategies include dietary changes to reduce bloating (e.g., promoting smaller, more frequent meals, encouraging consumption of liquids earlier in the meal, and possibly providing a percentage of calories, no more than 25% to 50%, in liquid form initially).58,95 Various prokinetic agents have been used to manage delayed gastric emptying in AN, although metoclopramide is difficult to tolerate in frequent or high dosage, domperidone is not available in the United States except in compounding pharmacies, and cisapride is no longer being manufactured. Moreover, existing data do not support a recommendation for their use for gastric motility complaints in AN.211 Some clinicians have reservations about treating the constipation that follows laxative abuse with laxatives. Although it does not make sense to reproduce purging behavior using cathartics to treat constipation in patients with chronic laxative abuse, some patients will benefit from a thoughtful bowel regimen to reduce discomfort and bloating that otherwise might induce relapse of laxative abuse. Increasing fluid intake, dietary fiber, and possibly the addition of stool softeners and bulk-forming laxatives would be reasonable and conservative initial treatment. Osmotic laxatives initially may be necessary for symptom relief in some cases.239 Management of constipation may require anorectal retraining if a result of anorectal dysfunction.121 Some patients may benefit from symptomatic relief of gastroesophageal reflux or esophagitis with antacids or H2 antagonists; proton pump inhibitors may be required for relief of more severe symptoms.58 Although this may be appropriate clinically, the underlying cause and exacerbation of the GI complaint should be made clear to the patient and also actively addressed in psychotherapeutic treatment when related to the eating disorder. Mild elevation of serum aminotransferase levels secondary to malnutrition in AN likely will likely remit with weight restoration. Elevated serum levels of liver enzymes in severely ill patients may be an indication of refeeding syndrome or reflect AN-related hypoperfusion, and require emergent evaluation and intervention.58,128 Although many GI symptoms may be related to restrictive eating, binge pattern eating, or purging, some GI complaints

will require diagnostic evaluation. Anecdotal reports of catastrophic GI complications of the eating disorders, as well as primary GI illness that arises coincidentally with an eating disorder or mimics an eating disorder, suggest that complaints should be evaluated in their specific clinical context. Acute gastric dilation may be unsuspected in the absence of clinical history of binge eating.130 If acute gastric dilation is confirmed in the setting of refeeding or in the presence of a history of an eating disorder with binge eating, nasogastric decompression and fluid resuscitation are necessary. If these are not effective, laparotomy may be necessary.113,114,130 In addition, symptoms that persist after nutritional rehabilitation may require additional diagnostic evaluation. Eating disorders commonly are associated with GI symptoms and severe eating disorders can be associated with serious GI complications. Patients with eating disorders commonly present to primary and specialty care settings with GI symptoms or illness. In such cases, patients should be engaged in the concept of team management of their eating disorder and associated medical and nutritional complications at the same time as their GI complaint is addressed. More clinical trial data are needed to clarify treatment strategies for GI complaints associated with eating disorders. However, management of GI symptoms in patients with an eating disorder can be guided by several key considerations. Primary GI illness should be excluded, and the possibility considered that an eating disorder is obscuring or mimicking a primary illness. If GI symptoms appear to be associated with the eating disorder, nutritional rehabilitation in combination with psychotherapeutic care should be considered as an initial step. Nutritional rehabilitation often will require inpatient-level care and monitoring for serious potential complications, such as refeeding syndrome and acute gastric dilation. During treatment of an eating disorder, resistance to weight gain, to eating normally, and to cessation of bingeing and purging is common, so the possibility that body image or emotional symptoms mediate GI complaints should be considered in the treatment plan.

KEY REFERENCES

American Psychiatric Association. Treatment of patients with eating disorders. 3rd edition. American Psychiatric Association. Am J Psychiatry 2006; 163(Suppl):4-54. (Ref 77.) American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed, text revision. Arlington, Va: American Psychiatric Association; 2000. (Ref 8.) Brownley KA, Berkman ND, Sedway JA, et al. Binge eating disorder treatment: A systematic review of randomized controlled trials. Int J Eat Disord 2007; 40:337-48. (Ref 157.) Bulik CM, Berkman ND, Brownley KA, et al. Anorexia nervosa: A systematic review of randomized controlled trials. Int J Eat Disord 2007; 40:310-20. (Ref 155.) Emmanuel AV, Stern J, Treasure J, et al. Anorexia nervosa in a gastrointestinal practice. Eur J Gastroenterol Hepatol 2004; 16:1135-42. (Ref 56.) Hadley SJ, Walsh BT. Gastrointestinal disturbances in anorexia nervosa and bulimia nervosa. Curr Drug Targets CNS Neurol Disord 2003; 2:1-9. (Ref 211.) Harris EC, Barraclough B. Excess mortality of mental disorder. Br J Psychiatry 1998; 173:11-53. (Ref 1.) Hudson JI, Hiripi E, Pope HG Jr, Kessler RC. The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication. Biol Psychiatry 2007; 61:348-58. (Ref 3.) Keel PK. Purging disorder: Subthreshold variant or full-threshold eating disorder? Int J Eat Disord 2007; 40(Suppl):S89-94. (Ref 69.) National Institute for Clinical Excellence (NICE). Eating disorders—core interventions in the treatment and management of anorexia nervosa, bulimia nervosa and related eating disorders. NICE Clinical Guideline No 9. London: NICE; 2004. (Ref 154.)

137

138

Section II  Nutrition in Gastroenterology Niego SH, Kofman MD, Weiss JJ, Geliebter A. Binge eating in the bariatric surgery population: A review of the literature. Int J Eat Disord 2007; 40:349-59. (Ref 226.) Shapiro JR, Berkman ND, Brownley KA, et al. Bulimia nervosa treatment: A systematic review of randomized controlled trials. Int J Eat Disord 2007; 40:321-36. (Ref 156.) Striegel-Moore RH, Franko DL, May A, et al. Should night eating syndrome be included in the DSM? Int J Eat Disord 2006; 39:544-9. (Ref 65.)

Winstead NS, Willard SG. Gastrointestinal complaints in patients with eating disorders. J Clin Gastroenterol 2006; 40:678-82. (Ref 53.) World Health Organization. The ICD-10 classification of mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva: WHO; 1992. (Ref 74.) Zerbe K. Integrated treatment of eating disorders. New York: WW Norton; 2008. (Ref 158.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

9

Food Allergies Hugh A. Sampson

CHAPTER OUTLINE Background, Definitions, and Prevalence  139 Definitions  139 Prevalence  139 Pathogenesis  139 Clinical Features  142 IgE-Mediated Disorders  143

BACKGROUND, DEFINITIONS, AND PREVALENCE The first recorded account of food allergy was provided by Hippocrates, but it was not until 1921 that the classic experiment of Prausnitz initiated investigation on a scientific level and established the immunologic basis of allergic reactions.1 In this experiment, Prausnitz injected serum from his patient, Kustner, who was allergic to fish, into his own skin; the next day he injected fish extract into the same areas and into control sites. Local reactions proved sensitivity could be transferred by a factor in serum from an allergic to a nonallergic person. In 1950, Loveless demonstrated the inaccuracy of diagnosing food allergy by history in her report of the first blinded, placebo-controlled food trials in patients with milk allergy.2 In the following three decades, standardized protocols for the evaluation of food allergy were developed, and the double-blind, placebo-controlled oral food challenge (DBPCFC) emerged as the accepted standard for the diagnosis of food allergy.3

DEFINITIONS

Terminology used by investigators in the field of food allergy differs slightly in different parts of the world. The following represents current terminology in the United States. An adverse food reaction is a generic term indicating any untoward reaction occurring after the ingestion of a food or food additive and may be the result of toxic or nontoxic reactions. Toxic reactions will occur in any exposed individual upon ingestion of a sufficient dose. Nontoxic reactions depend on individual susceptibilities and may be immune-mediated (food allergy or food hypersensitivity) or non–immune-mediated (food intolerance). Food intolerances comprise most adverse food reactions and are cate­ gorized as enzymatic, pharmacologic, or idiopathic food intolerances. Secondary lactase deficiency, an enzymatic intolerance, affects the vast majority of adults, whereas most other enzyme deficiencies are rare inborn errors of metabolism and thus primarily affect infants and children. Pharmacologic food intolerances are present in individuals who are abnormally reactive to substances such as vasoactive amines, which are normally present in some foods (e.g., tyramine in aged cheeses). Confirmed adverse food reactions for which the mechanism is not known are generally

Mixed IgE- and Non–IgE-Mediated Disorders  143 Non–IgE-Mediated Disorders  145 Diagnosis  146 Treatment and Natural History  148

classified as idiopathic intolerances. Food allergies usually are characterized as IgE-mediated or non–IgE-mediated; the latter are presumed to be cell-mediated.

PREVALENCE

About 6% of young children and 3.5% of adults in the United States have food allergies.4 The prevalence of food allergies is greatest in the first few years of life and decreases over the first decade. The most common food allergens in young children include milk (2.5%), egg (1.5%), peanut (0.8%), wheat (∼0.4%), and soy (∼0.4%). Other than peanut, most childhood food allergies are outgrown by the end of the first decade. Almost all infants who develop cow’s milk allergy do so in the first year of life, with about two thirds experiencing IgE-mediated reactions and 35% going on to develop other food allergies.5 Peanut, tree nut, sesame, and seafood allergies tend to be lifelong, but about 20% of young children with peanut allergy develop clinical tolerance.6 Food allergies may persist after childhood into adulthood or develop in adulthood, with the most common food allergies in adults consisting of shellfish (2%), peanut (0.6%), tree nuts (0.4%), and fish (0.4%).7 About 5% of the U.S. population experiences allergic reactions to raw fruits and vegetables. Most of these reactions are caused by crossreactivity between homologous proteins in pollens, such as ragweed, and certain fruits and vegetables, such as melons and bananas (oral allergy syndrome), in adolescents and adults who have seasonal allergic rhinitis. The prevalence of food allergies appears to be increasing.8 Studies from the United States and United Kingdom have indicated that the prevalence of peanut allergy has doubled in young children during the past decade.9,10 In addition, children with atopic disorders have a higher prevalence of food allergies; for example, 35% to 40% of children with moderate to severe atopic dermatitis have IgE-mediated food allergy.11

PATHOGENESIS Gut-associated lymphoid tissue (GALT), a component of the mucosal immune system, lies juxtaposed to the external environment and acts to differentiate organisms and foreign proteins that are potentially harmful from those that are not. Unlike the systemic immune system, which recognizes

139

140

Section II  Nutrition in Gastroenterology Table 9-1  Physiologic and Immunologic Barriers of the Gastrointestinal Tract Physiologic Barriers Block penetration of ingested antigens Epithelial cells—one cell layer of columnar epithelium Glycocalyx—coating of complex glycoprotein and mucins that traps particles Intestinal microvillus membrane structure—prevents penetration Tight junctions joining adjacent enterocytes—prevent penetration even of small peptides Intestinal peristalsis—flushes trapped particles out in the stool Break down ingested antigens Salivary amylases and mastication Gastric acid and pepsins Pancreatic enzymes Intestinal enzymes Intestinal epithelial cell lysozyme activity Immunologic Barriers Block penetration of ingested antigens Antigen-specific SigA in intestinal lumen Clear antigens penetrating the gastrointestinal barrier Serum antigen-specific IgA and IgG Reticuloendothelial system SigA, secretory immunoglobulin A; IgG, immunoglobulin G.

relatively small quantities of antigen and mounts a brisk inflammatory response to neutralize them, the mucosal immune system regularly encounters enormous quantities of antigen and generally functions to suppress immune reactivity to harmless foreign antigens (e.g., food proteins, commensal organisms), only mounting a brisk protective response to dangerous pathogens when appropriate. A single-cell layer of columnar intestinal epithelial cells (IECs) separates the external environment from the loosely organized lymphoid tissue of the lamina propria. A highly efficient gastrointestinal mucosal barrier has evolved, however, that provides an enormous surface area for proc­ essing and absorbing ingested food and discharging waste products.12 This barrier uses physiologic and immunologic barriers to prevent the penetration of foreign antigens (Table 9-1). The physiologic barrier is composed of the following: epithelial cells, joined by tight junctions and covered with a thick mucus layer that traps particles, bacteria, and viruses; trefoil factors (TFFs, 7 to 12 kd), protease-resistant proteins secreted by mucus-secreting cells of the stomach (TFF1, TFF2) and intestine (TFF3) that help strengthen and promote restoration of the barrier; and luminal and brush border enzymes, bile salts, and extremes of pH, all serving to destroy pathogens and render antigens nonimmunogenic. Innate (natural killer [NK] cells, polymorphonuclear leukocytes, macrophages, epithelial cells, and Toll-like receptors) and adaptive immune (intraepithelial and lamina propria lymphocytes, Peyer’s patches, secretory immunoglobulin A [S-IgA] and cytokines) responses provide an active barrier to foreign antigens. Developmental imma­ turity of various components of the intestinal barrier and immune system reduces the efficiency of the infant mucosal barrier. For example, the activity of various enzymes is suboptimal in the newborn period and the S-IgA system is not fully mature until four years of age. This immature state of the mucosal barrier may play a role in the increased prevalence of gastrointestinal infections and food allergies seen in the first few years of life. In addition, studies have shown that alteration of the physiologic barrier function, such as gastric acidity, can lead to increased IgE sensitization in children and adults.13

Despite the evolution of this complex mucosal barrier, about 2% of ingested food antigens are absorbed and transported throughout the body in an immunologically intact form, even through the normal mature intestine.12 In an elegant series of experiments performed more than 75 years ago, Walzer and colleagues used sera from food-allergic patients to sensitize volunteers passively and demonstrated that immunologically intact antigens cross the mucosal barrier and disseminate rapidly throughout the body.14-16 Increased gastric acidity and the presence of food in the intestine decrease antigen absorption, whereas hypochlorhydria (e.g., H2 blocker– and proton pump inhibitor– induced) and ingestion of alcohol increase antigen absorption.15 The immunologically intact proteins that elude the intestinal barrier usually do not provoke adverse reaction, because most individuals have developed tolerance, but in a sensitized individual, allergic reactions will occur. Although more common in the developing GALT of young children, it is clear that cellular and IgE-mediated allergic responses to foods can develop at any age. The dominant response in GALT is suppression, or tolerance. The means whereby the immune system is educated to avoid sensitization to ingested food antigens is not well understood, but studies have suggested that antigenpresenting cells, especially IECs and various dendritic cells, and regulatory T cells play a central role.12 High-dose tolerance is caused by lymphocyte anergy, resulting from antigen T cell receptor ligation in the absence of costimulatory signals, whereas low-dose tolerance is mediated by regulatory T cells. Five different regulatory T cells have been identified in conjunction with intestinal immunity: (1) Th3 cells, a population of CD4+ cells that secrete transforming growth factor-β (TGF-β); (2) Th1 cells, CD4+ cells that secrete interleukin-10 (IL-10); (3) CD4+,CD25+ regulatory T cells; (4) CD8+ suppressor T cells; and (5) γδ T cells.17 IECs have been shown to be nonprofessional antigen-presenting cells (APCs) that can process luminal antigen and present it to CD4+ T cells on class II major histocompatibility complex (MHC) molecules. IECs, however, lack a second signal necessary to activate T cells, thus suggesting there is another mechanism whereby these cells can induce tolerance to food antigens. Extracellular proteins that are internalized by professional APCs (e.g., monocytes, macrophages, dendritic cells) into vesicles are processed and displayed by class II MHC molecules to CD4+ T cells, whereas proteins in the cytosol of nucleated cells are processed and displayed by class I MHC molecules to CD8+ T cells. IECs also can present lipid and glycolipid antigens to CD8+ suppressor T cells by a nonclassic—that is, non-MHC, class I molecule (CD1d)—and other novel membrane molecules that interact with CD8+ T cells (see Fig. 9-1). In addition, dendritic cells residing within the lamina propria and noninflammatory environment of Peyer’s patches express IL-10 and IL-4, which favor the generation of tolerance. It has been suggested that T cells primed in the local mucosal environment induce tolerance, whereas T cells primed in the mesenteric lymph nodes, either from antigen reaching the nodes in lymph or carried there by circulating dendritic cells, differentiate and travel to the mucosa, where they induce local immune responses.18 Recently, the unique role of the oral mucosa and its relation to tolerance induction via Langerhans cells has been increasingly appreciated.19 It is likely that the commensal bowel flora also play a role in shaping the mucosal immune response. It is estimated that there are 1012 to 1014 bacteria/g of colonic tissue, which means that there are more bacteria in the colon than cells in the body.12 Bowel flora is largely established in the first 24 hours after birth, is dependent on maternal flora, genet-

Chapter 9  Food Allergies IgE-associated

Non–IgE-associated Soluble proteins

Particulate proteins

Glycocalyx

M cell

IEC IEC IEC IEC IEL

IgE receptor

Mast cell

T Mφ Peyer's patch

LPL

B

Lamina propria

Th3 cell

Mφ

Histamine B cell

IgE

B

Th cell

IgG

? TNF-α TNF-α IL-5 IL-4

ics, and local environment, and is relatively stable throughout life. The importance of bowel flora in the development of oral tolerance induction is suggested by the fact that mice raised in a germ-free environment from birth fail to develop normal tolerance.20 Studies in which lactating mothers and their offspring were fed Lactobacillus suggest that probiotics may be beneficial in preventing some atopic disorders, such as eczema,21 but results from other studies are not consistent. IECs also may play a central regulatory role in determining the rate and pattern of uptake of ingested antigens. Studies in sensitized rats have indicated that intestinal antigen transport proceeds in two phases.22 In the first phase, transepithelial transport occurs via endosomes, is antigen-specific and mast cell–independent, and occurs 10 times faster in sensitized rats compared with nonsen­ sitized control animals. Antigen-specific IgE antibodies bound to the mucosal surface of IECs via FcεRII are responsible for this accelerated allergen entry.23 In the second phase, paracellular transport predominates. Loosening of the tight junctions occurs as a result of factors released by mast cells activated in the first phase. Whereas the first antigen-specific pathway involves antibody, the second nonspecific pathway most likely involves cytokines. Consistent with this concept, IECs express receptors for a number of cytokines (IL-1, IL-2, IL-6, IL-10, IL-12, IL-15, granulocyte-monocyte colony-stimulating factor [GM-CSF], and interferon-γ [IFN-γ]), and have been shown to be functionally altered by exposure to these cytokines. Although the development and mechanistic features of non–IgE-mediated food-allergic responses are poorly understood, the development of IgE-mediated responses has been well characterized. Sensitivity to allergens (generally glycoproteins) is the result of a series of molecular and cellular interactions involving APCs, T cells, and B cells.24 APCs present small peptide fragments (T cell epitopes) in conjunction with MHC class II molecules to T cells. T cells bearing the appropriate complementary T cell receptor (TCR) will bind to the peptide-MHC complex. This interactive first signal leads to T cell proliferation and cytokine generation and the generation of a second signal (e.g., IL-4) that promotes an IgE response (Th2-like cell activation).

Figure 9-1.  Immunopathogenesis of food allergies. Massive quan­ tities of food proteins are processed in the intestinal tract to non­ immunogenic peptides and amino acids. As described in the text, however, a small amount of immunogenic protein passes through the intestinal barrier. Intestinal epithelial cells (IECs) normally process soluble proteins for presentation to appropriate T-helper (Th1 or Th2) and regulatory (Th3) T-cells. Protective IgA and IgG antibody responses are generated, and systemic T-cell responses are down-regulated. In IgE-associated disorders, food-specific IgEproducing B cells are activated. IgE antibodies adhere to the surface of mast cells, and release histamine and other mediators if surfacebound IgE encounters the food antigen. IgE also binds to FcεR on intestinal IECs, thereby expediting antigen transfer through IECs. In non–IgE-mediated disorders, antigen-presenting cells and/or T cells are activated to secrete TNF-α (dietary protein–induced enterocolitis syndrome) or IL-4 and/or IL-5 (allergic eosinophilic gastroenteritis). M cells overlying Peyer’s patches are believed to play a major role in processing particulate protein and pathogens. FcεR, Fc epsilon receptor; Ig, immunoglobulin; IL-4, -5, interleukin-4, -5; LPL, lamina propria lymphocyte; Mf, macrophage; TNF-α, tumor necrosis factor-α.

These cells and their products, in turn, interact with B cells bearing appropriate antigen-specific receptors, leading to isotype switching and the generation of antigen-specific IgE. At all stages, a number of specific cytokines are secreted that modulate the cell interactions. The antigen-specific IgE then binds to surface receptors of mast cells, basophils, macrophages, and other APCs, arming the immune system for an allergic reaction upon the next encounter with the specific antigen. A breakdown in mucosal integrity, caused by infection or other inflammatory processes, leads to increased intestinal permeability, which results in antigens bypassing the normal tolerogenic presentation by IECs and, under some circumstances, leads to allergic sensitization. Oral tolerance of humoral and cellular immunity has been demonstrated in rodents and humans. Feeding of keyhole limpet hemocyanin to human volunteers resulted in T cell tolerance but priming of B cells at both mucosal and systemic sites.25 The failure of human infants to develop oral tolerance or the breakdown of oral tolerance in older individuals results in the development of food allergy. Young infants are more prone to develop food-allergic reactions because of the immaturity of their immunologic system and, to some extent, the gastrointestinal (GI) tract (see Table 9-1). Exclusive breast-feeding promotes the development of oral tolerance and may prevent some food allergy and atopic dermatitis.26,27 The protective effect of breast milk appears to be the result of several factors, including decreased content of foreign proteins, the presence of S-IgA (which provides passive protection against foreign protein and pathogens), and the presence of soluble factors (e.g., prolactin), which may induce earlier maturation of the intestinal barrier and the infant’s immune response. The antibacterial activity of breast milk is well established, but the ability of breast milk S-IgA to prevent food antigen penetration is less clear. Low concentrations of food-specific IgG, IgM, and IgA antibodies commonly are found in the serum of normal persons. Food protein–specific IgG antibodies tend to rise in the first months following the introduction of a food and then generally decline, even though the food protein continues to be ingested. Persons with various inflammatory bowel disorders (e.g., celiac disease, food allergy) frequently have high levels of food-specific IgG and IgM antibodies,

141

142

Section II  Nutrition in Gastroenterology although there is no evidence that these antibodies are pathogenic. Increased lymphocyte proliferation or IL-2 production following food antigen stimulation in vitro is frequently seen in patients with inflammatory bowel disorders, but it also occurs in normal persons.28 Antigen-specific T-cell proliferation in vitro alone does not represent a marker of immunopathogenicity, but simply reflects response to antigen exposure. In genetically predisposed individuals, as noted, antigen presentation leads to excessive Th2 responsiveness (lymphocytes secreting IL-4, IL-5, IL-10, and IL-13), resulting in increased IgE production and expression of Fc epsilon (FcεI) receptors on a variety of cells.24 These IgE antibodies bind high-affinity FcεI receptors on mast cells, basophils, and dendritic cells, as well as low-affinity FcεII (CD23) receptors on macrophages, monocytes, lymphocytes, eosinophils, and platelets. When food allergens penetrate mucosal barriers and reach IgE antibodies bound to mast cells or basophils, the cells are activated and mediators (e.g., histamine, prostaglandins, and leukotrienes) are released that induce vasodilation, smooth muscle contraction, and mucus secretion, and lead to symptoms of immediate hypersensitivity. These activated mast cells also may release a variety of cytokines (e.g., IL-4, IL-5, IL-6, tumor necrosis factor-α [TNF-α], platelet-activating factor), which may induce the IgEmediated late-phase inflammation. Various symptoms have been associated with IgE-mediated allergic reactions— generalized: shock; cutaneous: urticaria, angioedema, and a pruritic morbilliform rash; oral and GI: lip, tongue, and palatal pruritus and swelling, vomiting, and diarrhea; and upper and lower respiratory: ocular pruritus and tearing associated with nasal congestion, laryngeal edema, and wheezing. A rise in the plasma histamine level has been associated with the development of these symptoms after blinded food challenges.29 In IgE-mediated GI reactions, endoscopic observation has revealed local vasodilation, edema, mucus secretion, and petechial hemorrhage.30 Cellmediated hypersensitivity reactions are believed responsible for allergic eosinophilic esophagitis and gastroenteritis (AEE, AEG). It is believed that activated T cells secrete IL-5 and other cytokines, attracting eosinophils and inducing the inflammatory response that causes the delayed onset of symptoms.31 Expansion studies of T cells from biopsy specimens of milk-induced AEG patients have revealed large numbers of CD4+ Th2 cells.32 In summary, the GI tract processes ingested food into a form that can be absorbed and used for energy and cell growth. During this process, nonimmunologic and immunologic mechanisms help destroy or block foreign antigens (e.g., bacteria, viruses, parasites, food proteins) from entering the body proper. Despite this elegant barrier, antigenically intact food proteins enter the circulation, but in the normal host are largely ignored by the immune system, which has become “tolerized” to these nonpathogenic substances.

CLINICAL FEATURES As depicted in Table 9-2, a number of GI food hypersensi­ tivity disorders have been described. Clinically, these disorders are generally divided into two main categories: IgE-Th2–mediated and non–IgE (cell)-mediated hypersen­ sitivities. There are a number of other disorders, however, that may result in symptoms similar to food-allergic reactions; these must be excluded during the evaluation (Table 9-3).

Table 9-2  Gastrointestinal Food Hypersensitivities IgE-Mediated Food Hypersensitivities Gastrointestinal allergy Infantile colic (minor subset) Oral allergy syndrome Mixed IgE- and Non–IgE-Mediated Hypersensitivities Allergic eosinophilic esophagitis Allergic eosinophilic gastritis Allergic eosinophilic gastroenteritis Allergic eosinophilic proctocolitis Non–IgE-Mediated Food Hypersensitivities Dietary protein-induced enterocolitis syndrome Dietary protein-induced enteropathy Celiac disease Dermatitis herpetiformis Mechanism Unknown Cow’s milk–induced occult gastrointestinal blood loss and iron deficiency anemia of infancy Gastroesophageal reflux disease Infantile colic (subset) Inflammatory bowel disease (?)

Table 9-3  Disorders That Must Be Differentiated from Food Hypersensitivities Food Intolerances Enterotoxigenic bacteria Vibrio cholerae, toxigenic Escherichia coli, Clostridium difficile Metabolic disorders Acrodermatitis enteropathica Hypo- or abetalipoproteinemia Primary carbohydrate malabsorption: lactase deficiency, sucrase deficiency Transient fructose and/or sorbitol malabsorption Postinfection malabsorption (secondary disaccharidase deficiency, villus atrophy, bile salt deconjugation) Bacterial: Shigella, Clostridium difficile Parasitic: Giardia, Cryptosporidium Viral: Rotavirus Anatomic Abnormalities Hirschsprung’s disease (especially with enterocolitis) Ileal stenosis Intestinal lymphangiectasia Short bowel syndrome Other Disorders Chronic nonspecific diarrhea of infancy Cystic fibrosis Inflammatory bowel disease Tumors Neuroblastoma Zollinger-Ellison syndrome (gastrinoma)

Long before IgE antibodies were identified, studies of food hypersensitivity focused on radiologic changes associated with immediate hypersensitivity reactions. In one of the first of these reports, hypertonicity of the transverse and pelvic colon and hypotonicity of the cecum and ascending colon were noted following wheat feeding to an allergic patient.33 In a later report, gastric retention, hypermotility of the small intestine, and colonic spasm were observed in 4 patients studied after administering barium containing specific food allergens.34 In a third study, fluoroscopy was used to compare the effect of barium contrast with and without food allergens in 12 food-allergic children35; gastric hypotonia and retention of the allergen test meal, prominent pylorospasm, and increased or decreased peristaltic activity of the intestines were noted.

Chapter 9  Food Allergies In the late 1930s, the rigid gastroscope was used to observe reactions in the stomachs of allergic patients. One study evaluated patients with GI food allergy or wheezing exacerbated by food ingestion and control subjects.36 Thirty minutes after a food allergen was placed on the gastric mucosa, patients with GI food allergy had markedly hyperemic and edematous patches of thick gray mucus and scattered petechiae at these sites, similar to those reported earlier by Walzer, in passively sensitized intestinal mucosal sites.15 Only mild hyperemia of the gastric mucosa was noted in patients with wheezing provoked by food ingestion. Subsequent studies confirmed these earlier observations and established an IgE-mediated mechanism for the reactions30; they demonstrated food-specific IgE antibodies and increased numbers of intestinal mast cells prior to challenge in food-allergic patients compared with normal controls, and significant decreases in stainable mast cells and tissue histamine content following a positive food challenge.

IgE-MEDIATED DISORDERS

The IgE-mediated food-induced GI allergic responses comprise two major symptom complexes: pollen-food allergy (oral allergy) syndrome and gastrointestinal allergy. These disorders are distinguished by their rapid onset, usually within minutes to an hour of ingesting the offending food. In addition, simple laboratory tests that detect food-specific IgE antibodies, such as prick skin tests and in vitro tests of serum food-specific IgE antibodies (e.g., ImmunoCAP; Phadia, Portage, Mich) often are useful in determining which foods are responsible for the patient’s symptoms.

Pollen-Food Allergy Syndrome

The pollen-food allergy syndrome (oral allergy syndrome) is a form of immediate contact hypersensitivity confined predominantly to the oropharynx and rarely involving other target organs.37 Symptoms include the rapid onset of pruritus and angioedema of the lips, tongue, palate and throat, generally followed by a rapid resolution of symptoms, and most commonly associated with the ingestion of various fresh (uncooked) fruits and vegetables. Symptoms result from local IgE-mediated reactions to conserved homologous proteins (sequences of amino acids in peptide backbones shared by plant pollens and fruit and vegetable proteins that remain unchanged through evolution) that are heat-labile (i.e., readily destroyed by cooking) and shared by certain fruits, vegetables, and some plant pollens.38 Patients with seasonal allergic rhinitis (hay fever) secondary to birch or ragweed pollen sensitivity often are afflicted with this syndrome. For example, in up to 50% of patients with ragweed-induced allergic rhinitis, ingestion of melons (e.g., watermelon, cantaloupe, honeydew) and bananas will provoke oral symptoms,39-41 whereas in birch pollen– allergic patients, symptoms may develop following the ingestion of raw potatoes, carrots, celery, apples, hazelnuts, and kiwi. Diagnosis is based on classic history and positive prick skin tests (e.g., prick and prick—pricking the fresh fruit or vegetable with a needle and then pricking the skin of the patient) with the implicated fresh fruits or vegetables.42

Gastrointestinal Allergy

Gastrointestinal allergy is a relatively common form of IgEmediated hypersensitivity, which generally accompanies allergic manifestations in other target organs (e.g., skin, airway) and results in a variety of symptoms.4 Symptoms typically develop within minutes to two hours of consum-

ing a food and consist of nausea, abdominal pain, cramps, vomiting, and/or diarrhea. In some infants, frequent ingestion of a food allergen appears to induce partial desensitization of gastrointestinal mast cells resulting in a subclinical reaction, with the only symptom reported being poor appetite and periodic abdominal pain. Diagnosis is established by clinical history, evidence of food-specific IgE antibodies (positive skin prick tests or serum food-specific IgE anti­ bodies), resolution of symptoms following complete elimination of the suspected food, and recurrence of symptoms following oral food challenges. GI allergy is common in IgEmediated food allergies, with more than 50% of children experiencing abdominal symptoms during double-blind, placebo-controlled food challenges.28

Infantile Colic

Infantile colic is an ill-defined syndrome of paroxysmal fussiness characterized by inconsolable agonized crying, drawing up of the legs, abdominal distention, and excessive gas. It generally develops in the first two to four weeks of life and persists through the third to fourth months of life.43 Various psychosocial and dietary factors have been implicated in the cause of infantile colic, but trials in bottle-fed and breast-fed infants have suggested that IgE-mediated hypersensitivity occasionally may be a pathogenic factor, possibly in 10% to 15% of colicky infants. Diagnosis of food-induced colic is established by the implementation of several brief trials of hypoallergenic formula. In infants with food allergen–induced colic, symptoms are generally shortlived, so prolonged restricted diets are generally unnecessary. Periodic rechallenges should be done every three to four months to determine when eliminated foods can be returned to the infant’s diet.

MIXED IgE- AND NON–IgE-MEDIATED DISORDERS Allergic eosinophilic esophagitis, gastroenteritis, and proctocolitis (AEE, AEG, AEP) may be caused by IgE- and/or non–IgE-mediated food allergies and are characterized by eosinophilic infiltration of the esophagus, stomach, and/or intestinal walls with peripheral eosinophilia in up to 50% of patients (see Chapter 27 for a more complete discussion).31,39-41 In the esophagus, basal hyperplasia and papillary lengthening are seen. The eosinophilic infiltrate may involve the mucosal, muscular, and/or serosal layers of the stomach or small intestine. Eosinophilic invasion of the muscular layer leads to thickening and rigidity of the stomach and small intestine, which may manifest as obstruction, whereas infiltration of the serosa commonly results in eosinophilic ascites. In most children with AEE-AEG, foodinduced IgE- and non–IgE-mediated reactions have been implicated in pathogenesis.44 Patients with IgE-mediated food-induced symptoms generally have atopic disease (atopic dermatitis, allergic rhinitis, and/or asthma), elevated serum IgE concentrations, positive skin prick tests to various foods and inhalants, peripheral blood eosinophilia, iron deficiency anemia, and hypoalbuminemia.

Allergic Eosinophilic Esophagitis

AEE manifests predominantly in young children, especially boys, with reflux or vomiting, irritability, food refusal, early satiety, and failure to thrive,45,46 whereas adults are more likely to present with reflux, epigastric or chest pain, dysphagia, and food impaction.43,47 Foodinduced AEE was first demonstrated in a group of 10 children with postprandial abdominal pain, early satiety or

143

144

Section II  Nutrition in Gastroenterology food refusal, vomiting or retching, failure to thrive, and refractoriness to standard medical therapy (4 of 10 had undergone Nissen fundoplication).44 Following six to eight weeks of an amino acid-based formula (Neocate) plus corn and apples, symptoms completely resolved in eight patients and were markedly improved in two others. Esophageal bio­psies revealed a marked reduction or clearing of the eosinophilic infiltrate and significant improvements in the basal zone hyperplasia and length of the vascular papillae. Symptoms could be reproduced with the introduction of certain foods. Some patients appear to have an association of pulmonary and esophageal inflammation, and some report seasonal esophageal symptoms.31,48 AEE appears to have increased in prevalence over the past decade, an observation some authors believe may be explained by the increased early use of antacids and pro­kinetic agents in young infants with symptoms of reflux. Because murine models of food-induced anaphylaxis require the use of antacids for sensitization,49,50 it is thought that antireflux medications may further compromise the young infant’s intestinal barrier function. In a cohort study of 152 adults using H2 blockers or proton pump inhibitors for three months, 10% of patients experienced an increase in foodspecific IgE and 15% developed de novo IgE to specific foods.51 Untersmayr and colleagues have noted that the use of antacid medications can lead to food sensitivity in children and adults.13 Diagnosis of AEE is based on a suggestive history, the demonstration of an eosinophilic infiltrate in the esophageal mucosa (>20 eosinophils/high-power field [×40]), and the absence of GERD, as evidenced by a normal pH monitoring study of the distal esophagus or lack of response to highdose proton pump inhibitors.52 Multiple biopsies are necessary because of the potential patchiness of the lesions; a single esophageal biopsy specimen has a sensitivity of 55%, whereas taking five biopsy specimens increases sensitivity to 100%.53 Esophagoscopy may reveal mucosal rings, furrowing, ulcerations, whitish papules (which represent eosinophilic abscesses), or strictures, but endoscopic findings are normal in at least one third of patients with AEE. There is some evidence suggesting that atopy patch testing may be useful in identifying foods responsible for the allergic inflammation, but further studies are necessary to confirm these early reports.54 Elimination of suspect foods for six to ten weeks should lead to resolution and normalization of esophageal histology, although clinical symptoms should improve substantially in three to six weeks.41,55,56 Challenges consist of reintroducing the suspected food allergen and evaluating for recurrence of symptoms and/or eosinophilic infiltrate on biopsy. If food allergens are not identified as provoking agents, oral glucocorticoids generally are required to alleviate symptoms. Although symptoms usually respond to glucocorticoid therapy, recurrence of symptoms is frequent when steroids are discontinued.57 Topical glucocorticoid therapy with swallowed fluticasone spray or viscous budesonide has been shown to induce remission in 50% to 80% of patients, but esophageal candidiasis may occur in up to 20% of patients using this form of treatment.58,59 If exacerbations recur, a daily regimen of low-dose prednisone or prednisolone or prednisone every other day may be successful in suppressing symptoms.60 Recent evidence suggests that anti–IL-5 may be useful in this disorder.61

Allergic Eosinophilic Gastroenteritis

AEG manifests with abdominal pain, nausea, vomiting, diarrhea, and weight loss.62 Generalized edema secondary to hypoalbuminemia may occur in some infants and young

children with marked protein-losing enteropathy, often in the presence of minimal gastrointestinal symptoms (e.g., occasional vomiting and diarrhea).63 Rarely AEG may manifest as pyloric stenosis in infants with outlet obstruction and postprandial projectile emesis.64 The immunopathogenesis of AEG is not known, but is believed to involve primarily cell-mediated mechanisms. A subset of patients have exacerbations of symptoms following the ingestion of food to which they have specific IgE antibodies, but most reactions do not appear to involve this mechanism. Peripheral blood T cells from all AEG patients evaluated have been shown to secrete excessive amounts of Th2 cytokines, IL-4, and IL-5 in vitro, compared with normal controls,65 and T cells expanded from duodenal biopsies of AEG patients express Th2 cytokines in vitro following antigen stimulation.32 The diagnosis of AEG is dependent on a suggestive history, gastrointestinal biopsy specimens demonstrating a prominent eosinophilic infiltration, and peripheral eosinophilia, which occurs in about 50% of patients. Lesions are not uniform; therefore, multiple biopsies are often necessary.62 Allergy skin testing may be helpful in some cases to identify causative foods, but often a therapeutic trial of an elemental diet for six to 10 weeks is necessary to determine whether food allergy is provoking the disorder. In a study of children with AEG and protein-losing enteropathy, institution of an amino acid-based formula therapy brought about resolution of symptoms and normalization of intestinal histology.63 As with AEE, if no sensitization is found, a trial of glucocorticoids is recommended, although relapses frequently occur when they are discontinued. The long-term prognosis of this disorder is not well characterized. In one series of children with AEG and protein-losing enteropathy, follow-up for 2.5 to 5.5 years revealed persistence of foodresponsive disease.

Allergic Eosinophilic Proctocolitis

AEP generally presents in the first few months of life and is most often secondary to cow’s milk or soy protein hypersensitivity. Over half of reported cases now occur in breastfed infants because of food antigens that are passed in maternal breast milk.66,67 Affected infants usually appear healthy, often have normally formed stools, and generally are evaluated because of the presence of gross or occult blood in their stools. Blood loss typically is minor but occasionally can produce anemia. Lesions generally are confined to the distal large bowel and consist of mucosal edema, with infiltration of eosinophils in the epithelium and lamina propria. In severe cases with crypt destruction, neutrophils also are prominent. The immunologic mechanism underlying this disorder is not known, but is believed to involve a cell-mediated reaction. There is no evidence that IgE antibodies are involved in this disorder and therefore skin prick testing or evaluation of food-specific IgE antibodies is not helpful. Diagnosis can be established when elimination of the responsible allergen leads to resolution of hematochezia, generally with dramatic improvement within 72 hours of appropriate food allergen elimination. Complete clearing and resolution of mucosal lesions may take up to one month. Reintroduction of the allergen leads to recurrence of symptoms within several hours to days. Sigmoidoscopic findings vary and range from areas of patchy mucosal injection to severe friability, with small aphthoid ulcerations and bleeding. Colonic biopsy reveals a prominent eosinophilic infiltrate in the crypt epithelia and lamina propria. Children with cow’s milk and soy protein–induced proctocolitis usually outgrow their protein sensitivity (i.e., become clinically tolerant

Chapter 9  Food Allergies within six months to two years of allergen avoidance), but occasionally refractory cases are seen.

NON–IgE-MEDIATED DISORDERS

Some gastrointestinal food-allergic disorders are clearly not IgE-mediated, and are believed to be the result of various cell-mediated mechanisms. Consequently, tests for evidence of food-specific IgE antibodies are of no value to identify the responsible food in these disorders. These non– IgE-mediated hypersensitivities are believed to result from different abnormal antigen processing and/or cell-mediated mechanisms, and may be divided into the following syndromes: dietary protein-induced enterocolitis and dietary protein-induced enteropathy.68

Dietary Protein-Induced Enterocolitis Syndrome

Dietary protein-induced enterocolitis syndrome is a dis­ order most commonly seen in young infants, presenting between one week and three months of age, with protracted vomiting and diarrhea that not infrequently results in dehydration.69,70 About one third of infants with severe diarrhea develop acidosis and transient methemoglobinemia. Cow’s milk and/or soy protein most often are responsible, but enterocolitis secondary to egg, wheat, rice, oat, peanut, nuts, chicken, turkey, and fish sensitivities has also been reported in older individuals.71 Breast-fed babies almost never develop symptoms while breast-feeding, but may be sen­ sitized through food proteins passed in the breast milk and experience a reaction on the first few feedings of the whole food.72,73 Similar reactions to seafood (e.g., shrimp, crab, lobster), with symptoms developing about two to four hours following ingestion, often are reported in adults. Stools frequently contain occult blood, polymorpho­ nuclear neutrophils, and eosinophils. Jejunal biopsies reveal flattened villi, edema, and increased numbers of lymphocytes, eosinophils, and mast cells. Food challenges generally result in vomiting and diarrhea within one to three hours, and result in hypotension in about 15% of cases. The immunopathogenesis of this syndrome remains unknown. Some studies suggest that food antigen-induced secretion of TNF-α from local mononuclear cells (e.g., macrophages, dendritic cells) may account for the reaction.74 Other studies indicate that the disorder may be caused by lower expression of type 1 TGF-β receptors than type 2 receptors, suggesting differential contributions of each receptor to the diverse biological activities of TGF-β in the intestinal epithelium.75 Some studies have suggested that atopy patch testing with the suspected food may be useful in distinguishing which children will develop symptoms following ingestion, but most such evidence is not convincing.76 Diagnosis can be established when elimination of the responsible allergen leads to resolution of symptoms within 72 hours and oral challenge provokes symptoms.72 Secondary disaccharidase deficiency may persist longer, however, and may result in ongoing diarrhea for up to two weeks. Oral food challenges consist of administering 0.3 to 0.6 g/kg body weight of the suspected protein allergen while monitoring the peripheral blood white cell count. Vomiting generally develops within one to four hours of administering the challenge food, whereas diarrhea or loose stools often develop after four to eight hours. In conjunction with a positive food challenge, the absolute neutrophil count in the peripheral blood will increase at least 3500 cells/mm3 within four to six hours of developing symptoms, and neutrophils and eosinophils may be found in the stools. About 15% of food antigen challenges lead to profuse vomiting, dehydration, and hypotension, so they must be performed under medical supervision.

Dietary Protein-Induced Enteropathy

Dietary protein-induced enteropathy (excluding celiac disease) frequently manifests in the first several months of life with diarrhea (mild to moderate steatorrhea in about 80%) and poor weight gain.67,77 Symptoms include protracted diarrhea, vomiting in up to two thirds of patients, failure to thrive, and malabsorption, demonstrated by the presence of reducing substances in the stools, increased fecal fat, and abnormal d-xylose absorption. Cow’s milk sensitivity is the most frequent cause of this syndrome, but it also has been associated with sensitivities to soy, egg, wheat, rice, chicken, and fish. The diagnosis is established by identifying and excluding the responsible allergen from the diet, which should result in resolution of symptoms within several days to weeks. On endoscopy, patchy villus atrophy is evident and biopsy reveals a prominent mononuclear round cell infiltrate and a small number of eosinophils, similar to celiac disease, but generally much less extensive. Colitic features such as mucus and gross or microscopic hematochezia usually are absent, but anemia occurs in about 40% of affected infants and protein loss occurs in most. Complete resolution of the intestinal lesions may require 6 to 18 months of allergen avoidance. Unlike celiac disease, loss of protein sensitivity and clinical reactivity frequently occurs, but the natural history of this disorder has not been well studied.

Celiac Disease

Celiac disease (CD) is a more extensive enteropathy leading to malabsorption (see details in Chapter 104). Total villus atrophy and extensive cellular infiltrate are associated with sensitivity to gliadin, the alcohol-soluble portion of gluten found in wheat, rye, and barley. CD is strongly associated with HLA-DQ2 (α1*0501, β1*0201), which is present in more than 90% of CD patients.78 The incidence of CD has been reported as 1 in 250 in the United States. The striking increase in CD in Sweden compared with genetically similar Denmark,79 and the variation in prevalence associated with changes in patterns of gluten feeding in Sweden,80 strongly implicate environmental factors (e.g., feeding practices) in the cause of this disorder.81 The intestinal inflammation in CD is precipitated by exposure to gliadin and is associated with increased mucosal activity of tissue transglutaminase (tTG), which deamidates gliadin in an ordered and specific fashion, creating epitopes that bind efficiently to DQ2 and are recognized by T cells.82 Initial symptoms often include diarrhea or frank steatorrhea, abdominal distention and flatulence, weight loss, and occasionally nausea and vomiting. Oral ulcers and other extraintestinal symptoms secondary to malabsorption are not uncommon. Villus atrophy of the small bowel is a characteristic feature of CD patients who are ingesting gluten. IgA antibodies to gluten are present in more than 80% of adults and children with untreated CD.83 In addition, patients generally have increased IgG antibodies to a variety of foods, presumably the result of increased food antigen absorption. Diagnosis has been dependent on demonstrating biopsy evidence of villus atrophy and an inflammatory infiltrate, resolution of biopsy findings after six to 12 weeks of gluten elimination, and recurrence of biopsy changes following gluten challenge. Revised diagnostic criteria have been proposed, however, that require greater dependency on serologic studies. Quantitation of IgA tTG antibodies may be used for screening in children older than two years. Diagnosis of CD, however, requires an intestinal biopsy showing clear-cut evidence of villus atrophy plus resolution of symptoms on a gluten-free diet, with serologic follow-up showing disappearance of the

145

146

Section II  Nutrition in Gastroenterology antibodies to confirm the diagnosis further.84,85 Once the diagnosis of CD is established, lifelong elimination of gluten-containing foods is necessary to control symptoms and possibly to avoid the increased risk of gastrointestinal malignancy.86

Dermatitis Herpetiformis

Dermatitis herpetiformis (DH) is a chronic blistering skin disorder associated with a gluten-sensitive enteropathy. It is characterized by a chronic, intensely pruritic, papulo­ vesicular rash symmetrically distributed over the extensor surfaces and buttocks.87,88 The histology of the intestinal lesion is almost identical to that seen in CD, although villus atrophy and the inflammatory infiltrate are generally milder and T-cell lines isolated from intestinal biopsy specimens of DH patients produce significantly more IL-4 than T cell lines isolated from CD patients.89 Although many patients have minimal or no gastrointestinal complaints, biopsy of the small bowel generally confirms intestinal involvement. Elimination of gluten from the diet generally leads to resolution of skin symptoms and normalization of intestinal findings over several months. Administration of sulfones, the mainstay of therapy, leads to rapid resolution of skin symptoms, but has almost no effect on intestinal symptoms.

Other Gastrointestinal Disorders

Several other disorders have been suggested to be caused by food protein hypersensitivity. Ingestion of pasteurized whole cow’s milk by infants younger than six months may lead to occult GI blood loss and occasionally to iron deficiency anemia.90,91 Substitution of heat-processed infant formula (including cow’s milk–derived formulas) for whole cow’s milk generally leads to resolution of symptoms within three days. Gastroesophageal reflux (GER) in young infants may be the result of food-induced AEE. In a study of 204 infants younger than one year with GER (diagnosed with a 24-hour esophageal pH test and esophageal biopsy),85 42% were diagnosed with cow’s milk–induced reflux by blinded milk challenges. These infants experienced resolution of GER and normalization of pH studies once cow’s milk was eliminated from the diet.92 Constipation also has been reported to be caused by milk allergy,93 although the underlying mechanism is not clear. Circumstantial evidence suggests a possible role of food allergy in inflammatory bowel disease (Crohn’s disease and ulcerative colitis), but convincing evidence of an immunopathogenic role remains to be established.

DIAGNOSIS The diagnosis of food allergy is a clinical exercise involving a careful history, physical examination, and selective laboratory studies. Various tests are used for the evaluation of food hypersensitivity (see Sicherer and Sampson4 and Sampson94). In some cases, the medical history may be useful in diagnosing food allergy (e.g., acute anaphylaxis after the isolated ingestion of peanuts). Fewer than 50% of reported food-allergic reactions, however, can be verified by a double-blind, placebo-controlled food challenge. Information useful in establishing that a food-allergic reaction has occurred and in constructing an appropriate oral food challenge includes the following: (1) food presumed to have provoked the reaction; (2) quantity of the suspected food ingested; (3) length of time between ingestion and development of symptoms; (4) type of symptoms provoked; and (5) whether similar symptoms developed on other occasions when the food was eaten. Although any food may induce an allergic reaction, a few foods are responsible for the vast majority of reactions (Table 9-4). Figure 9-2 depicts a standard approach for evaluating and managing adverse food reactions. If an IgE-mediated disorder is suspected, selected skin prick tests or quantification of food-specific IgE antibodies (e.g., ImmunoCAP) followed by an appropriate exclusion diet and blinded food challenge are warranted. If a non–IgE-mediated GI hypersensitivity disorder is suspected, laboratory and endoscopic studies (with or without oral food challenges) are required to arrive at the correct diagnosis (see earlier). Table 9-5 compares the Table 9-4  Foods Responsible for Most Food Hypersensitivity Disorders IgE-MEDIATED FOOD HYPERSENSITIVITIES*

NON–IgE-MEDIATED FOOD HYPERSENSITIVITIES†

Milk Egg Peanuts Shellfish Tree nuts Sesame Fish Soy Wheat

Barley Beef, lamb Egg Fish Milk Shellfish Soy Wheat White potato

*Listed in order of overall prevalence. † Listed alphabetically.

Table 9-5  Differentiating Non–IgE-Mediated Gastrointestinal Food Hypersensitivities

PARAMETER

AEE, AEG

DIETARY PROTEININDUCED ENTEROPATHY

Age of onset Duration Food proteins implicated

1 mo and older ≥1 yr Cow’s milk, egg, soy, wheat, barley

1-18 mo 18-36 mo Cow’s milk, soy, wheat, barley

2 wk-9 mo 9-36 mo Cow’s milk, soy

1 wk-3 mo 6-18 mo Cow’s milk, soy, breast milk*

Moderate to severe Prominent† Minimal Minimal to moderate

Moderate Variable Moderate Moderate

Moderate Prominent Severe Moderate

None None Rare Moderate to severe

Clinical Features Failure to thrive or weight loss Vomiting Diarrhea Hematochezia

*Food proteins in breast milk (most often cow’s milk or egg protein). † Retching or gastroesophageal reflux. AEE, allergic eosinophilic esophagitis; AEG, allergic eosinophilic gastroenteritis.

DIETARY PROTEININDUCED ENTEROCOLITIS

DIETARY PROTEININDUCED PROCTOCOLITIS

Chapter 9  Food Allergies Clinical History

Adverse food reaction likely (possible foods identified)

Toxic

Adverse food reaction unlikely (unless a non–IgE-mediated food hypersensitivity is possible)

Nontoxic

Educate

Food intolerance

Laboratory studies

Finished

Food hypersensitivity

Suspect Non–IgE-mediated

Suspect IgE-mediated Skin-prick tests

Laboratory studies and/or endoscopy

Suggestive Symptoms persist; look for other cause

Positive

Negative

History of anaphylaxis

Finished (unless a non–IgE-mediated food hypersensitivity is possible)

No

Yes

Restrict food

Elimination diet Symptoms improve

No symptoms occur

Resume regular diet

Finished

Symptoms recur Eliminate food from diet

Open challenge with foods least likely to provoke symptoms

Symptoms

Short list of food likely to provoke symptoms; blinded challenges

No symptoms; add food back

Symptoms; restrict food

Figure 9-2.  Algorithm for the evaluation and management of adverse food reactions.

main features of four non–IgE-mediated food-allergic disorders. An exclusion diet eliminating all foods suspected by history and/or skin testing (for IgE-mediated disorders) should be conducted for one to two weeks in suspected IgE-mediated disorders, food-induced enterocolitis, and benign eosinophilic proctocolitis. Exclusion diets may need to be extended for as long as 12 weeks in other suspected GI hypersensitivity disorders (e.g., food protein-induced enteropathy, AEE, or AEG) and may require the use of ele-

mental diets (e.g., Vivonex, Neocate One+, or EleCare) to exclude all antigens. If no improvement is noted and dietary compliance is ensured, it is unlikely that food allergy is involved. Before undertaking blinded food challenges (single- or double-blind), suspect foods should be eliminated from the diet for 7 to 14 days before challenge and even longer in some disorders when secondary disaccharidase deficiency may have developed, as noted earlier. Prescribing elimination diets, like prescribing medications,

147

148

Section II  Nutrition in Gastroenterology may have adverse effects (e.g., malnutrition or eating dis­ orders) and should not be done in the absence of evidence that they are likely to be beneficial.

TREATMENT AND NATURAL HISTORY Once the diagnosis of food hypersensitivity is established, strict elimination of the offending allergen is the only proven therapy. Patients must be taught to scrutinize food labels to detect potential sources of hidden food allergens.95 Drugs such as H1 and H2 antihistamines and glucocorticoids modify symptoms to food allergens but, overall, have minimal efficacy or unacceptable side effects. Anti–IL-5 antibodies have shown promise in the treatment of eosinophilic disorders.61 The prevalence of food hypersensitivity is greatest in the first few years of life, but most young children outgrow their food hypersensitivity within three to five years, except possibly for IgE-mediated hypersensitivities to peanuts, nuts, and seafood.28 Although younger children are more likely to outgrow food hypersensitivity, older children and adults also may lose their food hypersensitivity (i.e., develop clinical tolerance and be able to ingest the food without symptoms) if the responsible food allergen can be identified and eliminated from the diet for a period of time.96,97 Gastrointestinal food allergies affect about 4% of children younger than three years and about 1% of the general population. Current research in this field is providing new information regarding the pathogenesis of these disorders and should lead to the development of new diagnostic and therapeutic algorithms. In the interim, specific food hypersensitivities must be diagnosed carefully, and patients must be educated to avoid ingesting the responsible food allergens.

KEY REFERENCES

Breiteneder H, Clare Mills EN. Plant food allergens—structural and functional aspects of allergenicity. Biotechnol Adv 2005; 23:395-9. (Ref 38.)

Chehade M, Magid MS, Mofidi S, et al. Allergic eosinophilic gastroenteritis with protein-losing enteropathy: intestinal pathology, clinical course, and long-term follow-up. J Pediatr Gastroenterol Nutr 2006; 42:516-21. (Ref 63.) Chehade M, Mayer L. Oral tolerance and its relation to food hypersensitivities. J Allergy Clin Immunol 2005; 115:3-12. (Ref 12.) Furuta GT, Liacouras CA, Collins MH, et al. Eosinophilic esophagitis in children and adults: A systematic review and consensus recommendations for diagnosis and treatment. Gastroenterol 2007; 133:1342-63. (Ref 52.) Khan S, Orenstein SR. Eosinophilic gastroenteritis. Gastroenterol Clin North Am 2008; 37:333-48. (Ref 62.) Leffler DA, Kelly CP. Update on the evaluation and diagnosis of celiac disease. Curr Opin Allergy Clin Immunol 2006; 6:191-6. (Ref 78.) Novak N, Haberstok J, Bieber T, Allam JP. The immune privilege of the oral mucosa. Trends Mol Med 2008; 14:191-8. (Ref 19.) Nowak-Wegrzyn A, Sampson HA, Wood RA, Sicherer SH. Food proteininduced enterocolitis syndrome caused by solid food proteins. Pediatr 2003; 111:829-35. (Ref 71.) Rothenberg ME. Eosinophilic gastrointestinal disorders (EGID). J Allergy Clin Immunol 2004; 113:11-28. (Ref 31.) Sampson HA. Food allergy. Part 2: Diagnosis and management. J Allergy Clin Immunol 1999; 103:981-99. (Ref 94.) Sampson HA, Sicherer SH, Birnbaum AH. AGA technical review on the evaluation of food allergy in gastrointestinal disorders. American Gastroenterological Association. Gastroenterol 2001; 120:1026-40. (Ref 68.) Sicherer SH. Food protein-induced enterocolitis syndrome: Clinical perspectives. J Pediatr Gastroenterol Nutr 2000; 30:S45-9. 2000. (Ref 70.) Sicherer SH, Sampson HA. Food allergy. J Allergy Clin Immunol 2006; 117:S470-5. (Ref 4.) Vercelli D. Immunoglobulin E and its regulators. Curr Opin Allergy Clin Immunol 2001; 1:61-5. (Ref 24.) Wickens K, Black PN, Stanley TV, et al. A differential effect of 2 pro­ biotics in the prevention of eczema and atopy: A double-blind, randomized, placebo-controlled trial. J Allergy Clin Immunol 2008; 122:788-94 (Ref 21.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

10 Acute Abdominal Pain Frederick H. Millham

CHAPTER OUTLINE Anatomy  151 Visceral Pain  151 Somatic-Parietal Pain  152 Referred Pain  152 Clinical Evaluation  153 Approach to Acute Care  154 History  154 Physical Examination  155 Laboratory Data  156 Imaging Studies  156 Other Diagnostic Tests  156 Causes  156 Acute Appendicitis  157

Acute abdominal pain is a common complaint that brings patients to emergency departments. As many as 1 in 20 emergency department visits is for abdominal pain.1 Approximately half of these patients have nonspecific findings or “gastroenteritis.”2 The other half have a more serious disorder that warrants further evaluation and treatment. A small proportion of patients has a life-threatening disease. Therefore, the evaluation of acute abdominal pain must be efficient and lead to an accurate diagnosis early in the presentation so that the treatment of patients who are seriously ill is not delayed and patients with self-limited disorders are not overtreated. This chapter discusses the anatomic factors that determine how abdominal pain is perceived, a systematic approach to the evaluation of abdominal pain, and common and special circumstances encountered in evaluating patients with acute abdominal pain.

ANATOMY Physiologic determinants of pain include the nature of the stimulus, the type of receptor involved, the organization of the neural pathways from the site of injury to the central nervous system, and a complex interaction of modifying influences on the transmission, interpretation, and reaction to pain messages.3,4 Sensory neuroreceptors in abdominal organs are located in the mucosa and muscularis of hollow viscera, on serosal structures such as the peritoneum, and within the mesentery.5 In addition to nociception (the perception of noxious stimuli), sensory neuroreceptors are involved in the regulation of secretion, motility, and blood flow via local and central reflex arcs.6 Although sensory information conveyed in this manner usually is not perceived, disordered regulation of these gastrointestinal functions (secretion, motility, blood flow) can cause pain. For example, patients with irritable bowel syndrome perceive

Acute Biliary Disease  157 Small Bowel Obstruction  158 Acute Diverticulitis  158 Acute Pancreatitis  158 Perforated Peptic Ulcer  159 Acute Mesenteric Ischemia  159 Abdominal Aortic Aneurysm  160 Abdominal Compartment Syndrome  160 Other Intra-Abdominal Causes  160 Extra-Abdominal Causes  160 Special Circumstances  161 Pharmacologic Management  161

pain as a result of heightened sensitivity of intestinal afferent neurons to normal endogenous stimuli that results in altered gut motility and secretion (see Chapter 118).7 Abdominal pain is transmitted by two distinct types of afferent nerve fibers, unmyelinated C fibers and myelinated A-δ fibers. These two types of nerve fibers result in the perception of two different types of abdominal pain (visceral and somatic-parietal pain, respectively); interplay between the two systems results in a third type of pain, referred pain.

VISCERAL PAIN

Visceral pain is transmitted by C fibers that are found in muscle, periosteum, mesentery, peritoneum, and viscera. Most painful stimuli from abdominal viscera are conveyed by this type of fiber and tend to be dull, cramping, burning, poorly localized, and more gradual in onset and longer in duration than somatic pain. Because abdominal organs transmit sensory afferents to both sides of the spinal cord, visceral pain is usually perceived to be in the midline, in the epigastrium, periumbilical region, or hypogastrium (Fig. 10-1). Visceral pain is not well localized because the number of nerve endings in viscera is lower than that in highly sensitive organs such as the skin and because the innervation of most viscera is multisegmental. The pain is generally described as cramping, burning, or gnawing. Secondary autonomic effects such as sweating, restlessness, nausea, vomiting, perspiration, and pallor often accompany visceral pain. The patient may move about in an effort to relieve the discomfort. The afferent fibers that mediate painful stimuli from the abdominal viscera follow the distribution of the autonomic nervous system (Fig. 10-2). The cell bodies for these fibers are located in the dorsal root ganglia of spinal afferent nerves. On entering the spinal cord, these fibers branch into the dorsal horn and tract of Lissauer, where afferent nerves from adjacent spinal segments travel in a cephalad direction and caudally over one or two spinal segments

151

152

Section III  Symptoms, Signs, and Biopsychosocial Issues

A

1 2 3 1

2

Abdominal visceral nociceptors also respond to various chemical stimuli. Chemical nociceptors are contained mainly in the mucosa and submucosa of the hollow viscera. These receptors are activated directly by substances released in response to local mechanical injury, inflammation, tissue ischemia and necrosis, and noxious thermal or radiation injury. Such substances include H+ and K+ ions, histamine, serotonin, bradykinin and other vasoactive amines, substance P, calcitonin gene-related peptide, prostaglandins, and leukotrienes.11,12 Accumulation of nociceptor-reactive substances may change the microenvironment of the injured tissue and thereby reduce the pain threshold. The sensation of pain to a given stimulus is thus increased, and otherwise innocuous stimuli become painful. For example, the application of chemical irritants or pressure to normal gastric mucosa is not painful, whereas the application of the same stimuli to inflamed or injured gastric mucosa causes pain.

SOMATIC-PARIETAL PAIN 3

Figure 10-1.  Localization of visceral pain. Pain arising from organ areas depicted in 1, 2, and 3 is felt in the epigastrium, midabdomen, and hypogastrium, respectively, as shown in A. The arrow in A indicates biliary pain that is referred to the right scapular area.

before terminating on dorsal horn cells in laminae I and V. The dorsal horn cells within laminae I and V are the primary projection neurons for ascending pain pathways. From the dorsal horn, second-order neurons transmit nociceptive impulses via fibers that pass across the anterior commissure and ascend the spinal cord in the contralateral spinothalamic tract. These fibers project to the thalamic nuclei and the reticular formation nuclei of the pons and medulla. The thalamic nucleus sends third-order neurons to the somatosensory cortex, where the discriminative aspects of pain are perceived. The reticular formation nucleus sends neurons to the limbic system and frontal cortex, where the emotional aspects of pain are interpreted.8,9 Abdominal visceral nociceptors respond to mechanical and chemical stimuli. The principal mechanical signal to which visceral nociceptors are sensitive is stretch; cutting, tearing, or crushing of viscera does not result in pain. Visceral stretch receptors are located in the muscular layers of the hollow viscera, between the muscularis mucosa and submucosa, in the serosa of solid organs, and in the mesentery (especially adjacent to large vessels).5,10 Mechanoreceptor stimulation can result from rapid distention of a hollow viscus (e.g., intestinal obstruction), forceful muscular contractions (e.g., biliary pain or renal colic), and rapid stretching of solid organ serosa or capsule (e.g., hepatic congestion). Similarly, torsion of the mesentery (e.g., cecal volvulus) or tension from traction on the mesentery or mesenteric vessels (e.g., retroperitoneal or pancreatic tumor) results in stimulation of mesenteric stretch receptors.

Somatic-parietal pain is mediated by A-δ fibers that are distributed principally to skin and muscle. Signals from this neural pathway are perceived as sharp, sudden, welllocalized pain, such as that which follows an acute injury. These fibers convey pain sensations through spinal nerves. Stimulation of these fibers activates local regulatory reflexes mediated by the enteric nervous system and long spinal reflexes mediated by the autonomic nervous system, in addition to transmitting pain sensation to the central nervous system.13 Somatic-parietal pain arising from noxious stimulation of the parietal peritoneum is more intense and more precisely localized than visceral pain. An example of this difference occurs in acute appendicitis, in which the early vague periumbilical visceral pain is followed by the localized somatic-parietal pain at McBurney’s point that is produced by inflammatory involvement of the parietal peritoneum. Somatic-parietal pain is usually aggravated by movement or vibration. The nerve impulses that mediate such pain travel in somatic sensory spinal nerves. The fibers reach the spinal cord in the peripheral nerves that correspond to the cutaneous dermatomes of the skin, the sixth thoracic (T6) to first lumbar (L1) vertebra. Lateralization of the discomfort of parietal pain is possible because only one side of the nervous system innervates a given part of the parietal peritoneum. Reflexive responses, such as involuntary guarding and abdominal rigidity, are mediated by spinal reflex arcs involving somatic-parietal pain pathways. Afferent pain impulses are modified by inhibitory mechanisms at the level of the spinal cord. Somatic A-d fibers mediate touch, vibration, and proprioception in a dermatomal distribution that matches the visceral innervation of the injured viscera and synapse with inhibitory interneurons of the substantia gelatinosa in the spinal cord. In addition, inhibitory neurons that originate in the mesencephalon, periventricular gray matter, and caudate nucleus descend within the spinal cord to modulate afferent pain pathways. These inhibitory mechanisms allow cerebral influences to modify afferent pain impulses.9,14

REFERRED PAIN

Referred pain is felt in areas remote from the diseased organ and results when visceral afferent neurons and somatic afferent neurons from a different anatomic region converge on second-order neurons in the spinal cord at the same spinal segment. This convergence may result from the innervation, early in embryologic development, of adjacent

Chapter 10  Acute Abdominal Pain Heart Midbrain

Medulla

Larynx Trachea Bronchi Lungs

Vagus nerve

C1

Superior cervical ganglion

Esophagus Stomach

T1 Celiac ganglion

Abdominal blood vessels Liver Bile ducts Pancreas

Superior mesenteric ganglion L1 Inferior mesenteric ganglion

Adrenal Small intestine Large intestine

S1 Kidney Pelvic nerves

Bladder

Reproductive organs

structures that subsequently migrate away from each other. As such, referred pain can be understood to refer to an earlier developmental state. For example, the central tendon of the diaphragm begins its development in the neck and moves craniocaudad, bringing its innervation, the phrenic nerve, with it.15 Figure 10-3 shows how diaphragmatic irritation from a subphrenic hematoma or splenic rupture may be perceived as shoulder pain (Kehr’s sign).9

CLINICAL EVALUATION Effective evaluation of a patient with acute abdominal pain (an acute abdomen) requires careful but expeditious history taking and physical examination (often repeated serially) and, in many cases, informed use of imaging studies. When a carefully performed history and physical examination are paired with appropriate and timely imaging, an accurate diagnosis can often be determined relatively quickly. Inadequate clinical evaluation or poor selection of imaging methods leads to unnecessary delay, often resulting in a

Figure 10-2.  Pathways of visceral sensory innervation. The visceral afferent fibers that mediate pain travel with autonomic nerves to communicate with the central nervous system. In the abdomen, these fibers include vagal and pelvic parasympathetic nerves and thoracolumbar sympathetic nerves. Sympathetic fibers (red lines); parasympathetics (blue lines).

To brain C B Spinal cord A Visceral afferent first-order neuron A B Spinal cord second-order neuron Somatic afferent first-order neuron C

Figure 10-3.  Demonstration of the neuroanatomic basis of referred pain. Visceral afferent fibers that innervate the diaphragm can be stimulated by local irritation (e.g., subdiaphragmatic abscess [circle]). These visceral afferent fibers (A) synapse with second-order neurons in the spinal cord (B) as well as somatic afferent fibers (C) arising from the left shoulder area (cervical roots 3 to 5). The brain interprets the pain to be somatic in origin and localizes it to the shoulder.

153

Section III  Symptoms, Signs, and Biopsychosocial Issues poor outcome. Currently, common entities such as appendicitis, cholecystitis, and diverticulitis can be diagnosed with almost complete accuracy; patients with other diseases require an orderly and efficient evaluation and judicious selection of imaging studies.

APPROACH TO ACUTE CARE

When approaching a patient with acute abdominal pain, the physician should begin with a rapid assessment of the patient’s overall physiologic state. Quickly assessing the three domains according to the mnemonic “ABC” will help identify patients who are unstable and therefore require expedited treatment: A: Airway: Is the patient able to maintain an airway? Does an impaired sensorium endanger the patient’s airway or pose a risk for aspiration of vomit or oral secretions? B: Breathing: How effectively is the patient breathing? Are breaths rapid and shallow? Is the use of accessory muscles evident? Does the patient appear tachypneic? C: Circulation: Circulation encompasses three areas of assessment: (1) Is the patient in shock, as suggested by pallor, cyanosis, mottling, prostration, hypotension, tachycardia, or other signs of hypoperfusion? (2) Has intravenous access been established? (3) Is there evidence of active bleeding? If hemodynamic instability is apparent, including clinical evidence of shock, surgical consultation should be sought immediately, and consideration should be given to endotracheal intubation and resuscitation early in the encounter. The adage in acute care surgery that “death begins in radiology” should be a reminder that hemodynamic resuscitation should precede diagnostic imaging. Patients who are in shock demand urgent care and should not be sent for imaging studies without aggressive resuscitation and monitoring.

Chronology

The time courses of several common causes of acute abdominal pain are diagrammed in Figure 10-4. The rapidity of onset of pain is often a measure of the severity of the underlying disorder. Pain that is sudden in onset, severe, and well localized is likely to be the result of an intra-abdominal catastrophe such as a perforated viscus, mesenteric infarction, or ruptured aneurysm. Affected patients usually recall the exact moment of onset of their pain. Progression is an important temporal factor in abdominal pain. In some disorders, such as gastroenteritis, pain is self-limited, whereas in others, such as appendicitis, pain is progressive. Colicky pain has a crescendo-decrescendo pattern that may be diagnostic, as in renal colic. The duration of abdominal pain is also important. Patients who seek evaluation of abdominal pain that has been present for an extended period (e.g.,

D

B

C Severity

154

A

HISTORY

Despite the advances made in clinical imaging, history taking remains the most important component of the initial evaluation of the patient with acute abdominal pain.16 Characteristic features of the pain associated with various common causes of acute abdominal pain are shown in Table 10-1. Attention to these features can lead to a rapid clinical diagnosis or exclusion of important diseases in the differential diagnosis, thus enhancing the reliability and effectiveness of subsequent diagnostic testing.2

Time Figure 10-4.  Patterns of acute abdominal pain. A, Many causes of abdominal pain subside spontaneously with time (e.g., gastroenteritis). B, Some pain is colicky (i.e., the pain progresses and remits over time); examples include intestinal, renal, and biliary pain (colic). The time course may vary widely from minutes in intestinal and renal pain to days, weeks, or even months in biliary pain. C, Commonly, acute abdominal pain is progressive, as in acute appendicitis or diverticulitis. D, Certain conditions have a catastrophic onset, such as ruptured abdominal aortic aneurysm.

Table 10-1  Comparison of Common Causes of Acute Abdominal Pain Cause

ONSET

LOCATION

CHARACTER

DESCRIPTOR

RADIATION

Appendicitis

Gradual

None

++

Acute Acute Gradual Sudden

Constricting Boring Ache Burning

Scapula Midback None None

++ ++ to +++

Small bowel obstruction Mesenteric ischemia, infarction Ruptured abdominal aortic aneurysm Gastroenteritis Pelvic inflammatory disease Ruptured ectopic pregnancy

Gradual Sudden

Periumbilical area Periumbilical area

Diffuse early; localized late Localized Localized Localized Localized early, diffuse late Diffuse Diffuse

Ache

Cholecystitis Pancreatitis Diverticulitis Perforated peptic ulcer

Periumbilical area early; RLQ late RUQ Epigastrium, back LLQ Epigastrium

Cramping Agonizing

None None

++ +++

Sudden

Abdomen, back, flank

Diffuse

Tearing

None

+++

Gradual Gradual Sudden

Periumbilical area Either LQ, pelvis Either LQ, pelvis

Diffuse Localized Localized

Spasmodic Ache Sharp

None Upper thigh None

+ to ++ ++ ++

+ = mild; ++ = moderate; +++ = severe; LLQ = left lower quadrant; LQ = lower quadrant; RLQ = right lower quadrant; RUQ = right upper quadrant.

INTENSITY

+++

Chapter 10  Acute Abdominal Pain weeks) are less likely to have an acute life-threatening illness than patients who present within hours to days of the onset of their symptoms.

lent vomitus suggests more distal small bowel or colonic obstruction. A constellation of findings may indicate a particular disease entity.

Location

Past Medical History

The location of abdominal pain provides a clue to interpreting the cause. As noted, a given noxious stimulus may result in a combination of visceral, somatic-parietal, and referred pain, thereby creating confusion in interpretation unless the neuroanatomic pathways are considered. For example, the pain of diaphragmatic irritation from a left-sided subphrenic abscess may be referred to the shoulder and misinterpreted as pain from ischemic heart disease (see Fig. 10-3). Changes in location may represent progression from visceral to parietal irritation, as with appendicitis, or represent the development of diffuse peritoneal irritation, as with a perforated ulcer.

Intensity and Character

Acute abdominal pain usually follows one of three patterns. Pain that is prostrating, physically incapacitating the sufferer, is usually caused by a severe, life-threatening disease such as a perforated viscus, ruptured aneurysm, or severe pancreatitis. By contrast, patients with obstruction of a hollow viscus, as in intestinal obstruction, renal colic, or biliary pain, present with the gradual onset of cramping pain that follows a sinusoidal pattern of intense pain alternating with a period of relief. Nausea and vomiting are characteristic symptoms associated with this group of disorders. The obstructed viscus need not be the intestine for nausea or vomiting to occur, as in the case of a kidney stone. The third pattern is of gradually increasing discomfort, usually vague and poorly localized at the start, but becoming more localized as the pain intensifies. This picture is usually caused by inflammation, as with acute appendicitis or diverticulitis. Some disorders, such as acute cholecystitis, may start out as colicky pain but evolve into a constant pain as cystic duct obstruction leads to gallbladder inflammation. The clinician should be cautious, however, in assigning too much importance to a patient’s description of the pain; exceptions are common, and a given descriptor may be attributable to a number of conditions.

Aggravating and Alleviating Factors

The relationship of the pain to positional changes, meals, bowel movements, and stress may yield important diagnostic clues. Patients with peritonitis, for example, lie motionless, whereas those with renal colic may writhe in an attempt to find a comfortable position. Sometimes, certain foods exacerbate pain. A classic example is the relationship between the intake of fatty foods and the development of biliary pain. Pain associated with duodenal ulcer often is alleviated by meals. By contrast, patients with gastric ulcer or chronic mesenteric ischemia may report exacerbation of pain with eating. Patients often self-medicate to alleviate symptoms. A history of chronic antacid use or of nonsteroidal anti-inflammatory drug use, for example, may suggest the presence of peptic ulcer disease.

Associated Symptoms

Information regarding changes in constitutional symptoms (e.g., fever, chills, night sweats, weight loss, myalgias, arthralgias), digestive function (e.g., anorexia, nausea, vomiting, flatulence, diarrhea, constipation), jaundice, dysuria, changes in menstruation, and pregnancy should be solicited from the patient. A careful review of these symptoms may reveal important diagnostic information. For example, clear vomitus suggests gastric outlet obstruction, whereas fecu-

A careful review of the patient’s other medical problems often sheds light on the presentation of acute abdominal pain. Previous experience with similar symptoms suggests a recurrent problem. Patients with a history of partial small bowel obstruction, renal calculi, or pelvic inflammatory disease are likely to have recurrences. A patient whose presentation suggests intestinal obstruction, and who has no prior surgical history, deserves special attention because of the likelihood of surgical pathology, such as a hernia or malignancy. Patients with a systemic illness such as scleroderma, systemic lupus erythematosus, nephrotic syndrome, porphyria, or sickle cell disease often have abdominal pain as a manifestation of the underlying disorder. Abdominal pain also may arise as a side effect of a medication taken for another disease.

PHYSICAL EXAMINATION

The physical examination of the patient with acute abdom­ inal pain begins with an assessment of the patient’s appearance and airway, breathing, and circulation (ABC), as described earlier. The patient’s ability to converse, breathing pattern, position in bed, posture, degree of discomfort, and facial expression should be noted. A patient lying still in bed, in the fetal position and reluctant to move or speak, with a distressed facial expression, is likely to have peritonitis. On the other hand, a patient who writhes and changes position frequently likely has purely visceral pain, as in intestinal obstruction or gastroenteritis. Tachypnea may be a sign of metabolic acidosis caused by shock. Atrial fibrillation noted on physical examination or electro­ cardiogram may suggest mesenteric arterial embolus. All patients should undergo a careful, systematic examination, regardless of the differential diagnosis suggested by the history.

Abdominal Examination

Examination of the abdomen is central to the evaluation of a patient with acute abdominal pain and should begin with careful inspection. The entire abdomen, from the nipple line to the thighs, should be exposed. Obese patients should be asked whether the degree of protrusion of the abdominal wall is more than usual. Asthenic patients may feel distended but have relatively little apparent abdominal protrusion. Assessment for the presence of bowel sounds and their character should precede any maneuvers that will disturb the abdominal contents. Before concluding that an abdomen is silent, the examiner should listen for at least two minutes and in more than one quadrant of the abdomen. Experienced listeners will distinguish the high-pitched churning of a mechanical small intestinal obstruction from the more hollow sounds of toxic megacolon (like dripping in a cavern). The examiner should begin to palpate the abdomen with the head of the stethoscope while carefully watching the patient’s facial expression. If tenderness is detected, an assessment for rebound tenderness should be carried out next to look for evidence of peritonitis. Rebound tenderness may be elicited by jarring the patient’s bed or stretcher or by finger percussion. Palpation is performed next. If pain is emanating from one particular region, that area should be palpated last to detect involuntary guarding and muscular rigidity. Patients with a rigid abdomen rarely reveal any additional findings (such as a mass) on physical exam­ ination. Because these patients usually have a surgical

155

156

Section III  Symptoms, Signs, and Biopsychosocial Issues emergency, abdominal examination can be done more completely once the patient is under anesthesia, just before laparotomy.

Genital, Rectal, and Pelvic Examinations

The pelvic organs and external genitalia should be examined in every patient with acute abdominal pain. The rectum and vagina provide additional avenues for gentle palpation of pelvic viscera. Gynecologic pathology should be excluded in all women with acute abdominal pain.

LABORATORY DATA

The history and physical examination findings generally are not sufficient to establish a firm diagnosis in a patient with acute abdominal pain. All patients with acute abdominal pain should have a complete blood count, with a differential count, and urinalysis. The determination of serum electrolyte, blood urea nitrogen, creatinine, and glucose levels is useful for assessing the patient’s fluid and acid-base status, renal function, and metabolic state and should be done for every patient with acute abdominal pain who presents to an emergency department. Urine or serum pregnancy testing must be performed in all women of reproductive age with abdominal pain. Liver biochemical tests and serum amylase levels should be ordered for patients with upper abdominal pain or with jaundice. Leukocytosis, particularly when associated with immature band forms, is an important finding. Metabolic acidosis, an elevated serum lactate level, or a depressed bicarbonate level are all associated with tissue hypoperfusion and shock. Patients who manifest these findings are likely to require urgent surgical intervention or intensive care.

IMAGING STUDIES Computed Tomography

The development of high-speed helical computed tomog­ raphy (CT) scanning has revolutionized the evaluation of acute abdominal pain. In many conditions, such as appendicitis, CT scanning can almost eliminate diagnostic uncertainty. In the pre-CT era, history taking and physical examination alone had a specificity of approximately 80%; by contrast, the sensitivity and specificity of CT scanning for acute appendicitis are 94% and 95%, respectively.17 A negative CT scan in the setting of acute abdo­minal pain has considerable value in excluding common disorders. The question arises as to whether CT scanning should be a standard part of the evaluation in all patients with acute abdominal pain. Several arguments as to why CT should not be routine have been raised. First, CT scanning can be performed in a number of ways, and the most efficacious method must be chosen in any given clinical setting. For example, a patient with suspected renal colic should have a limited, non–contrast-enhanced, renal calculus protocol CT; obtaining a standard oral and intravenous contrast CT in this case may obfuscate rather than illuminate the pathology. Alternatively, a patient in whom arterial occlusive disease is suspected should undergo CT arteriography using a bolus intravenous contrast technique. A radiologist should be consulted regarding the selection of the most appropriate CT study in a given patient. Second, some diseases, such as acute cholecystitis and cholangitis, remain relatively invisible on CT. A patient with right upper quadrant pain who is suspected of having either of these diagnoses should undergo an ultrasound examination of the right upper quadrant as the primary diagnostic test. Third, as noted earlier, a patient who is unstable or exhibits signs of shock should be evaluated by a surgeon before any

imaging study is considered. In a patient with suspected trauma or hemoperitoneum, the focused abdominal sonogram for trauma (FAST; see later), which can be done at the bedside in the emergency department, is a preferable approach. The presence of shock and fluid in the abdomen is an indication for immediate laparotomy, and further diagnostic maneuvers, including CT, add little value to the patient’s care. A final consideration regarding the role of CT in the evaluation of acute abdominal pain is radiation exposure. Particularly for patients younger than 35 years and those who have required multiple examinations, abdominal CT may increase the lifetime risk of cancer.18 Additionally, unless a life-threatening condition is suspected, CT is best avoided in a pregnant patient, in whom ultrasound examination or magnetic resonance imaging (MRI) may provide a suitable alternative.

Ultrasonography

A FAST is a rapid, reliable, bedside test to detect fluid in the abdominal cavity. Although its main usefulness is for the evaluation of injured persons, this examination also aids in the diagnosis of any condition that results in free intraperitoneal fluid. Although not part of the formal FAST series, imaging of the aorta can be added, allowing a rapid assessment for aortic aneurysm. General beside ultrasound is likely to be used increasingly by nonradiologists in the future. A Swedish study has demonstrated that the diagnostic accuracy of emergency abdominal examinations by surgeons is increased significantly when an ultrasound examination is added to the evaluation.19

OTHER DIAGNOSTIC TESTS

Other diagnostic imaging modalities such as MRI and radionuclide scanning (e.g., 99mTc-labeled hydroxyl iminodiacetic acid [HIDA] scan) and endoscopy usually take a secondary role in the evaluation of the patient with acute abdominal pain. Use of these tests is generally guided by the results of CT or ultrasound. Angiography may be useful not only for establishing a diagnosis of visceral ischemia, but also for delivering therapy aimed at improving or reestablishing blood flow. Diagnostic peritoneal lavage, although seldom used now, is useful when a patient is too unstable from a cardiopulmonary standpoint to tolerate radiographic imaging. The finding of leukocytes in the lavage effluent in an unstable patient may, in extreme circumstances, constitute sufficient grounds for laparotomy. In a patient who is unstable and deteriorating and has signs of an acute abdomen, laparotomy as a diagnostic maneuver should be considered if imaging is considered prohibitively risky. An overall approach to the patient with acute abdominal pain is illustrated in Figure 10-5.

CAUSES Acute abdominal pain is usually defined as pain of less than one week in duration. Patients usually seek attention within the first 24 to 48 hours, although some may endure longer periods of abdominal discomfort. The most common reason for a patient to seek emergency department evaluation of abdominal pain is so-called nonspecific abdominal pain. Between 25% and 50% of all patients who visit an emergency department for abdominal pain will have no specific disease identified. The distribution of the causes of abdominal pain in patients who present to an emergency department is shown in Table 10-2.

Chapter 10  Acute Abdominal Pain Acute abdominal pain

Most likely diagnoses

Evaluation

ABC Prostration; hemodynamically unstable

Yes

No RLQ pain (gradual onset); RLQ tenderness, localized rebound tenderness

Yes

No Gradual onset of RUQ cramping pain; history of postprandial discomfort

Yes

Resuscitation Urgent surgical consultation Consider FAST examination Consider laparotomy

Perforated viscus Severe pancreatitis Ruptured spleen/ hemoperitoneum Ruptured AAA

Appendix protocol CT

Appendicitis*

In female patients, consider pelvic US or CT

Tubo-ovarian abscess Ovarian torsion Ectopic pregnancy Cholelithiasis Cholecystitis Bile duct obstruction Cholangitis

RUQ ultrasound

No Nausea, vomiting, obstipation, constipation, abdominal distention; prior surgery

Yes

Upright abdominal film or oral contrast CT scan

Small bowel obstruction

No Sudden onset, diffuse pain; involuntary guarding, rebound tenderness, peritonitis

Yes

Upright abdominal film or oral contrast CT scan

Perforated viscus Diverticulitis Mesenteric infarction Acute pancreatitis

Table 10-2  Causes of Acute Abdominal Pain in Patients Presenting to an Emergency Department CAUSE Nonspecific abdominal pain Appendicitis Bowel obstruction Urologic disease Biliary disease Diverticular disease Pancreatitis Medical illness Other

PATIENTS (%) 35 17 15 6 5 4 2 1 15

From Irvin TT. Causes of abdominal pain in 1190 patients admitted to a British Surgical Service. Br J Surg 1989; 76:1121-5.

ACUTE APPENDICITIS Acute appendicitis is a ubiquitous problem. In adult patients younger than 60 years, acute appendicitis accounts for 25% of admissions to the hospital from the emergency department for abdominal pain.20 The overall incidence of appendicitis is approximately 11/10,000 population, with a lifetime risk of 8.6% for men and 6.7% for women.21 Typically, acute appendicitis begins with prodromal symptoms of anorexia, nausea, and vague periumbilical pain. Within 6 to 8 hours, the pain migrates to the right lower quadrant and peritoneal signs develop. In uncomplicated appendi­ citis, a low-grade fever to 38°C and mild leukocytosis are usually present. A higher temperature and white blood cell

Figure 10-5.  An approach to the urgent evaluation of abdominal pain. Specific complaints and physical examination findings are coupled with appropriate radiologic imaging. AAA, abdominal aortic aneurysm; ABC, airway, breathing, circulation; CT, computed tomography; FAST, focused abdominal sonogram for trauma; RLQ, right lower quadrant; RUQ, right upper quadrant; US, ultrasound. *For left lower quadrant pain, the most likely diagnosis is diverticulitis.

count are associated with perforation and abscess formation. The mnemonic PANT can help the novice remember the classic progression of symptoms in appendicitis—pain followed by anorexia followed by nausea followed by temperature elevation. Uncommon presentations of acute appendicitis, however, are common, and the wary physician will not reject a diagnosis of acute appendicitis simply on the basis of the patient’s history and physical examination alone. Whereas plain abdominal radiographs are not diagnostic and have little role in the diagnosis of acute appendicitis, CT has dramatically improved the accuracy of diagnosis in patients with acute appendicitis. The finding of an appendiceal diameter larger than 6 mm has positive and negative predictive values of 98%.22 Other CT signs of acute appendicitis include periappendiceal fat inflammation, presence of fluid in the right lower quadrant, and failure of contrast dye to fill the appendix23; these findings have lower degrees of specificity. Traditionally, an erroneous diagnosis of appendicitis, reflected by the finding of normal pathology at surgical exploration, was as high as 33%.24 The addition of CT has reduced the false-negative rate to approximately 6% for men and 10% for women.25 As noted earlier, CT does entail radiation exposure,18 and some authorities advocate avoiding CT in children and adolescents,26 in whom a higher degree of diagnostic uncertainty is tolerated in favor of lower radiation exposure (see Chapter 116).

ACUTE BILIARY DISEASE

Biliary disease accounts for approximately 5% to 7% of emergency department visits for abdominal pain.2,20 Most

157

158

Section III  Symptoms, Signs, and Biopsychosocial Issues patients in this group present at some point on the spectrum between biliary pain and acute cholecystitis. Biliary pain is a syndrome of right upper quadrant or epigastric pain, usually postprandial, caused by transient obstruction of the cystic duct by a gallstone. Biliary pain is self-limited, generally lasting less than 6 hours. Acute cholecystitis is, in most cases, caused by persistent obstruction of the cystic duct by a gallstone. The pain of acute cholecystitis is almost indistinguishable from that of biliary pain, except that it is persistent. The pain usually is a dull ache and is localized to the right upper quadrant or epigastrium and may radiate around the back to the right scapula. Nausea, vomiting, and low-grade fever are common. On examination, right upper quadrant tenderness, guarding, and Murphy’s sign (inspiratory arrest on palpation of the right upper quadrant) are diagnostic of acute cholecystitis. The white blood cell count is usually mildly elevated but may be normal. Mild elevations in serum total bilirubin and alkaline phosphatase levels are typical. The role of gallstones in the etiology of biliary pain and acute cholecystitis makes ultrasound evaluation of the right upper quadrant the key diagnostic test. Demonstration of gallstones may suggest biliary pain, whereas the finding of stones with gallbladder wall thickening, pericholecystic fluid, and pain on compression of the gallbladder with the ultrasound probe (sonographic Murphy’s sign) is essentially diagnostic of acute cholecy­s­ titis, as is positive hepatobiliary scintigraphy (e.g., HIDA scan).27 Patients with acute cholecystitis are best managed with cholecystectomy within 48 hours.28-30 Patients who are diabetic, particularly those with a leukocyte count over 15,000/mm3, are at particular risk for gangrenous cholecystitis and should have immediate surgical consultation.31 These patients are likely to require an emergent open cholecystectomy. Patients who present with right upper quadrant pain with jaundice and signs of sepsis should be suspected of having obstruction of the bile duct by a gallstone. Right upper quadrant pain, fever and chills, and jaundice (Charcot’s triad) are suggestive of ascending cholangitis.32 These patients require intravenous fluids, antibiotics, and bile duct drainage, usually by endoscopy (see also Chapters 65, 66, and 67).

SMALL BOWEL OBSTRUCTION

Intestinal obstruction may occur in patients of all ages. In pediatric patients, intussusception, intestinal atresia, and meconium ileus are the most common causes. In adults, about 70% of cases are caused by postoperative adhesions; incarcerated hernias make up most of the remainder. Small bowel obstruction is characterized by sudden, sharp, periumbilical abdominal pain. Nausea and vomiting occur soon after the onset of pain and provide temporary relief of discomfort. Frequent bilious emesis with epigastric pain is suggestive of high (proximal) intestinal obstruction, whereas cramping periumbilical pain with infrequent feculent emesis is more typical of distal intestinal obstruction. Examination reveals an acutely ill, restless patient. Fever, tachycardia, and orthostatic hypotension are common. Abdominal distention is usual. Auscultation characteristically demonstrates hyperactive bowel sounds and audible rushes. The patient’s abdomen is diffusely tender to percussion and palpation, but peritoneal signs are absent, unless a complication such as ischemia or perforation has occurred. Leukocytosis and lactic acidosis suggest intestinal ischemia or infarction. Plain radiographs of the abdomen are diagnostic when they reveal dilated loops of small intestine with airfluid levels and decompressed distal small bowel and colon. Plain abdominal films can be misleading in a patient with

proximal jejunal obstruction, because dilated bowel loops and air-fluid levels may be absent. CT is superior for establishing the diagnosis and location of intestinal obstruction.33 In patients with partial small intestinal obstruction, initial treatment is with bowel rest, intravenous fluids, nasogastric decompression, and close observation. Surgery is required for patients who fail conservative management or have evidence of complete obstruction, especially if ischemia is suspected (see also Chapter 119).

ACUTE DIVERTICULITIS

Acute diverticulitis is a common disease. Approximately 80% of affected patients are older than 50 years,34 but the incidence may be increasing in younger persons.35 Patients with diverticulitis usually present with constant, dull, left lower quadrant pain and fever. They may complain of constipation or obstipation and usually are found to have a leukocytosis. Physical examination demonstrates left lower quadrant tenderness and, in some cases, a left lower quadrant mass. Localized peritoneal signs are frequent. In severe cases, generalized peritonitis may be present, making differentiation from other causes of a perforated viscus difficult. CT is reliable in confirming the diagnosis, with a sensitivity of 97%,36 and should be performed routinely in the emergency evaluation of patients with diverticulitis. Acute diverticulitis presents as a spectrum of disease from mild abdominal discomfort to gross fecal peritonitis, which is an acute surgical emergency. The severity of diverticulitis, as determined by CT, is best described using the Hinchey grading system (see Table 117-2).37 Patients with mild disease and no CT findings of perforation, in the absence of limiting comorbid disease, can generally be treated as an outpatient. Those with Hinchey grade I diverticulitis (localized pericolic abscess or inflammation) frequently require hospitalization for intravenous antibiotics. Patients with Hinchey grade II diverticulitis (pelvic, intraabdominal, or retroperitoneal abscess) should undergo CT-guided drainage of the abscess and receive a course of broad-spectrum intravenous antibiotics. Patients with Hinchey III (generalized purulent peritonitis) and IV (generalized fecal peritonitis) diverticulitis frequently require emergency surgery. Optimal surgical management of patients with Hinchey I or II diverticulitis is a matter of debate (see Chapter 117).

ACUTE PANCREATITIS

Hospital admissions for acute pancreatitis in the United States seem to be increasing. The incidence of acute pancreatitis in California rose from 33 to 43 cases/100,000 between 1994 and 2001.38 Acute pancreatitis typically begins as acute pain in the epigastrium that is constant, unrelenting, and frequently described as boring through to the back or left scapular region. Fever, anorexia, nausea, and vomiting are typical. Patients with pancreatitis usually are more comfortable sitting upright, leaning forward slightly, and are commonly found in this position in the emergency department. Physical examination reveals an acutely ill patient in considerable distress. Patients are usually tachycardic and tachypneic. Abdominal examination reveals hypoactive bowel sounds and marked tenderness to percussion and palpation in the epigastrium. Abdominal rigidity is a variable finding. In rare patients, flank or periumbilical ecchymoses (Grey-Turner’s or Cullen’s sign, respectively) develop in the setting of pancreatic necrosis with hemorrhage. Extremities are often cool and cyanotic, reflecting underperfusion. White blood cell counts of 12,000 to 20,000/mm3 are common. Elevated serum and urine amylase levels are usually present within the first few hours of

Chapter 10  Acute Abdominal Pain pain. Depending on the cause and severity of pancreatitis, serum electrolyte, calcium, and blood glucose levels and liver biochemical test and arterial blood gas results may be abnormal. Abdominal ultrasonography is useful for identifying gallstones as a potential cause of pancreatitis. CT is reserved for patients with severe or complicated pancreatitis. Although most cases of acute pancreatitis are self-limited, as many as 20% of patients have severe disease with local or systemic complications,39 including hypovolemia and shock, renal failure, liver failure, and hypocalcemia. Although a number of prognostic physiologic scales, such as the Sequential Organ Failure Assessment (SOFA) and Acute Physiologic Assessment and Chronic Health Evaluation (APACHE) II scores, have been advocated as measures of the severity of acute pancreatitis, the Ranson score, first published in 1974, remains a useful and widely used checklist for the early assessment of patients with acute pancreatitis.40 The Ranson score consists of five early and six late factors that indicate severe pancreatitis (see Table 58-2). A minority of patients with severe acute pancreatitis present with a profound intra-abdominal catastrophe, usually caused by thrombosis of the middle colic artery or right colic artery, which travels in proximity to the head of the pancreas, with resulting colonic infarction. This process may not be seen clearly on CT scans obtained early in the course of disease and should be suspected in any case marked by rapid hemodynamic collapse. Such patients require immediate laparotomy (see Chapter 58).

PERFORATED PEPTIC ULCER

The epidemiology of peptic ulcer disease continues to change. The overall incidence of peptic ulcer disease has declined significantly since the late 1970s,41,42 and the number of patients requiring hospital admission for severe and complicated peptic ulcer disease has also decreased.42 Better therapeutic modalities, including proton pump inhibitors, eradication of Helicobacter pylori, and endoscopic methods for control of hemorrhage, have reduced the number of patients with peptic ulcer disease who require surgical intervention,43 although the incidence of complicated disease has increased in older adults, in whom morbidity and mortality related to surgery are also increased.42 Patients with a perforated peptic ulcer typically present with the sudden onset of severe diffuse abdominal pain. These patients may be able to specify the precise moment of the onset of symptoms. In the usual case, the afflicted patient presents acutely with excruciating abdominal pain. Abdominal examination reveals peritonitis, with rebound tenderness, guarding, or abdominal muscular rigidity. In such cases, distinguishing perforated ulcer from other causes of a perforated viscus, such as a perforated colonic diverticulum or perforated appendicitis, may not be possible. Older or debilitated patients may present with less dramatic symptoms, with perforation detected by the presence of free intraperitoneal air on an upright abdominal film or CT scan. A perforated peptic ulcer should be suspected in any patient with the sudden onset of severe abdominal pain who presents with abdominal rigidity and free intraperitoneal air. Pneumoperitoneum is identified on an abdominal radiograph in 75% of patients (Fig. 10-6). In equivocal cases, CT of the abdomen usually suggests the diagnosis by demonstrating edema in the region of the gastric antrum and duodenum, associated with extraluminal air. CT may not be diagnostic, however, and patients with diffuse peritonitis or hemodynamic collapse should be explored surgically. Lapa-

Figure 10-6.  This upright chest film of an 80-year-old man with the acute onset of severe epigastric pain demonstrates free intra-abdominal air under the right hemidiaphragm. The patient has pneumoperitoneum as a result of a perforated viscus. At surgery, an anterior duodenal ulcer perforation was found.

rotomy is acceptable as the primary diagnostic maneuver in such patients. Endoscopy is not advisable when the diagnosis of a perforated peptic ulcer is suspected. Insufflation of the stomach can convert a sealed perforation into a free perforation. Survival following emergency surgery for complications of peptic ulcer disease is surprisingly poor. Patients who require surgery for a complication of peptic ulcer disease are generally older and more medically ill than those seen in the past. Sarosi and colleagues have reported a 23% in-hospital mortality rate in a Veterans Administration population,44 and Imhof and associates,45 reporting on a series of German patients with perforated peptic ulcer, found an in-hospital mortality rate of 12.1%, a one-year mortality rate of 28.7%, and a five-year mortality rate of 46.8% (see also Chapters 52 and 53).

ACUTE MESENTERIC ISCHEMIA

Acute mesenteric ischemia can result from occlusion of a mesenteric vessel arising from an embolus, which may emanate from an atheroma of the aorta or cardiac mural thrombus, or from primary thrombosis of a mesenteric vessel, usually at a site of atherosclerotic stenosis. Embolic occlusion is more common in the superior mesenteric artery than the celiac or inferior mesenteric artery, presumably because of the less acute angle of the superior mesenteric artery off the abdominal aorta. Nonocclusive mesenteric ischemia results from inadequate visceral perfusion and can also lead to intestinal ischemia and infarction. Such cases are usually consequent to catastrophic systemic illnesses such as cardiogenic or septic shock. Acute mesenteric embolism, mesenteric thrombosis, and nonocclusive mesenteric ischemia each account for approximately one third of cases of acute mesenteric ischemia and have a combined mortality rate of 60% to 100%.46 The hallmark of the diagnosis of acute mesenteric ischemia is the abrupt onset of intense cramping epigastric and periumbilical pain out of proportion to the findings on abdominal examination. Other symptoms may include diarrhea, vomiting, bloating, and melena. On physical examination, most patients appear acutely ill, but the presentation may be subtle. Shock is present in about 25% of cases.

159

160

Section III  Symptoms, Signs, and Biopsychosocial Issues CT is the best initial diagnostic test. Mesenteric angio­g­ raphy may be useful for determining the cause of intestinal ischemia and defining the extent of vascular disease. Patients with acute embolic or thrombotic intestinal ische­ mia should be referred for immediate revascularization and bowel resection.47 Patients with nonocclusive mesenteric ischemia are best managed by treatment of the underlying shock state. For those with persistent symptoms, laparotomy for resection of infarcted intestine may be necessary. Transcatheter vasodilator therapy may be helpful for patients who are found to have vasospasm on visceral arteriography (see also Chapter 114).47

ABDOMINAL AORTIC ANEURYSM

Rupture of an abdominal aortic aneurysm is heralded by the sudden onset of acute, severe abdominal pain localized to the midabdomen or paravertebral or flank areas. The pain is tearing in nature and associated with prostration, lightheadedness, and diaphoresis. If the patient survives transit to the hospital, shock is the most common presen­ tation. Physical examination reveals a pulsatile, tender abdominal mass in about 90% of cases. The classic triad of hypotension, a pulsatile mass, and abdominal pain is present in 75% of cases and mandates immediate surgical intervention.47

ABDOMINAL COMPARTMENT SYNDROME

Although not usually presenting as acute abdominal pain, abdominal compartment syndrome (ACS) warrants consideration in any patient with an abdominal emergency. First reported in the setting of massive intra-abdominal trauma, ACS, defined as pathologic elevation of intraabdominal pressure, is now recognized as a frequent complication of many severe disease processes. An elevated intra-abdominal pressure may develop in a patient who survives massive volume resuscitation with resulting visceral edema or who has a disease such as severe pancreatitis that can cause visceral or retroperitoneal edema. The elevated intra-abdominal pressure in turn compromises visceral perfusion, with resulting injury and additional edema. The kidney is particularly prone to underperfusion in this setting, and kidney failure may be the first sign of ACS.48 Intra-abdominal pressure can be measured simply by connecting a transducer to a urinary catheter, with the zero reference point at the midaxillary line in a supine patient. The World Society for Abdominal Compartment Syndrome has established a consensus grading scheme for ACS based on the measured bladder pressure. A normal value for bladder pressure is less than 7 mm Hg. Grade I ACS is defined as a pressure of 12 to 15 mm Hg, grade II as 16 to 20 mm Hg, grade III as 21 to 25 mm Hg, and grade IV as greater than 25 mm Hg. Nonsurgical options for treating low-grade ACS include gastric decompression, sedation, neuromuscular blockade, placing the patient in a reverse Trendelenburg position while allowing the hips to remain in a neutral position, and diuretics. In a patient with highgrade ACS, particularly when renal and respiratory function is com­promised, laparotomy and creation of an open abdomen is most effective. Management of the open abdomen requires specific surgical expertise usually found in referral medical centers.49

OTHER INTRA-ABDOMINAL CAUSES

Other intra-abdominal causes of acute abdominal pain include the following: gynecologic conditions such as endometritis, acute salpingitis with or without tubo-ovarian

abscess, ovarian cysts or torsion, and ectopic pregnancy; spontaneous bacterial peritonitis (Chapter 91); functional dyspepsia and peptic ulcer disease (Chapters 13 and 52); infectious gastroenteritis (Chapters 107 and 108); viral hepatitis and other liver infections (Chapters 77 to 82); pyelonephritis; cystitis; mesenteric lymphadenitis; inflammatory bowel disease (Chapters 111 and 112); and functional abnormalities such as irritable bowel syndrome (Chapter 118) and intestinal pseudo-obstruction (Chapter 120).

EXTRA-ABDOMINAL CAUSES

Acute abdominal pain may arise from disorders involving extra-abdominal organs and systemic illnesses. Examples are listed in Table 10-3. Surgical intervention for patients with acute abdominal pain arising from an extra-abdominal or systemic illness is seldom required except in cases of pneumothorax, empyema, and esophageal perforation. Esophageal perforation may be iatrogenic, result from blunt or penetrating trauma, or occur spontaneously (Boerhaave’s syndrome; see also Chapter 45).

Table 10-3  Extra-Abdominal Causes of Acute Abdominal Pain Cardiac Congestive heart failure Endocarditis Myocardial ischemia and infarction Myocarditis Thoracic Esophageal rupture (Boerhaave’s syndrome) Esophageal spasm Empyema Esophagitis Pleurodynia (Bornholm’s disease) Pneumonitis Pneumothorax Pulmonary embolism and infarction Hematologic Acute leukemia Hemolytic anemia Henoch-Schönlein purpura Sickle cell anemia Metabolic Acute adrenal insufficiency (Addison’s disease) Diabetes mellitus (especially with ketoacidosis) Hyperlipidemia Hyperparathyroidism Porphyria Uremia Toxins Hypersensitivity reactions (e.g., to insect bites, reptile venoms) Lead poisoning Infections Herpes zoster Osteomyelitis Typhoid fever Neurologic Abdominal epilepsy Radiculopathy, spinal cord or peripheral nerve tumors, degenerative arthritis of spine, herniated vertebral disk Tabes dorsalis Miscellaneous Familial Mediterranean fever Heat stroke Muscle contusion, hematoma, tumor Narcotic withdrawal Psychiatric disorders

Chapter 10  Acute Abdominal Pain SPECIAL CIRCUMSTANCES Extremes of Age

Evaluation of acute abdominal pain in patients at the extremes of age is a challenge. Historical information and physical examination findings are often difficult to elicit or are unreliable. Similarly, laboratory data may be misleadingly normal in the face of serious intra-abdominal pathology. For these reasons, patients at the extremes of age often are diagnosed late in the course of the disease, thereby resulting in increased morbidity. For example, the perforation rate for appendicitis in the general population averages 10% but exceeds 50% in infants. A carefully obtained history, thorough physical examination, and high index of suspicion are the most useful diagnostic aids. The occurrence of acute abdominal conditions is highly variable in these populations, and a high index of suspicion is required. In the pediatric population, the causes of acute abdominal pain vary with age. In infancy, intussusception, pyelonephritis, gastroesophageal reflux, Meckel’s diverticulitis, and bacterial or viral enteritis are common. In children, Meckel’s diverticulitis, cystitis, pneumonitis, enteritis, mesenteric lymphadenitis, and inflammatory bowel disease are prevalent. In adolescents, pelvic inflammatory disease, inflammatory bowel disease, and the common adult causes of acute abdominal pain predominate. In children of all ages, two of the most common causes of pain are acute appendicitis and abdominal trauma secondary to child abuse. In the older adult population, biliary tract disease accounts for almost 25% of cases of acute abdominal pain and is followed in frequency by nonspecific abdominal pain, malignancy, intestinal obstruction, complicated peptic ulcer disease, and incarcerated hernia. Appendicitis, although rare in older patients, usually manifests late in its course and is associated with high morbidity and mortality rates.

Pregnancy

The gravid woman with acute abdominal pain presents a difficult diagnostic dilemma. Pregnant women develop acute appendicitis and cholecystitis at the same rate as their nonpregnant counterparts. A number of additional diag­ noses, such as placental abruption and pain related to tension on the broad ligament, must be distinguished from nonobstetric diagnoses. Furthermore, the risk of radiation injury to the developing fetus must be considered when imaging studies are planned. Surgery in pregnancy is not rare; approximately 1 in 500 pregnancies will be associated with a nonobstetric general surgical intervention.50 Primary consideration is given to the health of the mother. The middle three months of gestation are optimal for abdominal surgical intervention, because this period presents the lowest risk for teratogenicity and spontaneous labor. Emergency interventions during pregnancy carry a risk of fetal loss that varies with the type of intervention and the age of gestation. Appendicitis occurs in approximately 1 in 2000 pregnancies and is equally distributed among the three trimesters. In later stages of pregnancy, the appendix may be displaced cephalad, with consequent displacement of the signs of peritoneal irritation away from McBurney’s point. Ultrasound or, in challenging cases, MRI may be useful for establishing a diagnosis in this setting. Biliary tract disease is also common during pregnancy. Open or laparoscopic management of these diseases is safe but is associated with a rate of preterm delivery of approximately 12% for appendectomy and 11% for cholecystectomy.51

Immunocompromised Hosts

In addition to diseases that occur in the general population, such as appendicitis and cholecystitis, a number of diseases unique to immunocompromised hosts may manifest with acute abdominal pain, including neutropenic enterocolitis, drug-induced pancreatitis, graft-versus-host disease, pneumatosis intestinalis, and cytomegalovirus (CMV) and fungal infections. Patients infected with human immunodeficiency virus (HIV) can present a particular challenge. When advanced, HIV infection is associated with a number of other diseases that may present as acute abdominal pain. One of the most common abdominal disorders seen in immunocompromised persons in the developing world is primary peritonitis. Affected patients have suppurative peritonitis without a definable source. Spontaneous intes­ tinal perforation, usually secondary to CMV infection, is also common in patients with advanced HIV infection. Tuberculous peritonitis is a consideration in patients from areas in which tuberculosis is common.52 In general, immunocompromised patients may lack the definitive signs of an acute abdominal crisis usually seen in immunocompetent persons; an elevated temperature, peritoneal signs, and leukocytosis may be absent in these cases.

PHARMACOLOGIC MANAGEMENT An unfortunate practice in the care of patients with acute abdominal pain is the delay in administration of narcotics pending definitive surgical assessment. Sir Zachary Cope stated that “Morphine does little or nothing to stop serious intra-abdominal disease, but it puts an efficient screen in front of the symptoms.”53 The practice of delaying relief of pain in a suffering patient, however, does not appear to withstand careful clinical scrutiny. Six studies in which the early administration of analgesia was compared with administration of placebo in patients with acute abdominal pain have shown that the patients who receive analgesics are more comfortable and do not experience a delay in diagnosis.54 Patients with acute abdominal pain frequently are suffering the most intense pain that they have ever experienced and should receive appropriate opioid analgesics early in their care. Patients with acute abdominal processes frequently require antibiotic treatment for peritonitis. When appro­ priate, antibiotic therapy aimed at the likely causative pathogens should be given as soon as a putative diagnosis is reached; little benefit is derived from treating an immunocompetent patient with broad-spectrum antibiotics before a likely source is identified. Patients who are immunocompromised or neutropenic are an exception to this rule. They should receive broad-spectrum antibiotics early in the course of management for acute abdominal pain (see Chapters 37 and 91).

KEY REFERENCES

An G, West M. Abdominal compartment syndrome: A concise clinical review. Crit Care Med 2008; 36:1304-10. (Ref 49.) Addiss DG, Shaffer N, Fowler B, Tauxe RV. The epidemiology of appendicitis and appendectomy in the United States. Am J Epidemiol 1990; 132:910-25. (Ref 21.) Birnbaum BA, Wilson SR. Appendicitis at the millennium. Radiology 2000; 215:337-48. (Ref 23.) Bohner H, Yang Q, Franke C, et al. Simple data from history and phy­ sical examination help to exclude bowel obstruction and to avoid radiographic studies in patients with acute abdominal pain. Eur J Surg 1998; 164:777-84. (Ref 2.)

161

162

Section III  Symptoms, Signs, and Biopsychosocial Issues Cotton M. The acute abdomen and HIV. Trop Doct 2006; 36:198-200. (Ref 52.) Diaz JJ, Bokhari F, Mowery NT, et al. Guidelines for management of small bowel obstruction. J Trauma 2008; 64:1651-64. (Ref 33.) Frossard JL, Steer ML, Pastor CM. Acute pancreatitis. Lancet 2008; 371:143-52. (Ref 39.) Jacobs DO. Diverticulitis. N Engl J Med 2007; 357:2057-66. (Ref 34.) Maerz L, Kaplan LJ. Abdominal compartment syndrome. Crit Care Med 2008; 36:S212-15. (Ref 48.) Manterola C, Asutdillo P, Losada H, et al. Analgesia in patients with acute abdominal pain. Cochrane Database Syst Rev 2007; (3):CD005660. (Ref 54.) McGory ML, Zingmond DS, Nanayakkara D, et al. Negative appendectomy rate: Influence of CT scans. Am Surg 2005; 71:803-8. (Ref 25.)

Parangi S, Levine D, Henry A, et al. Surgical gastrointestinal disorders during pregnancy. Am J Surg 2007; 193:223-32. (Ref 50.) Peng WK, Sheikh Z, Nixon SJ, Paterson-Brown S. Role of laparoscopic cholecystectomy in the early management of acute gallbladder disease. Br J Surg 2005; 92:586-91. (Ref 30.) Silen W. Cope’s early diagnosis of the acute abdomen, 18th ed. New York: Oxford University Press; 1991. p 301. (Ref 16.) Terasawa T, Blackmore CC, Brent S, Kohlwes RJ. Systematic review: Computed tomography and ultrasonography to detect acute appen­ dicitis in adults and adolescents. Ann Intern Med 2004; 141:537-46. (Ref 17.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

11 Chronic Abdominal Pain Joseph C. Yarze and Lawrence S. Friedman

CHAPTER OUTLINE Definition and Clinical Approach  163 Abdominal Wall Pain  164 Anterior Cutaneous Nerve Entrapment and Myofascial Pain Syndromes  164 Slipping Rib Syndrome  165 Thoracic Nerve Radiculopathy  165 Functional Abdominal Pain Syndrome  165 Epidemiology  165

The evaluation of any patient with a complaint of abdom­ inal pain is challenging. Abdominal pain can be benign and self-limited or a harbinger of a serious life-threatening disease (see Chapter 10). Chronic abdominal pain poses a particularly challenging clinical problem. Not only is the management of chronic abdominal pain a frequently daunt­ ing task, but also the possibility of overlooking a structural or organic disorder is always a concern. Many disorders discussed elsewhere in this text can produce chronic abdominal pain (Table 11-1). Many of these diagnoses require careful consideration and clinical interrogation, in addition to appropriate diagnostic testing, to discern whether the entity is indeed the cause of the patient’s pain. Diagnosis of a functional gastrointestinal disorder is gener­ ally considered once potential causes of organic chronic abdominal pain have been confidently excluded. Although the causes of chronic abdominal pain are varied, the pathophysiologic pathways that produce chronic pain are common to many of them. This chapter focuses on the neuromuscular causes of chronic abdominal pain and the functional abdominal pain syndrome (FAPS). FAPS serves as a model to illustrate many of the complex issues involved in caring for patients with chronic abdominal pain.

DEFINITION AND CLINICAL APPROACH Abdominal pain is considered chronic when it has been occurring constantly or intermittently over at least six months. Abdominal pain is considered acute when it has been occurring for several days and subacute when it has been occurring more than several days but less than 6 months. These arbitrary definitions are often helpful when formulating a differential diagnosis. The clinician initially must adopt a broad-based approach, which necessarily becomes more focused as the evaluation ensues. Impor­ tantly, although typical patterns of presentation are useful to remember, some patients, especially immunosuppressed and older persons, may present with atypical features. As for acute abdominal pain (see Chapter 10), the initial step in evaluating a patient with chronic abdominal pain is

Pathophysiology  165 Clinical Features  168 Diagnosis and Differential Diagnosis  169 Treatment  169 Role of Laparoscopy with Lysis of Adhesions  170

to elicit a detailed history from the patient. The chronology of the pain, including its abruptness of onset and duration, and its location and possible radiation should be deter­ mined. Visceral pain emanating from the digestive tract is perceived in the midline, because of the relatively sym­ metrical bilateral innervation of the organs, but is diffuse and poorly localized.1 Referred pain is ordinarily located in the cutaneous dermatomes that share the same spinal cord level as the affected visceral inputs.2 The patient should be questioned about the intensity and character of the pain, with the understanding that these parameters are subjective. The patient’s perception of precipitating, exacerbating, or mitigating factors may be useful when diagnostic possibil­ ities are considered. When initially attempting to determine whether the patient’s pain is caused by an organic or functional process, the clinician should search for clues in the patient’s history and physical examination that support or refute the diagnosis of a progressive, serious, chronic underlying illness. Such features in the history include fever, night sweats, appetite change, weight loss, and nocturnal awakening. A complete physical examination is indicated to look for evidence of a systemic disease. The abdominal examination should use a combination of inspection, auscultation, per­ cussion, and palpation. The most critical step for a patient with an acute exacerbation of chronic abdominal pain is to ascertain promptly whether a surgical abdomen is present (see Chapter 10). Although most causes of chronic abdomi­ nal pain do not require immediate surgical treatment, a complication related to a disease process ordinarily asso­ ciated with chronic abdominal pain may present acutely (e.g., intestinal perforation in a patient with inflammatory bowel disease). Furthermore, a patient who has experienced chronic abdominal pain may present with acute pain related to another disease process (e.g., acute mesenteric ischemia in a patient with underlying irritable bowel syndrome [IBS]). The abdomen should be auscultated to detect an abdominal bruit, because the presence of a bruit may suggest chronic mesenteric ischemia (intestinal angina). Abdominal palpation for the presence of organomegaly, masses, and ascites and examination for hernias are particularly per­ tinent. Other physical findings that suggest an underlying

163

164

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 11-1  Differential Diagnosis of Chronic or Recurrent Abdominal Pain Structural (or Organic) Disorders Inflammatory Appendicitis (Chapter 116) Celiac disease (Chapter 104) Eosinophilic gastroenteritis (Chapter 27) Fibrosing mesenteritis (mesenteric panniculitis) (Chapter 37) Inflammatory bowel diseases (Chapters 111 and 112) Pelvic inflammatory diseases Primary sclerosing cholangitis (Chapter 68) Vascular Celiac artery syndrome (Chapter 36) Mesenteric ischemia (Chapter 114) Superior mesenteric artery syndrome (Chapter 14) Metabolic Diabetic neuropathy Familial Mediterranean fever (Chapter 35) Hereditary angioedema Porphyria (Chapter 76) Neuromuscular Anterior cutaneous nerve entrapment syndrome Myofascial pain syndrome Slipping rib syndrome Thoracic nerve radiculopathy Other Abdominal adhesions (Chapter 119) Abdominal neoplasms (Chapters 29-32, 46, 54, 60, 69, 94, 121-123) Anaphylaxis (Chapter 9) Chronic pancreatitis (Chapter 59) Endometriosis (Chapter 124) Gallstones (Chapter 65) Hernias (Chapter 24) Intestinal malrotation (Chapter 96) Intestinal obstruction (Chapter 119) Lactose intolerance (Chapter 101) Peptic ulcer disease (Chapter 52) Small intestinal and pelvic lipomatosis (Chapter 37) Functional Gastrointestinal Disorders Biliary pain (gallbladder or sphincter of Oddi dysfunction)  (Chapter 63) Functional abdominal pain syndrome Functional (nonulcer) dyspepsia (Chapter 13) Gastroparesis (Chapter 48) Irritable bowel syndrome (Chapter 118) Levator ani syndrome (Chapter 125)

organic illness include signs of malnutrition (e.g., muscle wasting or edema), vitamin deficiencies, or extraintestinal processes (e.g., arthropathy or skin changes). Although not entirely specific, the closed eyes sign is often seen in patients with FAPS (see later). Similarly, Carnett’s sign and the hover sign (described later) may be seen in persons with abdominal wall pain. The laboratory evaluation can be helpful, but the clinician must first distill pertinent facets of the history and physical examination to focus the laboratory assessment. Injudicious use of laboratory testing is costly and can confuse the clinical picture and even lead to complications. It is worth emphasizing that an abnormal laboratory test result does not necessarily prove causality in relation to a patient’s chronic pain syndrome. The clinician must exercise the utmost discretion when ordering and interpreting the results of laboratory tests. Endoscopic and imaging studies have important roles in diagnosing and excluding many causes of chronic abdom­ inal pain. Upper endoscopy and colonoscopy, as well as capsule endoscopy, may be indicated in selected cases. Available imaging investigations include barium and radio­

nuclide studies, ultrasonography, computed tomography, magnetic resonance imaging, positron emission tomography (PET), and conventional angiography. The indications for each of these radiologic investigations differ, as do their potential to clarify an individual clinical situation. Endo­ scopic and radiologic testing in specific disorders is dis­ cussed in detail elsewhere in this text.

ABDOMINAL WALL PAIN ANTERIOR CUTANEOUS NERVE ENTRAPMENT AND MYOFASCIAL PAIN SYNDROMES

Anterior cutaneous nerve entrapment syndrome (ACNES) and myofascial pain syndrome (MFPS) are common causes of chronic abdominal wall pain. These syndromes share clinical, diagnostic, and treatment characteristics. The importance of recognizing these syndromes rests in provid­ ing the patient with an accurate diagnosis and effective treatment, as well as avoiding further expensive investiga­ tion and unnecessary surgical intervention. The abdominal wall should be suspected as the cause of symptoms when there is a complaint of chronic and unremitting abdominal pain that is unrelated to eating or bowel function but clearly related to movement. Although ACNES was initially described in the 1970s, it remains a frequently overlooked cause of chronic abdom­ inal pain.3,4 In ACNES, the pain is believed to occur when there is entrapment of a cutaneous branch of a sensory nerve that is derived from a neurovascular bundle emanating from spinal levels T7 to T12. The nerve entrapment may be related to pressure from an intra- or extra-abdominal lesion or to another localized process, such as fibrosis or edema. Pain emanating from the abdominal wall is discrete and localized, in contrast to pain originating from an intraabdominal source, which is diffuse and poorly localized. Patients usually point to the location of their pain with one finger, and the examiner can often localize the area of maximal tenderness to a region less than 2 cm in diameter. During physical examination, the patient often guards the affected area from the examiner’s hands (hover sign).5 Patients often note that activities associated with tightening of the abdominal musculature are associated with an exac­ erbation of pain and, during physical examination, the clini­ cian will note increased localized tenderness to palpation when the patient tenses the abdominal muscles (Carnett’s sign).6 In contrast, an increase in tenderness during relax­ ation of the abdominal musculature suggests an intraabdominal source of pain. In MFPS, pain emanates from myofascial trigger points in skeletal muscle.7 Causative factors include musculo­ skeletal trauma, vertebral column disease, intervertebral disc disease, osteoarthritis, overuse, psychological distress, and relative immobility. The exact pathophysiology of pain in MFPS remains unclear. Chronic abdominal wall pain may occur in patients with MFPS. Pain may be referred from another site, and the identification of trigger points (including those remote from the site of pain) is a useful physical finding. When attempting to identify a trigger point, the examiner uses a single finger to palpate a tender area. This is most often located in the central portion of a muscle belly, which may feel indurated or taut to palpation, and elicits a jump sign.8 This finding refers to a patient’s response by wincing, jerking away, or crying out as the myofascial trigger point is detected. Less commonly, trigger points may be located at sites such as the xiphoid process, costochondral junctions, or ligamentous and tendinous insertions.

Chapter 11  Chronic Abdominal Pain Treatment of ACNES and MFPS, when successful, not only improves symptoms, but also confirms the diagno­ sis.9,10 The treatment strategy depends on the severity of the symptoms. With mild and intermittent symptoms that are reproducibly precipitated by certain movements, simple reassurance and a recommendation to avoid such move­ ments may suffice. Non-narcotic analgesics, nonsteroidal anti-inflammatory drugs (NSAIDs), and application of heat can be used during exacerbations. Physical therapy may be beneficial, although no randomized studies have supported this treatment modality. For severe and persistent symp­ toms, injection therapy with a local anesthetic, with or without a glucocorticoid, is recommended.9-11 In one study of 136 patients in whom the history and physical examina­ tion suggested abdominal wall pain, and in whom benefit was noted with injection therapy, the diagnosis remained unchanged after a mean follow-up of four years in 97% of cases.12 In carefully selected patients with symptoms refrac­ tory to injection therapy, a prospective nonrandomized investigation has suggested that diagnostic laparoscopy with open exploration of abdominal trigger points may be beneficial.13 In this study, after intra-abdominal adhesions in close proximity to trigger points were lysed, subcutane­ ous nerve resection was performed. After a median postop­ erative follow-up of 37 months, 23 of 24 patients (96%) believed that this approach was beneficial in managing their previously intractable pain.

SLIPPING RIB SYNDROME

The slipping rib syndrome (SRS), which was described ini­ tially in the early 20th century,14,15 is an uncommonly rec­ ognized cause of chronic lower chest and upper abdominal pain. SRS ordinarily causes unilateral, sharp, often lancinat­ ing pain in the subcostal region. The acute pain may be followed by a more protracted aching sensation. The syn­ drome is associated with hypermobility of the costal car­ tilage at the anterior end of a false rib (rib 8, 9, or 10), with slipping of the affected rib behind the superior adjacent rib during contraction of the abdominal musculature. This slipping causes pain by a variety of potential mechanisms, including costal nerve impingement and localized tissue inflammation. The key to diagnosis is clinical awareness of the syndrome, in conjunction with use of the hooking maneuver; the clinician hooks his or her examining fingers underneath the patient’s lowest rib and, as the rib is moved anteriorly, the pain is reproduced and an audible pop or click is often heard.16 Conservative therapeutic measures often suffice but, on occasion, costochondral nerve block­ ade, response to which supports the diagnosis, or even surgical rib resection is required.17

THORACIC NERVE RADICULOPATHY

Disease related to thoracic nerve roots T7 through T12 may be responsible for abdominal pain. The disease processes that may cause this problem include neuropathy related to back and spine disorders, diabetes mellitus, and herpes zoster infection.18,19 Obtaining a complete history and per­ forming a careful physical examination of the patient, with attention to the possibility of a systemic disease and abnor­ mal neurologic and dermatologic findings, should lead to the correct diagnosis in most instances. Treatment depends on the specific underlying disease process.

FUNCTIONAL ABDOMINAL PAIN SYNDROME FAPS is a distinct medical disorder. Evidence suggests that the syndrome relates to central nervous system (CNS)

Table 11-2  Rome III Criteria for Functional Abdominal Pain Syndrome* Must include all the following: 1. Continuous or almost continuous abdominal pain 2. No or only occasional relationship of pain with physiologic events (e.g., eating, defecation, menses) 3. Some loss of daily functioning 4. Pain is not feigned (e.g., no malingering) 5. Insufficient symptoms to meet criteria for another functional gastrointestinal disorder that would explain the pain *Criteria fulfilled for the past three months with symptom onset at least six months prior to diagnosis.

amplification of normal regulatory visceral signals, rather than functional abnormalities in the gastrointestinal tract.20,21 The disorder is characterized by continuous, almost con­ tinuous, or at least frequently recurrent abdominal pain that is poorly related to bowel habits and often not well local­ ized. FAPS is properly understood as abnormal perception of normal (regulatory) bowel function rather than a motility disorder. The syndrome appears to be closely related to alterations in endogenous pain modulation systems, includ­ ing dysfunction of descending and cortical pain modulation circuits.21 The Rome III diagnostic criteria for FAPS are shown in Table 11-2.20,21 Studies that included patients who meet diagnostic criteria for FAPS have revealed that only rarely is an organic cause of chronic abdominal pain found during long-term follow-up.22,23 FAPS is commonly associated with other unpleasant somatic symptoms, and, when it persists or dominates the patient’s life, it usually is associated with chronic pain behaviors and comorbid psychological disturbances.24 Patients with FAPS typically define their illness as medical, and their symptoms tend to be more severe and associated with greater functional impairment than those of patients with IBS.24 Psychological disturbances, if present, must be considered as comorbid features of FAPS rather than as part of a primarily psychiatric problem.25 When compared with patients who have chronic back pain, those with chronic abdominal pain report significantly better physical func­ tioning, yet their overall perception of health is significantly worse.26

EPIDEMIOLOGY

Although the epidemiology of FAPS is incompletely known, in the U.S. Householder Survey of Functional Gastrointestinal Disorders, FAPS was estimated to be present in 2% of the sample and was less frequent than IBS (9%).27 A female predominance was noted (F:M = 1.5). Patients with FAPS missed more work days because of illness and had more physician visits than those without abdominal symptoms. A substantial proportion of patients are referred to gastroenterology practices and medical centers; they have a disproportionate number of health care visits and often undergo numerous diagnostic procedures and treatments.

PATHOPHYSIOLOGY

Chronic pain is a multidimensional (sensory, emotional, cognitive) experience explained by abnormalities in neuro­ physiologic functioning at the afferent, spinal, and CNS levels. Unlike acute pain arising from peripheral or visceral injury or disease, chronic functional pain is not associated with increased afferent visceral stimuli from structural abnormalities and tissue damage. FAPS is considered what is termed a biopsychosocial disorder related to dysfunction

165

166

Section III  Symptoms, Signs, and Biopsychosocial Issues MCC

Early life Genetics Environment

Primary somatosensory cortex

Limbic system

Thalamus

ACC Insula

Psychosocial factors Life stress Psychologic state Coping Social support CNS

Reticulothalamic tract Spinothalamic tract

Outcome

ENS

Medications Health provider visits Daily function Quality of life

Physiology Motility Sensation

FAPS Symptom experience Behavior Figure 11-1.  Biopsychosocial model of functional abdominal pain syndrome (FAPS). Consistent with a biopsychosocial model of illness, a person may be predisposed to FAPS because of factors (e.g., genetic, environmental) in early life. The patient’s symptoms and behavioral responses result from the interaction between psychosocial factors (e.g., life stress, social support) and gastrointestinal physiology (i.e., motility and sensation). FAPS relates to dysfunction in the brain-gut neuraxis with abnormal modulation of afferent signals from the gut that influences the symptoms experienced by the patient and that leads to increased use of health care resources and reduced quality of life. CNS, central nervous system; ENS, enteric nervous system.

of the brain-gut axis (see Chapter 21).25 As shown in Figure 11-1, the clinical expression of FAPS is derived from psy­ chological and intestinal physiologic input that interacts via the CNS-gut neuraxis. This model integrates the clinical, physiologic, and psychosocial features of FAPS into a com­ prehensible form, providing the basis for understanding psychological influences and application of psychopharma­ cologic treatments. Research relating to the pathophysiology of painful func­ tional gastrointestinal disorders has focused on the concepts of visceral hypersensitivity and alterations of brain-gut interactions. Visceral hypersensitivity is facilitated by upregulation of mucosal nociceptors and sensitization of vis­ ceral afferent nerves.28 Dysregulation of the brain-gut axis can be manifested as central enhancement of afferent vis­ ceral signals.29 The brain-gut dysregulation can, in turn, be initiated or modified by a variety of events. In a large-scale, prospective, controlled investigation of the development of chronic abdominal pain in women undergoing gynecologic surgery for nonpainful indications, pain developed signi­ ficantly more frequently in the surgical group (15%) than in a nonsurgical control group (4%). The development of chronic abdominal pain in the postoperative setting was predicted only by psychosocial, and not surgical, variables, implying that the development of pain is associated closely with central registration and amplification of the afferent signal. This study lends strong support to the biopsychosocial model, documenting the importance of cognitive and emotional input during the development of postoperative FAPS.

Spinoreticular tract Colon Figure 11-2.  Neuroanatomic pathways that mediate visceral pain sensation. The afferent transmission of visceral abdominal pain involves firstorder neurons that innervate the viscera and subsequently synapse in the dorsal horn of the spinal cord. Second-order neurons ascend from the dorsal horn of the spinal cord via the spinothalamic tract and the spinoreticular and reticulothalamic tracts to link in the thalamus with third-order neurons that then synapse in the limbic system, which contains the insula and anterior cingulate cortex (ACC), and in the primary somatosensory cortex. MCC, midcingulate cortex.

Ascending Visceral Pain Transmission

The afferent transmission of visceral abdominal pain involves first-order neurons that innervate the viscera, carry information to the thoracolumbar sympathetic nervous system, and subsequently synapse in the dorsal horn of the spinal cord. Second-order neurons cross and ascend from the dorsal horn via the spinothalamic and spinoretic­ ular tracts. These second-order neurons synapse in the thalamus with third-order neurons that synapse with the somatosensory cortex (sensory-discriminative component), which is involved in the somatotypic or point-specific local­ ization and intensity of afferent signals, and with the limbic system (motivational-affective component), which contains the anterior cingulate cortex (ACC; Fig. 11-2; see also Chapter 21). The insular cortex receives input from the sensory thalamus and the nucleus tractus solitarius and integrates visceral sensory and emotional information.31 The limbic system serves as a modulator of the pain experience, based on the individual’s emotional state, prior experiences, and cognitive interpretation of the signal. This multicompo­ nent integration of nociceptive information in the CNS explains the variability in the experience and reporting of pain.32 Motivational-affective regions of the CNS are important contributors to the chronic pain experience by modulating afferent sensory information from the intestine. This conceptual scheme of pain modulation has been demonstrated through PET imaging with the use of radiola­ beled oxygen.33 In a group of healthy subjects who immersed their hands in hot water, half were hypnotized to experience the immersion as painful and the other half as not painful or even pleasant. The changes in cortical activation were compared between the two groups, and no difference was found in activity in the somatosensory cortex; however, those who experienced pain had significantly greater activation of the ACC of the limbic system, which is involved in the affective component of the pain experience. Func­ tional brain imaging studies comparing patients with functional gastrointestinal disease and normal controls

Chapter 11  Chronic Abdominal Pain Thalamus Limbic system

ACC

PAG Locus coeruleus Caudal raphe nucleus Noradrenergic pathway Serotonergic pathway

Amygdala Rostral ventral medulla

Opioidergic pathway

Colon Figure 11-3.  The descending endorphin- or enkephalin-mediated inhibitory system. This network includes connections from the sensory cortex and limbic system (via the amygdala and thalamus), which have major links to the midbrain periaqueductal gray (PAG) matter, locus coeruleus, and medullary caudal raphe nucleus. Connections then project to neurons in the dorsal horn of the spinal cord. When activated, this system inhibits afferent impulses from peripheral nociceptive sites (e.g., the colon) to the brain. Endorphin activity, which has opioidergic properties, is facilitated by release of serotonin (serotonergic pathway) and possibly norepinephrine (noradrenergic pathway). ACC, anterior cingulate cortex.

have shown abnormal brain activation mainly in the motivational-affective pain regions, including the prefrontal cortex, ACC, amygdala, and insula.34 These regions gener­ ally show increased activation in patients with chronic pain, thereby suggesting abnormal afferent input as well as central modulation, which could be caused in part by increased attention to visceral stimuli, abnormal cognitive or affective processing of afferent input, or comorbid psy­ chiatric disorders.

Descending Modulation of Pain

According to the gate control theory, afferent transmission of visceral pain can be modulated by descending impulses from the cortex down to the visceral nerves.32 In this model, the central descending control of the gating system occurs primarily through the descending inhibitory system.35 This system is an endorphin- or enkephalin-based neural network that originates from the cortex and limbic system and descends to the spinal cord, with major links in the mid­ brain (periaqueductal gray) and medulla (caudal raphe nucleus; Fig. 11-3). This system inhibits nociceptive pro­ jection directly on the second-order neurons or indirectly via inhibitory interneurons in the spinal cord. Then, the dorsal horn of the spinal cord acts as a gate to modulate (i.e., increase or decrease) transmission of afferent impulses from peripheral nociceptive sites to the CNS. In effect, this descending pain modulation system determines the amount of peripheral afferent input from the gut that is allowed to ascend to the brain. Descending inhibitory systems can be diffuse and, when activated, inhibit pain sensitivity throughout the body—so-called diffuse noxious inhibitory control (DNIC). Patients with chronic pain syn­ dromes, including FAPS, appear to have an impaired ability to activate DNIC.36

Visceral Sensitization

Recurrent peripheral stimulation is thought to up-regulate afferent signals or inhibit descending pain control mecha­

nisms, thereby sensitizing the bowel and producing a state of visceral hyperalgesia (increased pain response to a noxious signal) and chronic pain. Several clinical studies have supported this concept, and the increase in pain appears to occur to a greater degree in patients with func­ tional gastrointestinal disorders than in healthy subjects.37 Furthermore, preoperative treatment with local or regional anesthesia or NSAIDs reduces the severity of postoperative pain,38 suggesting that the CNS response to peripheral injury can be modified by prior reduction of afferent input to the spinal cord and CNS. Conversely, recurrent peripheral injury, such as repeated abdominal operations, may sensi­ tize intestinal receptors, thereby making perception of even baseline afferent activity more painful (allodynia). Visceral sensitization may develop through different mechanisms at one or more levels of the neuraxis, including the mucosal level (via afferent silent nociceptors) and spinal level (spinal hyperexcitability). Patients with IBS may also experience hyperal­gesia. Studies of rectal balloon distention in patients with IBS have demonstrated that a greater proportion of patients report discomfort to balloon distention than normal volunteers at a given volume of inflation; in addition, the intensity of the discomfort in patients is higher than in the normal volunteers.39 Rectal hypersensitivity induced by repetitive painful rectal distention is seen in patients with IBS, but not FAPS.40 This observation supports the conten­ tion that IBS and FAPS are distinct functional gastrointes­ tinal disorders.

Biochemical Mechanisms of Sensitization

The biochemical basis of visceral sensitization is under active study, and this research may identify future targets for therapy. Serotonin (5-hydroxytryptamine [5-HT]) has received considerable attention because the gastrointestinal tract is its main source within the body.41 5-HT is found primarily in mucosal enterochromaffin cells, where it appears to serve as a neurotransmitter of the enteric nervous system (ENS) and as a paracrine molecule that signals other (e.g., vagal) neural activity. 5-HT mediates numerous gas­ trointestinal functions, and modulation of various receptor subtypes, such as 5-HT1, 5-HT3, and 5-HT4, and of 5-HT reuptake affects gastrointestinal sensorimotor function.

Role of the Central Nervous System

Although peripheral sensitization may influence the onset of pain, the CNS is critically involved in the predisposition to and perpetuation of chronic pain. In FAPS, the preemi­ nent role of the CNS is evident by the lack of peripheral motor or sensory abnormalities and the strong association with psychosocial disturbances. In addition, comorbid psy­ chiatric diagnoses, major life stressors, a history of sexual or physical abuse, poor social support, and maladaptive coping all are associated with more severe chronic abdomi­ nal pain and poorer health outcomes.31,42,43 These factors in patients with FAPS and other functional gastrointestinal pain conditions may impair or diminish descending inhibi­ tory pain pathways that act on dorsal horn neurons or may amplify visceral afferent signals.25,36,44 Prospective studies of patients with postinfection IBS (see Chapter 118) and post­ operative FAPS support the importance of the brain in the experience of gastrointestinal pain.30,45 Functional brain imaging has been useful in clarifying brain-gut interaction and has demonstrated that links between emotional distress and chronic pain may be medi­ ated through impairment in the ability of the limbic system to modulate visceral signals. The motivational-affective component of the central pain system, specifically the ACC (see Figs. 11-2 and 11-3), is dysfunctional in patients with

167

168

Section III  Symptoms, Signs, and Biopsychosocial Issues IBS and other chronic painful conditions. Functional mag­ netic resonance imaging (MRI) and PET brain imaging in response to rectal distention in patients with IBS have shown differential activation of the ACC in patients com­ pared with normal subjects46 and increased activation of the thalamus.46,47 Similar results have been found in patients with a history of abuse, somatization, and post-traumatic stress disorder. Furthermore, the return of ACC activity to baseline in depressed patients is associated with clinical improvement48 and predicts response to antidepressant treatment.49 As the pain and emotional distress of a patient with IBS improve, the activity within the ACC changes cor­ respondingly.50 A study of patients with IBS and an abuse history used functional brain MRI to show that during aver­ sive visceral stimulation (rectal balloon distention), differ­ ential activation of regions of the ACC occurred51; areas involved in pain facilitation (posterior and middle cingulate subregions) were stimulated, whereas activity in a region usually associated with pain inhibition (supragenual ante­ rior cingulate) was reduced. This study confirmed a strong association between visceral pain reporting and brain acti­ vation in predetermined brain regions involved in the affective and motivational aspects of the pain experience. The observed synergistic effect of IBS and abuse history on differential ACC activation suggests a mechanism to explain how afferent processing in the CNS can be associ­ ated with reporting of greater pain severity and poorer outcomes in this patient population. This and other research29,52-55 has suggested that dysregulation of central pain modulation is critical and may occur in various medical and psychological conditions. The challenge remains to alter this dysregulated afferent processing network repro­ ducibly and to reverse the findings on functional brain imaging studies (by pharmacologic, psychological, or other therapeutic means), with a concomitant improvement in patient outcomes.

Clinical Implications

The concept of FAPS as a dysregulation of CNS–enteric nervous system function at varying levels of the neuraxis, rather than a purely psychiatric or structural gastrointesti­ nal disorder, suggests that chronic pain results from enhanced pain perception as a result of combinations of the following: (1) activation of silent nociceptors; (2) dorsal horn transmission of impulses stimulated by release of cyto­ kines or other substances; and (3) chronic or frequently recurring psychosocial stresses that influence central pain modulation. By linking psychosocial factors to the patho­ physiology of chronic abdominal pain, this conceptual scheme alters the therapeutic approach from one that is purely psychiatric in nature to one that encompasses a broader array of potential therapies. Early pharmacologic and psychological treatment ultimately may be proven to prevent the development of a subsequent chronic pain syndrome.

CLINICAL FEATURES History

Typically, patients with FAPS are middle-aged and female. The history is one of chronic abdominal pain, often for more than 10 years, and the patient is often in distress at the time of initial consultation. The pain is frequently described as severe, constant, and diffuse. Pain is often a focal point in the patient’s life, may be described in emotional or bizarre terms (e.g., as nauseating or like a knife stabbing), and is not influenced by eating or defecation. The abdominal pain may be one of several painful symptoms or part of a continuum

of painful experiences often beginning in childhood and recurring over time.22 FAPS sometimes coexists with other disorders, and the clinician must determine the degree to which one of these other conditions contributes to the FAPS. Frequently, FAPS will evolve in a patient who has had another well-defined gastrointestinal disorder, but who has been operated on one or more times and, following these operations, has developed chronic abdominal pain. Repetitive surgery in such patients is often performed for alleged intestinal obstruction caused by adhesions. Patients with FAPS often have a psychiatric diagnosis of anxiety, depression, or somatization.24 They may minimize the role of psychological factors, possibly having learned in childhood that attention is more likely received when reporting illness but not emotional distress. A history of unresolved losses is a common feature.56 Symptoms fre­ quently worsen soon after these events and recur on their anniversaries or during holiday seasons. A history of sexual and physical abuse is frequent and is predictive of poor health, refractoriness to medical care, and a high number of diagnostic and therapeutic procedures and health care visits.43 Because patients do not usually volunteer an abuse history, physicians should inquire about this possibility, particularly in those with refractory symptoms.57 Finally, patients with FAPS may report poor social net­ works and exhibit ineffective coping strategies. They feel unable to decrease their symptoms and may “catastrophize”— that is, view their condition in pessimistic and morbid ways without any sense of control over the consequences. These cognitions are associated with greater pain scores that lead to a cycle of more illness reporting, more psychological distress, and poorer clinical outcomes.58 For many, the illness provides social support via increased attention from friends, family, and physicians.

Patient Behavior

Certain behavioral traits are common in patients with FAPS. Often, these patients demand that the physician not only diagnose the problem promptly, but also relieve their chronic symptoms rapidly. They similarly deny a relation­ ship between their problem and psychologically disturbing issues and often attribute depression to pain rather than recognizing it as a primary factor. Frequently, an accompa­ nying spouse or parent takes responsibility for reporting the patient’s history, an observation that suggests the possibility of family dysfunction. A history of narcotic use is not uncommon, as is a request by the patient for such medica­ tion during the initial visit. This type of behavior reflects the patient’s consideration of his or her situation as an acute condition requiring immediate symptom relief, rather than as a chronic condition in which treatment must be directed toward enhancing coping and adaptive strategies.

Physical Examination

Certain physical findings help support a diagnosis of FAPS, yet none is perfectly sensitive or specific. Abdominal palpa­ tion should begin at an area remote from the perceived site of maximal intensity. The patient’s behavior during abdomi­ nal palpation should be noted, with an emphasis on whether a change is noted during distracting maneuvers. Patients with FAPS usually lack signs of autonomic arousal. The presence of multiple abdominal surgical scars without clearly understood indications may suggest chronic pain behaviors that have led to unnecessary procedures. The closed eyes sign may be noted59; when the abdomen is pal­ pated, the patient with FAPS may wince, with her or his eyes closed, whereas those with acute pain caused by organic pathology tend to keep their eyes open in fearful

Chapter 11  Chronic Abdominal Pain anticipation of the examination. Often, the stethoscope sign (i.e., gentle, distracting compression on a painful site of the abdomen with the diaphragm of the stethoscope), elicits a diminished behavioral response in a patient with FAPS, thereby affording a more accurate appraisal of the complaint of pain.

DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS

After obtaining a complete history, performing a thorough physical examination, and paying appropriate attention to psychosocial factors in the patient’s life, the scenario will often point the physician toward a diagnosis of FAPS. A physical examination that does not suggest evidence of organic intra-abdominal pathology, as well as normal results of a battery of routine laboratory tests, lends support to the contention that the patient’s pain is not the result of an identifiable structural disease. Recognition of the diagnostic criteria for FAPS (see Table 11-2), and failure to find evi­ dence of another cause of chronic abdominal pain (see Table 11-1), should lead the physician to a diagnosis of FAPS. If the features of FAPS are absent or atypical, or if concerning abnormalities are found on physical examina­ tion (e.g., abdominal mass, enlarged liver) or on screening laboratory studies (e.g., anemia, hypoalbuminemia), another diagnosis should be considered and pursued accordingly. Not uncommonly, nonspecific abnormalities are found (e.g., a liver cyst) and require determination of their relevance to the patient’s symptoms.

TREATMENT Establishing a Successful Patient-Physician Relationship

Once other diagnoses have been excluded, formation of a successful relationship between the physician and patient with FAPS is necessary for effective management. Several factors must be taken into account to help establish this relationship and move toward successful treatment. An understanding of the psychosocial background is helpful, because a detailed knowledge of this aspect of the patient’s life aids in selecting the most useful treatment strategies. Having an appreciation of the degree of the patient’s under­ standing of the illness is also important, particularly for enhancing the success of a treatment plan. Early in the development of the patient-physician rela­ tionship, it is important to determine whether there are abnormal illness behaviors and associated psychiatric diag­ noses, which are often present in patients with FAPS. The role of the family in relation to the patient’s illness should also be understood. Normally, family experiences with illness lead to emotional support and a focus on recovery. With dysfunctional family interactions, stresses are not managed in an optimal fashion, and diverting attention toward illness serves to reduce family distress.60 Dysfunc­ tion is seen when family members indulge the patient, assume undue responsibility in the patient’s management, or become the spokesperson for the patient. If such family dysfunction is observed, counseling may help the family develop more useful coping strategies. Cultural belief systems must also be understood, because patients may not comply with treatments that are inconsistent with their cultural values. It is important to gain knowledge of the patient’s psychosocial resources (i.e., the availability of social networks) that may assist in buffering the adverse effects of stress and improve the outcome. It is essential for the physician to convey validation of illness to the patient by acknowledging the patient’s illness and the effect it has had on his or her life in a nonjudgmen­ tal fashion. This step is important in ensuring that the

patient understands that the physician considers FAPS to be a medical illness. Empathy is primary, because it acknowledges the reality and distress associated with the patient’s pain. Providing an empathetic approach can provide benefit by improving adherence to a treatment plan, patient satisfaction, and clinical outcomes.61 It does not, however, equate with overreacting to the patient’s wish for a rapid diagnosis and overmedication or performing unnec­ essary diagnostic studies. Education is provided by eliciting the patient’s knowledge of the syndrome, addressing any concerns, explaining the nature of the symptoms, and ensuring understanding in all matters that have been dis­ cussed. It is helpful to reiterate that FAPS is a medical disorder and that symptoms can be attenuated by pharma­ cologic or psychological treatments that modify the regula­ tion of pain control. Reassurance should be provided, because patients may fear serious disease. After the evalu­ ation is complete, the physician should respond to the patient’s concerns in a clear, objective, and nondismissive manner. Both patient and physician must then negotiate the treatment. This approach will enable the patient to contrib­ ute to and take some responsibility for the treatment plan. Within the context of the patient’s prior experience, inter­ ests, and understanding, the physician should provide choices rather than directives. Adherence to a treatment plan is more likely when the patient has confidence that it will benefit him or her and its rationale is understood. Finally, the physician must set reasonable limits in relation to time and effort expended. The key to success is to maintain a trusting relationship, while setting proper boundaries.

Instituting a Treatment Plan

Successful treatment rests on formulating a plan that encom­ passes ongoing interviews to ensure that the patient does not expect a cure. The physician should explain that a real­ istic treatment goal is to attenuate the symptoms and improve daily function. The patient should increase his or her responsibility for the illness by identifying the circum­ stances surrounding episodes of pain, including emotional and cognitive responses. This technique helps the patient achieve insight into aggravating factors and also character­ izes the patient’s coping style. Such information helps identify a strategy for behavioral treatment. The treatment chosen should be based on the severity of symptoms and degree of associated disability. Symptoms that are intermit­ tent and less severe and those that are clearly linked to psychological distress are frequently amenable to psycho­ logical treatment. If the pain is continuous and severe, phar­ macotherapy targeted to achieve central analgesia may be helpful.

Pharmacotherapy

There is a paucity of evidence from prospective, random­ ized, controlled trials to support the use of drug therapy in FAPS. Drug development in the area of functional gastroin­ testinal disorders, particularly FAPS, has been slow. A major reason for this slow progress is the rather empirical process for experimental testing that necessarily occurs in a symptom-based syndrome.62 Pharmacologic brain imaging approaches hold promise as a means to accelerate drug discovery and subsequent development.63 Despite these limitations, some specific medications have been used in the treatment of FAPS (see later). Peripherally acting anal­ gesics (e.g., acetaminophen, aspirin, other NSAIDs) offer little benefit to patients with FAPS, given the pathophysiol­ ogy of the disorder (i.e., a biopsychosocial disorder related to dysfunction of the brain-gut neuraxis). Moreover, narcot­

169

170

Section III  Symptoms, Signs, and Biopsychosocial Issues ics and benzodiazepines should not be prescribed for treat­ ment of FAPS, because of the potential for increased pain sensitivity and a lowering of the pain threshold, respec­ tively. Furthermore, the omnipresent potential for drug dependency with these types of medications must be borne in mind. Importantly, prescribing such medications subor­ dinates the development of more comprehensive treatment strategies to that of providing medication, can be counter­ productive by leading to narcotic-induced potentiation of visceral pain, and can thus result in the narcotic bowel syndrome.63 Narcotic bowel syndrome may occur in patients with other gastrointestinal disorders or in patients with no other structural intestinal disease who have been exposed to high doses of narcotic medication. The clinical scenario is dominated by chronic abdominal pain that continues to worsen, despite the use of escalating doses of narcotics. The keys to successful management of this disorder include the timely recognition of the syndrome followed by the estab­ lishment of an effective physician-patient relationship and graded tapering of the narcotic, with simultaneous insti­ tution of medical therapy to mitigate the effects of opiate withdrawal.63 As in the treatment of other chronic pain disorders, tricy­ clic antidepressants (TCAs) can be helpful in FAPS.64-66 The benefit of these medications is derived from their ability to improve pain directly and treat associated depression. In general, TCAs have been shown to be effective but can cause anticholinergic effects, hypotension, sedation, and cardiac arrhythmias. They can be given in dosages lower than those used to treat major depression (e.g., desipramine, 25 to 100 mg/day at bedtime) to reduce side effects. However, dosage increases may be needed, particularly if the patient has psychiatric comorbidity. There is less evi­ dence for the use of selective serotonin reuptake inhibitors (SSRIs) in FAPS. These medications may cause agitation, sleep disturbance, vivid dreams, and diarrhea but are much safer than TCAs if taken in an overdose. In most cases, administration of a single daily dose (e.g., 20 mg of fluox­ etine, paroxetine, or citalopram) will suffice. Although the efficacy of SSRIs for pain control is not well established, this class of drugs has additional benefits because they are anxiolytic and helpful for patients with social phobia, posttraumatic stress disorder, panic disorder, and obsessional thoughts related to their condition. Drug combinations (e.g., TCAs with SSRIs) have little support for their use in patients with functional gastrointestinal disorders.67 Anticonvulsants such as carbamazepine and gabapentin have been evaluated in other chronic pain syndromes but have no proven efficacy in FAPS. These drugs may find a role as adjunctive agents in the future. As is the case for other peripherally acting analgesics, topical capsaicin would not be expected to be helpful in the management of FAPS.68 Leuprolide acetate may be of benefit for premen­ strual females with FAPS,69 but the consequent reproduc­ tive hormonal effects of this therapy have dampened enthusiasm for this approach. To enhance compliance, especially in the case of TCA use, the physician should explain that these medications work as central analgesics and are not simply being used to treat a psychiatric condition. Investing the time to explain that these drugs induce neurotransmitter changes in the brain and thereby alter pain perception, and that the dosage is usually lower than that typically chosen for treatment of psychiatric disorders, is often helpful. Further, it may be beneficial to emphasize that the lag time for clinical effect may be several weeks; most side effects diminish after a few days and can be reduced by temporarily lowering the dose of the drug.

Mental Health Referral and Psychological Treatments

Patients may be reluctant to see a psychologist or psychia­ trist because they lack knowledge of the benefits of referral, feel stigmatized for being thought to have a psychiatric problem, or see referral as a rejection by the medical physi­ cian. Psychological interventions are best presented as vehi­ cles that are orchestrated in parallel with medical visits and are used to help manage pain and reduce the psychological distress caused by the symptoms. The mental health consultant may recommend any of several types of psychological treatments for pain manage­ ment.21,70 Cognitive-behavioral treatment, which identifies maladaptive thoughts, perceptions, and behaviors, may be beneficial.65 Evidence from functional brain imaging sug­ gests that this psychological intervention decreases activa­ tion from rectal stimulation in the central emotional regions that are typically hyperactive in chronic pain, such as the amygdala, ACC, and frontal cortex.71 Hypnotherapy has been investigated primarily in IBS, where the focus is on relaxation of the gut. A randomized, controlled trial in chil­ dren that included 31 patients with FAPS has concluded that hypnotherapy is superior to standard medical therapy in reducing pain at one year of follow-up.72 Dynamic or interpersonal psychotherapy and relaxation training have less evidence to support their use in FAPS.

ROLE OF LAPAROSCOPY WITH LYSIS OF ADHESIONS The value of laparoscopy with lysis of adhesions (adhesioly­ sis) in patients with chronic abdominal pain continues to be debated. Relevant studies generally have often been retrospective and nonrandomized, with varying criteria for selecting patients and durations of follow-up. Therefore, the role of adhesiolysis is difficult to assess. Prospective observational investigations have shown improvement in 45% to 90% of patients.73-76 Perhaps most provocative is a prospective, blinded, randomized investigation performed by Swank and colleagues in which patients who were found at laparoscopy to have adhesions were randomized to undergo adhesiolysis or no treatment.77 At 12 months of follow-up, patients in both groups reported substantial pain relief and improved quality of life; however, there were no differences between the groups. The authors concluded that laparoscopic adhesiolysis could not be recommended in this setting. Given these somewhat conflicting data, it seems reasonable to withhold laparoscopy in most patients with chronic abdominal pain, with the understanding that, on occasion, the procedure may be of some benefit. The chal­ lenge for the future will be to define which patients will benefit from such intervention.

ACKNOWLEDGMENT

The authors thank Dr. Douglas A. Drossman for his expert discussion of FAPS in previous editions of this text.

KEY REFERENCES

Bixquert-Jiménez M, Bixquert-Pla L. Antidepressant therapy in func­ tional gastrointestinal disorders. Gastroenterol Hepatol 2005; 28:48592. (Ref 66.) Clouse RE, Mayer EA, Aziz Q, et al. Functional abdominal pain syn­ drome. In: Drossman DA, Corazziari E, Delvaux M, et al, editors. Rome III. The functional gastrointestinal disorders. 3rd ed. McLean, Va: Degnon Associates: 2006. p 557. (Ref 21.) Drossman DA, Ringel Y, Vogt B, et al. Alterations in brain activity associated with resolution of emotional distress and pain in a case of severe IBS. Gastroenterology 2003; 124:754-61. (Ref 50.)

Chapter 11  Chronic Abdominal Pain Drossman DA. Brain imaging and its implications for studying centrally targeted treatments in irritable bowel syndrome: A primer for gastro­ enterologists. Gut 2005; 54:569-73. (Ref 29.) Drossman DA. Functional abdominal pain syndrome. Clin Gastroenterol Hepatol 2004; 2:353-65. (Ref 24.) Kuan LC, Li YT, Chen FM, et al. Efficacy of treating abdominal wall pain by local injection. Taiwan J Obstet Gynecol 2006; 45:239-43. (Ref 10.) Lackner JM, Lou Coad M, Mertz HR, et al. Cognitive therapy for irritable bowel syndrome is associated with reduced limbic activity, GI symp­ toms, and anxiety. Behav Res Ther 2006; 44:621-38. (Ref 71.) Mayer EA, Naliboff BD, Craig AD. Neuroimaging of the brain-gut axis: From basic understanding to treatment of functional GI disorders. Gastroenterology 2006; 131:1925-42. (Ref 54.) Peterson LL, Cavanaugh DL. Two years of debilitating pain in a football spearing victim: Slipping rib syndrome. Med Sci Sports Exerc 2003; 35:1634-7. (Ref 17.)

Ringel Y, Drossman DA, Leserman JL, et al. Effect of abuse history on pain reports and brain responses to aversive visceral stimulation: An FMRI study. Gastroenterology 2008; 134:396-404. (Ref 51.) Ringel Y. New directions in brain imaging research in functional gas­ trointestinal disorders. Dig Dis 2006; 24:278-85. (Ref 52.) Sperber AD, Morris CB, Greemberg L, et al. Development of abdominal pain and IBS following gynecological surgery: A prospective, con­ trolled study. Gastroenterology 2008; 134:75-84 (Ref 30.) Swank DJ, Swank-Bordewijk SCG, Hop WCJ, et al. Laparoscopic adhesiolysis in patients with chronic abdominal pain: A blinded randomized controlled multi-centre trial. Lancet 2003; 361:1247-51. (Ref 77.) Vlieger AM, Menko-Frankenhuis C, Wolfkamp SCS, et al. Hypnotherapy for children with functional abdominal pain or irritable bowel syndrome: A randomized controlled trial. Gastroenterology 2007; 133:1430-6. (Ref 72.) Full references for this chapter can be found on www.expertconsult.com.

171

CHAPTE R

12 Symptoms of Esophageal Disease Kenneth R. DeVault

CHAPTER OUTLINE Dysphagia  173 Pathophysiology  173 Differential Diagnosis and Approach  174 Odynophagia  176 Globus Sensation  176 Pathophysiology and Approach  176 Hiccups  177

Symptoms related to the esophagus are among the most common in general medical as well as gastroenterologic practice. For example, dysphagia becomes more common with aging and affects up to 15% of persons age 65 or older.1 Heartburn, regurgitation, and other symptoms of gastroesophageal reflux disease (GERD) also are common. A survey of healthy subjects in Olmsted County, Minnesota, found that 20% of persons, regardless of gender or age, experienced heartburn at least weekly.2 Mild symptoms of GERD rarely indicate severe underlying disease but must be addressed, especially if they have occurred for many years. Frequent or persistent dysphagia or odynophagia suggests an esophageal problem that necessitates investigation and treatment. Other less specific symptoms of possible esophageal origin include globus sensation, chest pain, belching, hiccups, rumination, and extraesophageal complaints, such as wheezing, coughing, sore throat, and hoarseness, especially if other causes have been excluded. A major challenge in the evaluation of esophageal symptoms is that the degree of esophageal damage often does not correlate well with the patient’s or physician’s impression of symptom severity. This is a particular problem in older patients, in whom the severity of gastroesophageal reflux–induced injury to the esophageal mucosa is increased despite an overall decrease in the severity of symptoms.3

DYSPHAGIA Dysphagia, from the Greek dys (difficulty, disordered) and phagia (to eat), refers to the sensation that food is hindered in its passage from the mouth to the stomach. Most patients complain that food sticks, hangs up, or stops, or they feel that the food “just won’t go down right.” Occasionally they complain of associated pain. If asked, “Do you have trouble swallowing?” some patients with dysphagia in the lower esophagus will actually say “no” in that they may only think of swallowing as the transfer of food from the mouth to the esophagus. Dysphagia always indicates malfunction of

Chest Pain of Esophageal Origin  177 Pathophysiology and Approach  177 Heartburn and Regurgitation  178 Pathophysiology and Approach  179 Extraesophageal Symptoms of Gastroesophageal Reflux Disease  179

some type in the oropharynx or esophagus, although associated psychiatric disorders can amplify this symptom.

PATHOPHYSIOLOGY

The inability to swallow is caused by a problem with the strength or coordination of the muscles required to move material from the mouth to the stomach or by a fixed obstruction somewhere between the mouth and the stomach. Occasional patients may have a combination of the two processes. The oropharyngeal swallowing mechanism and the primary and secondary peristaltic contractions of the esophageal body that follow usually transport solid and liquid boluses from the mouth to the stomach within 10 seconds (see Chapter 42). If these orderly contractions fail to develop or progress, the accumulated bolus of food distends the esophageal lumen and causes the discomfort that is associated with dysphagia. In some patients, particularly older adults, dysphagia is the result of low-amplitude primary or secondary peristaltic activity that is insufficient to clear the esophagus. Other patients have a primary or secondary motility disorder that grossly disturbs the orderly contractions of the esophageal body. Because these motor abnormalities may not be present with every swallow, dysphagia may wax and wane (see Chapter 42). Mechanical narrowing of the esophageal lumen may interrupt the orderly passage of a food bolus despite adequate peristaltic contractions. Symptoms vary with the degree of luminal obstruction, associated esophagitis, and type of food ingested. Although minimally obstructing lesions cause dysphagia only with large, poorly chewed boluses of foods such as meat and dry bread, lesions that obstruct the esophageal lumen completely lead to symptoms with solids and liquids. GERD may produce dysphagia related to an esophageal stricture, but some patients with GERD clearly have dysphagia in the absence of a demonstrable stricture, and perhaps even without esophagitis.4 Abnormal sensory perception in the esophagus may lead to the perception of dysphagia, even when the bolus has cleared the esophagus. Because some normal subjects experience the sensation of dysphagia when the distal esophagus

173

174

Section III  Symptoms, Signs, and Biopsychosocial Issues is distended by a balloon, as well as by other intraluminal stimuli, an aberration in visceral perception could explain dysphagia in patients who have no definable cause.5 This mechanism also may apply to the amplification of symptoms in patients with spastic motility disorders, in whom the frequency of psychiatric disorders is increased.6

DIFFERENTIAL DIAGNOSIS AND APPROACH

When faced with a patient who complains of dysphagia, the physician should approach the problem in a systematic way. Most patients can localize dysphagia to the upper or lower portion of the esophagus, although occasional patients with a distal esophageal cause of dysphagia will present with symptoms referred only to the suprasternal notch or higher. The approach to dysphagia can be divided into oropharyngeal and esophageal dysphagia, although con­ siderable overlap may occur in certain groups of patients. In addition, an attempt should be made to determine whether the patient has difficulty only with solid boluses or with liquids and solids.

Oropharyngeal Dysphagia

Processes that affect the mouth, hypopharynx, and upper esophagus produce a distinctive type of dysphagia. The patient often is unable to initiate a swallow and repeatedly has to attempt to swallow. Patients frequently describe coughing or choking when they attempt to eat. The inability to propel a food bolus successfully from the hypopharyngeal area through the upper esophageal sphincter (UES) into the esophageal body is called oropharyngeal, or transfer, dysphagia. The patient is aware that the bolus has not left the oropharynx and locates the site of symptoms specifically to the region of the cervical esophagus. Dysphagia that occurs immediately or within one second of swallowing suggests an oropharyngeal abnormality. At times, a liquid bolus may enter the trachea or nose rather than the esophagus. Some patients describe recurrent bolus impactions that require manual dislodgment. In severe cases, saliva cannot be swallowed, and the patient drools. Abnormalities of speech such as dysarthria or nasal speech may be associated with oropharyngeal dysphagia. Oral pathology should be considered as well. For example, poor teeth or poorly fitting dentures may disrupt mastication and result in an attempt to swallow an overly large or poorly chewed bolus. Loss of salivation—caused by medications, radiation, or primary salivary dysfunction—may result in a bolus that is difficult to swallow. Recurrent bouts of pulmonary infection may reflect spillover of food into the trachea because of inadequate laryngeal protection. Hoarseness may result from recurrent laryngeal nerve dysfunction or intrinsic muscular disease, both of which cause ineffective vocal cord movement. Weakness of the soft palate or pharyngeal constrictors causes dysarthria and nasal speech as well as pharyngonasal regurgitation. Swallowing associated with a gurgling noise may be described by patients with Zenker’s diverticulum. Finally, unexplained weight loss may be the only clue to a swallowing disorder; patients avoid eating because of the difficulties encountered. Potential causes of oropharyngeal dysphagia are shown in Table 12-1. After an adequate history is obtained, the initial test is a carefully conducted barium radiographic examination, which is optimally performed with the assistance of a swallowing therapist (modified barium swallow). If the study is normal with liquid barium, the examination is repeated after the patient is fed a solid bolus in an attempt to bring out the patient’s symptoms and thereby aid in localizing any pathology. If the oropharyngeal portion of the study is

Table 12-1  Causes of Oropharyngeal Dysphagia Neuromuscular Causes* Amyotrophic lateral sclerosis (ALS, Lou Gehrig’s disease) CNS tumors (benign or malignant) Idiopathic UES dysfunction Manometric dysfunction of the UES or pharynx† Multiple sclerosis Muscular dystrophy Myasthenia gravis Parkinson’s disease Polymyositis or dermatomyositis Postpolio syndrome Stroke Thyroid dysfunction Structural Causes Carcinoma Infections of pharynx or neck Osteophytes and other spinal disorders Prior surgery or radiation therapy Proximal esophageal web Thyromegaly Zenker’s diverticulum *Any disease that affects striated muscle or its innervation may result in dysphagia. † Many manometric disorders (hypertensive and hypotensive UES, abnormal coordination, and incomplete UES relaxation) have been described, although their true relationship to dysphagia is often unclear. CNS, central nervous system; UES, upper esophageal sphincter.

normal, the remainder of the esophagus should be examined. This single test usually identifies the problem and directs initial therapy.

Esophageal Dysphagia

Most patients with esophageal dysphagia localize their symptoms to the lower sternum or, at times, the epigastric region. A smaller number of patients will describe a sensation in the suprasternal notch or higher, even though the bolus stops in the lower esophagus. Esophageal dysphagia frequently can be relieved by various maneuvers, including repeated swallowing, raising the arms over the head, throwing the shoulders back, and using the Valsalva maneuver. Motility disorders or mechanical obstructing lesions can cause esophageal dysphagia. To clarify the origin of symptoms of esophageal dysphagia, the answers to three questions are crucial: 1. What type of food or liquid causes symptoms? 2. Is the dysphagia intermittent or progressive? 3. Does the patient have heartburn? On the basis of these answers, distinguishing the several causes of esophageal dysphagia (Table 12-2) as a mechanical or a neuromuscular defect and postulating the specific cause are often possible (Fig. 12-1). Patients who report dysphagia with solids and liquids are more likely to have an esophageal motility disorder than mechanical obstruction. Achalasia is the prototypical esophageal motility disorder in which, in addition to dysphagia, many patients complain of bland regurgitation of undigested food, especially at night, and of weight loss. By contrast, patients with spastic motility disorders such as diffuse esophageal spasm may complain of chest pain and sensitivity to hot or cold liquids. Patients with scleroderma of the esophagus usually have Raynaud’s phenomenon and severe heartburn. In these patients, mild complaints of dysphagia can be caused by a motility disturbance or esophageal inflammation, but severe dysphagia almost always signals the presence of a peptic stricture (see Chapters 35 and 43).

Chapter 12  Symptoms of Esophageal Disease Oropharyngeal (see Table 12-1)

Esophageal (see Table 12-2)

Type of dysphagia

Video swallow study

Abnormal Address specific cause

Type of bolus

Normal Consider other causes (including esophageal dysphagia)

Solids only

Solids and liquids

Character

Character

Progressive

No weight loss

Age >50 or weight loss

Caustic stricture Diverticula Peptic stricture

Carcinoma

Intermittent

Progressive

Intermittent

Eosinophilic esophagitis Esophageal ring

Achalasia Chagas’ disease Scleroderma

Less specific motility disorder

Table 12-2  Common Causes of Esophageal Dysphagia Motility (Neuromuscular) Disorders Primary Disorders Achalasia Diffuse esophageal spasm Hypertensive LES Ineffective esophageal motility Nutcracker (high-pressure) esophagus Secondary Disorders Chagas’ disease Reflux-related dysmotility Scleroderma and other rheumatologic disorders Structural (Mechanical) Disorders Intrinsic Carcinoma and benign tumors Diverticula Eosinophilic esophagitis Esophageal rings and webs (other than Schatzki ring) Foreign body Lower esophageal (Schatzki) ring Medication-induced stricture Peptic stricture Extrinsic Mediastinal mass Spinal osteophytes Vascular compression LES, lower esophageal sphincter.

In patients who report dysphagia only after swallowing solid foods and never with liquids alone, a mechanical obstruction is suspected. A luminal obstruction of sufficiently high grade, however, may be associated with dysphagia for solids and liquids. If food impaction develops,

Figure 12-1.  Diagnostic algorithm for patients with dysphagia. For details of the approach to each type of dysphagia, see the text and tables. Less specific motility disorders include nutcracker esophagus, diffuse esophageal spasm, and other disorders of ineffective esophageal motility. (Modified from Castell DO, Donner MW. Evaluation of dysphagia: A careful history is crucial. Dysphagia 1987; 2:65-71.)

the patient frequently must regurgitate for relief. If a patient continues to drink liquid after the bolus impaction, large amounts of that liquid may be regurgitated. In addition, hypersalivation is common during an episode of dysphagia, thereby providing even more liquid to regurgitate. Episodic and nonprogressive dysphagia without weight loss is characteristic of an esophageal web or a distal esophageal (Schatzki) ring. The first episode typically occurs during a hurried meal, often with alcohol. The patient notes that the bolus of food sticks in the lower esophagus; it often can be passed by drinking large quantities of liquids. Many patients finish the meal without difficulty after the obstruction is relieved. The offending food frequently is a piece of bread or steak—hence the term steakhouse syndrome.7 Initially, an episode may not recur for weeks or months, but subsequent episodes may occur frequently. Daily dysphagia, however, is likely not caused by a lower esophageal ring (see Chapter 41). If solid food dysphagia is clearly progressive, the differential diagnosis includes peptic esophageal stricture and carcinoma. Benign esophageal strictures develop in some patients with GERD. Most of these patients have a long history of associated heartburn. Weight loss seldom occurs in patients with a benign lesion, because these patients have a good appetite and convert their diet to high-calorie soft and liquid foods to maintain weight (see Chapter 43). Patients with carcinoma differ from those with peptic stricture in several ways. As a group, the patients with carcinoma are older and present with a history of rapidly progressive dysphagia. They may or may not have a history of heartburn, and heartburn may have occurred in the past but not the present. Most have anorexia and weight loss (see Chapter 46). True dysphagia may be seen in patients with

175

176

Section III  Symptoms, Signs, and Biopsychosocial Issues pill, caustic, or viral esophagitis; however, the predominant complaint of patients with these acute esophageal injuries is usually odynophagia. Patients may present with food bolus impaction, and eosinophilic esophagitis should be considered in the differential diagnosis of all patients (particularly those who are young) who present with dysphagia (see Chapter 27).8 After a focused history of the patient’s symptoms is obtained, a barium radiograph, including a solid bolus challenge, is often advocated as the first test. Alternatively, many experts have advocated endoscopy as the first test, especially in patients with intermittent dysphagia for solid food suggestive of a lower esophageal ring or with pronounced reflux symptoms. The choice of the initial test should be based on local expertise and the preference of the individual health care provider. If the barium examination demonstrates an obstructive lesion, endoscopy is usually done for confirmation and biopsy. Endoscopy also permits dilation of strictures, rings, and neoplasms. Empirical dilation of the esophagus is often performed in patients with a suggestive history and normal endoscopic examination,9 but the safety and efficacy of this approach have been questioned.10 If the barium examination is normal, esophageal manometry is often performed to look for a motility disorder. Some patients with reflux symptoms and dysphagia, a normal barium study or endoscopy, or both, will respond to a trial of gastric acid suppressive therapy.

Table 12-3  Causes of Odynophagia Caustic Ingestion Acid Alkali Pill-Induced Injury Alendronate and other bisphosphonates Aspirin and other NSAIDs Emepronium bromide Iron preparations Potassium chloride (especially slow-release form) Quinidine Tetracycline and its derivatives Zidovudine Infectious Esophagitis Viral Cytomegalovirus Epstein-Barr virus Herpes simplex virus Human immunodeficiency virus Bacterial Mycobacteria (tuberculosis or Mycobacterium avium complex) Fungal Candida albicans Histoplasmosis Protozoan Cryptosporidium Pneumocystis Severe Reflux Esophagitis Esophageal Carcinoma NSAIDs, nonsteroidal anti-inflammatory drugs.

ODYNOPHAGIA Like dysphagia, odynophagia, or painful swallowing, is specific for esophageal involvement. Odynophagia may range from a dull retrosternal ache on swallowing to a stabbing pain with radiation to the back so severe that the patient cannot eat or even swallow his or her own saliva. Odynophagia usually reflects an inflammatory process that involves the esophageal mucosa or, in rare instances, the esophageal muscle. The most common causes of odynophagia include caustic ingestion, pill-induced esophagitis, radiation injury, and infectious esophagitis (Candida, herpesvirus, and cytomegalovirus; Table 12-3). In these diseases, dysphagia also may be present, but pain is the dominant complaint. Odynophagia is an infrequent complaint of patients with GERD and, when present, usually is associated with severe ulcerative esophagitis. In rare cases, a nonobstructive esophageal carcinoma can produce odynophagia. Because many of the diseases that cause odynophagia have associated symptoms and signs, a carefully taken history can often lead directly to a diagnosis. For example, a teenager who takes tetracycline for acne and in whom odynophagia develops most likely has pill dysphagia, an immunocompromised patient with odynophagia is likely to have an infectious cause, and a patient with GERD is likely to have severe peptic esophagitis. On the other hand, gastrointestinal endoscopy to visualize and obtain biopsies of the esophageal mucosa is required to confirm a specific diagnosis in most patients with odynophagia.

GLOBUS SENSATION Globus sensation is a feeling of a lump or tightness in the throat, unrelated to swallowing. Up to 46% of the general population experience globus sensation at one time or another.11 The sensation can be described as a lump, tight-

ness, choking, or strangling feeling, as if something is caught in the throat. Globus sensation is present between meals, and swallowing of solids or large liquid boluses may give temporary relief. Frequent dry swallowing and emotional stress may worsen this symptom. Globus sensation should not be diagnosed in the presence of dysphagia or odynophagia.

PATHOPHYSIOLOGY AND APPROACH

The detection of physiologic and psychological abnormalities in patients with globus sensation has been inconsistent and controversial. Although frequently suggested, manometrically identifiable UES dysfunction has not been iden­ tified directly as the cause of globus sensation. The UES also does not appear to be hyperresponsive to esophageal distention, acidification, or mental stress.12 Furthermore, esophageal distention can cause globus sensation unrelated to any rise in UES pressure, and stress can induce an increase in UES pressure that is not associated with globus sensation in normal subjects and in patients who complain of globus sensation. Heartburn has been reported in up to 90% of patients with globus sensation,13 yet documentation of esophagitis or abnormal gastroesophageal reflux by esophageal pH monitoring is found in fewer than 25%. Balloon distention of the esophagus produces globus sensation at lower balloon volumes in globus sufferers than in controls; this finding suggests that the perception of esophageal stretch may be heightened in some patients with globus sensation. Psychological factors may be important in the genesis of globus sensation. The most common associated psychiatric diagnoses include anxiety, panic disorder, depression, hypochondriasis, somatization, and introversion.14 Indeed, globus sensation is the fourth most common symptom in patients with somatization disorders.15 A combination of biological factors, hypochondriacal traits, and learned fear

Chapter 12  Symptoms of Esophageal Disease after a choking episode provides a framework for misinterpretation of the symptoms and intensifies the symptoms of globus or the patient’s anxiety.16 The approach to globus sensation involves excluding a more sinister underlying disorder and then offering symptom-driven therapy. A nasal endoscopy to rule out pharyngeal pathology and a barium swallow to rule out a fixed pharyngeal lesion are often helpful.17 If these studies are negative, trials of acid suppression with a proton pump inhibitor, medications directed at visceral sensitivity, or other psychologically based therapies are reasonable. If a patient has heartburn, then acid suppressive therapy is the first step, but reflux may be the cause of globus sensation, even in the absence of heartburn. A trial of a proton pump inhibitor (usually given twice daily, before meals) is diagnostic and therapeutic in some patients. Ambulatory reflux monitoring may show acid or nonacid reflux in some patients.18 Alternatively, if the patient has obvious anxiety and has already failed a trial of acid suppression, therapy directed toward the psychological component of the problem should be considered.

HICCUPS The symptom of hiccups (hiccoughs, singultus) is caused by a combination of diaphragmatic contraction and glottic closure. Therefore, it is not classically an esophageal symptom but is a common complaint in primary care and gastroenterology. Most cases of hiccups are idiopathic, but the symptom has been associated with many conditions (trauma, masses, infections) that affect the central nervous system, thorax, or abdomen. Gastrointestinal causes include GERD, achalasia, gastropathies, and peptic ulcer. Hiccups associated with uremia may be particularly difficult to control. They often occur after a large meal. Because most cases are self-limited, intervention is not usually required. The evaluation of chronic or difficult cases should include selected tests to exclude esophageal, thoracic, or systemic diseases. Because GERD has been associated with hiccups, a trial of acid suppressive therapy may be reasonable in some patients.19 Many agents have been used to suppress hiccups with varying success, including chlorpromazine, nifedipine, haloperidol, phenytoin, metoclopramide, bac­ lofen, and gabapentin.20 Alternative modalities, including acupuncture, also have been tried in refractory cases.21

CHEST PAIN OF ESOPHAGEAL ORIGIN Chest pain of esophageal origin may be indistinguishable to patients and their health care providers from angina pectoris. The esophagus and heart are anatomically adjacent and share innervation. In fact, once cardiac disease is excluded, esophageal disorders are probably the most common causes of chest pain. Of the approximately 500,000 patients in the United States who undergo coronary angio­ graphy yearly for presumed cardiac pain, almost 30% have normal epicardial coronary arteries; of these patients, esophageal diseases may account for the symptoms in 18% to 56%.22 Esophageal chest pain usually is described as a squeezing or burning substernal sensation that radiates to the back, neck, jaw, or arms. Although it is not always related to swallowing, the pain can be triggered by ingestion of hot or cold liquids. It may awaken the patient from sleep and can

worsen during periods of emotional stress. The duration of pain ranges from minutes to hours, and the pain may occur intermittently over several days. Although the pain can be severe, causing the patient to become ashen and to perspire, it often abates spontaneously and may be eased with antacids. Occasionally, the pain is so severe that narcotics or nitroglycerin are required for relief. Close questioning reveals that most patients with chest pain of esophageal origin have other esophageal symptoms; however, chest pain is the only esophageal complaint in about 10% of cases.23 The clinical history does not enable the physician to distinguish reliably between a cardiac and esophageal cause of chest pain. In fact, gastroesophageal reflux may be triggered by exercise24 and cause exertional chest pain that mimics angina pectoris, even during treadmill testing. Symptoms suggestive of esophageal origin include pain that continues for hours, retrosternal pain without lateral radiation, pain that interrupts sleep or is related to meals, and pain relieved with antacids. The presence of other esophageal symptoms helps establish an esophageal cause of pain. As many as 50% of patients with cardiac pain, however, also have one or more symptoms of esophageal disease.25 Furthermore, relief of pain with sublingual nitroglycerin has been shown not to be specific for a coronary origin of pain.26 Cardiac and esophageal disease increase in frequency as people grow older, and both problems may not only coexist but also interact to produce chest pain.

PATHOPHYSIOLOGY AND APPROACH

The specific mechanisms that produce esophageal chest pain are not well understood. Chest pain that arises from the esophagus has commonly been attributed to the stimulation of chemoreceptors (by acid, pepsin, or bile) or mechanoreceptors (by distention or spasm); thermoreceptors (stimulated by cold) also may be involved. Gastroeso­phageal reflux causes chest pain primarily through acid-sensitive esophageal chemoreceptors (see later). Acid-induced dysmotility may be a cause of esophageal pain. Older studies have shown that perfusion of acid into the esophagus in patients with gastroesophageal reflux increases the amplitude and duration of esophageal contractions and induces simultaneous and spontaneous contractions, with the occurrence of pain.27 Diffuse esophageal spasm also has been demonstrated during spontaneous acid reflux. Subsequent studies with modern equipment have shown that such changes in motility are rare.28 In addition, studies using 24-hour ambulatory esophageal pH and motility monitoring have shown that the association between abnormal motility and pain is uncommon, and that spontaneous acid-induced chest pain is rarely associated with abnormalities in esophageal motility.29,30 Patients with chest pain suspected to be esophageal in origin have an increased frequency of esophageal contractions of high amplitude and a slightly increased frequency of simultaneous contractions when compared with a normal control population.31 In addition, intraluminal ultrasound has been able to identify abnormal sustained contractions of the longitudinal smooth muscle in a subset of patients with chest pain.32 How these contractions cause pain is unknown. One possible explanation is that pain occurs when high intramural esophageal tension resulting from altered motility inhibits blood flow to the esophagus for a critical period of time (i.e., myoischemia). MacKenzie and coworkers have found that rates of esophageal rewarming are decreased after infusions of cold water into the esophagus of patients with symptomatic esophageal motility disorders as compared with age-matched controls.33 Because

177

178

Section III  Symptoms, Signs, and Biopsychosocial Issues the rate of rewarming after cold water infusion in patients with Raynaud’s phenomenon correlates directly with blood flow, the authors theorized that esophageal ischemia is the cause of the reduced rate of rewarming. None of the patients with a symptomatic esophageal motility disorder, however, experienced chest pain during the study. Furthermore, the extensive arterial and venous blood supply to the esophagus makes it unlikely that blood flow is compromised after even the most abnormal esophageal contractions. Complicating the relationship between esophageal chest pain and abnormal esophageal contractions is the consistent observation that most of these patients are asymptomatic when the contraction abnormalities are identified. In addition, amelioration of chest pain does not correlate predictably with reduction in the amplitude of esophageal contractions.34 The possibility exists that chest pain–associated motility changes represent an epiphenomenon of a chronic pain syndrome rather than the direct cause of the pain. Other potential causes of esophageal chest pain include excitation of temperature receptors and luminal distention. The ingestion of hot or cold liquids can produce severe chest pain. This association was previously thought to be related to esophageal spasm, but subsequent studies have shown that cold-induced pain produces esophageal aperistalsis and dilatation, not spasm.35 This observation suggests that the cause of esophageal chest pain may be activation of stretch receptors by acute distention. Esophageal distention and pain are experienced during an acute food impaction, drinking of carbonated beverages (in some patients), and dysfunction of the belch reflex.36 In susceptible persons, esophageal chest pain can be reproduced by distention of an esophageal balloon to volumes lower than those that produce pain in asymptomatic persons.37 Therefore, altered pain perception may contribute to the patient’s reaction to a painful stimulus. Panic disorder is a commonly overlooked coexisting condition in patients with chest pain38 and should be sought specifically during history taking. The observation that anxiolytics and antidepressants can raise pain thresholds, as well as improve mood states, may explain why these medications may improve esophageal chest pain in the absence of manometric changes,39,40 The approach to patients with esophageal chest pain has evolved over the years. Before the esophagus is considered to be the cause of chest pain, a cardiac cause must be excluded. Appropriate testing may include an exercise stress test, noninvasive cardiac imaging studies, and coronary angiography. Insufficiency of coronary blood flow with normalappearing epicoronary arteries (microvascular angina) has been suggested as a cause of chest pain in some patients.41 Diagnosing microvascular angina on the basis of a thera­ peutic trial is difficult because the medications reported to improve this condition also have effects on the esophagus; however, the prognosis of patients with microvascular angina is thought to be good. The recognition that chest pain is often associated with GERD has been a major advance in our understanding of esophageal chest pain. Ambulatory pH testing can document pathologic amounts of acid reflux or a correlation between acid reflux and chest pain in up to 50% of patients in whom a cardiac cause has been excluded.42 In addition, a trial of therapy with a proton pump inhibitor produces symptomatic improvement in many such patients.43 The association between chest pain and GERD is easy to recognize when the patient has coexisting reflux symptoms but not so clear when typical reflux symptoms are absent. A 10- to 14-day trial of an oral proton pump inhibitor taken twice daily has been shown to be sensitive and specific for

the diagnosis of esophageal chest pain when compared with ambulatory intraesophageal pH testing (see later).44 If a patient fails this trial, the next practical approach may be a trial of agents such as imipramine or trazodone that raise the pain threshold. Some authorities recommend esophageal testing with stationary manometry at this point to exclude a motility disorder and ambulatory pH testing to exclude reflux unresponsive to the initial trial of the proton pump inhibitor therapy. The advent of a tube-free system for reflux monitoring allows a longer and more comfortable monitoring period, which increases the likelihood of observing a correlation between pain and an acid event.45 If reflux is confirmed by ambulatory pH testing, an additional trial of acid suppressive therapy is warranted. If a spastic motility disorder is discovered on manometry, an attempt at lowering esophageal pressure with nitrates or a calcium channel blocker is appropriate (see Chapter 42).

HEARTBURN AND REGURGITATION Heartburn (pyrosis) is one of the most common gastrointestinal complaints in Western populations.46 In fact, it is so common that many people assume it to be a normal part of life and fail to report the symptom to their health care providers. They seek relief with over-the-counter antacids, which accounts for most of the $1 billion/year sales of these nonprescription drugs. Despite its high prevalence, the term heartburn is frequently misunderstood. It has many synonyms, including indigestion, acid regurgitation, sour stomach, and bitter belching. The physician should listen for these descriptors if the patient does not volunteer a complaint of heartburn. A study from Europe has suggested that using a word-picture description of “a burning feeling rising from the stomach or lower chest up toward the neck” increases the ability to identify patients with reflux.47 The burning sensation often begins inferiorly and radiates up the entire retrosternal area to the neck, occasionally to the back, and rarely into the arms. Heartburn caused by acid reflux may be relieved, albeit only transiently, by the ingestion of antacids, baking soda, or milk. Interestingly, the severity of esophageal damage (esophagitis or Barrett’s esophagus) does not correlate with the severity of heartburn (e.g., patients with severe heartburn may have a normalappearing esophagus, and those with severe esophagitis or Barrett’s esophagus may, at times, have mild or even no symptoms; see Chapters 43 and 44).48 Heartburn is most frequently noted within one hour after eating, particularly after the largest meal of the day. Sugars, chocolate, onions, carminatives, and foods high in fats may aggravate heartburn by decreasing lower esophageal sphincter (LES) pressure. Other foods commonly associated with heartburn—including citrus products, tomato-based foods, and spicy foods—irritate the inflamed esophageal mucosa because of acidity or high osmolarity.49 Beverages, including citrus juices, soft drinks, coffee, and alcohol, also may cause heartburn. Many patients have exacerbation of heartburn if they retire shortly after a late meal or snack, and others say that their heartburn is more pronounced while they lie on their right side.50 Weight gain frequently results in the development of new GERD symptoms and in the worsening of GERD symptoms in patients with preexisting symptoms.51 Activities that increase intra-abdominal pressure, including bending over, straining at stool, lifting heavy objects, and performing isometric exercises, may aggravate heartburn. Running also may aggravate heartburn, and stationary

Chapter 12  Symptoms of Esophageal Disease bike riding may be a better exercise for those with GERD.52 Because nicotine and air swallowing relax LES pressure, cigarette smoking exacerbates the symptoms of reflux.53 Emotions such as anxiety, fear, and worry may exacerbate heartburn by lowering visceral sensitivity thresholds rather than by increasing the amount of acid reflux.54,55 Some heartburn sufferers complain that certain drugs may initiate or exacerbate their symptoms by reducing LES pressure and peristaltic contractions (e.g., theophylline, calcium channel blockers) or by irritating the inflamed esophagus (e.g., aspirin, other nonsteroidal anti-inflammatory drugs, bisphosphonates). Heartburn may be accompanied by the appearance of fluid in the mouth, either a bitter acidic material or a salty fluid. Regurgitation describes return of bitter acidic fluid into the mouth and, at times, the effortless return of food, acid, or bilious material from the stomach. Regurgitation is more common at night or when the patient bends over. The absence of nausea, retching, and abdominal contractions suggests regurgitation rather than vomiting. Water brash is an uncommon and frequently misunderstood symptom that should be used to describe the sudden filling of the mouth with clear, slightly salty fluid. This fluid is not regurgitated material but is secreted from the salivary glands as part of a protective, vagally mediated reflex from the distal esophagus.56 Regurgitation and symptoms similar to water brash can occur in patients with achalasia, who may be misdiagnosed as having GERD. Regurgitation must be distinguished from the syndrome of rumination (see Chapter 14). Rumination is a clinical diagnosis and is best described by the Rome III diagnostic criteria. Patients must have persistent or recurrent regurgitation (not preceded by retching) of recently ingested food into the mouth, with subsequent remastication and swallowing. Supportive criteria include absence of nausea, cessation of the process when the regurgitated material becomes acidic, and content consisting of recognizable food with a pleasant taste in the regurgitant.57 Rumination is essentially a diagnosis of exclusion when there is clinical suspicion. Nocturnal reflux symptoms have particular significance. In a survey of patients with frequent reflux symptoms, 74% reported nocturnal symptoms.58 These nighttime symptoms interrupt sleep and health-related quality of life to a greater degree than daytime reflux symptoms alone. Patients who have prolonged reflux episodes at night also are at increased risk of complications of GERD, including severe reflux esophagitis and Barrett’s esophagus.

PATHOPHYSIOLOGY AND APPROACH

The physiologic mechanisms that produce heartburn remain poorly understood. Although the reflux of gastric acid is most commonly associated with heartburn, the same symptom may be elicited by esophageal balloon distention,59 reflux of bile salts,60 and acid-induced motility disturbances. The best evidence that the pain mechanism is probably related to the stimulation of mucosal chemoreceptors is the sensitivity of the esophagus to acid that is perfused into the esophagus or acid reflux, demonstrated by monitoring of pH. The location of these chemoreceptors is not known. One suggestion is that the esophagus is sensitized by repeated acid exposure, resulting in the production of symptoms from smaller boluses after repeated exposure to acid. This hypersensitivity has been reported to resolve with acid suppressive therapy.61 The correlation between discrete episodes of acid reflux and symptoms, however, is poor. For example, postprandial gastroesophageal reflux is common in healthy people, but

symptoms are uncommon. Intraesophageal pH monitoring of patients with endoscopic evidence of esophagitis typically shows excessive periods of acid reflux, but fewer than 20% of these reflux episodes are accompanied by symptoms.62 Moreover, one third of patients with Barrett’s esophagus, the most extreme form of GERD, are acid-insensitive.63 As patients age, their sensitivity to esophageal acid seems to decline; this finding may explain the common observation that mucosal damage is fairly severe but symptoms are minimal in older patients.64 Therefore, the development of symptoms must require more than esophageal contact with acid. Mucosal disruption and inflammation may be a contributing factor but, on endoscopy, the esophagus appears normal in most symptomatic patients. Other factors that possibly influence the occurrence of heartburn include the acid clearance mechanism, salivary bicarbonate concentration, volume of acid refluxed, as measured by the duration and proximal extent of reflux episodes, frequency of heartburn, and interaction of pepsin with acid (see Chapter 43). In addition, studies in which acid reflux is monitored for more than 24 hours have demonstrated considerable daily variability in esophageal acid exposure.65,66 As noted, heartburn strongly suggests gastroesophageal acid reflux, but peptic ulcer disease, delayed gastric emptying, and even gallbladder disease can produce symptoms similar to those caused by reflux. Regurgitation is not quite as specific for acid reflux as heartburn, and the differential diagnosis of regurgitation should include an esophageal obstruction (e.g., ring, stricture, or achalasia) or a gastric emptying problem (e.g., gastroparesis or gastric outlet obstruction). Some patients have overlap among symptoms of gastroesophageal reflux, dyspepsia, and irritable bowel syndrome (see Chapters 13, 43, and 118).67 The approach to patients with heartburn and regurgitation is discussed extensively in Chapter 43. In brief, published guidelines support an initial trial of acid suppressive therapy, generally with a proton pump inhibitor, as a diagnostic and therapeutic maneuver.68 This concept is cost-effective but plagued by limitations in sensitivity and specificity.69 If the cause of symptoms remains uncertain after a therapeutic trial, ambulatory intraesophageal pH testing is the best test to document pathologic esophageal acid exposure. Endoscopy of the esophagus is reserved for patients with symptoms suggestive of a complication (e.g., dysphagia, weight loss, signs of bleeding), but the predictive value of using a symptom profile to predict esophageal damage is questionable at best. Although not without controversy, most guidelines also suggest endoscopy to screen for Barrett’s esophagus in patients with chronic reflux symptoms70; the risk is particularly increased in older and obese patients.71,72

EXTRAESOPHAGEAL SYMPTOMS OF GASTROESOPHAGEAL REFLUX DISEASE Extraesophageal symptoms of GERD are listed in Table 12-4. Although these symptoms may be caused by esophageal motility disorders, they are most frequently associated with GERD. In patients with extraesophageal symptoms, the classic reflux symptoms of heartburn and regurgitation often are mild or absent (see Chapter 43). Gastroesophageal reflux is thought to cause chronic cough and other extraesophageal symptoms as a result of recurrent microaspiration of gastric contents, a vagally mediated neural reflex or, in many patients, a combination of both. Although bronchodilators lower LES pressure, most persons

179

180

Section III  Symptoms, Signs, and Biopsychosocial Issues Possible extraesophageal manifestation of GERD

Table 12-4  Extraesophageal Manifestations of Gastroesophageal Reflux Disease Asthma Chronic cough Excess mucus or phlegm Globus sensation Hoarseness Laryngitis Pulmonary fibrosis Sore throat

with asthma have gastroesophageal reflux with or without bronchodilator therapy. In animal studies, the instillation of small amounts of acid in the trachea or on the vocal cords73 can produce marked changes in airway resistance, as well as vocal cord ulcers. Direct evidence for aspiration is more difficult to identify in adults and rests primarily on the presence of fat-filled macrophages in sputum,74 radioactivity in the lungs after a tracer is placed in the stomach overnight,75 and a high degree of esophageal or hypopharyngeal acid reflux recorded by 24-hour pH monitoring with dual probes.76 Data from animal and human studies suggest that a neural reflex is another pathophysiologic basis for these symptoms. Acid perfusion into the distal esophagus increases airway resistance in all subjects, but the changes are most marked in patients with both asthma and heartburn.77 Abnormal amounts of acid reflux recorded by prolonged intraesophageal pH monitoring have been identified in 35% to 80% of asthmatic adults.78 Symptoms that suggest refluxinduced asthma include the onset of wheezing in adulthood in the absence of a history of allergies or asthma, nocturnal cough or wheezing, asthma that is worsened after meals, by exercise, or in the supine position, and asthma that is exacerbated by bronchodilators or that is glucocorticoiddependent. In patients with reflux, symptoms strongly suggestive of aspiration include nocturnal cough and heartburn, recurrent pneumonia, unexplained fevers, and an associated esophageal motility disorder. Ear, nose, and throat complaints associated with gastroesophageal reflux include postnasal drip, voice changes, hoarseness, sore throat, persistent cough, otalgia, halitosis, dental erosion, and excessive salivation. Many patients with GERD complain of only head and neck symptoms. Examination of the vocal cords may help in evaluating patients with suspected acid reflux–related extraesophageal problems. Some patients have redness, hyperemia, and edema of the vocal cords and arytenoids. In more severe cases, vocal cord ulcers, granulomas, and even laryngeal cancer, all secondary to GERD, have been reported. Normal results of a laryngeal examination, however, are not incompatible with acid reflux–related extraesophageal symptoms, nor are the aforementioned laryngeal signs specific for a GERD-related pathogenesis. The options in a patient with suspected extraesophageal GERD are to study them with an ambulatory intraesophageal pH test or to initiate a trial of therapy to confirm the diagnosis and treat the symptom (Fig. 12-2). Either approach is reasonable, but many experts favor an initial trial of acid suppressive therapy with a proton pump inhibitor twice daily.79 Ambulatory pH testing is then reserved for those who fail the initial trial, although it is not clear whether pH testing should be done with the patient continuing or discontinuing acid-suppressive therapy (see Chapter 43).

Test first strategy

Exclude underlying cardiac, thoracic, and head/neck disease

Treat first strategy

24-hour pH study

Trial of twice daily PPI

Positive?

Successful?

Yes

No

Trial of twice daily PPI

Successful?

Yes

Yes

24-hour pH study

Positive?

No No

GERD maintenance therapy

Yes

No

Maximize medical therapy or consider antireflux surgery

Consider other diagnosis Figure 12-2.  Suggested approach to patients with extraesophageal manifestations of reflux disease, including noncardiac chest pain. The approach to the exclusion of underlying disease varies, depending on the symptom under evaluation (see text). A proton pump inhibitor (PPI) is given before breakfast and before the evening meal. The duration of the trial depends on the symptom. For example, a 10- to 14-day trial may be sufficient for noncardiac chest pain, whereas a three-month trial may be needed for chronic cough. GERD, gastroesophageal reflux disease.

The association between reflux and extraesophageal symptoms, particularly laryngeal symptoms, has been challenged. In one study, pH monitoring of the hypopharynx and proximal and distal esophagus was performed in patients with presumed acid reflux–related endoscopic laryngeal findings.80 An abnormal result was noted in only 15% of hypopharyngeal probes, 9% of proximal esophageal probes, and 29% of distal esophageal probes, thereby indicating that most patients (70%) with symptoms and signs of laryngeal reflux do not have documentable abnormal acid exposure. That preliminary study was followed by a randomized, placebo-controlled trial of esomeprazole, 40  mg twice daily, in the same patients, with response rates of 42% in those treated with esomeprazole and 46% in those treated with placebo.81 In addition, a randomized controlled trial of therapy with a proton pump inhibitor in asthmatics produced similar results.82 Despite the contradictory data, an early trial of proton pump inhibitor therapy in patients

Chapter 12  Symptoms of Esophageal Disease with symptoms suggestive of extraesophageal GERD is reasonable; however, the patient and physician should not be surprised if this therapy fails.

KEY REFERENCES

Avidan B, Sonnenberg A, Schnell TG, Sontag SJ. There are no reliable symptoms for erosive oesophagitis and Barrett’s oesophagus: Endoscopic diagnosis is still essential. Aliment Pharmacol Ther 2002; 16:735-42. (Ref 48.) Corley DA, Kubo A, Levin TR, et al. Abdominal obesity and body mass index as risk factors for Barrett’s esophagus. Gastroenterology 2007; 133:34-41. (Ref 72.) DeVault KR, Castell DO. Updated guidelines for the diagnosis and treatment of gastroesophageal reflux disease. Am J Gastroenterol 2005; 100:190-200. (Ref 68.) Farup C, Kleinman L, Sloan S, et al. The impact of nocturnal symptoms associated with gastroesophageal reflux disease on health-related quality of life. Arch Intern Med 2001; 161:45-52. (Ref 58.) Fass R, Naliboff BD, Fass SS, et al. The effect of auditory stress on perception of intraesophageal acid in patients with gastroesophageal reflux disease. Gastroenterology 2008; 13:696-705. (Ref 55.) Furuta G, Liacouras C, Collins M, et al. Eosinophilic esophagitis in children and adults: A systematic review and consensus recommendations for diagnosis and treatment. Gastroenterology 2007; 133:134263. (Ref 8.)

Henrikson CA, Howell EE, Bush DE, et al. Chest pain relief by nitroglycerin does not predict active coronary artery disease. Ann Intern Med 2003; 139:979-86. (Ref 26.) Johnson DA, Fennerty MD. Heartburn severity underestimates erosive esophagitis severity in elderly patients with gastroesophageal reflux disease. Gastroenterology 2004; 126:660-4. (Ref 3.) Kiljander TO, Harding SM, Field SK, et al. Effects of esomeprazole 40 mg twice daily on asthma: A randomized, placebo-controlled trial. Am J Resp and Crit Care Med 2006; 173:1091-7. (Ref 82.) Numans ME, Lau J, de Witt NJ, Bonis PA. Short-term treatment with proton-pump inhibitors as a test for gastroesophageal reflux disease. A meta-analysis of diagnostic test characteristics. Ann Intern Med 2004; 140:518-27. (Ref 69.) Pandolfino JE, Richter JE, Ours T, et al. Ambulatory esophageal pH monitoring using a wireless system. Am J Gastroenterol 2003; 98:7409. (Ref 66.) Prakash C, Clouse R. Wireless pH monitoring in patients with noncardiac chest pain. Am J Gastroenterol 2006; 101:446-52. (Ref 45.) Rey E, Moreno-Elola-Olaso C, Artalejo FR, et al. Association between weight gain and symptoms of gastroesophageal reflux in the general population. Am J Gastroenterol 2006; 101:229-33. (Ref 51.) Vaezi MF, Richter JE, Stasney CR, et al. Treatment of chronic posterior laryngitis with esomeprazole. Laryngoscope 2006; 116:254-60. (Ref 81.) Full references for this chapter can be found on www.expertconsult.com.

181

CHAPTE R

13 Dyspepsia Jan Tack

CHAPTER OUTLINE Definition  183 Organic Causes of Dyspepsia  183 Intolerance to Food or Drugs  184 Peptic Ulcer Disease  184 Gastroesophageal Reflux Disease  184 Gastric and Esophageal Cancer  184 Pancreatic and Biliary Tract Disorders  184 Other Gastrointestinal and Systemic Disorders  184 Functional Dyspepsia  185 The Dyspepsia Symptom Complex  185 Epidemiology  187 Pathophysiology  187 Pathogenic Factors  188

DEFINITION Dyspepsia is derived from the Greek words dys and pepse and literally means “difficult digestion.” In current medical terminology, dyspepsia refers to a heterogeneous group of symptoms located in the upper abdomen. Dyspepsia is often broadly defined as pain or discomfort centered in the upper abdomen1,2 and may include multiple and varying symptoms such as epigastric pain, postprandial fullness, early satiation (also called early satiety), anorexia, belching, nausea and vomiting, upper abdominal bloating, and even heartburn and regurgitation. Patients with dyspepsia commonly report several of these symptoms.3 Several consensus definitions of dyspepsia and functional dyspepsia have been proposed. The overlap between symptoms of gastric origin and symptoms of presumed esophageal origin (especially those associated with gastroesophageal reflux disease [GERD]) has remained an area of controversy. With time, definitions of dyspepsia have evolved to become more restrictive and more focused on symptoms thought to arise from the gastroduodenal region and not the esophagus. Earlier definitions considered dyspepsia to consist of all upper abdominal and retrosternal sensations—in effect, all symptoms considered to be referable to the pro­ximal gastrointestinal tract.4 The Rome I and II consensus committees both defined dyspepsia as pain or discomfort centered in the upper abdomen.1,2 Discomfort includes postprandial fullness, upper abdominal bloating, early satiation, epigastric burning, belching, nausea, and vomiting. Heartburn may occur as part of the symptom constellation, but the Rome II committee decided that when heartburn is the predominant symptom, the patient should be con­sidered to have GERD and not dyspepsia. The most recent consensus committee, Rome III, has defined dyspepsia as the presence of symptoms considered by the physician to originate from the gastroduodenal region (Table 13-1).5 Only four symptoms (bothersome postpran-

Approach to Uninvestigated Dyspepsia  189 History and Physical Examination  189 Laboratory Testing  189 Initial Management Strategies  189 Additional Investigations  191 Treatment of Functional Dyspepsia  191 General Measures  191 Pharmacologic Treatment  191 Psychological Interventions  193 Recommendations  193

dial fullness, early satiation, epigastric pain, and epigastric burning) are now considered to be specific for a gastroduodenal origin, although many other symptoms are acknowledged to coexist with dyspepsia. In patients with dyspepsia, additional clinical investigations may identify an underlying organic disease that is the likely cause of the symptoms. In these persons, dyspeptic symptoms are attributable to an organic cause of dyspepsia (Table 13-2). In most people with dyspeptic symptoms, however, no organic abnormality is identified by a routine clinical evaluation, and patients who have undergone a diagnostic investigation (including endoscopy) and have not been found to have an obvious specific cause of their symptoms are said to have functional dyspepsia. The term uninvestigated dyspepsia refers to dyspeptic symptoms in persons in whom no diagnostic investigations have yet been performed and in whom a specific diagnosis that explains the dyspeptic symptoms has not been determined.

ORGANIC CAUSES OF DYSPEPSIA The most important identifiable causes underlying dyspeptic symptoms are peptic ulcer disease and GERD. Malignancies of the upper gastrointestinal tract and celiac disease are less common but important causes of dyspeptic symptoms (see Table 13-2).6-10 The investigation of choice in persons with dyspeptic symptoms is endoscopy, which may identify erosive esophagitis, peptic ulcer, or gastric or esophageal cancer. Systematic studies indicate that approximately 20% of patients with dyspeptic symptoms have erosive esophagitis, 20% have endoscopy-negative GERD, 10% have peptic ulcer, 2% have Barrett’s esophagus, and 1% or less have malignancy.6 Minor findings such as duodenitis or gastritis do not seem to correlate with the presence or absence of dyspeptic symptoms.

183

184

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 13-1  Dyspeptic Symptoms and Their Definitions* Symptom More Specific Postprandial fullness Early satiation Epigastric pain

Epigastric burning Less Specific Bloating in the upper abdomen Nausea Vomiting Belching

Definition An unpleasant sensation perceived as the prolonged persistence of food in the stomach A feeling that the stomach is overfilled soon after starting to eat, out of proportion to the size of the meal being eaten, so that the meal cannot be finished. Previously, the term early satiety was used, but satiation is the correct term for the disappearance of the sensation of appetite during food ingestion. Epigastric refers to the region between the umbilicus and lower end of the sternum, within the midclavicular lines. Pain refers to a subjective, unpleasant sensation; some patients may feel that tissue damage is occurring. Epigastric pain may or may not have a burning quality. Other symptoms may be extremely bothersome without being interpreted by the patient as pain. Epigastric refers to the region between the umbilicus and lower end of the sternum, within the midclavicular lines. Burning refers to an unpleasant subjective sensation of heat. An unpleasant sensation of tightness located in the epigastrium. Bloating should be distinguished from visible abdominal distention Queasiness or sick sensation; a feeling of the need to vomit Forceful oral expulsion of gastric contents associated with contraction of the abdominal and chest wall muscles. Vomiting is usually preceded by and associated with retching, repetitive contractions of the abdominal wall without expulsion of gastric contents. Venting of air from the stomach or the esophagus

*According to the Rome III committee. Adapted from Tack J, Talley NJ, Camilleri M, et al. Functional gastroduodenal disorders. In: Drossman DA, Corazziari E, Delvaux M, et al, editors. Rome III. The Functional Gastrointestinal Disorders. 3rd ed. McLean, Va: Degnon Associates; 2006. p 422.

INTOLERANCE TO FOOD OR DRUGS

Contrary to popular beliefs, ingestion of specific foods such as spices, coffee, or alcohol, or of excessive amounts of food, has never been established as causing dyspepsia.11,12 Although ingestion of food often aggravates dyspeptic symptoms, this effect probably is related to the sensorimotor response to food rather than to specific food intolerances or allergies. Studies have shown that acute ingestion of capsaicin induces dyspeptic symptoms in healthy persons and in those with functional dyspepsia, with greater intensity in the latter group.13 Dyspepsia is a common side effect of many drugs, including iron, antibiotics, narcotics, digitalis, estrogens and oral contraceptives, theophylline, and levodopa. Medications may cause symptoms through direct gastric mucosal injury, changes in gastrointestinal sensorimotor function, provo­ cation of gastroesophageal reflux, or idiosyncratic mech­ anisms. Nonsteroidal anti-inflammatory drugs (NSAIDs) have received the most attention because of their potential to induce ulceration in the gastrointestinal tract. Chronic use of aspirin and other NSAIDs may provoke dyspeptic symptoms in up to 20% of persons, but the occurrence of dyspepsia correlates poorly with the presence of an ulcer. In controlled trials, dyspepsia develops in 4% to 8% of persons treated with NSAIDs, with odds ratios ranging from 1.1 to 3.1 compared with placebo; the magnitude of this effect depends on the dose and type of NSAID.14 Compared with NSAIDs, selective cyclooxygenase-2 inhibitors are associated with a lower frequency of dyspepsia and peptic ulceration.15

PEPTIC ULCER DISEASE

Peptic ulcer is a well-established cause of dyspeptic symptoms and is an important consideration for clinicians in the management of patients who present with dyspepsia. The frequency of peptic ulcer in patients with dyspepsia, however, is only 5% to 10%.6,10,14 Increasing age, NSAID use, and Helicobacter pylori infection are the main risk factors for peptic ulcer (see Chapters 50 and 52).

GASTROESOPHAGEAL REFLUX DISEASE

Erosive esophagitis is a diagnostic marker for GERD, but many patients with symptoms that are attributable to the

reflux of stomach contents into the esophagus have no endoscopic signs of esophageal erosion; this is referred to as nonerosive GERD. Erosive esophagitis is found in approximately 20% of dyspeptic patients, and a similar number of patients may have nonerosive GERD (see Chapter 43).6,10

GASTRIC AND ESOPHAGEAL CANCER

The risk of gastric or esophageal malignancy in patients with dyspeptic symptoms is estimated to be less than 1%.9 The risk of gastric cancer is increased among persons with H. pylori infection, persons with a family history of gastric malignancy, persons with a previous history of gastric surgery, and immigrants from areas endemic for gastric malignancy. The risk of esophageal cancer is increased in men, smokers, persons with a high consumption of alcohol, and those with a long-standing history of heartburn (see Chapters 46 and 54).

PANCREATIC AND BILIARY TRACT DISORDERS

Despite the high prevalence of dyspepsia and gallstones in adults, epidemiologic studies have confirmed that cholelithiasis is not associated with dyspepsia. Therefore, patients with dyspepsia should not be investigated routinely for cholelithiasis, and cholecystectomy in patients with cholelithiasis is not indicated for dyspepsia alone. The clinical presentation of biliary pain is easily distinguishable from that of dyspepsia (see Chapter 65). Pancreatic disease is less prevalent than cholelithiasis, but symptoms of acute or chronic pancreatitis or of pan­ creatic cancer may initially be mistaken for dyspepsia. Pancreatic disorders, however, are usually associated with more severe pain and are often accompanied by anorexia, rapid weight loss, or jaundice (see Chapters 58 to 60).

OTHER GASTROINTESTINAL AND SYSTEMIC DISORDERS

Several gastrointestinal disorders may cause dyspepsia-like symptoms. These include infectious (e.g., Giardia lamblia and Strongyloides stercoralis parasites, tuberculosis, fungal infections, syphilis), inflammatory (e.g., celiac disease, Crohn’s disease, sarcoidosis, lymphocytic gastritis, eosino-

Chapter 13  Dyspepsia Table 13-2  Causes of Dyspepsia Luminal Gastrointestinal Tract Chronic gastric volvulus Chronic gastric or intestinal ischemia Food intolerance Functional dyspepsia Gastroesophageal reflux disease Gastric or esophageal neoplasms Gastric infections (e.g., cytomegalovirus, fungus, tuberculosis, syphilis) Gastroparesis (e.g., diabetes mellitus, postvagotomy, scleroderma, chronic intestinal pseudo-obstruction, postviral, idiopathic) Infiltrative and inflammatory gastric disorders (e.g., Crohn’s disease, eosinophilic gastroenteritis, sarcoidosis, amyloidosis) Irritable bowel syndrome Ménétrier’s disease Peptic ulcer disease Parasites (e.g., Giardia lamblia, Strongyloides stercoralis) Medications Acarbose Aspirin, other nonsteroidal anti-inflammatory drugs (including cyclooxygenase-2 selective agents) Colchicine Digitalis preparations Estrogens Ethanol Gemfibrozil Glucocorticoids Iron Levodopa Niacin Narcotics Nitrates Orlistat Potassium chloride Quinidine Sildenafil Theophylline Pancreaticobiliary Disorders Biliary pain—cholelithiasis, choledocholithiasis, sphincter of Oddi dysfunction Chronic pancreatitis Pancreatic neoplasms Systemic Conditions Adrenal insufficiency Congestive heart failure Diabetes mellitus Hyperparathyroidism Intra-abdominal non-gastrointestinal malignancy Myocardial ischemia Pregnancy Renal insufficiency Thyroid disease

philic gastroenteritis), and infiltrative (e.g., lymphoma, amyloidosis, Ménétrier’s disease) disorders of the upper gastrointestinal tract. Most of these causes will be identi­ fiable by upper gastrointestinal endoscopy with mucosal biopsies. Recurrent gastric volvulus and chronic mesenteric or gastric ischemia may present with dyspeptic symptoms (see Chapters 27, 29, 35, 47, 104, 109, 111, and 114). The symptom pattern associated with gastroparesis (idiopathic, drug-induced, or secondary to metabolic, systemic, or neurologic disorders) is similar to dyspepsia, and the distinction between idiopathic gastroparesis and functional dyspepsia with delayed gastric emptying (see later) is not well-defined (see Chapter 48). Finally, dyspepsia may be the presenting or accom­ panying symptom of acute myocardial ischemia, pregnancy, acute or chronic kidney disease, thyroid dysfunction, adrenal insufficiency, and hyperparathyroidism (see Chapters 35 and 38).

FUNCTIONAL DYSPEPSIA According to the Rome III criteria, functional dyspepsia is defined as the presence of early satiation, postprandial fullness, epigastric pain, and epigastric burning in the absence of organic, systemic, or metabolic disease that is likely to explain the symptoms.

THE DYSPEPSIA SYMPTOM COMPLEX Pattern and Heterogeneity

The dyspepsia symptom complex is broader than the four cardinal symptoms that constitute the Rome III definition. It includes multiple symptoms such as epigastric pain, bloating, early satiation, fullness, epigastric burning, belching, nausea, and vomiting. Although often chronic, the symptoms in patients with functional dyspepsia are mostly intermittent, even during highly symptomatic episodes.3,16 In persons with functional dyspepsia who present to a tertiary care center, the most frequent symptoms are postprandial fullness and bloating, followed by epigastric pain, early satiation, nausea, and belching.17-20 Heterogeneity of symptoms is considerable, however, as shown, for example, in the number of symptoms that patients report (Fig. 13-1). In the general population, the most frequent dyspeptic symptoms are postprandial fullness, early satiation, upper abdominal pain, and nausea.21-23 Weight loss is traditionally considered an alarm symptom, pointing toward potentially serious organic disease. Patients with functional dyspepsia who present to a tertiary care center also have a high frequency of unexplained weight loss,17,18 and population-based studies in Australia and in Europe have established an association between uninve­s­ tigated dyspepsia and unexplained weight loss.22,23

Subgroups

The heterogeneity of the dyspepsia symptom complex is well accepted. Factor analyses of dyspepsia symptoms in the general population and in patients who present to a tertiary care center have not supported the existence of functional dyspepsia as a homogeneous (i.e., unidimensional) condition.22-24 These studies confirmed the hetero­ geneity of the dyspepsia symptom complex but did not provide clinically meaningful subdivisions of the syndrome. Several attempts have been made to identify clinically meaningful dyspepsia subgroups to simplify the intricate heterogeneity of the dyspepsia symptom complex and to guide management. The Rome II consensus committee proposed a classification based on a predominant symptom of pain or discomfort. Although correlations were found between the two subdivisions and the presence or absence of H. pylori infection, the absence or presence of delayed gastric emptying, and response or lack of response to gastric acid suppressive therapy,25,26 the subdivisions have been criticized because of the difficulty in distinguishing pain from discomfort, the lack of a widely accepted definition of predominant, uncertainty concerning overlap between the symptom subgroups, the lack of an association with putative pathophysiologic mechanisms and, especially, the lack of stability of the predominant symptom over short time periods.5,27-30 The Rome III consensus committee has proposed different subdivisions (Fig. 13-2). Studies of patients referred to a tertiary care center and of patients with uninvestigated dyspepsia in the general population have revealed that between 40% and 75% of dyspeptic persons report aggravation of symptoms after ingestion of a meal.23,31,32 Assuming

185

Section III  Symptoms, Signs, and Biopsychosocial Issues 100 90 80 Prevalence (% of patients)

186

70 60 50 40

Absent Mild Moderate Severe

30 20 10 0 Fullness

Bloating

Pain

Nausea

Early satiety

Belching Epigastric Vomiting burning

Figure 13-1.  Prevalence of symptoms and their severity ratings in 674 patients with functional dyspepsia seen at a tertiary referral center. (Unpublished, University Hospital Gasthuisberg, Leuven, Belgium). Uninvestigated dyspepsia (postprandial fullness, early satiation, epigastric pain, epigastric burning) Endoscopy, other investigations

Functional dyspepsia

Postprandial distress syndrome (PDS): Meal-related FD – Early satiation – Postprandial fullness

Organic dyspepsia (e.g., ulcer, esophagitis)

Epigastric pain syndrome (EPS): Meal-unrelated FD – Epigastric pain – Epigastric burning

Figure 13-2.  Classification of uninvestigated dyspepsia, functional dys­ pepsia (FD), and subtypes of functional dyspepsia, according to the Rome III criteria.

that a distinction between meal-related and meal-unrelated symptoms might be pathophysiologically and clinically relevant, the Rome III consensus committee proposed that functional dyspepsia be used as an umbrella term and that postprandial distress syndrome (PDS; meal-related dyspeptic symptoms, characterized by postprandial fullness and early satiation) be distinguished from the epigastric pain syndrome (EPS; meal-unrelated dyspeptic symptoms, characterized by epigastric pain and epigastric burning; Table 13-3).5 Few studies have evaluated the Rome III–based classification of functional dyspepsia. One study of postprandial symptom patterns in persons with functional dyspepsia has provided some support for the distinction between EPS and PDS,32 and a population-based study confirmed the existence of the two distinct subgroups, with less

Table 13-3  Classification of and Diagnostic Criteria for Functional Dyspepsia, Postprandial Distress Syndrome, and Epigastric Pain Syndrome* Functional Dyspepsia† Includes one or more of the following: 1. Bothersome postprandial fullness 2. Early satiation 3. Epigastric pain 4. Epigastric burning and No evidence of structural disease (including at upper endoscopy) that is likely to explain the symptoms Postprandial Distress Syndrome† Must include one or both of the following: 1. Bothersome postprandial fullness, occurring after ordinary-sized meals, at least several times per week 2. Early satiation that prevents finishing a regular meal, at least several times per week Supportive Criteria 1. Upper abdominal bloating or postprandial nausea or excessive belching can be present 2. Epigastric pain syndrome may coexist Epigastric Pain Syndrome† Must include all of the following: 1. Pain or burning localized to the epigastrium of at least moderate severity, at least once per week 2. Pain is intermittent 3. Not generalized or localized to other abdominal or chest regions 4. Not relieved by defecation or passage of flatus 5. Not fulfilling criteria for gallbladder or sphincter of Oddi disorders Supportive Criteria 1. Pain may be of a burning quality, but without a retrosternal component 2. Pain is commonly induced or relieved by ingestion of a meal, but may occur while fasting 3. Postprandial distress syndrome may coexist *According to the Rome III committee. † Criteria fulfilled for the previous 3 months with symptom onset at least 6 months prior to diagnosis. Adapted from Tack J, Talley NJ, Camilleri M, et al. Functional gastroduodenal disorders. In: Drossman DA, Corazziari E, Delvaux M, et al, editors. Rome III. The Functional Gastrointestinal Disorders. 3rd ed. McLean, Va: Degnon Associates; 2006; pp 427-428.

Chapter 13  Dyspepsia than anticipated overlap between EPS and PDS.33 On the other hand, an open-access endoscopy-based study found considerable overlap in endoscopic findings between patients with EPS or PDS and a large group of dyspeptic patients who were not classified with either.34 The validity of the Rome III classification will have to be assessed in additional ongoing and future studies.

Overlap with Heartburn and Irritable Bowel Syndrome

The issue of overlap of dyspepsia with GERD has been a challenging one. Although earlier investigators considered a group of patients with reflux-like dyspepsia,4 the Rome committees did not consider heartburn to arise primarily from the gastroduodenal region, and this symptom was thus excluded from the definition of dyspepsia.2,5 Heartburn commonly occurs along with dyspeptic symptoms, however, both in the general population and in those with a diagnosis of functional dyspepsia.23,27,35 Nevertheless, separating GERD from dyspepsia is hampered by a number of confounding factors, such as the presence of dyspepsia-type symptoms in many patients with GERD36 and difficulties in recognizing heartburn by patients and physicians.37,38 The Rome II consensus committee stated that patients with typical heartburn as a dominant complaint almost invariably have GERD and should be distinguished from patients with dyspepsia.2 Although this distinction is probably valid, it has become clear that the predominant symptom approach does not reliably identify or exclude patients with GERD. The Rome III consensus committee has proposed identification of patients with frequent heartburn, and the suggestion has been made that a word-picture questionnaire be used to facilitate recognition of heartburn by patients and to identify patients with functional dyspepsia who may respond to acid suppressive therapy or in whom pathologic esophageal acid exposure can be demonstrated.39,40 Whereas the Rome II definition for functional dyspepsia excluded patients with predominant heartburn and was unclear about those with nonpredominant heartburn, the Rome III definition stated that heartburn is not a gastroduodenal symptom, although it often occurs simultaneously with symptoms of functional dyspepsia and its presence does not exclude the diagnosis of functional dyspepsia.5 Similarly, the frequent co-occurrence of functional dyspepsia and irritable bowel syndrome (IBS)41 is explicitly recognized and does not exclude a diagnosis of functional dyspepsia.

EPIDEMIOLOGY

Dyspeptic symptoms are common in the general population, with frequencies ranging from 10% to 45%.11,16,23,27,42-44 The frequency of dyspepsia is slightly higher in women, and the influence of age varies among studies. The results of prevalence studies are strongly influenced by the criteria used to define dyspepsia, and several studies included patients with typical symptoms of GERD or did not take into account the presence of dyspepsia-like symptoms in many patients with GERD. When heartburn is excluded, the frequency of uninvestigated dyspepsia in the general population is in the range of 5% to 15%.43,44 Long-term follow-up studies have suggested improvement or resolution of symptoms in approximately half of patients. The annual incidence rate of dyspepsia has been estimated to range from 1% to 6%. Quality of life is significantly affected by dyspepsia, especially functional dyspepsia. Although most patients do not seek medical care, a significant proportion will eventually proceed with a consultation, constituting a major impact on the cost of care.16,45-47 Factors that influence health care–

seeking are the severity of symptoms, fear of underlying serious disease, psychological distress, and lack of adequate psychosocial support (see later).48

PATHOPHYSIOLOGY

Several pathophysiologic mechanisms have been suggested to underlie functional dyspeptic symptoms. These suggested mechanisms include delayed gastric emptying, impaired gastric accommodation to a meal, hypersensitivity to gastric distention, altered duodenal sensitivity to lipids or acid, abnormal intestinal motility, and central nervous system dysfunction.3 The heterogeneity of functional dyspepsia seems to be confirmed in the contribution of one or more of these disturbances in subgroups of patients. The studies that have investigated the pathophysiologic mechanisms of functional dyspepsia predate the Rome III consensus committee and classification. Therefore, most studies define functional dyspepsia according to the Rome I and II consensus definitions.

Delayed Gastric Emptying

Several studies have investigated gastric emptying and its relationship to the pattern and severity of symptoms in patients with functional dyspepsia. The frequency of delayed gastric emptying has ranged from 20% to 50%.3,5 In a meta-analysis of 17 studies involving 868 dyspeptic patients and 397 control subjects, a significant delay in gastric emptying of solids was present in almost 40% of patients with functional dyspepsia.49 Most of the studies, however, were performed in small groups of patients with small groups of control subjects. In the largest studies, gastric emptying of solids was delayed in about 30% of the patients with functional dyspepsia. Most studies failed to find a convincing relationship between delayed gastric emptying and the pattern of symptoms. Three large-scale single-center studies from Europe have shown that patients with delayed gastric emptying for solids are more likely to report postprandial fullness, nausea, and vomiting,20,50,51 although two other large multicenter studies in the United States found no or a very weak association.52,53 Whether delayed gastric emptying causes symptoms or is an epi­ phenomenon is a matter of ongoing controversy.

Impaired Gastric Accommodation

The motor functions of the proximal and distal stomach differ remarkably. Whereas the distal stomach regulates gastric emptying of solids by grinding and sieving the contents until the particles are small enough to pass the pylorus, the proximal stomach serves mainly as a reservoir during and after ingestion of a meal. Accommodation of the stomach to a meal results from a vagally mediated reflex relaxation of the proximal stomach that provides the meal with a reservoir and enables the stomach to handle large intragastric volumes without a rise in intragastric pressure.54 Studies using a gastric barostat, scintigraphy, ultraso­ nography, single photon emission computed tomography (SPECT), or noninvasive surrogate markers (e.g., satiety drinking test) have all suggested the presence of impaired gastric accommodation in approximately 40% of patients with functional dyspepsia.3,5,17,19,54 Insufficient accommodation of the proximal stomach during and after the ingestion of a meal may be accompanied by increased intragastric pressure and activation of mechanoreceptors in the gastric wall, thus inducing symptoms. Although a number of studies found associations between impaired accommodation and both early satiation and weight loss, other studies failed to find such associations. In addition, the mechanisms whereby impaired accommodation can be a cause of

187

188

Section III  Symptoms, Signs, and Biopsychosocial Issues symptoms is still unclear; meal ingestion in the absence of proper relaxation of the proximal stomach may be accompanied by activation of tension-sensitive mechano­ receptors in the proximal stomach. On the other hand, insufficient accommodation of the proximal stomach may force the meal into the distal stomach, thereby causing activation of tension-sensitive mechanoreceptors in a distended antrum.

Hypersensitivity to Gastric Distention

Visceral hypersensitivity, defined as abnormally enhanced perception of visceral stimuli, is considered one of the major pathophysiologic mechanisms in the functional gastrointestinal disorders (see Chapters 11, 21, and 118).55 Several studies have established that as a group, patients with functional dyspepsia are hypersensitive to isobaric gastric distention.3,5,18 The level at which visceral hypersensitivity is generated is unclear, and evidence exists for involvement of tension-sensitive mechanoreceptors as well as alterations at the level of visceral afferent nerves or of the central nervous system.56-58

Altered Duodenal Sensitivity to Lipids or Acid

In healthy subjects and in persons with functional dyspepsia, duodenal perfusion with nutrient lipids, but not glucose, enhances the perception of gastric distention through a mechanism that requires lipid digestion and subsequent release of cholecystokinin.59-61 Duodenal infusion of hydrochloric acid induces nausea in persons with functional dyspepsia but not in healthy subjects, thereby suggesting duodenal hypersensitivity to acid.62 Duodenal pH monitoring using a clipped pH electrode has revealed increased postprandial duodenal acid exposure in patients with functional dyspepsia compared with controls, and this difference has been attributed to impaired clearance of acid.63 On the basis of these observations, it has been proposed that increased duodenal sensitivity to lipids or acid may contribute to the generation of symptoms in patients with functional dyspepsia, but further research in this area is needed.

Other Mechanisms

One study has reported a high frequency of rapid gastric emptying in patients with functional dyspepsia; rapid gastric emptying was correlated with postprandial symptom intensity.64 Other studies that evaluated rapid emptying, however, failed to find such a high frequency or correlation.20,32 Phasic fundic contractions induce transient increases in gastric wall tension, which can be perceived by patients with functional dyspepsia.57 One study has reported lack of suppression of phasic contractility of the proximal stomach after a meal in a subset of patients with functional dyspepsia.65 Evidence also exists that abnormalities in the control of gastric myoelectrical activity, as measured by cutaneous electrogastrography, are found in up to two thirds of patients with functional dyspepsia.66,67 No correlation was found between the symptom pattern and presence of electrogastrographic findings. Small intestinal motor alterations, usually hypermotility with burst activity or clusters and an increased proportion of duodenal retrograde contractions (see Chapter 97), have been reported in patients with functional dyspepsia, but no clear correlation with symptoms has been found.68

PATHOGENIC FACTORS

The cause of symptoms in patients with functional dyspepsia has not been established, but evidence exists for genetic

susceptibility, infectious factors, and psychological factors. The relationship between potential pathogenic factors and putative pathophysiologic mechanisms has not been addressed in depth.

Genetic Predisposition

Population studies have suggested that genetic factors contribute to functional dyspepsia. The risk of dyspepsia is increased in first-degree relatives of patients compared with their spouses.69 In a case-control study, polymorphisms of the GNB3 gene that encodes guanine nucleotide binding protein, beta polypeptide 3 (especially the homozygous GNB3 825C state), were associated with symptoms of functional dyspepsia in blood donors and patients.70 Subsequently, a community-based study in the United States reported that both homozygous variants (CC and TT) of GNB3 were associated with meal-unrelated dyspepsia.71

Infection

Helicobacter pylori Infection Depending on the region and population studied, a variable proportion of patients with functional dyspepsia are infected with H. pylori.3,5 Although H. pylori is associated with a number of organic causes of dyspepsia, little evidence supports a causal relationship between H. pylori infection and functional dyspepsia.72 No consistent differences in the pattern of symptoms or putative pathophysiologic mechanisms have been found between H. pylori–positive and H. pylori–negative subjects.73 The best evidence in support of a role for H. pylori in functional dyspepsia is the small but statistically significant beneficial effect of eradication therapy on symptoms (see later).74 Postinfection Functional Dyspepsia Postinfection functional dyspepsia has been proposed as a possible clinical entity on the basis of a large retrospective study from a tertiary referral center.19 Compared with patients who had functional dyspepsia of unspecified onset, patients with a history suggestive of postinfection functional dyspepsia were more likely to report symptoms of early satiation, weight loss, nausea, and vomiting and had a significantly higher frequency of impaired accommodation of the proximal stomach, which was attributed to dysfunction at the level of gastric nitrergic (nitroxidergic) neurons.19 In a prospective cohort study, development of functional dyspepsia was increased fivefold in patients one year after acute Salmonella gastroenteritis compared with subjects who did not have gastroenteritis.75 Additional studies are required to identify the underlying pathophysiology and risk factors and to determine the long-term prognosis.

Psychosocial Factors

Review of the literature clearly reveals an association between psychosocial factors and functional dyspepsia.3,5,76-80 The most common psychiatric comorbidities in patients with functional dyspepsia are anxiety disorders, depressive disorders, somatoform disorders, and a recent or remote history of physical or sexual abuse.79,80 Psychological distress has long been assumed to be a feature of health care–seeking behavior in patients with functional bowel disorders, including functional dyspepsia. Studies have confirmed an association between dyspeptic symptoms in the general population and psychosocial factors such as somatization, anxiety, and life event stress; this association argues against a mere health care–seeking effect.31,81 Furthermore, the severity of symptoms in patients with functional dyspepsia seen in a tertiary care center is more strongly

Chapter 13  Dyspepsia related to psychosocial factors (especially depression, abuse history, and somatization) than to abnormalities of gastric sensorimotor function (see Chapter 21).82 Although these observations show a close interaction between different psychosocial variables and the presence and severity of symptoms of functional dyspepsia, they do not establish whether the psychosocial factors and dyspeptic symptoms are manifestations of a common predisposition or whether the psychosocial factors play a causal role in the pathophysiology of dyspeptic symptoms. The relationship is unlikely to be simple. A factor analysis of symptoms of functional dyspepsia and their relationship with pathophysiology and psychopathology has clearly demonstrated the heterogeneity and complexity of these interactions. It identified four separate functional dyspeptic symptom factors, of which the factor consisting of epigastric pain was associated with visceral hypersensitivity, several psychosocial dimensions, including somatization and neuroticism, and low health-related quality of life.24 These observations suggest a relationship between psychosocial factors and visceral hypersensitivity in particular. Acutely induced anxiety in healthy volunteers, however, was not associated with increased visceral sensitivity but with decreased gastric compliance and a significant inhibition of meal-induced accommodation.83 In patients with functional dyspepsia, a correlation between anxiety and gastric sensitivity was found in the subgroup of hypersen­ sitive patients, but not in the group as a whole.84 A history of physical or sexual abuse was associated with visceral hypersensitivity in patients with functional dyspepsia.85 Clearly, the role of psychosocial factors in the generation and severity of symptoms, especially in terms of their impact on clinical management, requires further study.

APPROACH TO UNINVESTIGATED DYSPEPSIA Considering the high prevalence of dyspepsia and the large number of persons who present to a physician for their symptoms, the initial aim of management is to decide which patients can be treated empirically and which patients should undergo additional diagnostic evaluation.

HISTORY AND PHYSICAL EXAMINATION

A complete clinical history should be obtained and a physical examination performed in all patients with dyspepsia. The nature of the symptoms, as well as their frequency and chronicity, should be ascertained, particularly with regard to their relationship to the ingestion of meals and the possible influence of specific dietary factors. The onset of symptoms—acute with a gastroenteritis-like episode or more chronic and gradual—is also of interest. The presence and amount of weight loss, if present, needs to be assessed, as should other alarm symptoms, such as anemia, blood loss, and dysphagia. Symptom subgroupings according to the Rome II or III classification have not proved to be of clinical usefulness. In cases of long-standing symptoms, the reason that the patient is seeking health care at this time should be elicited, so that specific fears and concerns can be addressed. Further assessment of symptoms or signs of systemic disorders (e.g., diabetes mellitus, cardiac disease, thyroid disorders) and of the patient’s family and personal history will indicate whether the patient is at risk for specific organic diseases that may present as dyspepsia. Physical findings, such as an abdominal mass or organomegaly, ascites, or fecal occult blood necessitate further evaluation. Specific attention should be given to a history of heartburn, and a word-picture questionnaire may help the

patient recognize the typical symptom pattern.37 Burning pain confined to the epigastrium is a cardinal symptom of dyspepsia and is not considered to be heartburn unless the pain radiates retrosternally. The presence of frequent and typical reflux symptoms should lead to a provisional diagnosis of GERD rather than dyspepsia, and the patient should initially be managed as a patient with GERD (see Chapter 43). On the other hand, overlap of GERD with dyspepsia is probably frequent and needs to be considered when symptoms do not respond to appropriate management of GERD. The possible presence of overlapping IBS should also be assessed, and symptoms that improve with bowel movements or are associated with changes in stool frequency or consistency should lead to a presumptive diagnosis of IBS. The use of prescription and nonprescription medications should be reviewed, and medications commonly associated with dyspepsia (especially NSAIDs) should be discontinued, if possible. In patients in whom NSAIDs cannot be discontinued, a trial of a proton pump inhibitor can be considered, although some guidelines recommend endoscopic evaluation to exclude peptic ulcer (see later).

LABORATORY TESTING

The cost-effectiveness of routine laboratory testing, especially in younger patients with uncomplicated dyspepsia, has not been established. Nevertheless, most clinicians will consider routine tests (complete blood count, routine electrolytes, serum calcium level, liver biochemical tests, and thyroid function tests) in patients older than 45 to 55 years. Other studies such as the serum amylase level, antibodies for celiac disease, stool testing for ova and parasites or Giardia antigen, and a pregnancy test may be considered in certain cases.

INITIAL MANAGEMENT STRATEGIES

In most cases, the patient’s history and physical examination will allow dyspepsia to be distinguished from symptoms suggestive of esophageal, pancreatic, or biliary disease. The history and physical findings, and even the presence of alarm symptoms, are unreliable for distinguishing functional from organic causes of dyspepsia by primary care physicians and by gastroenterologists.6,9,10,86,87 Therefore, most guidelines and recommendations advocate prompt endoscopy when risk factors such as NSAID use, age above a certain threshold, or alarm symptoms are present.88-90 The optimal management strategy for most patients who do not have a risk factor for an organic cause of dyspepsia remains a matter of debate and controversy, and several approaches have been proposed. The options include the following: (1) prompt diagnostic endoscopy, followed by targeted medical therapy; (2) noninvasive testing for H. pylori infection, followed by treatment based on the result (test and treat strategy); and (3) empirical antisecretory therapy. In the two latter strategies, endoscopy is performed in patients who do not respond to treatment or who experience recurrent symptoms after treatment.

Prompt Endoscopy and Directed Treatment

Diagnostic upper gastrointestinal endoscopy allows direct recognition of organic causes of dyspepsia such as peptic ulcer, erosive esophagitis, or malignancy. Endoscopy before any therapy has been instituted is still considered the diagnostic gold standard for patients with an upper gastrointestinal disorder.91 The procedure may also have a reassuring effect on physicians and patients.92-94 Gastric mucosal biopsies allow the diagnosis of H. pylori infection, with subsequent eradication therapy if the result is positive. Endoscopy

189

190

Section III  Symptoms, Signs, and Biopsychosocial Issues is claimed to permit diagnosis of early gastric cancer at a curable stage, but detecting early gastric cancer in a symptomatic individual is relatively rare, and evidence for the claim is weak.95-97 On the other hand, endoscopy is expensive and invasive and may not have as major an impact on treatment as hoped. Patients found to have a peptic ulcer or erosive esophagitis will receive antisecretory therapy. In those with negative upper endoscopy findings, functional dyspepsia and non­ erosive GERD are the likely diagnoses, and both are treated empirically with antisecretory therapy. Still, it is argued that initial empirical antisecretory therapy will only delay endoscopy, because functional dyspepsia and GERD are likely to recur after discontinuation of antisecretory therapy, at which time the patient will be referred for endoscopy anyway. A number of randomized controlled trials have compared prompt endoscopy with empirical noninvasive management strategies. A meta-analysis of five trials that compared initial endoscopy with a test and treat strategy has concluded that initial endoscopy may be associated with a small reduction in the risk of recurrent dyspeptic symptoms but that this gain is not cost-effective.98 Most relevant studies found that the direct and indirect costs are higher with prompt endoscopy and that these costs are not completely offset by a reduction in medication use or the number of subsequent physician visits.99-101 The available data, therefore, do not support early endoscopy as a cost-effective initial management strategy for all patients with uncomplicated dyspepsia. Nevertheless, most available practice guidelines advocate initial endoscopy in all patients above a certain age threshold, usually 45 to 55 years, to detect a potentially curable upper gastrointestinal malignancy.88-90 The rationale for this approach is that the vast majority of gastric malignancies occur in patients older than 45 years and that the rate of cancer detection rises in persons with dyspepsia older than age 45.95-97 Most patients with newly diagnosed gastric cancer, however, are already incurable at the time of diagnosis, and many will have some alarm features that would have warranted immediate endoscopy.97 Early endoscopy is also recommended in patients younger than 45 years who have a family history of gastric cancer, emigrated from a country with a high rate of gastric cancer, or had a prior partial gastrectomy.

Test and Treat for H. pylori Infection

H. pylori infection is causally associated with most peptic ulcers and is the most important risk factor for gastric cancer.102 Because of the involvement of H. pylori in peptic ulcer disease, several consensus panels have advocated noninvasive testing for H. pylori in young patients (younger than 45 to 55 years) with uncomplicated dyspepsia. Patients with a positive test result receive eradication therapy (a proton pump inhibitor and two antibiotics, such as amoxicillin and clarithromycin, taken for 7 to 14 days (see Chapter 50), whereas patients with a negative test result are treated empirically, usually with a proton pump inhibitor. The benefits of this test and treat strategy are the cure of peptic ulcer disease or prevention of future peptic ulcers and symptom resolution in a small subset (approximately 7% above the rate with placebo) of patients with functional dyspepsia who are infected with H. pylori.74,103 Eradication of H. pylori eliminates chronic gastritis and, in theory, may thereby contribute to a reduction in the risk of H. pylori– associated gastric cancer.104 On the other hand, in Western countries, the prevalence of H. pylori infection in patients with uninvestigated dys-

pepsia is rapidly declining, and infection rates are especially low in persons younger than 30 years (10% to 30%). Widespread use of antibiotics has the disadvantages of inducing resistance and occasionally causing allergic reactions. Whether eradication of H. pylori causes or worsens GERD has not been proved and is an ongoing matter of debate.102,105 Furthermore, the accuracy of noninvasive testing for H. pylori depends on the prevalence of H. pylori in the population as well as the sensitivity and specificity of the test. Serologic tests are the least expensive but also the least accurate. If the prevalence of H. pylori in a population is less than 60%, the fecal antigen and urea breath tests for H. pylori are preferred, because their higher accuracy rates lead to a reduction in inappropriate treatment for patients without H. pylori infection (see Chapter 50).106 Randomized placebo-controlled trials have shown only a modest reduction in symptoms of dyspepsia after a test and treat approach in primary care.107-109 A meta-analysis of studies that compared a test and treat strategy with empirical antisecretory therapy in persons with uninvestigated dyspepsia has found little difference in symptom resolution or costs between the two approaches.110 Although earlier models that assumed a higher prevalence rate of H. pylori infection suggested a greater benefit to a test and treat approach,111-113 subsequent economic models have suggested that the test and treat strategy may be equally or less cost-effective than empirical antisecretory therapy.114,115 The test and treat strategy as an initial approach is most likely to be beneficial in areas where the H. pylori infection rate is high.

Empirical Antisecretory Therapy

Initial empirical antisecretory therapy is widely used in primary care for patients with uninvestigated dyspepsia. This approach is attractive because it controls symptoms and heals lesions in most patients with underlying GERD or peptic ulcer disease, and may provide symptomatic benefit in up to one third of patients with functional dyspepsia.116,117 Proton pump inhibitors provide symptomatic relief superior to that of histamine H2 receptor agonists, and the response usually occurs within two weeks of initiating therapy.99 Disadvantages of empirical proton pump inhibitor therapy are a rapid relapse in symptoms after cessation of therapy and the potential for rebound gastric hypersecretion when therapy is discontinued.115,116 Many patients, therefore, will continue to take proton pump inhibitor therapy chronically. As noted earlier, a meta-analysis of studies that compared a test and treat strategy with empirical antisecretory therapy in persons with dyspepsia found little difference in symptom resolution or costs between the two strategies.110 Empirical antisecretory therapy may be equally or more cost-effective.114,115

Recommendations

The optimal cost-effective approach to the initial management of uncomplicated dyspepsia remains unclear, and clinical decisions should take into account specific aspects of a patient’s case and weigh several risk-benefit factors. In a young dyspeptic patient (younger than 45 to 55 years) without alarm features, initial endoscopy cannot be recommended because the yield is low and the test is unlikely to lead to improved outcomes. This position can be reconsidered if the patient is worried about an underlying disease, has a family history of cancer, or has emigrated from an area with a high incidence of gastric or esophageal cancer.

Chapter 13  Dyspepsia In a population with a high prevalence rate (>20%) of H. pylori infection, the test and treat approach remains attractive because it will cure patients with peptic ulcer disease. The tests of choice for H. pylori infection are the urea breath test or the fecal antigen test. H. pylori–positive patients should be given a 7- to 14-day course of H. pylori eradication therapy. In those who are negative for H. pylori, a proton pump inhibitor can be prescribed for one to two months. In populations in which the prevalence of H. pylori infection is low, empirical antisecretory therapy (a proton pump inhibitor for one to two months) appears to be the preferred option. Patients who fail to respond to these initial approaches, and possibly those with recurrent symptoms after cessation of antisecretory therapy, should undergo endoscopy, although the yield is still likely to be low. In patients older than 45 to 55 years without alarm features, most guidelines recommend initial diagnostic endoscopy, although a benefit in the detection of early-stage malignancies remains unproved. In these cases, management will depend on the endoscopic findings and detection of H. pylori infection, but proton pump inhibitor therapy is likely to be prescribed to most patients.

ADDITIONAL INVESTIGATIONS

Additional investigations may be pursued in patients with progressive or refractory dyspepsia that does not respond to the initial management approaches described earlier. Testing for celiac disease and Giardia infection is useful for patients with refractory symptoms, especially when accompanied by weight loss. In patients with severe pain or weight loss, abdominal ultrasonography or computed tomography scans can be used to rule out pancreaticobiliary disease and to screen for stenosis of large abdominal arteries. In cases of severe postprandial fullness, and especially in cases of refractory nausea and vomiting, gastric emptying testing using scintigraphy or a breath test can be considered (see Chapter 48). In cases of a severe delay in gastric emptying, a small bowel series can rule out mechanical obstruction as a contributing factor. In cases of refractory intermittent pain or epigastric burning, esophageal pH with impedance monitoring is useful for diagnosing atypical manifestations of GERD that is not responsive to empirical antisecretory therapy (see Chapter 43). Psychological or psychiatric assessment is recommended in cases of long-standing refractory or debilitating symptoms. Electrogastrography, barostat studies, or simple nutrient challenge tests have been used in pathophysiologic studies but have no established role in the clinical management of dyspeptic patients.

might be advisable.59,60 Similarly, consumption of spicy foods containing capsaicin and other irritants is often discouraged.15 Coffee may aggravate symptoms in some cases119 and, if implicated, should be avoided. Cessation of smoking and alcohol consumption is suggested to be helpful, with no convincing evidence of efficacy.120 The avoidance of aspirin and other NSAIDs is commonly recommended and seems sensible, although not of established value.14,15 If a patient has an apparent coexisting anxiety disorder or depression, appropriate treatment should be considered (see later).

PHARMACOLOGIC TREATMENT

For many but not all patients, pharmacotherapy will be considered. The efficacy of pharmacologic treatments for functional dyspepsia is limited, however.

Acid Suppressive Drugs

In patients with gastroesophageal reflux, a trial of antisecretory therapy often has therapeutic and diagnostic value. Based on meta-analyses of therapeutic outcomes in patients with functional dyspepsia, the efficacy of antacids, sucralfate, and misoprostol has not been demonstrated.121 A metaanalysis of 12 randomized placebo-controlled trials that evaluated the efficacy of H2 receptor antagonists in patients with functional dyspepsia reported a significant benefit over placebo, with a relative risk reduction of 23% and a number needed to treat of 7.121 H2 receptor blockers thus appear to be efficacious in functional dyspepsia. Many of these trials, however, probably included patients with GERD under a broad interpretation of functional dyspepsia, thereby accounting for much of the benefit. A meta-analysis of eight placebo-controlled, randomized trials of proton pump inhibitors for functional dyspepsia also confirmed that this class of agents was superior to placebo, with a number needed to treat of 10 (Table 13-4).121,122 The relative risk reduction (13%) was lower than that for H2 receptor blockers, probably reflecting more stringent entry criteria and better exclusion of patients with GERD. No difference in efficacy was found between halfdose and full-dose proton pump inhibitors, and a double dose of a proton pump inhibitor was also not superior to a single dose. H. pylori status did not affect the response to proton pump inhibitor therapy. Subgrouping of patients with functional dyspepsia using Rome definitions showed a trend for proton pump inhibitor therapy to be most effective in the group with overlapping dyspepsia and reflux, less effective in those with only epigastric pain, and ineffective in those with dysmotility.

Eradication of H. pylori Infection TREATMENT OF FUNCTIONAL DYSPEPSIA GENERAL MEASURES

Reassurance and education are of primary importance in patients with functional dyspepsia. In spite of normal findings at endoscopy, the patient should be given a confident and positive diagnosis. In patients with IBS, a positive physician-patient interaction can reduce health care– seeking behavior, and this approach is probably also valid for patients with functional dyspepsia.118 Lifestyle and dietary measures are usually prescribed to patients with functional dyspepsia, but the impact of dietary interventions has not been studied systematically.11 Having patients eat more frequent, smaller meals seems logical. Because the presence of lipids in the duodenum enhances gastric sensitivity, avoiding meals with a high fat content

A Cochrane meta-analysis has reported a 10% pooled relative risk reduction in dyspepsia for therapy to eradicate H. pylori infection, compared with placebo, at 12 months of follow-up, with a number needed to treat of 14 (Table 13-5).74 Arguments against eradication therapy are the low number of responders and the delayed occurrence of a demonstrable symptomatic benefit. On the other hand, H. pylori eradication can induce sustained remission in dyspepsia, albeit in a small minority of patients.123 Other arguments in favor of the use of eradication therapy are protection against peptic ulcer, presumed protection against gastric cancer, and short-term nature and relatively low cost of the treatment.

Prokinetic Agents

Gastric prokinetic agents are a heterogeneous class of compounds that act through different types of receptors. The

191

192

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 13-4  Meta-Analysis of 10 Randomized Controlled Trials of Proton Pump Inhibitor (PPI) Therapy in Patients with Functional Dyspepsia STUDY/YEAR Blum 2000 Bolling-Stemevald 2002 Farup 1999 Gerson 2005 Peura 2004 (M96) Peura 2004 (M97) Talley 1998 (BOND) Talley 1998 (OPERA) van Zanten 2006 Wong 2002

PPI n/N

PLACEBO n/N

274/395 71/100 6/14 16/21 165/261 164/249 242/423 277/403 49/109 231/301

171/203 79/96 8/10 9/19 104/131 109/133 81/110 71/102 62/115 107/152

RELATIVE RISK 95% CI

RELATIVE RISK 95% CI 0.82 [0.75, 0.90] 0.86 [0.74, 1.01] 0.54 [0.27, 1.06] 1.61 [0.95, 2.74] 0.80 [0.70, 0.90] 0.80 [0.71, 0.91] 0.78 [0.68, 0.89] 0.99 [0.85, 1.14] 0.83 [0.64, 1.09] 1.09 [0.97, 1.23]

Total (95% CI) 2276 1071 Total events: 1495 (PPI), 801 (Placebo) Test for heterogeneity chi-square = 31.33, df = 9, P = 0.0003, I2 = 71.3% Test for overall effect z = 279, P = 0.005

0.87 [0.80, 0.96]

0.1

0.2 0.5 10 1 2 5 Favors PPI Favors placebo

*For each trial, n/N represents the proportion of nonresponders (n) over the total number of patients in that group (N). CI, confidence interval. From Moayyedi P, Soo S, Deeks J, et al. Pharmacological interventions for non-ulcer dyspepsia. Cochrane Database Syst Rev 2006; (4):CD001960.

Table 13-5  Meta-Analysis of 12 Randomized Controlled Trials of Helicobacter pylori Eradication in Patients with Functional Dyspepsia STUDY/YEAR Blum (OCAY) 1998 Froehlich 2001 Gisbert 2004 Gonzalez Carro 2004 Hsu 2001 Kcetz 2003 Koskenpato 2001 Malfertheiner 2003 Martinek 2005 Mazzoleni 2006 McColl 1998 Miwa 2000 Ruiz 2005 Talley (ORCHID) 1999 Talley (USA) 1999 van Zanten 2003 Varannes 2001

TREATMENT n/N

CONTROL n/N

119/164 31/74 13/34 22/47 34/81 67/89 61/77 338/534 5/20 39/46 121/154 33/48 46/79 101/133 81/150 45/75 74/129

130/164 34/70 8/16 31/46 36/80 73/92 63/74 177/266 12/20 40/43 143/154 28/37 64/79 111/142 72/143 55/82 86/124

RELATIVE RISK 95% CI

RELATIVE RISK 95% CI 0.92 [0.81, 1.03] 0.86 [0.60, 1.24] 0.76 [0.40, 1.46] 0.69 [0.48, 1.00] 0.93 [0.66, 1.33] 0.95 [0.81, 1.11] 0.93 [0.80, 1.08] 0.95 [0.85, 1.06] 0.42 [0.18, 0.96] 0.91 [0.79, 1.06] 0.85 [0.77, 0.93] 0.91 [0.70, 1.18] 0.72 [0.58, 0.89] 0.97 [0.85, 1.11] 1.07 [0.86, 1.34] 0.89 [0.70, 1.14] 0.83 [0.68, 1.00]

Total (95% CI) 1934 1632 Total events: 1230 (Treatment), 1163 (Control) Test for heterogeneity chi-square = 17.69, df = 16, P = 0.34, I2 = 9.5% Test for overall effect z = 4.58, P < 0.00001

0.90 [0.86, 0.94]

0.1 0.2 0.5 1 2 5 10 Favors Treatment Favors Control *For each trial, n/N represents the proportion of nonresponders (n) over the total number of patients in that group (N). CI, confidence interval. From Moayyedi P, Soo S, Deeks J, et al. Eradication of Helicobacter pylori for non-ulcer dyspepsia. Cochrane Database Syst Rev 2006; (2):CD002096.

efficacy of available prokinetic agents in functional dyspepsia has been controversial.121,124,125 A meta-analysis, based mainly on studies of domperidone and cisapride, suggested superiority of prokinetic agents over placebo in patients with functional dyspepsia, with a relative risk reduction of 33% and a number needed to treat of six121; separate analy-

ses have also suggested efficacy for cisapride and domperidone individually.124 Metoclopramide and domperidone are dopamine receptor agonists with a stimulatory effect on upper gastrointestinal motility. Unlike metoclopramide, which may cause serious neurologic adverse effects, domperidone—which is not approved by the U.S. Food

Chapter 13  Dyspepsia and Drug Administration—does not cross the blood-brain barrier. Cisapride facilitates the release of acetylcholine in the myenteric plexus via 5-hydroxytryptamine4 (5-HT4) receptor agonism and accelerates gastric emptying. The available trials with these drugs, however, were often of poor quality, concerns were raised about publication bias, and cisapride has been withdrawn from the market because of cardiac safety concerns.121 Unfortunately, more recent studies with other types of prokinetic agents have generally not demonstrated symptomatic relief in patients with functional dyspepsia.125 The motilin receptor agonist, ABT-229, was actually found to worsen symptoms compared with placebo.126 Mosapride, which like cisapride is a mixed 5-HT4 receptor agonist and 5-HT3 receptor antagonist, demonstrated no benefit when compared with placebo in a large European study.127 The 5-HT4 receptor agonist tegaserod, 6 mg twice daily, was evaluated in two phase 3 randomized controlled trials in women with dysmotility-like functional dyspepsia. The two primary endpoints were the percentage of days with satisfactory symptom relief and the symptom severity on a composite average daily severity score. Statistical significance for both endpoints was obtained in one study but not in the other, and the overall therapeutic gain was small.128 The drug was well tolerated in this program but was withdrawn from the market because of an increased frequency of cardiovascular ischemic events. Itopride is a dopamine D2 antagonist and acetylcholinesterase inhibitor that was intensively studied in functional dyspepsia. A phase 2 placebo-controlled trial found significantly more responders to itopride, based on a global efficacy measure.129 No significant improvement in symptoms compared with placebo was observed, however, in two subsequent phase 3 trials.130

Antidepressants

Antidepressants are commonly used for the treatment of functional gastrointestinal disorders that do not respond to initial conventional approaches. Although systematic reviews suggest that anxiolytics and antidepressants, especially tricyclic antidepressants, may have some benefit in patients with functional gastrointestinal disorders, including functional dyspepsia (pooled relative risk reduction of 45%), the available trials are small and of poor quality, and publication bias cannot be excluded.131,132 A multicenter controlled trial of the tricyclic antidepressant desipramine in patients with functional bowel disorders failed to show benefit in an intention-to-treat analysis, but symptomatic improvement was obtained in a per-protocol analysis.133 Most of the enrolled patients, however, seemed to have IBS, and the number of patients with functional dyspepsia in the trial is unclear. The mechanism of action of antidepressants is also unclear; symptomatic relief from these medications appears to be independent of the presence of depression,133 and no significant effects of antidepressants on visceral sensitivity have been established in functional dyspepsia.134,135 The selective serotonin reuptake inhibitor paroxetine enhanced gastric accommodation in healthy subjects, but clinical studies evaluating this class of agents in functional dyspepsia are lacking. A large controlled trial with the selective serotonin and norepinephrine reuptake inhibitor venlafaxine in functional dyspepsia failed to show any benefit.135

Other Pharmacotherapeutic Approaches

Based on a meta-analysis of four trials, bismuth salts seemed efficacious, but the analysis had marginal statistical signi­ ficance.121 Simethicone was superior to placebo in one

controlled trial.136 Various studies reported an improvement in symptoms during treatment with mixed herbal prep­ arations, Chinese herbal preparations, or artichoke leaf extract.137-139 The data suggest that some of these preparations are effective, but the basis for the improvement remains to be determined. One study reported that the chronic administration of red pepper was more effective than placebo in decreasing the intensity of dyspeptic symptoms in patients with functional dyspepsia.140

New Drug Development

Fundic relaxants and visceral analgesics to reverse impaired gastric accommodation and visceral hypersen­sitivity are other attractive approaches for treating sensorimotor dis­ orders of the upper gastrointestinal tract. Although nitrates, sildenafil, and sumatriptan can relax the proximal stomach, they seem less suitable for therapeutic application in functional dyspepsia.54,125 A number of serotonergic drugs are also able to enhance gastric accommodation, including 5-HT1A, 5-HT3, and 5-HT4 receptor agonists.54,125 A clinical trial with a newly developed 5-HT1A receptor agonist R137696 in functional dyspepsia failed to show any symptomatic benefit.141 Acotiamide (Z-338) is a novel compound that enhances acetylcholine release via antagonism of the M1 and M2 muscarinic receptors (see Chapter 49). In a pilot study, acotiamide showed potential to improve symptoms and quality of life through a mechanism that may involve enhanced accommodation.142 Visceral hypersensitivity is another attractive target for drug development. The principal drug classes under evaluation are neurokinin receptor antagonists and peripherally acting kappa opioid receptor agonists. The kappa opioid agonist fedotozine showed potential efficacy in functional dyspepsia, but development of this drug was discontinued.143 More recently, asimadoline, another kappa opioid receptor agonist, failed to improve symptoms in a small pilot study.144

PSYCHOLOGICAL INTERVENTIONS

Studies have shown that patients with functional dyspepsia have a higher prevalence of psychosocial comorbidities, although the role of psychosocial factors in symptom generation remains unclear. Based in part on these comorbidities, psychological interventions such as group support with relaxation training, cognitive therapy, psychotherapy, and hypnotherapy have been used in patients with functional dyspepsia. A systematic review of clinical trials of psychological interventions for functional dyspepsia found that all trials claimed benefit from psychological interventions, with effects persisting for longer than one year, but all studies were limited by inadequate statistical analysis.145 The authors concluded that the evidence to confirm the efficacy of psychological interventions in functional dyspepsia is insufficient.

RECOMMENDATIONS In patients with functional dyspepsia who have mild or intermittent symptoms, reassurance, education, and some dietary changes may be sufficient (Figs. 13-3 and 13-4). Drug therapy can be considered for patients with more severe symptoms or those who do not respond to reassurance and lifestyle changes. Testing for H. pylori infection is recommended and, if the results are positive, eradication therapy can be prescribed. An immediate impact on symptoms is unlikely, however, and any potential benefit is observed mainly over longer follow-up. Both proton pump inhibitors

193

194

Section III  Symptoms, Signs, and Biopsychosocial Issues Uninvestigated dyspepsia

Clinical evaluation: History and physical examination Determine reason for presentation If patient has any of the following: Age >45-55 years Alarm features: Unexplained weight loss Bleeding Unexplained anemia Dysphagia Protracted vomiting Change in character of chronic symptoms Fear of cancer or organic disease

Consider: Dietary indiscretion Medication-induced dyspepsia Cardiac disease Conditions associated with gastroparesis Hepatobiliary disorders Other systemic disease

If patient has: Age 3 perception score). (From Serra J, Azpiroz F, Malagelada J-R. Impaired transit and tolerance of intestinal gas in the irritable bowel syndrome. Gut 2001; 48:14-9.)

similar volumes of bowel gas in patients with IBS and healthy controls.2,70,71 Although some studies using plain abdominal radiography72,73 found that intra-abdominal gas content was greater in patients with IBS than in healthy subjects (by 54% to 118%), no significant correlations were observed between intestinal gas content and bloating. The most compelling evidence for the lack of an association between bloating and the volume of bowel gas was provided by a study in which a validated CT technique was used to show that most patients with bloating have normal volumes of bowel gas. Excessive gas was observed, however, in patients with severe motility disorders.65,74 Although the volume of bowel gas seemingly is normal in persons with bloating, multiple studies using intestinal gas infusion have shown consistently that these patients have impaired handling of the infused gas. In response to an exogenous gas load, these patients exhibit gas retention, abdominal symptoms, or both (Fig. 16-5).2,70,71 These abnor­ malities apparently reflect impaired reflex control of gas transit40,51,75,76 as well as the frequently demonstrated hyper­ sensitivity to bowel distention that characterizes patients with IBS.77-79 Therefore, gas transit studies seem to provide a sensitive method of identifying subtle intestinal motor disturbances not detectable by conventional diagnostic tests. Although total gas volume is not increased in patients with IBS, disturbances in gas propulsion may lead to local­ ized gas accumulations (e.g., the splenic flexure syndrome) that cause symptoms in the hyperreactive intestines of patients with bloating. A study of patients with IBS in which CT images were compared under basal conditions and during an episode of severe distention has demonstrated that the sensation of distention is associated with an increase in the anteropos­

Because patients with bloating and abdominal distention seem to have a common variant of IBS, the basic approach to treatment should be similar to that prescribed for IBS (see Chapter 118). These patients, however, may have a disorder resulting from a number of altered pathophysiologic mecha­ nisms. A hypersensitive gut, for example, may be associated with impaired anal evacuation, particularly in patients with constipation-predominant IBS, and symptoms will worsen if gas production is increased. In these patients, the treat­ ment strategy may need to be modified.82 Many therapies have been claimed to relieve bloating, but the few wellcontrolled studies were directed toward symptoms of IBS in general, not bloating specifically. Nonpharmacologic Therapies Intestinal clearance of perfused gas is increased by mild exercise and the erect posture, which may explain anecdotal observations that activity (as opposed to resting in the supine posture) improves bloating symptoms.39,83,84 Although intestinal gas volumes appear to be normal in patients with bloating, the sensitivity of their intestines to normal volumes of bowel contents suggests that limiting gas production to a minimum may be beneficial. Therefore, dietary manipula­ tions to reduce gas production (described earlier) may be beneficial. Low-fiber85 and low-residue diets also have been reported to improve symptoms.86 More than 20 reports on the use of probiotics to reduce symptoms of IBS have been published. The results are promising but inconsistent, possibly depending on the bacterial species used, doses, duration of treatment, and endpoints used for evaluation. Hypnosis has been reported to reduce symptoms of IBS, including bloating.87 Pharmacologic Therapies Although studies have suggested that antibiotics, particu­ larly rifaximin, can reduce symptoms of IBS,55 IBS may first appear after antibiotic therapy.88,89 Until more data are avail­ able, using antibiotics to treat bloating seems inadvisable. Simethicone has defoaming properties that eliminates bubbles that might trap gas,90 but it does not reduce the volume of gas. The effectiveness of this compound in the treatment of gas symptoms remains controversial.91 Neostigmine, a potent prokinetic agent, has been reported to reduce abdominal symptoms resulting from an intestinal infusion of gas. Chronic administration of pyridostigmine improves symptoms in patients complaining of bloating but has only marginal effects on intestinal gas content.71,74 In placebo-controlled trials, the prokinetic agents metoclo­ pramide and the restricted drug cisapride have produced statistically significant reductions in complaints of abdomi­ nal distention.92,93 Tegaserod (no longer available in the United States) also has been shown to reduce bloating and distention in some, but not all, controlled trials carried out with this agent.94

239

240

Section III  Symptoms, Signs, and Biopsychosocial Issues As noted, patients with bloating have reduced tolerance to gas infused in the intestine, and the perception of intes­ tinal distention by healthy subjects is reduced when intes­ tinal motor activity is inhibited by glucagon. Therefore, drugs that reduce intestinal motility theoretically could enhance the tolerance of the hypersensitive gut to normal volumes of gas. Anticholinergic drugs such as hyoscyamine, dicyclomine, or scopolamine inhibit intestinal motility and may reduce the response to bowel gas. A meta-analysis of the efficacy of smooth muscle relaxants in the treatment of IBS46,95 has concluded that these drugs are superior to placebo in the management of symptoms, specifically abdominal pain and distention. These drugs, however, may enhance gas retention, and anecdotal evidence suggests that their administration may actually worsen abdominal disten­ tion in some patients. Peppermint oil has an antispasmodic effect on the gastrointestinal tract because of the calcium channel blocker activity of its active constituent, menthol.96 Although some studies have reported improvement in abdominal distention and reduction in flatus emission with peppermint oil, a meta-analysis97 has indicated that its benefit in IBS is questionable. Drugs that act on the efferent nerves of the intestine also might be effective. For example, low-dose tricyclic antidepressants and possibly selective serotonin reuptake inhibitors have proven useful in the treatment of abdominal pain and symptoms of IBS, an effect probably related to their antinociceptive action (see Chapter 118).98

PNEUMATOSIS CYSTOIDES INTESTINALIS

Pneumatosis cystoides intestinalis and coli is a condition characterized by the presence of gas-filled cysts in the wall of the small bowel, colon, or both (see Chapter 124). The cysts may be asymptomatic or associated with diarrhea, bloating, or abdominal pain. Many patients with pneuma­ tosis have extremely high breath H2 concentrations, a finding indicative of high luminal concentrations of H2.99,100 The feces of three patients with pneumatosis of the colon were found to have unusually low concentrations of H2consuming organisms, and a patient with pneumatosis limited to the small intestine had small bowel contents that produced but could not consume H2. Therefore, the high luminal H2 of these subjects appears to reflect H2 production that is relatively unopposed by H2 consumption. An association between pneumatosis and chronic adminis­ tration of chloral hydrate seemingly is explained by the ability of chloral hydrate to inhibit H2 consumption by intestinal flora.101 How a high luminal H2 tension results in pneumatosis is controversial. One proposal is that the high luminal H2 results in supersaturation of tissue with H2. As a result, H2 bubbles form via a process similar to that which results in tissue collections of gas in deep sea divers.102 A second theory proposes that small intramural gas collections nor­ mally occur with some frequency, but are quickly absorbed

into the circulation. In the presence of high H2 production, rapid diffusion of luminal H2 into the cyst dilutes other cyst gases (e.g., N2).100 Thus, the cyst N2 tension remains lower than or equal to that in the blood. As a result, N2 in the cyst cannot be absorbed and the cyst persists. The most effective treatment to eliminate the cysts is the administration of high concentrations of O2 via inhalation.103 This maneuver reduces the blood N2 tension to a value below that of the cyst, allowing N2 to diffuse from the cyst into the blood, with resolution of the cyst. Other forms of therapy that may be effective are heliox (a low-density gas mixture), antibiotics that inhibit H2 production (ciprofloxacin has been used successfully in a patient with small intestinal bacterial overgrowth and pneumatosis cystoides intesti­ nalis), and dietary manipulations, such as lactose restric­ tion, that reduce the delivery of fermentable substrate to the colonic bacteria.

KEY REFERENCES

Accarino A, Perez F, Azpiroz F, et al. Intestinal gas and bloating: Effect of prokinetic stimulation. Am J Gastroenterol 2008; 103:2036-42. (Ref 74.) Agrawal A, Houghton LA, Lea R, et al. Bloating and distention in irri­ table bowel syndrome: The role of visceral sensation. Gastroenterol­ ogy 2008; 134:1882-9. (Ref 63.) Azpiroz F, Malagelada J-R. Abdominal bloating. Gastroenterology 2005; 129:1060-78. (Ref 60.) Bredenoord AJ, Smout AJ. Physiologic and pathologic belching. Clin Gastroenterol Hepatol 2007; 5:772-5. (Ref 5.) Houghton LA, Lea R, Agrawal A, et al. Relationship of abdominal bloat­ ing to distention in irritable bowel syndrome and effect of bowel habit. Gastroenterology 2006; 131:1003-10. (Ref 62.) Levitt MD, Furne J, Aeolus MR, et al. Evaluation of an extremely flatu­ lent patient: Case report and proposed diagnostic and therapeutic approach. Am J Gastroenterol 1998; 11:2276-81. (Ref 50.) Levitt MD, Furne J, Olsson S. The relation of passage of gas an ab­ dominal bloating to colonic gas production. Ann Intern Med 1996; 124:422-4. (Ref 19.) Levitt M, Olsson S. Pneumatosis cystoides intestinalis and high breath H2 excretion: Insights into the role of H2 in this condition. Gastroen­ terology 1995; 108:1560-5. (Ref 100.) Passos MC, Tremolaterra F, Serra J, et al. Impaired reflex control of intestinal gas transit in patients with abdominal bloating. Gut 2005; 54:344-8. (Ref 75.) Perez F, Accarino A, Azpiroz F, et al. Gas distribution within the human gut: Effect of meals. Am J Gastroenterol 2007; 102:842-9. (Ref 8.) Posserud I, Stotzer PO, Bjornsson ES, et al. Small intestinal bacterial overgrowth in patients with irritable bowel syndrome. Gut 2007; 56:802-8. (Ref 67.) Salvioli B, Serra J, Azpiroz F, et al. Origin of gas retention and symptoms in patients with bloating. Gastroenterology 2005; 128:574-9. (Ref 51.) Suarez F, Furne J, Springfield J, et al. Insights into human colonic physiology obtained from the study of flatus composition. Am J Physiol 1997; 272:G1028-33. (Ref 14.) Suarez FL, Springfield J, Levitt MD. Identification of gases responsible for the odour of human flatus and evaluation of a device purported to reduce this odour. Gut 1998; 43:100-4. (Ref 31.) Tremolaterra F, Villoria A, Azpiroz F, et al. Impaired viscerosomatic reflexes and abdominal wall dystony associated with bloating. Gas­ troenterology 2006; 130:1062-8. (Ref 81.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

17 Fecal Incontinence Satish S. C. Rao

CHAPTER OUTLINE Epidemiology  241 Health Care Burden  241 Pathophysiology  242 Functional Anatomy and Physiology of the Anorectum  242 Pathogenic Mechanisms  243 Evaluation  245 Clinical Features  245

Fecal incontinence is usually defined as the involuntary passage of fecal matter through the anus or the inability to control the discharge of bowel contents. The severity of incontinence can range from occasional unintentional elimination of flatus to the seepage of liquid fecal matter or the complete evacuation of bowel contents. Consequently, the problem has been difficult to characterize from an epidemiologic and pathophysiologic standpoint, but undoubtedly causes considerable embarrassment, a loss of self-esteem, social isolation, and a diminished quality of life.1

EPIDEMIOLOGY Fecal incontinence affects people of all ages, but its prevalence is disproportionately higher in middle-aged women, older adults, and nursing home residents. Estimates of the prevalence of fecal incontinence vary greatly and depend on the clinical setting, definition of incontinence, frequency of occurrence, and influence of social stigma and other factors.2 Both the embarrassment and social stigma attached to fecal incontinence make it difficult for subjects to seek health care; consequently, treatment is often delayed for several years. Fecal incontinence not only causes significant morbidity in the community, but also consumes substantial health care resources. In a U.S. householder survey, frequent leakage of stool or fecal staining for more than one month were reported by 7.1% and 0.7% of the population, respectively.3 In the United Kingdom, two or more episodes of fecal incontinence per month were reported by 0.8% of patients who presented to a primary care clinic.4 In an older self-caring population (older than 65 years), fecal incontinence occurred at least once a week in 3.7% of subjects and in more men than women (ratio of 1.5 : 1).5 The frequency of fecal incontinence increases with age, from 7% in women younger than 30 years to 22% in women in their seventh decade.6,7 By contrast, 25% to 35% of institutionalized patients and 10% to 25% of hospitalized geriatric patients have fecal incontinence.1 In the United States, fecal incontinence

Physical Examination  246 Diagnostic Testing  246 Treatment  250 Supportive Measures  250 Specific Therapies  252 Treatment of Subgroups of Patients  256

is the second leading reason for placement in a nursing home. In a survey of 2570 households, comprising 6959 individuals, the frequency of at least one episode of fecal incontinence during the previous year was 2.2%; among affected persons, 63% were women, 30% were older than 65 years, 36% were incontinent of solid stool, 54% were incontinent of liquid stool, and 60% were incontinent of flatus.1 Furthermore, in another prospective survey of patients who attended either a gastroenterology or a primary care clinic, over 18% reported fecal incontinence at least once a week.8 Only one third had ever discussed the problem with a physician. When stratified for the frequency of episodes, 2.7% of patients reported incontinence daily, 4.5% weekly, and 7.1% monthly.8 In another survey, fecal incontinence was associated with urinary incontinence in 26% of women who attended a urology-gynecology clinic.9 A high frequency of mixed fecal and urinary incontinence was also reported in nursing home residents. Persons with incontinence were 6.8 times as likely to miss work or school, and missed an average of 50 work or school days per year, compared with those without incontinence or other functional gastrointestinal symptoms.3

HEALTH CARE BURDEN The cost of health care related to fecal incontinence includes measurable components such as the evaluation, diagnostic testing, and treatment of incontinence, the use of disposable pads and other ancillary devices, skin care, and nursing care. Approximately $400 million/year is spent on adult diapers,8 and between $1.5 and $7 billion/year is spent on care for incontinence among institutionalized older patients.1,2,10 In a long-term facility, the annual cost for a patient with mixed fecal and urinary incontinence was $9,711.11 In the outpatient setting, the average cost per patient (including evaluation) has been estimated to be $17,166.12 In addition, these persons incur costs that cannot be easily measured and that result from their impaired quality of life and social dysfunction.7 Fecal incontinence

241

242

Section III  Symptoms, Signs, and Biopsychosocial Issues is most likely to affect a person’s quality of life significantly and lead to increased use of health care, predominantly in women with moderate to severe symptoms.

PATHOPHYSIOLOGY FUNCTIONAL ANATOMY AND PHYSIOLOGY OF THE ANORECTUM

A structurally and functionally intact anorectal unit is essential for maintaining normal continence of bowel contents (see Chapters 98 and 125).13 The rectum is a hollow muscular tube composed of a continuous layer of longitudinal muscle that interlaces with the underlying circular muscle. This unique muscle arrangement enables the rectum to serve both as a reservoir for stool and as a pump for emptying stool. The anus is a muscular tube 2 to 4 cm in length that at rest forms an angle with the axis of the rectum (Fig. 17-1). At rest, the anorectal angle is approximately 90 degrees; with voluntary squeeze, the angle becomes more acute, approximately 70 degrees; and during defecation the angle becomes obtuse, about 110 to 130 degrees. The anal sphincter consists of two muscular components—the internal anal sphincter (IAS), a 0.3- to 0.5-cm thick expansion of the circular smooth muscle layer of the rectum, and the external anal sphincter (EAS), a 0.6- to 1.0-cm thick expansion of the levator ani muscles. Morphologically, both sphincters are separate and heterogenous.14 The IAS is composed predominantly of slow-twitch, fatigueresistant smooth muscle and generates mechanical activity with a frequency of 15 to 35 cycles/min as well as ultraslow waves at 1.5 to 3 cycles/min.13 The IAS contributes approximately 70% to 85% of the resting anal sphincter pressure but only 40% of the pressure after sudden distention of the rectum and 65% during constant rectal distention; the remainder of the pressure is provided by the EAS.15 Therefore, the IAS is responsible chiefly for maintaining anal continence at rest. The anus is normally closed by the tonic activity of the IAS. This barrier is reinforced during voluntary squeeze by the EAS. The anal mucosal folds, together with the expansive anal vascular cushions (see later), provide a tight seal.16 These barriers are augmented by the puborectalis muscle, which forms a flap-like valve that creates a forward pull and reinforces the anorectal angle.13

Anorectal angle Levator ani muscle Puborectalis shelf

Symphysis pubis

Coccyx Internal anal sphincter muscle External anal sphincter muscle Posterior Anterior Figure 17-1.  Sagittal diagrammatic view of the anorectum. (From Rao SSC. Pathophysiology of adult fecal incontinence. Gastroenterology 2004; 126:S14-22.)

The anorectum is richly innervated by sensory, motor, and autonomic nerves and by the enteric nervous system. The principal nerve to the anorectum is the pudendal nerve, which arises from the second, third, and fourth sacral nerves (S2, S3, S4), innervates the EAS, and subserves sensory and motor function.17 A pudendal nerve block creates a loss of sensation in the perianal and genital skin and weakness of the anal sphincter muscle but does not affect rectal sensation.15 A pudendal nerve block also abolishes the rectoanal contractile reflexes (see later), an observation that suggests that pudendal neuropathy may affect the rectoanal contractile reflex response. The sensation of rectal distention is most likely transmitted along the S2, S3, and S4 parasympathetic nerves. These nerve fibers travel along the pelvic splanchnic nerves and are independent of the pudendal nerve.13 How humans perceive stool contents in the anorectum is not completely understood. Earlier studies failed to demonstrate rectal sensory awareness.13 Subsequent studies have confirmed that balloon distention is perceived in the rectum and that such perception plays a role in maintaining continence.16,18 Furthermore, sensory conditioning can improve hyposensitivity19,20 and hypersensitivity21 of the rectum. Mechanical stimulation of the rectum can produce cerebral evoked responses,22 thereby confirming that the rectum is a sensory organ. Although organized nerve endings are not present in the rectal mucosa or myenteric plexus, myelinated and unmyelinated nerve fibers are present.13 These nerves most likely mediate the distention or stretch-induced sensory responses as well as the viscerovisceral,22 rectoanal inhibitory, and rectoanal contractile reflexes. The sensation of rectal distention is most likely transmitted via the parasympathetic nervi erigentes along the S2, S3, and S4 splanchnic nerves. Rectal sensation and the ability to defecate can be abolished completely by resection of the nervi erigentes.23 If parasympathetic innervation is absent, rectal filling is perceived only as a vague sensation of discomfort. Even persons with paraplegia or sacral neuronal lesions may retain some degree of sensory function, but almost no sensation is felt if lesions occur in the higher spine.15,18,24 Therefore, the sacral nerves are intimately involved in the maintenance of continence. The suggestion has been made that bowel contents are sensed periodically by anorectal sampling,25 the process whereby transient relaxation of the IAS allows the stool contents from the rectum to come into contact with specialized sensory organs in the upper anal canal. Specialized afferent nerves may exist that subserve sensations of touch, temperature, tension, and friction, but the mechanisms are incompletely understood.13 Incontinent patients appear to sample rectal contents less frequently than continent subjects. The likely role of anal sensation is to facilitate discrimination between flatus and feces and the fine-tuning of the continence barrier, but its precise role has not been well characterized. Rectal distention is associated with a fall in anal resting pressure known as the rectoanal inhibitory reflex. The amplitude and duration of this relaxation increases with the volume of rectal distention. This reflex is mediated by the myenteric plexus and is present in patients in whom the hypogastric nerves have been transected and in those with a spinal cord lesion. The reflex is absent after transection of the rectum, but it may recover.18 Although the rectoanal inhibitory reflex may facilitate discharge of flatus, rectal distention is also associated with a rectoanal contractile response, a subconscious reflex effort to prevent release of rectal contents, such as flatus.26,27 This contractile response involves contraction of the EAS and is

Chapter 17  Fecal Incontinence mediated by the pelvic splanchnic and pudendal nerves. The amplitude and duration of the rectoanal contractile reflex also increases with rectal distention, up to a maximum volume of 30 mL. Abrupt increases in intra-abdominal pressure, as caused by coughing or laughing, are associated with an increase in anal sphincter pressure. A number of mechanisms, including reflex contraction of the puborectalis, may be involved. The blood-filled vascular tissue of the anal mucosa also plays an important role in producing optimal closure of the anus. An in vitro study has shown that even during maximal involuntary contraction, the internal sphincter ring is unable to close the anal orifice completely, and a gap of approximately 7 mm remains. This gap is filled by the anal cushions, which may exert pressures of up to 9 mm Hg and thereby contribute 10% to 20% to the resting anal pressure.26

PATHOGENIC MECHANISMS

Fecal incontinence occurs when one or more mechanisms that maintain continence is disrupted to the extent that other mechanisms are unable to compensate. Therefore, fecal incontinence is often multifactorial.2,27 In a prospective study, 80% of patients with fecal incontinence had more than one pathogenic abnormality (Fig. 17-2).13 Although the pathophysiologic mechanisms often overlap, they can be categorized under four broad groups, as summarized in Table 17-1.

cushions may lead to a poor seal and an impaired sampling reflex. These changes may cause passive incontinence or fecal seepage (see later), often under resting conditions. Both sphincters may be defective in many patients. The extent of muscle loss can influence the severity of incontinence.13 The most common cause of anal sphincter disruption is obstetric trauma, which may involve the EAS, IAS, or pudendal nerves. Why most women who have sustained an obstetric injury in their 20s or 30s typically do not present with fecal incontinence until their 50s, however, is unclear. In a prospective study, 35% of primiparous (normal antepartum) women showed evidence of anal sphincter disrup-

Anal sphincter dysfunction Pudendal neuropathy Impaired rectal sensation Poor rectal compliance Other

Abnormal Anorectal and Pelvic Floor Structures

Anal Sphincter Muscles Disruption or weakness of the EAS muscle causes urgerelated or diarrhea-associated fecal incontinence. In contrast, damage to the IAS muscle or anal endovascular

0

20

40

60

80

100

Frequency (%) Figure 17-2.  Relative frequencies of the common mechanisms that lead to fecal incontinence.

Table 17-1  Mechanisms, Causes, and Pathophysiology of Fecal Incontinence mechanism

CAUSES

Abnormal Anorectal or Pelvic Floor Structures Anal sphincter muscle Hemorrhoidectomy, neuropathy, obstetric injury Puborectalis muscle Aging, excessive perineal descent, trauma Pudendal nerve Excessive straining, obstetric or surgical injury, perineal descent Nervous system, spinal cord, autonomic Avulsion injury, spine surgery, diabetes mellitus, nervous system head injury, multiple sclerosis, spinal cord injury, stroke Rectum Aging, inflammatory bowel disease, irritable bowel syndrome, prolapse, radiation Abnormal Anorectal or Pelvic Floor Function Impaired anorectal sensation Autonomic nervous system disorders, central nervous system disorders, obstetric injury Fecal impaction Dyssynergic defecation Altered Stool Characteristics Increased volume and loose consistency Drugs, bile salt malabsorption, infection, inflammatory bowel disease, irritable bowel syndrome, laxatives, metabolic disorders Hard stools, retention Drugs, dyssynergia Miscellaneous Physical mobility, cognitive function Aging, dementia, disability Psychosis Willful soiling Drugs* Anticholinergics Antidepressants Caffeine Laxatives Muscle relaxants Food intolerance Fructose, lactose, or sorbitol malabsorption *Pathophysiology is noted for each class of drugs.

PATHOPHYSIOLOGY Sphincter weakness, loss of sampling reflex Obtuse anorectal angle, sphincter weakness Sphincter weakness, sensory loss, impaired reflexes Loss of sensation, impaired reflexes, secondary myopathy, loss of accommodation Loss of accommodation, loss of sensation, hypersensitivity Loss of stool awareness, rectoanal agnosia Fecal retention with overflow, impaired sensation Diarrhea and urgency, rapid stool transport, impaired accommodation Fecal retention with overflow Multifactorial changes Multifactorial changes Constipation Altered sensation, cconstipation Relaxation of sphincter tone Diarrhea Relaxation of sphincter tone Diarrhea, flatus

243

244

Section III  Symptoms, Signs, and Biopsychosocial Issues tion after vaginal delivery.28,29 Other important risk factors include a forceps-assisted delivery, prolonged second stage of labor, large birth weight, and occipitoposterior presentation.13 A prospective study of 921 primiparous women has shown that the frequencies of fecal incontinence at 6 weeks and 6 months postpartum are 27% and 17%, respectively, in subjects with vaginal delivery and a sphincter tear; 11% and 8%, respectively, in subjects with vaginal delivery but without a tear; and 10% and 7.6%, respectively, in subjects who underwent cesarean section.30 This study showed clearly that the occurrence and severity of fecal incontinence were attributable to an anal sphincter tear that occurred at the time of vaginal delivery. Episiotomy is believed to be a risk factor for anal sphincter disruption. In one study, medial episiotomy was asso­ ciated with a ninefold higher risk of anal sphincter dysfunction.31 Regardless of the type of delivery, however, incontinence of feces or flatus occurred in a surprisingly large percentage of middle-aged women, thereby suggesting that age-related changes in the pelvic floor may predispose to fecal incontinence. Aging affects anal sphincter function.32 In men and women older than 70 years, sphincter pressures decrease by 30% to 40% compared with younger persons.33 Also, in all age groups, anal squeeze pressure is lower in women than men,33 with a rapid fall after menopause.34 Estrogen receptors have been identified in the human striated anal sphincter, and ovariectomy in rats leads to atrophy of the striated anal sphincter muscle.13,35 These observations suggest that the strength and vigor of the pelvic floor muscles are influenced by hormones. Pudendal nerve terminal motor latency (PNTML) is prolonged in older women, and pelvic floor descent is excessive on straining.36 These mechanisms may contribute to progressive damage to the striated anal sphincter muscle. Aging is also associated with increased thickness and echogenicity of the IAS.37 Other causes of anatomic disruption include anorectal surgery for hemorrhoids, fistulas, and fissures. Anal dilation or lateral sphincterotomy may result in incontinence because of fragmentation of the anal sphincters.38 Hemorrhoidectomy can cause incontinence by inadvertent damage to the IAS39 or loss of endovascular cushions. Accidental perineal trauma or a pelvic fracture may also cause direct sphincter trauma that leads to fecal incontinence,40 but anoreceptive intercourse is not associated with anal sphincter dysfunction.41 Finally, IAS dysfunction may also occur because of myopathy, degeneration, or radiotherapy.13 Puborectalis Muscle The puborectalis muscle is also important for maintaining continence by forming a flap valve mechanism.42 Studies using three-dimensional ultrasound have shown that 40% of women with fecal incontinence have major abnormalities, and another 32% have minor abnormalities of the puborectalis muscle, compared with 21% and 32%, respectively, of asymptomatic parous controls.43 Also, assessment of puborectalis function by a perineal dynamometer revealed impaired puborectalis (levator ani) contraction in patients with fecal incontinence, and this finding was an independent risk factor for and correlated with the severity of fecal incontinence.44 Furthermore, improvement in puborectalis strength following biofeedback therapy was associated with clinical improvement, in part because the upper portion of the puborectalis muscle receives its innervations from branches of the S3 and S4 sacral nerves rather than the pudendal nerve. Therefore, the puborectalis muscle and EAS have separate neurologic innervations. Consequently, pudendal blockage does not abolish voluntary con-

traction of the pelvic floor45 but completely abolishes EAS function.15 Nervous System Intact innervation of the pelvic floor is essential for maintaining continence. Sphincter degeneration secondary to pudendal neuropathy and obstetric trauma may cause fecal incontinence in women.28 The neuropathic injury is often sustained during childbirth, probably as a result of stretching of the nerves during elongation of the birth canal or direct trauma during the passage of the fetal head. The nerve damage is more likely to occur when the fetal head is large, the second stage of labor is prolonged, or forceps are applied, especially with a high-forceps delivery or prolonged labor. The role of extrinsic autonomic innervation is somewhat controversial. Animal studies have shown that the pelvic nerves convey fibers that relax the rectum.46 Consequently, these nerves may play a role in accommodating and storing feces and gas. Damage to the pelvic nerves may lead to impaired accommodation and rapid transit through the rectosigmoid region, thereby overwhelming the continence barrier mechanisms. Sympathetic efferent activity, as studied by stimulating the presacral sympathetic nerves, tends to relax the IAS, whereas parasympathetic stimulation may cause contraction of the anal sphincter. The upper motor neurons for voluntary sphincter muscle lie close to those that innervate the lower limb muscles in the parasagittal motor cortex, adjacent to the sensory representation of the genitalia and perineum in the sensory cortex.13 Consequently, damage to the motor cortex from a central nervous system (CNS) lesion may lead to incontinence. In some patients with neurogenic incontinence, the sensory and motor nerve fibers may be damaged, resulting in sensory impairment.47 This damage can impair conscious awareness of rectal filling as well as the associated reflex responses in the striated pelvic floor sphincter muscles. Approximately 10% of patients with fecal incontinence may have a lesion more proximal than the intrapelvic or perianal nerves. The primary abnormality in these patients is cauda equina nerve injury,48 which may be occult and not evident through clinical evaluation. These patients have a prolongation of nerve conduction along the cauda equina nerve roots without an abnormality in PNTML.49 In a minority of patients, however, a combination of peripheral and central lesions is present. Other disorders such as multiple sclerosis, diabetes mellitus, and demyelination injury (or toxic neuropathy from alcohol or traumatic neuropathy) may also lead to incontinence.13 Rectum The rectum is a compliant reservoir that stores stool until social conditions are conducive to its evacuation.2 If rectal wall compliance is impaired, a small volume of stool material can generate a high intrarectal pressure that can overwhelm anal resistance and cause incontinence.50 Causes include radiation proctitis, ulcerative colitis, or Crohn’s disease, infiltration of the rectum by tumor, and radical hysterectomy.51 Similarly, rectal surgery, particularly pouch surgery,52 and spinal cord injury53 may be associated with loss of rectal compliance.

Abnormal Anorectal and Pelvic Floor Function

Impaired Anorectal Sensation An intact sensation not only provides a warning of imminent defecation, but also helps distinguish among formed stool, liquid feces, and flatus. Older persons,54 those who are physically and mentally challenged, and children with

Chapter 17  Fecal Incontinence fecal incontinence55 often show blunted rectal sensation. Impaired rectal sensation may lead to excessive accumulation of stool, thereby causing fecal impaction, megarectum (extreme dilatation of the rectum), and fecal overflow. Causes of impaired sensation include neurologic damage such as multiple sclerosis, diabetes mellitus, and spinal cord injury.53 Less well known is that analgesics (particularly opiates) and antidepressants also may impair rectal sensation and produce fecal incontinence. The importance of the rectum in preserving continence has been demonstrated conclusively through surgical studies in which preservation of the distal 6 to 8 cm of the rectum, along with its parasympathetic nerve supply, helped subjects avoid incontinence.56 By contrast, rectal sensation and the ability to defecate can be abolished completely by resection of the nervi erigentes (see earlier).23 An intact sampling reflex allows an individual to choose whether to discharge or retain rectal contents. Conversely, an impaired sampling reflex may predispose a subject to incontinence.25 The role of the sampling reflex in maintaining continence, however, remains unclear. In children who have undergone colonic pull-through surgery (see Chapter 113), some degree of sensory discrimination is preserved.57 Because the anal mucosal sensory zone is absent in these children, the suggestion has been made that sensory receptors, possibly located in the puborectalis muscle, may play a role in facilitating sensory discrimination. Also, traction on the muscle is a more potent stimulus for triggering defecation and a sensation of rectal distention. Because abolition of anal sensation by the topical application of 5% lidocaine does not reduce resting sphincter pressure (although it affects voluntary squeeze pressure but does not affect the ability to retain saline infused into the rectum), the role of anal sensation in maintaining fecal continence has been questioned.13

of large-volume liquid stools, which often transit the hindgut rapidly, continence can only be maintained through intact sensation and a strong sphincteric barrier. Similarly, in patients with bile salt malabsorption, lactose or fructose intolerance, or rapid dumping of osmotic material into the colon, colonic transit of gaseous and stool contents is too rapid and can overwhelm the continence mechanisms (see Chapters 15 and 101).2

Dyssynergic Defecation and Incomplete Stool Evacuation In some patients, particularly older adults, prolonged retention of stool in the rectum or incomplete evacuation may lead to seepage of stool or staining of undergarments.54 Most of these patients show obstructive or dyssynergic defecation,58 and many of them also exhibit impaired rectal sensation, whereby anal sphincter and pudendal nerve function is intact but the ability to evacuate a simulated stool is impaired. Similarly, in older adults and in children with functional incontinence, the prolonged retention of stool in the rectum can lead to fecal impaction. Fecal impaction may also cause prolonged relaxation of the IAS, thereby allowing liquid stool to flow around impacted stool and to escape through the anal canal (see Chapter 18).55

CLINICAL FEATURES

Descending Perineum Syndrome In women with long-standing constipation and a history of excessive straining for many years (perhaps even without prior childbirth), excessive straining may lead to progressive denervation of the pelvic floor muscles.59 Most of these patients demonstrate excessive perineal descent and sphincter weakness, which may lead to rectal prolapse; however, fecal incontinence is not an inevitable consequence. Whether or not incontinence develops will depend on the state of the pelvic floor and the strength of the sphincter muscles.

Altered Stool Characteristics

The consistency, volume, and frequency of stool and the presence or absence of irritants in stool also may play a role in the pathogenesis of fecal incontinence.2 In the presence

Miscellaneous Mechanisms

Various medical conditions and disabilities may predispose to fecal incontinence, particularly in older adults. Immo­ bility and lack of access to toileting facilities are primary causes of fecal incontinence in this population.60 Several drugs may inhibit sphincter tone. Some are used to treat urinary incontinence and detrusor instability, including anticholinergics such as tolterodine tartarate (Detrol) and oxybutynin (Ditropan) and muscle relaxants such as baclofen (Lioresal), and cyclobenzaprine (Flexeril). Stimulants such as caffeinated products, fiber supplements, or laxatives may produce fecal incontinence by causing diarrhea.13

EVALUATION The evaluation of the patient with fecal incontinence includes a detailed clinical assessment and appropriate physiologic and imaging tests of the anorectum. These three sources of information are complementary and should provide useful data regarding the severity of the problem, underlying causative factors, and impact on the patient’s quality of life. On the basis of this information, appropriate treatment strategies can be designed. The first step in the evaluation of a patient with fecal incontinence is to establish a trusting relationship with the patient and assess the duration and nature of the symptoms, with specific attention to whether the leakage consists of flatus, liquid stool, or solid stool and to the impact of the symptoms on the quality of the patient’s life (Table 17-2). Because many people misinterpret fecal incontinence as diarrhea or urgency,61 a detailed characterization of the symptom(s) is important. The clinician should ask about the use of pads or other devices and the patient’s ability to discriminate between formed or unformed stool and gas (the lack of such discrimination is termed rectal agnosia).2 An obstetric history; history of coexisting conditions such as diabetes mellitus, pelvic radiation, neurologic problems, or

Table 17-2  Features of the History That Should Be Elicited from a Patient with Fecal Incontinence Onset and precipitating event(s) Duration and timing Severity Stool consistency and rectal urgency History of fecal impaction Coexisting problems (e.g., diarrhea, inflammatory bowel disease) Drugs, caffeine, diet Past history—spine surgery, urinary incontinence, back injury, diabetes mellitus, neurologic disorders Clinical subtypes—passive or urge incontinence or fecal seepage Obstetric history—use of forceps, tears, presentation of the infant, repairs

245

246

Section III  Symptoms, Signs, and Biopsychosocial Issues spinal cord injury; dietary history; and history of coexisting urinary incontinence are important. A prospective stool diary can be useful. The circumstances under which incontinence occurs should also be determined. Such a detailed inquiry may facilitate the recognition of the following types of fecal incontinence: 1. Passive incontinence, the involuntary discharge of fecal matter or flatus without any awareness. This pattern suggests a loss of perception or impaired rectoanal reflexes, with or without sphincter dysfunction. 2. Urge incontinence, the discharge of fecal matter or flatus despite active attempts to retain these contents. The predominant causes of this pattern are disruption of the sphincter function and a decrease in rectal capacity to retain stool. 3. Fecal seepage, the undesired leakage of stool, often after a bowel movement, with otherwise normal continence and evacuation. This condition results primarily from incomplete evacuation of stool or impaired rectal sensation.54,58 Sphincter function and pudendal nerve function are mostly intact. Although overlap exists among these three types, by determining the predominant pattern, useful insights can be gained regarding the underlying mechanism(s) and preferred management. Symptom assessment, however, may not correlate well with manometric findings (see later). In one study, leakage had a sensitivity of 98.9%, specificity of 11%, and positive predictive value of 51% for detecting a low resting anal sphincter pressure on manometry.62 The positive predictive value for detecting a low anal squeeze pressure was 80%. Therefore, for an individual patient with fecal incontinence, the history and clinical features alone are insufficient to define the pathophysiology, and objective testing is essential63,64 (see later). On the basis of the clinical features, several grading systems have been proposed. A modification of the Cleveland Clinic grading system65 has been validated by the St. Mark’s investigators66 and provides an objective method of quantifying the degree of incontinence. It can also be useful for assessing the efficacy of therapy. This grading system is based on seven parameters that include the following: (1-3) the character of the anal discharge as solid, liquid, or flatus; (4) the degree of alterations in lifestyle; (5, 6) the need to wear a pad or take antidiarrheal medication; and (7) the ability to defer defecation. The total score ranges from 0 (continent) to 24 (severe incontinence). As noted earlier, however, clinical features alone are insufficient to define the pathophysiology. The use of validated questionnaires such as the SCL-90R (Symptom Checklist-90-R) and SF-36 (Short-Form 36) surveys may provide additional information regarding psychosocial issues and the impact of fecal incontinence on the patient’s quality of life.

PHYSICAL EXAMINATION

A detailed physical examination, including a neurologic examination, should be performed in any patient with fecal incontinence, because incontinence may be secondary to a systemic or neurologic disorder. The focus of the examination is on the perineum and anorectum. Perineal inspection and digital rectal examination are best performed with the patient lying in the left lateral position and with good illumination. On inspection, the presence of fecal matter, prolapsed hemorrhoids, dermatitis, scars, skin excoriations, or a gaping anus and the absence of perianal creases may be noted. These features suggest sphincter weakness or chronic skin irritation and provide clues regarding the underlying cause.2 Excessive perineal descent or rectal prolapse can be

demonstrated by asking the patient to attempt defecation. An outward bulge that exceeds 3 cm is usually defined as excessive perineal descent (see Chapter 18).67 Perianal sensation should be checked. The anocutaneous reflex examines the integrity of the connections between the sensory nerves and the skin; the intermediate neurons in spinal cord segments S2, S3, and S4; and the motor innervation of the external anal sphincter. This reflex can be assessed by gently stroking the perianal skin with a cotton bud in each perianal quadrant. The normal response consists of a brisk contraction of the external anal sphincter (“anal wink”). An impaired or absent anocutaneous reflex suggests either afferent or efferent neuronal injury.2 After inserting a lubricated, gloved index finger into the anus and rectum, one should assess the resting sphincter tone, length of the anal canal, strength of the puborectalis sling, acuteness of the anorectal angle, strength of anal sphincter squeeze, and elevation of the perineum during voluntary squeeze. Also, the presence of a rectocele or impacted stool may be noted. The accuracy of the digital rectal examination has been assessed in several studies. In one study of 66 patients, digital rectal examination by an experienced surgeon correlated somewhat with resting sphincter pressure (r = 0.56; P < 0.001) or maximum squeeze pressure (r = 0.72; P < 0.001).68 In a study of 280 patients with various anorectal disorders, a reasonable correlation was reported between digital examination and manometric findings, but the sensitivity, specificity, and positive predictive values of digital examination were low.69 In another study of 64 patients, the agreement between digital rectal examination and resting or squeeze pressure was 0.41 and 0.52, respectively.70 These data suggest that digital rectal examination provides only an approximation of sphincter strength. The findings are influenced by many factors, including the size of the examiner’s finger, technique used, and cooperation of the patient. One study has shown that trainees lack adequate skills for recognizing the features of fecal incontinence on digital rectal examination.71 Therefore, digital rectal examination is not reliable and is prone to interobserver differences. Digital rectal examination can identify patients with fecal impaction and overflow but is not accurate for diagnosing sphincter dysfunction and should not be used as the basis for decisions regarding treatment.2

DIAGNOSTIC TESTING

The first step in assessing a patient with fecal incontinence is to determine whether the incontinence is secondary to diarrhea or independent of stool consistency. If diarrhea coexists with incontinence, appropriate tests should be performed to identify the cause of the diarrhea (see Chapter 15). Such testing may include flexible sigmoidoscopy or colonoscopy to exclude colonic mucosal inflammation, a rectal mass, or stricture and stool studies for infection, volume, osmolality, electrolytes, fat content, and pancreatic function. Biochemical tests should be performed to look for thyroid dysfunction, diabetes mellitus, and other metabolic disorders. Breath tests may be considered for lactose or fructose intolerance or small intestinal bacterial overgrowth.2 A history of cholecystectomy may suggest bile salt malabsorption and prompt a therapeutic trial of a bile salt– binding agent. Specific tests are available for defining the underlying mechanisms of fecal incontinence and are often used in complementary fashion. The most useful tests are anorectal manometry, anal endosonography, the balloon expulsion test, and PNTML.2,72-74

Chapter 17  Fecal Incontinence Anorectal Manometry and Sensory Testing

Anorectal manometry is a simple and useful method for assessing IAS and EAS pressures (Fig. 17-3) as well as rectal sensation, rectoanal reflexes, and rectal compliance. Several types of probes and pressure recording devices are available. Each system has distinct advantages and drawbacks. A water-perfused probe with multiple closely spaced sensors is commonly used.2 Increasingly, a solid-state probe with microtransducers or air-filled miniaturized balloons is used. A novel solid-state probe with 36 circumferential sensors spaced at 1-cm intervals, with a 4.2-mm outer diam100

Squeeze

Rest

Anal canal 1.0 cm (mm Hg)

Anal canal 2.5 cm (mm Hg)

Rectum (mm Hg)

80

A

60 40 20 0 100 80 60 40 20 0 100 80

eter (Sierra Scientific Instruments, Los Angeles) has been reported to provide higher resolution than older style probes.75 This device uses a novel pressure transduction technology (TactArray) that allows each of the 36 pressure sensing elements to detect pressure over a length of 2.5 mm and in each of 12 radially dispersed sectors. The data can be displayed in isobaric contour plots that can provide a continuous dynamic representation of pressure changes, although anal sphincter pressures are higher than those recorded with water-perfused manometry. A high-definition manometry system with 256 circumferentially arrayed sensors in a 5-cm probe76 has become available and may provide anal sphincter pressure profiles and topographic changes of even higher fidelity (Fig. 17-4). Anal sphincter pressures can be measured by stationary or station pull-through techniques.73,74 Resting anal sphincter pressure predominantly represents IAS function, and voluntary anal squeeze pressure predominantly represents EAS function. Patients with fecal incontinence have low resting and low squeeze pressures (see Figs. 17-3 and 17-4), indicating IAS and EAS weakness.2,69 The duration of sustained squeeze pressure provides an index of sphincter muscle fatigue. The ability of the EAS to contract reflexively can be assessed during abrupt increases in intra-abdominal pressure, as when the patient coughs. This reflex response causes the anal sphincter pressure to rise above that of the rectal pressure to preserve continence. The response may be triggered by receptors in the pelvic floor and mediated through a spinal reflex arc. In patients with a spinal cord

60 Posterior Maximal & Sustained Squeeze

40 20

Anterior

Posterior

0 13:29:30 100

13:30:00 Squeeze

13:30:30

13:31:00

Rest

Anal canal 1.0 cm (mm Hg)

Anal canal 2.5 cm (mm Hg)

Rectum (mm Hg)

80

B

60 40

A

20 0 100

Maximal Squeeze

80 60 40 20 0 100 80 60

B

40 20 0 13:29:30

13:30:00

13:30:30

13:31:00

Figure 17-3.  Anorectal manometry profiles in (A) a healthy normal subject in whom resting (internal anal sphincter) and squeeze (external anal sphincter) pressures are normal and (B) a patient with fecal incontinence in whom resting and squeeze pressures are weak. Upper tracings, Rectal pressure activity; middle tracings, anal pressure activity at 2.5 cm; lower tracings, anal pressure activity at 1.0 cm from the anal margin.

low pressure area

3D: SAGITTAL VIEW 2D: UNFOLDED VIEW Figure 17-4.  High dynamic anal sphincter vector topography showing pressure changes during maximal squeeze in three-dimensional (3D) sagittal view (left) and two-dimensional (2D) unfolded view (right). A, Changes in a healthy control subject. B, Changes in a subject with fecal incontinence. The subject with incontinence has significant weakness of the anal sphincter, with an asymmetrical squeeze and a change in some vectors (pre­ dominantly yellow and green), whereas the healthy subject shows a robust squeeze (orange and red) and symmetrical decrease in sphincter diameter.

247

248

Section III  Symptoms, Signs, and Biopsychosocial Issues lesion above the conus medullaris, this reflex response is preserved, even though voluntary squeeze may be absent, whereas in patients with a lesion of the cauda equina or sacral plexus, both the reflex and voluntary squeeze responses are absent.2,77,78

Rectal Sensory Testing

Rectal balloon distention with air or water can be used to assess sensory responses and compliance of the rectal wall. By distending a balloon in the rectum with incremental volumes, the thresholds for first perception, first desire to defecate, and urgent desire to defecate can be assessed. A higher threshold for sensory perception indicates reduced rectal sensitivity.2,77,79 The balloon volume required for partial or complete inhibition of anal sphincter tone also can be assessed. The volume required to induce reflex anal relaxation is lower in incontinent patients than in controls.80 Because sampling of rectal contents by the anal mucosa may play an important role in maintaining continence,25 quantitative assessment of anal perception using electrical or thermal stimulation has been advocated but is not used clinically.2 Rectal compliance can be calculated by assessing the changes in rectal pressure during balloon distention with air or fluid.73,81 Rectal compliance is reduced in patients with colitis,50 patients with a low spinal cord lesion, and diabetic patients with incontinence but is increased in those with a high spinal cord lesion. Anorectal manometry can provide useful information regarding anorectal function.72,73,82 The American Motility Society has provided consensus guidelines and minimal standards for manometry testing.74 Although there are insufficient data regarding normal values, overlap betwen healthy subjects and patients with fecal incontinence,69 and large confidence intervals for test reproducibility,83 manometry testing can be useful for the individual patient with fecal incontinence.74 Manometric tests of anorectal function may also be useful for assessing objective improvement following drug therapy, biofeedback therapy, or surgery.84-86

Imaging the Anal Canal

Anal Endosonography Anal endosonography is performed by using a 7- to 15-mHz rotating transducer with a focal length of 1 to 4 cm.87 The test provides an assessment of the thickness and structural integrity of the EAS and IAS and can detect scarring, loss of muscle tissue, and other local pathology (Fig. 17-5).88 Higher frequency (10- to 15-mHz) probes that provide

better delineation of the sphincter complex have become available.88 After vaginal delivery, anal endosonography has revealed occult sphincter injury in 35% of primipara women; most of these lesions were not detected clinically. In another study, sphincter defects were detected in 85% of women with a third-degree perineal tear compared with 33% of subjects without a tear.89 In studies that compared electromyography (EMG; see later) mapping with anal endosonography, the concordance rate for identifying a sphincter defect was high.90,91 The technique is, however, operatordependent and requires training and experience.73 Although endosonography can distinguish internal from external sphincter injury, it has a low specificity for demonstrating the cause of fecal incontinence.2 Because anal endoso­ nography is more widely available, less expensive, and certainly less painful than EMG, which requires needle insertion, it is the preferred technique for examining the morphology of the anal sphincter muscles. Magnetic Resonance Imaging Endoanal magnetic resonance imaging (MRI) has been shown to provide superior imaging with excellent spatial resolution, particularly for defining the anatomy of the EAS.92,93 One study,94 but not another,92 has shown that MRI is less accurate than anal endosonography. A major contribution of anal MRI has been the recognition of external sphincter atrophy, which may adversely affect sphincter repair95 (see later). Atrophy also may be present without pudendal neuropathy.96 The addition of dynamic pelvic MRI using fast imaging sequences or MRI colpocystography, which involves filling the rectum with ultrasound gel as a contact agent and having the patient evacuate while lying inside the magnet, may define the anorectal anatomy more precisely.97 The use of an endoanal coil significantly enhances the resolution and allows more precise definition of the sphincter muscles. Comparative studies of costs, availability, technical factors, clinical utility, and role in treatment decision making are warranted.

Defecography

Defecography uses fluoroscopic techniques to provide morphologic information about the rectum and anal canal.98 It is used to assess the anorectal angle, measure pelvic floor descent and length of anal canal, and detect the presence of a rectocele, rectal prolapse, or mucosal intussusception. Approximately 150 mL of contrast material is placed into the rectum, and the subject is asked to squeeze or cough and

EAS

Figure 17-5.  Anal endosonograms. A, Normal healthy subject with intact, hypoechoic internal anal sphincter (IAS) and intact, thicker, and hyperechoic external anal sphincter (EAS). B, Subject with fecal incontinence secondary to an obstetric injury causing a large anterior sphincter defect that involves the IAS and EAS and spans the circumference between the 10 and 2 o’clock positions (arrows).

IAS

A

B

Chapter 17  Fecal Incontinence

Normal

100 µV O

Time

A

P

Amplitude (µV)

P

O

Time

3 msec

Patient

100 µV

3 msec

B

Figure 17-6.  Pudendal nerve terminal motor latency time in a normal subject (A) and a patient with fecal incontinence and pudendal neuropathy (B). Compared with the pudendal terminal motor nerve latency time in the normal subject, the patient’s tracing shows a delayed onset (O) and peak (P). µV, microvolts; msec, milliseconds.

expel the contrast. Although defecography can detect a number of abnormalities, these abnormalities can also be seen in otherwise asymptomatic persons,73,99 and their presence correlates poorly with impaired rectal evacuation. Agreement between observers in the measurement of the anorectal angle is also poor. Whether one should use the central axis of the rectum or the posterior wall of the rectum when measuring the angle is unclear. The functional significance of identifying morphologic defects has been questioned. Although defecography can confirm the occurrence of incontinence at rest or during coughing, it is most useful for demonstrating rectal prolapse2,100 or poor rectal evacuation (see Chapter 18). In selected patients, magnetic resonance defecography may evaluate evacuation and identify coexisting problems such as a rectocele, enterocele, cystocele, or mucosal intussusception.88

Balloon Expulsion Test

Normal subjects can expel a 50-mL water-filled balloon101 or a silicone-filled artificial stool from the rectum in less than one minute.2 Most patients with fecal incontinence have little or no difficulty with evacuation, but patients with fecal seepage58 and many older persons with fecal incontinence secondary to fecal impaction54 demonstrate impaired evacuation. In these patients, a balloon expulsion test may help identify coexisting dyssynergia or a lack of coordination between the abdominal, pelvic floor, and anal sphincter muscles during defecation. One study has shown a high frequency of dyssynergia in residents of nursing homes (see Chapter 18).102

Neurophysiologic Testing

Electrical recording of the muscle activity from the anal sphincter (EMG) is a useful technique for identifying sphincter injury as well as denervation-reinnervation potentials that can indicate neuropathy.22,73 EMG can be performed using a fine wire needle electrode or a surface electrode, such as an anal plug. Abnormal EMG activity, such as fibrillation potentials and high-frequency spontaneous discharges, provides evidence of chronic denervation, which commonly is seen in patients with fecal incontinence secondary to pudendal nerve injury or cauda equina syndrome.103 The PNTML measures the neuromuscular integrity between the terminal portion of the pudendal nerve and the anal sphincter. Injury to the pudendal nerve leads to denervation of the anal sphincter muscle and muscle

weakness. Therefore, measurement of the nerve latency time can help distinguish muscle injury from nerve injury as the cause of a weak sphincter muscle. A disposable electrode (St. Mark’s electrode; Dantec, Denmark) is used to measure the latency time.104 A prolonged nerve latency time suggests pudendal neuropathy (Fig. 17-6). Women who have delivered vaginally with a prolonged second stage of labor or have had forceps-assisted delivery have been found to have a prolonged PNTML compared with women who delivered by cesarean section or spontaneously.105,106 An American Gastroenterological Association technical review did not recommend PNTML,73 although an expert review has noted that patients with pudendal neuropathy generally have a poor surgical outcome.107 A normal PNTML does not exclude pudendal neuropathy, because the presence of a few intact nerve fibers can lead to a normal result, whereas an abnormal latency time is significant. PNTML may be useful in the assessment of patients prior to anal sphincter repair and is particularly helpful in predicting the outcome of surgery. The integrity of the peripheral component of efferent motor pathways that control anorectal function can also be assessed by recording the motor evoked potentials (MEPs) of the rectum and anal sphincter in response to magnetic stimulation of the lumbosacral nerve roots (translumbar magnetic stimulation [TLMS] and transsacral magnetic stimulation [TSMS]).22,108,109 The technique is based on Faraday’s principle, which states that in the presence of a changing electrical field, a magnetic field is generated. Consequently, when a current is discharged rapidly through a conducting coil, a magnetic flux is produced around the coil. The magnetic flux causes stimulation of neural tissue. Magnetic stimulation of the lumbosacral roots (TLMS and TSMS) may allow more precise localization of the motor pathways between the brain and the anal sphincter as well as subcomponent analysis of the efferent nervous system between the brain and sphincter. Electrical or magnetic stimulation of the lumbosacral nerve roots facilitates measurement of the conduction time within the cauda equina and can diagnose sacral motor radiculopathy as a possible cause of fecal incontinence.110,111 One study has shown that translumbar MEP and transsacral MEP of the rectum and anus provides delineation of peripheral neuromuscular injury in subjects with fecal incontinence108 (Fig. 17-7) and can reveal hitherto undetected changes in patients with back injury.

249

Section III  Symptoms, Signs, and Biopsychosocial Issues LEFT

RIGHT Spinal cord injury

Spinal cord injury Amplitude

250

Healthy

Healthy

Time Figure 17-7.  Anal motor evoked potential (MEP) responses following translumbar magnetic stimulation in a subject with fecal incontinence and a history of spinal cord injury (upper tracings) and in a healthy normal subject (lower tracings). The MEP responses on the left and right sides are shown separately. When compared with the healthy subject, the incontinent subject with spinal cord injury shows an MEP response with a prolonged onset time on the left and right sides and a smaller amplitude of the MEP response on the right side. These features indicate that bilateral lumbospinal neuropathy is the cause of fecal incontinence.

Saline Infusion Test

The saline infusion test assesses the overall capacity of the defecation unit to maintain continence during conditions that simulate diarrhea.72,80,82 With the patient lying on the bed, a 2-mm plastic tube is introduced approximately 10 cm into the rectum and taped in position. Next, the patient is transferred to a commode. The tube is connected to an infusion pump and 800 mL of warm saline (37°C) is infused into the rectum at a rate of 60 mL/min. The patient is instructed to hold the liquid for as long as possible. The volume of saline infused at the onset of first leak (defined as a leak of at least 15 mL) and the total volume retained at the end of infusion are recorded. Most normal subjects should retain most of the infused volume without leakage, whereas patients with fecal incontinence or patients with impaired rectal compliance, such as those with ulcerative colitis,112 leak at much lower volumes. The test is also useful for assessing objective improvement of fecal incontinence after biofeedback therapy.85

Clinical Utility of Tests for Fecal Incontinence

In one prospective study, history taking alone could detect an underlying cause in only 9 of 80 patients (11%) with fecal incontinence, whereas physiologic tests revealed an abnormality in 44 patients (55%).113 In a large retrospective study of 302 patients with fecal incontinence, an underlying pathophysiologic abnormality was identified, but only after manometry, EMG, and rectal sensory testing were performed.114 Most patients had more than one pathophysiologic abnormality. In another large study of 350 patients, incontinent patients had lower resting and squeeze sphincter pressures, a smaller rectal capacity, and earlier leakage following saline infusion in the rectum.82 Nevertheless, results of a single test or a combination of three different tests (anal manometry, rectal capacity, saline continence test) provided a low discriminatory value between continent and incontinent patients. This finding emphasizes the wide range of normal values and the ability of the body to compensate for the loss of any one mechanism involved in fecal incontinence. In a prospective study, anorectal manometry with sensory testing not only confirmed a clinical impression, but also provided new information that was not detected clinically.72 Furthermore, the diagnostic information obtained from these studies can influence both the management and outcome of patients with incontinence. A single abnorma­ lity was found in 20% of patients, whereas more than one abnormality was found in 80% of patients. In another study,

abnormal sphincter pressure was found in 40 patients (71%), whereas altered rectal sensation or poor rectal compliance was present in 42 patients (75%).113 These findings were confirmed by another study, which showed that phy­ siologic tests provided a definitive diagnosis in 66% of patients with fecal incontinence.114 Still, on the basis of the test results alone, it is not possible to predict whether an individual patient is continent or incontinent. Consequently, an abnormal test result must be interpreted in the context of the patient’s symptoms and the results of other complementary tests. Tests of anorectal function provide objective data and define the underlying pathophysiology. Table 17-3 summarizes the key tests, information gained from them, and evidence to support their clinical use.

TREATMENT The goal of treatment for patients with fecal incontinence is to restore continence and improve their quality of life. Strategies that include supportive and specific measures may be used. An algorithmic approach to the evaluation and management of patients with fecal incontinence is presented in Figure 17-8.

SUPPORTIVE MEASURES

Supportive measures such as avoiding offending foods, ritualizing bowel habit, improving skin hygiene, and instituting lifestyle changes may serve as useful adjuncts to the management of fecal incontinence. Obtaining a comprehensive history (see Table 17-1), performing a detailed physical examination, and requesting that the patient keep a prospective stool diary2,73 can provide important clues regarding the severity and type of incontinence as well as predisposing conditions, such as fecal impaction, dementia, neurologic disease, inflammatory bowel disease, or dietary factors (e.g., carbohydrate intolerance). If present, these conditions should be treated or corrected. In the management of older or institutionalized patients with fecal incontinence, the availability of personnel experienced in the treatment of fecal incontinence, timely recognition of soiling, and immediate cleansing of the perianal skin are of paramount importance.60 Hygienic measures such as changing undergarments, cleaning the perianal skin immediately following a soiling episode, use of moist tissue paper (baby wipes) rather than dry toilet paper, and use of barrier creams such as zinc oxide and calamine lotion

Evaluates presence of fecal retention; inexpensive and widely available Simple, inexpensive, bedside assessment of ability to expel a simulated stool; identifies dyssynergic defecation

Colonic transit study with radiopaque markers Balloon expulsion test (BET)

Directly visualizes the colon to exclude mucosal lesions (e.g., solitary rectal ulcer syndrome, inflammation, malignancy)

Identifies megacolon, megarectum, stenosis, diverticulosis, extrinsic compression, and intraluminal masses

Simultaneously evaluates global pelvic floor anatomy and dynamic motion; reveals sphincter morphology and pathology outside the anorectum Identifies excessive amount of stool in the colon; simple, inexpensive, widely available

Invasive, risks related to procedure (perforation, bleeding) and sedation

Lack of standardization of interpretation, lack of controlled studies Lack of standardization, embarrassment, radiation exposure, lack of controlled studies

Interobserver bias; scars difficult to identify Radiation exposure, embarrassment, availability, interobserver bias, inconsistent methodology Expensive, lack of standardization, availability

Minimally invasive, low sensitivity, interobserver differences Lack of standardization, training, controlled studies, and availability Inconsistent methodology, validity has been questioned Lack of standardization

Invasive, painful; not widely available Inaccurate, frequent artifacts

Lack of standardization

WEAKNESSES

*Evidence-based summary. EAS, external anal sphincter; EMG, electromyography; IAS, internal anal sphincter; MRI, magnetic resonance imaging.

Endoscopy Flexible sigmoidoscopy and colonoscopy

Barium enema

Plain abdominal film

MRI

Defecography

Imaging Anorectal ultrasonography

Visualizes IAS and EAS defects, thickness, and atrophy and puborectalis muscles Detects prolapse, intussusception, obtuse anorectal angle, and pelvic floor weakness, as well as rectoceles and megarectum

Quantifies nerve conduction time of entire spinoanal and spinorectal pathways; minimally invasive

Translumbar and transsacral motor evoked potentials

Surface EMG

Needle EMG

Pudendal nerve terminal motor latency (PNTML)

clinical use and STRENGTHS

Clinical Use

Quantifies EAS and IAS pressures; identifies rectal hyposensitivity, rectal hypersensitivity, impaired rectal compliance, dyssynergic defecation Quantifies spike potentials and re-innervation pattern indicating neuropathy or myopathy Displays EMG activity; can provide information on normal or weak muscle tone Measures latency of terminal portion of pudendal nerve, simple to perform

Physiologic Anorectal manometry

TEST

Table 17-3  Diagnostic Tests for Fecal Incontinence*

Poor

Poor

Poor

Fair

Fair

Good

Good

Good

Fair

Fair

Fair

Fair

Good

quality of EVIDENCE

Indicated in patients with unexplained diarrhea and seepage and in subjects older than age 50 yr

Not recommended for routine evaluation but useful in older adults and children with incontinence and fecal impaction Not recommended as part of routine evaluation

Used as an adjunct to other tests

Useful and complementary with other tests

Most widely available

Useful for identifying patients with fecal seepage and older persons with impaction Normal BET does not exclude dyssynergia; should be interpreted in the context of other anorectal test results

Promising noninvasive test; more objective and higher yield than PNTML

Conflicting data; correlation with other tests and surgical outcome unclear

Used largely for neuromuscular training

Useful but used largely in research laboratories

Useful for detecting anal sphincter weakness, altered rectal sensation and accommodation, dyssynergia

COMMENTS

Chapter 17  Fecal Incontinence 251

252

Section III  Symptoms, Signs, and Biopsychosocial Issues History, physical examination (including digital rectal examination)

Diarrhea + incontinence

Obstetric, surgical, neurologic injury

Local anorectal problem Appropriate treatment (see Chapter 125)

Flexible sigmoidoscopy, colonoscopy, and/or barium enema + routine blood tests

Suspected rectal prolapse Clinically confirmed

Not confirmed

Positive

Defecography + MRI

All test results negative Trial of loperamide, diphenoxylate and atropine, or other antidiarrheal agent Improved

Not improved

Anorectal manometry + Anal endosonography + − Balloon expulsion test + − Neurophysiology tests (EMG/PNTML/MEP testing)

Weak sphincter or sphincter defect + No or mild neuropathy Neuromuscular training Figure 17-8.  Algorithm for the evaluation and management of patients with fecal incontinence. EMG, electromyography; MEP, motor evoked potential; MRI, magnetic resonance imaging; PNTML, pudendal nerve terminal motor latency.

Weak sphincter or sphincter defect + neuropathy Neuromuscular training or colostomy

Surgery

Normal ? Factitious incontinence

Impaired sensation

Dyssynergic defecation + impaired evacuation

Neuromuscular training

Neuromuscular training

If ineffective Sphincteroplasty or sphincter repair Sacral nerve stimulation Artificial bowel sphincter Colostomy

(Calmoseptine; Calmoseptine, Huntington Beach, Calif) may help prevent skin excoriation. Perianal fungal infections should be treated with topical antifungal agents. More importantly, scheduled toileting with a commode at the bedside or bedpan and supportive measures to improve the general well-being and nutritional status of the patient may prove effective. Stool deodorants (e.g., Bedside Care Perineal Wash, Minneapolis; Derifil, Integra, Plainsboro, NJ; Devrom, Parthenon, Salt Lake City) can help disguise the smell of feces. In an institutionalized patient, ritualizing the bowel habit and instituting cognitive training may prove beneficial. Using these measures, short-term (3- to 6-month) success rates of up to 60% have been reported in case series.115 Patients in whom these measures fail have been shown to have a higher mortality rate than those without incontinence and than those with incontinence who respond to these measures.116 Other supportive measures include dietary modifications, such as reducing caffeine or fiber intake. Caffeinecontaining coffee enhances the gastrocolic (or gastroileal) reflex, increases colonic motility,117 and induces fluid secretion in the small intestine.118 Therefore, reducing caffeine consumption, particularly after meals, may help lessen postprandial urgency and diarrhea. Brisk physical activity, particularly after meals or immediately after waking, may precipitate fecal incontinence because these physiologic events are associated with increased colonic motility.119

Decreased rectal reservoir Rectal augmentation surgery

Acute exercise can enhance colonic motor activity and transit.120 A food and symptom diary may identify dietary factors that cause diarrheal stools and incontinence; frequent culprits are lactose and fructose, which may be malabsorbed.121 Eliminating food items containing these constituents may prove beneficial.2 Fiber supplements such as psyllium are often advocated in an attempt to increase stool bulk and reduce watery stools. In a single casecontrolled study, psyllium led to a modest improvement,122 but fiber supplements can potentially worsen diarrhea by increasing colonic fermentation of unabsorbable fiber.

SPECIFIC THERAPIES Pharmacologic Therapy

The antidiarrheal agents loperamide hydrochloride (Imodium) and diphenoxylate and atropine sulfate (Lomotil ) remain the mainstays of drug treatment for fecal incontinence, although other drug treatments have been proposed.2,123 In placebo-controlled studies, loperamide, 4 mg three times daily, has been shown to reduce the frequency of incontinence, improve stool urgency, and increase colonic transit time,84 as well as increase anal resting sphincter pressure124 and reduce stool weight. Clinical improvement was also reported with diphenoxylate and atropine,125 but objective testing showed no improvement in the ability of the patient to retain saline or spheres in the

Chapter 17  Fecal Incontinence rectum. Although most patients benefit from antidiarrheal agents temporarily, many report cramping, lower abdominal pain, or difficulty with evacuation after a few days. Therefore, careful titration of the dose is required to produce the desired result. Idiopathic bile salt malabsorption may be an important underlying cause of diarrhea and fecal incontinence (see Chapter 15).126 Patients with this problem may benefit from titrated doses of ion exchange resins such as cholestyramine (Questran) or colestipol (Colestid). Alosetron (Lotronex), a 5-hydroxytryptamine3 receptor antagonist used for the treatment of irritable bowel syndrome and diarrhea, may serve as an adjunct to the therapy of fecal incontinence, but use of the drug is restricted because of side effects (see Chapter 118).127 Postmenopausal women with fecal incontinence may benefit from estrogen replacement therapy.128 An openlabeled study has shown that oral amitriptyline, 20 mg, is useful in the treatment of patients with urinary or fecal incontinence without evidence of a structural defect or neuropathy.129 Suppositories or enemas may also have a role in the treatment of incontinent patients with incomplete rectal evacuation or in those with postdefecation seepage. In some patients, constipating medications alternating with periodic enemas may provide more controlled evacuation of bowel contents, but these interventions have not been tested prospectively.

Neuromuscular Training

Pressure (mm Hg)

Neuromuscular training, usually referred to as biofeedback therapy, improves symptoms of fecal incontinence, restores quality of life, and improves objective parameters of ano­ rectal function. Biofeedback training is useful in patients with a weak sphincter or impaired rectal sensation. The method is based on operant conditioning techniques whereby an individual acquires a new behavior through a learning process of repeated reinforcement and instant feedback.2,130 The goals of neuromuscular training in a patient with fecal incontinence are as follows: (1) to improve the strength of the anal sphincter muscles; (2) to improve the coordination between the abdominal, gluteal, and anal 100 90 80 70 60 50 40 30 20 10 0 100 90 80 70 60 50 40 30 20 10 0

Squeeze

Rest

100 90 80 70 60 50 40 30 20 10 0 100 90 80 70 60 50 40 30 20 10 0

14:09:00

A

Time

sphincter muscles during voluntary squeeze and following rectal perception; and (3) to enhance anorectal sensory perception. Because each goal requires a specific method of training, the treatment protocol should be customized for each patient on the basis of the underlying pathophysiologic mechanism(s). Neuromuscular training often is performed using visual, auditory, or verbal feedback techniques, and the feedback is provided by a manometry or EMG probe placed in the anorectum.2,130 When a patient is asked to squeeze, the anal sphincter contraction is displayed as an increase in anal pressure or EMG activity. This visual cue provides instant feedback to the patient. The aim of rectoanal coordination training is to achieve a maximum voluntary squeeze in less than two seconds after a balloon is inflated in the rectum. In reality, this maneuver mimics the arrival of stool in the rectum and prepares the patient to react appropriately by contracting the right group of muscles.2,130 Patients are taught how to squeeze their anal muscles selectively without increasing intra-abdominal pressure or inappropriately contracting their gluteal or thigh muscles. Also, this maneuver identifies sensory delay and trains the individual to use visual clues to improve sensorimotor coordination.131,132 Sensory training of the rectum educates the patient to perceive a lower volume of balloon distention but with the same intensity as they had felt earlier with a higher volume. This goal is achieved by repeatedly inflating and deflating a balloon in the rectum. These neuromuscular training techniques must be used together with pelvic muscle strengthening (modified Kegel exercises) and other supportive measures to achieve sustained improvement of bowel function. A component analysis—muscle training, sensory training, or both—is most effective; whether Kegel exercises alone are more effective than the use of multiple approaches has not been determined. Predicting how many neuromuscular treatment sessions will be required is often difficult. Most patients seem to require between four and six training sessions (Fig. 17-9).2,85,130 Studies that used a fixed number of treatment

Squeeze Rest

9:20:00

B

Time

Figure 17-9.  Anal manometric pressure tracings in a patient with fecal incontinence before (A) and after (B) neuromuscular training (biofeedback) while squeezing and at rest. Before neuromuscular training, the patient has a weak and poorly sustained squeeze and makes multiple ineffective attempts to squeeze. After six sessions of training, the ability to generate and sustain the squeeze has improved significantly.

253

254

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 17-4  Outcome of Neuromuscular Training (Biofeedback Therapy) and/or Exercises for Fecal Incontinence in Adults* REFERENCE

SUBJECTS (F/M)

TREATMENT

CONTROL

OUTCOME

Manometric BFB + rectal sensory training + coordination training (weekly, 4 wk) BFB + electrical stimulation (augmented) (weekly, 12 wk)

Sham training (crossover design)

Treatment improved symptoms

Vaginal manometric biofeedback

Greater symptom improvement in treated group than control group (P < 0.001) Treatment improved symptoms more than PFMT alone (77% vs. 41%; P = 0.001) NSD between groups

131

17/8

137

40/0

138

83/25

BFB + PFMT + sensory training (biweekly, 12 wk)

PFMT

142

60/0

BFB

143

49/0

BFB (weekly, 12 wk) + electrical stimulation BFB + home exercises

144

159/12

Four groups:

NA

1. Education + advice 2. As per group 1 + PFMT 3. As per group 2 + manometric

Electrical stimulation

Both groups improved; NSD between groups in symptoms and QOL ~54% improved in all groups NSD between groups in symptoms and QOL

BFB

4. As per group 3 + home BFB 145

107/13

(biweekly, 6 sessions, 3 mo) Three groups: 1. PFMT 2. PFMT + anal ultrasound BFB 3. PFMT + manometric BFB (monthly, 5 sessions)

NA

NSD between groups in symptoms, QOL, and manometry changes

*Selected randomized controlled trials. BFB, biofeedback (using electromyography probe unless otherwise specified); F, females; M, males; NA, not applicable; NSD, no significant difference; PFMT, pelvic floor muscle training; QOL, quality of life. Adapted from Norton C. Fecal incontinence and biofeedback therapy. Gastroenterol Clin North Am 2008; 37:587-604.

sessions, often less than three, showed a less favorable improvement response than those that titrated the number of sessions on the basis of the patient’s performance.133,134 In one study, periodic reinforcement with neuromuscular training at six weeks, three months, and six months was thought to confer additional benefit85 and long-term improvement.135 In the literature on fecal incontinence,136-146 the terms improvement, success, or cure have been used interchangeably, and the definition of each term has been inconsistent. In uncontrolled studies, subjective improvement has been reported in 40% to 85% of patients.2,133 Table 17-4 summarizes selected randomized controlled trials of neuromuscular training in patients with fecal incontinence.130,131,137,138,142-145 A Cochrane review of 11 randomized, controlled trials has concluded that no method of training is better than any other method.147 Whether biofeedback is superior to conservative management is also unclear. In the most recent randomized controlled trial,138 108 patients were randomized to receive either six sessions of EMG biofeedback (n = 44) or Kegel exercises (n = 64) plus supportive therapy. After treatment, 77% of patients who received biofeedback reported adequate relief of symptoms compared with 41% of those who did Kegel exercises (P < 0.001). The number of episodes of incontinence was not different between groups in an intention-to-treat analysis, but a trend toward improvement (P = 0.042) was observed in a per-protocol analysis.138 This study suggests that biofeedback is superior to Kegel exercises. The technique of neuromuscular training has not been standardized, and the use of this treatment is largely restricted to specialized centers. The manometric parameters obtained at baseline do not appear to predict the clinical

response to biofeedback treatment.148 Similarly, the patient’s age, presence of sphincter defects, or presence of neuropathy do not predict outcome.149 Therefore, criteria used for selection, motivation of the individual patient, enthusiasm of the therapist, and severity of incontinence each may affect the outcome.2,130,133,134,144 Despite the lack of a uniform approach and the inconsistencies in the reported outcomes of randomized controlled trials, neuromuscular training seems to confer benefit (see Table 17-4). Therefore, neuromuscular training should be offered to all patients with fecal incontinence who have failed supportive measures and especially to older patients, patients with comorbid illnesses, and those for whom reconstructive surgery is being considered. Severe fecal incontinence, pudendal neuropathy, and an underlying neurologic disorder are associated with a poor response to biofeedback therapy.150-152 One study has suggested that neuromuscular training may be most beneficial in patients with urge incontinence.153 Biofeedback also seems to be useful for patients who have undergone anal sphincteroplasty,154 postanal repair (see later),155 or low-anterior resection156 and children who have undergone correction of a congenital anorectal anomaly.157

Plugs, Sphincter Bulkers, and Electrical Stimulation

Disposable anal plugs have been used to help occlude the anal canal temporarily.158 Unfortunately, many patients are unable to tolerate prolonged insertion of the device.159,160 A plug may be useful for patients with impaired anal canal sensation, those with neurologic disease,161 and those who are institutionalized or immobilized. In some patients with fecal seepage, insertion of an anal plug made of cotton wool may prove beneficial162; the recommended wear

Chapter 17  Fecal Incontinence Table 17-5  Success Rates of Surgical Interventions for Fecal Incontinence PROCEDURE Current Anal sphincter repair Sacral nerve stimulation Dynamic gracilis neosphincter Artificial bowel sphincter Fecal diversion Evolving Injection of biomaterials Radiofrequency therapy (Secca procedure) Rectal augmentation

OUTCOME MEASURES

SUCCESS RATE (%)

quality OF EVIDENCE

Clinical, physiologic Complete continence Improvement in continence by ≥50% Restoration of continence Full continence NA

50-66* 40-75 75-100 42-85 50-100§ No data

Good† Very good

Cessation of leakage or improvement in continence Improvement in continence by ≥50% Avoidance of stoma

66 (short term)¶

Poor

84 64

Poor Poor

Poor‡ Good NA

*5-year success rates fall to 50%. † Derived from a Cochrane review, but in some cases data were extrapolated from only one study. ‡ Based on a systematic review of case series; no comparative studies available. § Explantation rates in case series of approximately 50%. ¶ No difference in continence scores compared with preoperative scores on long-term follow-up. NA, not available. Adapted from Gladman MA. Surgical treatment of patients with constipation and fecal incontinence. Gastroenterol Clin North Am 2008; 37:605-25, with permission.

time (although not formally tested) is up to 12 hours.130 Diapers generally are thought to be unsatisfactory for providing security or comfort, protecting the skin, or disguising odor. Many people with fecal incontinence choose not to wear a pad. Small anal dressings may be useful for people with minor soiling contained between the buttocks but can become costly if several dressings are needed each day. Bulking the anal sphincter to augment its surface area and thereby provide a better seal for the anal canal has been attempted with a variety of agents, including autologous fat,163 glutaraldehyde-treated collagen,164 and synthetic macromolecules.165 These materials usually are injected submucosally at the site where the sphincter is deficient or circumferentially if the whole muscle is degenerated or fragmented. Studies have shown definite improvement in the short term in patients with passive fecal incontinence. The experience with these techniques, however, is limited, and controlled and long-term outcome studies have not been done. Newer and better designed anal plugs are currently being tested. Electrical stimulation of striated muscle at a frequency sufficient to produce a tonic involuntary contraction (usually 30 to 50 Hz) can increase muscle strength, conduction rate of the pudendal nerve, and size of motor units, encourage neuronal sprouting, and promote local blood flow.166,167 Stimulation at lower frequencies (typically 5 to 10 Hz) can modulate autonomic function, including sensation and overactivity. Studies of electrical stimulation for fecal incontinence, however, generally have been small and uncontrolled and have been confounded by the effects of exercise, biofeedback, or other interventions. A Cochrane review of four randomized controlled trials with 260 participants concluded that electrical stimulation may have some effect.168 One study has found that anal electrical stimulation with anal biofeedback produces short-term benefits greater than those with biofeedback alone,137 whereas another study found no additional benefit to electrical stimulation over exercises and biofeedback alone.142 Also, patients have been shown to improve equally with stimulation at 1 and 35 Hz.130 Two randomized controlled trials have reported that biofeedback and electrical stimulation are equally effective.169,170 Therefore, whether electrical stimulation by itself is helpful remains unclear.

Surgical Therapy Surgery should be considered for selected patients who have failed conservative measures or biofeedback therapy. The choice of surgical procedure must be tailored to the need of the individual patient and can be described under four broad clinical categories: (1) simple structural defects of the anal sphincters; (2) weak but intact anal sphincters; (3) complex disruption of the anal sphincter complex; and (4) extrasphincteric abnormalities. Table 17-5 summarizes success rates for these surgical procedures.171 In most subjects, particularly those with obstetric trauma, overlapping sphincter repair is often sufficient. The torn ends of the sphincter muscle are plicated together and to the puborectalis muscle. Overlapping sphincter repair, as described by Parks and McPartlin,172 involves a curved incision anterior to the anal canal with mobilization of the external sphincter, which is divided at the site of the scar; the scar tissue is preserved to anchor the sutures, and overlap repair is carried out using two rows of sutures. If an internal anal sphincter defect is identified, a separate imbrication (overlapping repair) of the internal anal sphincter may be undertaken. Symptom improvement with a frequency in the range of 70% to 80% has been reported, although one study reported an improvement rate of only approximately 50%.172-176 Furthermore, some patients may experience problems with evacuation after surgery. In patients with incontinence caused by a weak but intact anal sphincter, postanal repair has been tried.177 The anorectal angle is made more acute via an intersphincteric approach, thereby improving continence. The long-term success of this approach ranges from 20% to 58%.178 In patients with severe structural damage of the anal sphincter and significant incontinence, construction of a neosphincter has been attempted using two approaches: (1) use of autologous skeletal muscle, often the gracilis and rarely the gluteus107,179; and (2) use of an artificial bowel sphincter (ABS).180 The technique of stimulated gracilis muscle transposition (dynamic graciloplasty) has been tested in many centers.181,182 This technique uses the principle that a fast-twitch, fatigable skeletal muscle, when stimulated over a long period of time, can be transformed into a slow-twitch fatigable muscle that can provide a sustained, sphincter-like muscle response. Such continuous

255

256

Section III  Symptoms, Signs, and Biopsychosocial Issues stimulation is maintained by an implanted pacemaker. When the subject has to defecate or expel gas, an external magnetic device is used to switch off the pacemaker tem­ porarily. Rates of clinical improvement with this approach have ranged from 38% to 90% (mean 67%).171 The other approach to neosphincter construction has been to implant an ABS. The ABS consists of an implanted inflatable cuffed device that is filled with fluid from an implanted balloon reservoir, which is controlled by a subcutaneous pump. The cuff is deflated to allow defecation. In one series of 24 carefully selected patients, almost 75% reported satisfactory results, although some had the device explanted.183 Both approaches (dynamic graciloplasty and ABS) require major surgery and are associated with revision rates that approach 50%. At medium-term follow up, 50% to 70% of patients have a functioning new sphincter. Several groups have reported their experiences with the ABS in small numbers of patients with overall improvement in continence in approximately 50% to 75% of patients.184,185 A randomized controlled trial has demonstrated that ABS is better than conservative treatment in improving continence.186 Longterm outcome studies, with median follow-up periods of approximately seven years, however, have documented success rates of less than 50%, explantation rates as high as 49%, and infection rates of up to 33%.131,187 Additionally, evacuation problems occur in 50% of patients. Rectal augmentation is a novel approach to correct the physiologic abnormalities in a subgroup of patients with intractable fecal incontinence secondary to reservoir or rectal sensorimotor dysfunction.188 Candidates have low rectal compliance and heightened rectal sensation (rectal hypersensitivity). The procedure involves the creation of a side-to-side ileorectal pouch, or ileorectoplasty, that involves incorporating a 10-cm patch of ileum on its vascular pedicle into the anterior rectal wall to increase rectal capacity and compliance.189 In 11 subjects, at medium-term follow up (4.5 years), rectal capacity was increased, with an associated improvement in bowel symptoms (increased ability to defer defecation and reduced frequency of episodes of incontinence) and in patients’ quality of life.190

Other Procedures

Radiofrequency energy can be delivered deep to the mucosa of the anal canal via multiple needle electrodes with use of a specially designed probe (Secca System; Rayfield Technology, Houston) inserted into the anal canals of patients with fecal incontinence.191 The proposed mechanism of action is heat-induced tissue contraction and remodeling of the anal canal and distal rectum. In one study, symptomatic improvement was sustained at two and five years after treatment.192 A multicenter trial has confirmed the improvements in continence and quality of life, at least in the short term (at six months). Complications include ulceration of the mucosa and delayed bleeding.193 Interestingly, no changes were seen in the results of anorectal manometry, PNTML measurement, or anal endosonography. Results of a randomized controlled trial of this method completed in the United States are pending. The Malone, or antegrade continent, enema procedure194 consists of fashioning a cecostomy button or appendicostomy195 to allow periodic antegrade washout of the colon. This approach may be suitable for children and for patients with neurologic disorders.195-197 If none of these techniques is suitable or all have failed, a colostomy remains a safe, although aesthetically less preferable, option for many patients.107,198-200 It is particularly suitable for patients with spinal cord injury, immobilized patients, and those with severe skin problems or other com-

plications. A colostomy should not be regarded as a failure of medical or surgical treatment.171 For many patients with fecal incontinence, the restoration of a normal quality of life and amelioration of symptoms can be rewarding. The use of a laparoscopic-assisted approach, a trephine colostomy, may help to fashion a stoma with minimal morbidity for the patient.201 In one study, the total direct costs were estimated to be $31,733 for a dynamic graciloplasty, $71,576 for a colostomy including stoma care, and $12,180 for conventional treatment of fecal incontinence.202 No controlled studies have compared surgical management with pharmacologic therapy or biofeedback therapy. Similarly, no controlled studies of the different surgical approaches have been published. Because the outcome of most procedures ranges from significant improvement initially to a less satisfactory result in the long term, no single procedure is universally accepted. In the future, a better understanding of the underlying pathophysiology and development of safer and better techniques, followed by prospective controlled trials, may allow selection of younger patients with well-defined sphincter defects for appropriate surgery.

Sacral Nerve Stimulation

Sacral nerve stimulation (SNS) has emerged as a useful treatment option in selected patients, although how SNS improves fecal incontinence remains unclear.203 The benefit may relate to direct effects peripherally on colorectal sensory or motor function or to central effects at the level of the spinal cord or brain.204 Earlier studies were performed in subjects with a morphologically intact anal sphincter, but subsequent reports have described the treatment in patients with EAS defects,205 IAS defects,206 and cauda equina syndrome207 or spinal injuries.208 The technique of SNS consists of two phases. The first phase is a temporary trial phase of two weeks during which electrodes are implanted in the second or third sacral nerve roots and the nerves are stimulated with a neurostimulator device. If the patient reports satisfactory improvement of symptoms, a permanent neurostimulator device is placed in the second phase (Fig. 17-10). Initial reports of SNS have described marked improvements in clinical symptoms and quality of life and marginal effects on physiologic parameters.176,209 The results of multicenter studies of SNS have reported marked and sustained improvement in fecal incontinence and quality of life.210-212 A randomized controlled trial has found SNS to be superior to supportive therapy (pelvic floor exercises, bulking agents, and dietary manipulation),213 but long-term outcomes are not yet available. A morphologically intact anal sphincter may not be a pre­ requisite for success with SNS, and patients with EAS defects of less than 33% can be treated effectively with this method.214 A systematic review of the published outcomes of trials of SNS has revealed that 40% to 75% of patients achieve complete continence, and 75% to 100% experience improvement, with a low (10%) frequency of adverse events.215 An evidence-based summary of current therapies for fecal incontinence is shown in Table 17-6.

TREATMENT OF SUBGROUPS OF PATIENTS Patients with Spinal Cord Injury

Patients with a spinal cord injury demonstrate delayed colonic motility or anorectal dysfunction that may manifest as incontinence, seepage, difficulty with defecation, or rectal hyposensitivity.216 Anal sphincter pressures and rectal compliance are low in these patients, but the correla-

Chapter 17  Fecal Incontinence soiling with stool, followed by the periodic administration of enemas or the use of laxatives or lavage solutions at convenient intervals.2 A cecostomy procedure also may be appropriate.217 In some patients, colostomy may be the best option.198

Patients with Fecal Seepage

Because patients with fecal seepage show dyssynergic defecation with impaired rectal sensation, neuromuscular conditioning with biofeedback techniques to improve dyssynergia can be useful (see Chapter 18).58,218 Therapy that consists of sensory conditioning and rectoanal coordination of the pelvic floor muscles to evacuate stools more completely has been shown to reduce the number of fecal seepage events substantially and to improve bowel function and anorectal function by objective measures.58

Older Patients

Figure 17-10.  Plain abdominal film showing a nerve stimulator device located in the right lower quadrant along with electrodes (radiopaque) permanently implanted into the sacral nerves. This patient presented with fecal incontinence and underwent a colonic transit study that revealed significant retention of radiopaque markers, which were located mostly in the distal colon, suggesting anorectal outlet dysfunction.

Table 17-6  Treatment Options for Fecal Incontinence* TREATMENT Pharmacologic treatment Loperamide Diphenoxylate and atropine Amitriptyline Cholestyramine Neuromuscular training (biofeedback) Surgical treatment Sphincteroplasty Dynamic graciloplasty Artificial bowel sphincter Colectomy Novel treatments Anal plugs Sphincter bulking agents Sacral nerve stimulation Radiofrequency therapy (Secca procedure)

quality OF EVIDENCE Fair Fair Poor Poor Good Fair Fair Fair Poor Poor Poor Good Poor

*Evidence-based summary.

tion between manometric findings and bowel dysfunction is poor. Studies of translumbar and transsacral MEPs have shown profound neuromuscular dysfunction affecting the entire spinoanal and spinorectal pathways.109 Patients with a spinal cord injury may have fecal incontinence because of a supraspinal lesion or lesion of the cauda equina.77,78 In the former group, the sacral neuronal reflex arc is intact, and the cough reflex is preserved. Therefore, reflex defecation is possible through digital stimulation or with suppositories. In patients with a low spinal cord or cauda equina lesion, digital stimulation may not be effective because the defecation reflex is often impaired. In these cases, management consists of antidiarrheal agents to prevent continuous

Fecal incontinence is a common problem in older adults and may be a marker of declining health and increased mortality in patients in nursing homes.60 In one study, fecal incontinence developed in 20% of nursing home residents during a 10-month period after admission, and long-lasting incontinence was associated with reduced survival.116 In one report, immobility, dementia, and the use of restraints that precluded a patient from reaching the toilet in time were the most important risk factors for the development of fecal incontinence.219 Usual mechanisms of incontinence include impaired anorectal sensation, weak anal sphincter, and weak pelvic floor muscles. Decreased mobility and lowered sensory perception are common causes of incon­ tinence.220 Many of these patients have fecal impaction and overflow.54,221 Fecal impaction, a leading cause of fecal incontinence in institutionalized older adults, results largely from a person’s inability to sense and respond to the presence of stool in the rectum. A retrospective screening of 245 permanently hospitalized geriatric patients222 has revealed that fecal impaction (55%) and laxatives (20%) are the most common causes of diarrhea and that immobility and fecal incontinence are strongly associated with fecal impaction and diarrhea. One study has shown that impaired anal sphincter function (a risk factor for fecal incontinence), decreased rectal sensation, and dyssynergia are seen in up to 75% of nursing home residents with fecal incontinence.36,223 Stool softeners, saline laxatives, and stimulant laxatives are frequently administered as prophylactic treatment to prevent constipation and impaction. In a study of institutionalized older patients, the use of a single osmotic agent with a rectal stimulant and weekly enemas to achieve complete rectal emptying reduced the frequency of fecal incontinence by 35% and the frequency of soiling by 42%.224 If fecal impaction is not relieved by laxatives and better toileting, a regimen of manual disimpaction, tap water enemas two or three times weekly, and rectal suppositories should be considered.225 In the presence of impaired sphincter function and decreased rectal sensation, however, liquid stools may be counterproductive. Similarly, neuromuscular training to improve dyssynergia in older adults, ritualizing the patient’s bowel habit, improving mobility, and cognitive training may be useful.60

Children

Incontinence is seen in 1% to 2% of otherwise healthy 7-year-old children.226 It is caused by functional fecal retention (previously described as encopresis), functional nonretentive fecal incontinence,227 congenital anomalies, developmental disability, or mental retardation.

257

258

Section III  Symptoms, Signs, and Biopsychosocial Issues In children with functional fecal retention, the bowel movements are irregular, often large, bulky, and painful. Consequently, when the child experiences an urge to defecate, he or she assumes an erect posture, holds the legs close together, and forcefully contracts the pelvic and gluteal muscles. Over time, this conscious suppression of defecation leads to excessive rectal accommodation, loss of rectal sensitivity, and loss of the normal urge to defecate. The retained stools become progressively more difficult to evacuate, thereby leading to a vicious cycle. The ultimate result is overflow incontinence, with seepage of mucus or liquid stool around an impacted fecal mass. This aberrant behavior may lead to the unconscious contraction of the external sphincter during defecation and cause dyssynergic defecation.221,228 By contrast, functional nonretentive fecal incontinence represents the repeated and inappropriate passage of stool at a place other than the toilet by a child older than four years with no evidence of fecal retention. According to criteria established by the Rome III consensus committee,227 children with functional nonretentive fecal incontinence often pass stools daily in the toilet but in addition have almost complete stool evacuations in their underwear more than once a week. They have no palpable abdominal or rectal fecal mass nor evidence of fecal retention on an abdominal x-ray, and colonic radiopaque marker studies are normal.229 The frequency of daytime and nighttime enuresis is higher (40% to 45%) in children with functional nonretentive fecal incontinence than in those with fecal retention. Children with functional nonretentive fecal incontinence have significantly more behavioral problems and more externalizing or internalizing of psychosocial problems than controls. The goals of treatment are to remove any fecal impaction, restore a normal bowel habit, including passage of soft stools without discomfort, and ensure self-toileting and passage of stools at appropriate places.229 Disimpaction is best accomplished with oral medication or enemas. High doses of polyethylene glycol 3350 (1 to 1.5 g/kg/day for three days) have been shown to be effective.230 Once disimpaction has been achieved, the treatment should focus on preventing a recurrence through dietary interventions, behavioral modification, and laxatives. Treatment of functional nonretentive fecal incontinence is based on education, a nonaccusatory approach, regular toilet use with rewards, and referral to a psychologist. Successful resolution of symptoms may require prolonged treatment and follow-up.231,232 Resolving parental conflicts and psycho­ social stressors and alleviating the fear of painful bowel movements may be critical to a successful outcome.218,233 The most common congenital anomalies are neural tube defects, such as meningomyelocele or spina bifida, and anal atresia (imperforate anus; see Chapter 96). Children with a

neural defect or malformation may benefit from behavioral therapy, including a stimulated defecation program (see earlier).234 Anal atresia is best treated by surgery, but about 20% may have unsatisfactory results.229 Surprisingly, children with anorectal malformations seem to cope well with their illness.235 Children with mental retardation or those with a developmental delay may be slow or never achieve full bowel control and require life-long supportive therapy.

ACKNOWLEDGMENT

I am most grateful for the excellent secretarial assistance of Ms. Kimberly Klein. This work was supported in part by Grant 1RO1 DK57100 from the National Institutes of Health.

KEY REFERENCES

Bharucha AE, Fletcher JG, Harper CM, et al. Relationship between symptoms and disordered continence mechanisms in women with idiopathic fecal incontinence. Gut 2005; 54:546-55. (Ref 63.) Bharucha AE, Zinsmeister AR, Locke GR, et al. Prevalence and burden of fecal incontinence: A population-based study in women. Gastroenterology 2005; 129:42-9. (Ref 7.) Borello-France D, Burgio KL, Richter HE, et al. Fecal and urinary incontinence in primiparous women. Obstet Gynecol 2006; 108:863-72. (Ref 30.) Diamant NE, Kamm MA, Wald A, et al. AGA technical review on anorectal testing techniques. Gastroenterology 1999; 116:735-60. (Ref 73.) Gladman MA. Surgical treatment of patients with constipation and fecal incontinence. Gastroenterol Clin North Am 2008; 37:605-25. (Ref 171.) Kamm MA. Obstetric damage and faecal incontinence. Lancet 1994; 344:730-3. (Ref 28.) Leung FW, Schnelle JF. Urinary and fecal incontinence in nursing home residents. Gastroenterol Clin North Am 2008; 37:697-707. (Ref 60.) Norton C, Cody JD, Hosker G. Biofeedback and/or sphincter exercises for the treatment of faecal incontinence in adults. Cochrane Database System Rev 2006; (3):CD002111. (Ref 147.) Rao SSC. Practice guidelines: Diagnosis and management of fecal incontinence. Am J Gastroenterol 2004; 99:1585-604. (Ref 2.) Rao SSC. Pathophysiology of adult fecal incontinence. Gastroenterology 2004; 126:S14-22. (Ref 13.) Read NW, Abouzekry L, Read MG, et al. Anorectal function in elderly patients with fecal impaction. Gastroenterology 1985; 89:959-66. (Ref 54.) Remes-Troche J, Rao SSC. Neurophysiological testing in anorectal disorders. Gastroenterol Hepatol 2008; 2:323-35. (Ref 22.) Savoye-Collet C, Koning E, Dacher J. Radiologic evaluation of pelvic floor disorders. Gastroenterol Clin North Am 2008; 37:553-67. (Ref 88.) Scott SM, Gladman MA. Manometric, sensorimotor, and neurophysiologic evaluation of anorectal function. Gastroenterol Clin North Am 2008; 37:511-38. (Ref 79.) Tjandra JJ, Chan MK, Yeh CH, et al. Sacral nerve stimulation is more effective than optimal medical therapy for severe fecal incontinence: A randomized, controlled study. Dis Colon Rectum 2008; 51:494-502. (Ref 213.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

18 Constipation Anthony J. Lembo and Sonal P. Ullman

CHAPTER OUTLINE Definition and Presenting Symptoms  259 Epidemiology  260 Prevalence  260 Incidence  260 Public Health Perspective  260 Risk Factors  260 Gender  260 Age  260 Ethnicity  262 Socioeconomic Class and Education Level  262 Diet and Physical Activity  262 Medication Use  262 Colonic Function  262 Luminal Contents  262 Absorption of Water and Sodium  263 Diameter and Length  263 Motor Function  263 Innervation and the Interstitial Cells of Cajal  263 Defecatory Function  264 Size and Consistency of Stool  264 Classification  264 Pathophysiology  264 Normal-Transit Constipation  264 Slow-Transit Constipation  265 Defecatory Disorders  265 Disorders of the Anorectum and Pelvic Floor  266 Rectocele  266 Descending Perineum Syndrome  266 Diminished Rectal Sensation  267 Rectal Prolapse and Solitary Rectal Ulcer Syndrome  267

Constipation affects a substantial portion of the Western population and is particularly prevalent in women, chil­ dren, and older adults. Many persons with constipation do not seek medical attention, but because constipation affects between 2% and 28% of the population, it results in over $6.9 billion in medical costs annually and is one of the most common reasons for an office visit to a physician. For most affected persons, constipation is intermittent and requires no or minimal intervention, such as fiber sup­ plements or other dietary modifications. For others, con­ stipation can be challenging to treat and have a negative impact on quality of life. In these cases, specific causes of constipation, such as systemic or structural diseases, must be excluded, although constipation most commonly results from disorders of function of the colon or rectum. An under­ standing of the pathophysiology of constipation is funda­ mental to effective management. Treatment of chronic constipation begins with lifestyle modifications, if appropriate, and therapy with fiber. Osmotic and stimulant laxatives, stool softeners, emollients,

Systemic Disorders  268 Hypothyroidism  268 Diabetes Mellitus  268 Hypercalcemia  268 Nervous System Disease  268 Loss of Conscious Control  268 Parkinson’s Disease  268 Multiple Sclerosis  268 Spinal Cord Lesions  268 Structural Disorders of the Colon, Rectum, Anus, and Pelvic Floor  269 Obstruction  269 Disorders of Smooth Muscle  269 Disorders of Enteric Nerves  269 Medications  270 Psychological Disorders  270 Depression  270 Eating Disorders  271 Denied Bowel Movements  271 Clinical Assessment  271 History  271 Physical Examination  271 Diagnostic Tests  272 Tests to Exclude Systemic Disease  272 Tests to Exclude Structural Disease of the Intestine  272 Physiologic Measurements  272 Treatment  274 General Measures  274 Specific Therapeutic Agents  276 Other Forms of Therapy  282

and enemas sometimes are required to treat refractory con­ stipation. Newer agents and nonpharmacologic approaches offer further options for the treatment of constipation.

DEFINITION AND PRESENTING SYMPTOMS The definition of constipation varies among people, and it is important to ask patients what they mean when they say “I am constipated.” Most persons are describing a percep­ tion of difficulty with bowel movements or a discomfort related to bowel movements. The most common terms used by young healthy adults to define constipation are straining (52%), hard stools (44%), and the inability to have a bowel movement (34%).1 The definition of constipation also varies among physi­ cians and other health care providers. The traditional medical definition of constipation, based on the 95% lower confidence limit for healthy adults in North America and

259

260

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 18-1  Rome III Criteria for Functional Constipation Two or more of the following six must be present*: Straining during at least 25% of defecations Lumpy or hard stools in at least 25% of defecations Sensation of incomplete evacuation for at least 25% of defecations Sensation of anorectal obstruction/blockage for at least 25% of defecations Manual maneuvers to facilitate at least 25% of defecations (e.g., digital evacuation, support of the pelvic floor) Fewer than three defecations/wk *Criteria fulfilled for the previous three months with symptom onset at least six months prior to diagnosis. In addition, loose stools should rarely be present without the use of laxatives, abdominal pain is not required, and there should be insufficient criteria for irritable bowel syndrome. These criteria may not apply when the patient is taking laxatives.

the United Kingdom,2 has been three or fewer bowel move­ ments/week. Reports of stool frequency, however, are often inaccurate and do not correlate with complaints of constipa­ tion.3 In an attempt to standardize the definition of constipa­ tion, a consensus definition was initially developed by international experts in 1992 (Rome I criteria)4 and was revised in 1999 and in 2006 (Rome II and III criteria, respec­ tively; Table 18-1).5,6 The Rome criteria incorporate the multiple symptoms of constipation, of which stool frequency is only one, and require that a minimum of two symptoms be present at least 25% of the time. Unlike the Rome I criteria, the Rome II criteria include symptoms suggestive of outlet obstruction (e.g., a sensation of anorectal blockage or obstruction and use of maneuvers to facilitate defecation). The Rome III criteria allow patients to have occasional loose stools and require that symptoms be present during the previous three months, with an onset at least six months earlier. When abdominal pain or discomfort is the predominant symptom, irritable bowel syndrome (IBS), rather than constipation, should be considered to be the diagnosis (see Chapter 118). Intermittently loose stools unrelated to laxative use also suggest a diagnosis of IBS. Although distinguishing IBS from constipation alone is important, the symptoms and pathophysiology of these entities overlap substantially.

EPIDEMIOLOGY PREVALENCE

The prevalence of constipation ranges from 2% to 28% of the population in Western countries (Table 18-2)7-17 and varies depending on the demographics of the population, definition of constipation (e.g., self-reported symptoms, fewer than three bowel movements/week, or the Rome cri­ teria), and method of questioning (e.g., postal questionnaire, interview). Some studies have attempted to identify subcat­ egories of constipation based on the symptom pattern. In general, the prevalence is highest when constipation is self-reported9 and lowest when the Rome II criteria for con­ stipation are applied. When the Rome II criteria are used to diagnose constipation, the effects of gender, race, socioeco­ nomic status, and level of education on the prevalence of constipation are reduced.10

INCIDENCE

Little is known about the incidence of constipation in the general population. Talley and colleagues studied 690 non­ elderly residents of Olmsted County, Minnesota, at baseline

and after 12 to 20 months.18 Constipation, defined as fre­ quent straining at stool and passing hard stool, a stool fre­ quency of fewer than three stools/week, or both, was present in 17% of respondents on the first survey and 15% on the second survey. The rate of new constipation in this study was 50/1000 person-years, whereas the disappearance rate was 31/1000 person-years. Robson and colleagues found that 12.5% of older persons (mean age, 83 years) entering a nursing home had constipation and that constipation devel­ oped in 7% over three months of follow-up.19

PUBLIC HEALTH PERSPECTIVE

Constipation results in more than 2.5 million physician visits, 92,000 hospitalizations, and several hundred million dollars of laxative sales/year in the United States.20 Eightyfive percent of physician visits for constipation lead to a prescription for laxatives or cathartics.21 The cost of testing alone in patients with constipation has been estimated to be $6.9 billion annually.22 Among patients with constipa­ tion seen in a tertiary referral center, the average cost of a medical evaluation was $2,252, with the greatest cost attrib­ uted to colonoscopy.23 In an analysis of physician visits for constipation in the United States between 1958 and 1986, 31% of patients who required medical attention were seen by general and family practitioners, followed by internists (20%), pediatricians (15%), surgeons (9%), and obstetricians-gynecologists (9%). Only 4% of patients were seen by gastroenterologists, sug­ gesting that few such patients were deemed to need advice from a specialist.20,21 In a National Canadian Survey, 34% of persons who reported constipation had seen a physician for their symptoms.9

RISK FACTORS Risk factors for constipation in the United States include female gender, advanced age, nonwhite ethnicity, low levels of income and education, and low level of physical activ­ ity.3,8,11,24 Other risk factors include use of certain medica­ tions and particular underlying medical disorders (see later). Diet and lifestyle also may play a role in the develop­ ment of constipation (Table 18-3).

GENDER

The prevalence of self-reported constipation is two to three times higher in women than in men,10-12,16 and infrequent bowel movements (e.g., once a week) are reported almost exclusively by women.25 In one study of 220 normal sub­ jects eating their normal diets, 17% of women, but only 1% of men, passed less than 50 g of stool daily.26 The reason for the female predominance is unknown. A reduction in levels of steroid hormones has been observed in women with severe idiopathic constipation, although the clinical signifi­ cance of this finding is dubious.27 An overexpression of progesterone receptors on colonic smooth muscle cells has been reported to down-regulate contractile G proteins and up-regulate inhibitory G proteins.28 In addition, overexpression of progesterone receptor B on colonic muscle cells, thereby making them more sensitive to physiologic concentrations of progesterone, has been pro­ posed as an explanation for severe slow-transit constipation in some women.29

AGE

The prevalence of self-reported constipation among older adults ranges from 15% to 30%, with most,7,21,24,30,31 but not

Telephone interview Mailed questionnaire Mailed questionnaire

United States8 Canada9 Spain* 10,018 1,149 349

5,430

690

328

1,897

14,407 42,375 835

15,014 563

SAMPLE SIZE

RII, FC; RII, OD SR SR, RI, RII

RI, FC; RI, D

Straining and hard stools or frequency < three/wk SR; RI, FC; RI, OD

SR SR Straining and hard stools or frequency < three/wk Stool type and frequency

SR SR

DEFINITION OF CONSTIPATION

12.5, SR; 18.3, FC; 11.0, OD 3.6, FC; 13.8, D 4.6, FC; 4.6, OD 27.2, SR; 16.7, RI; 14.9, RII 29.5, SR; 19.2, RI; 14.0, RII

15 to >45 (mean, 49) 18 to >70 18 to >65 18-65

24.1

—

— 3.5 17.4

12.8 7.3

prevalence (%)

30-64

65-93

25-69

25-74 80 30-64

12-74 Mean, 24 (65% students)

AGE range (yr)

FC: M, 17; F, 16 OD M, 6; F, 17 FC: M, 2.4 F, 4.8; D: M, 11.5; F, 16 — — SR: M, 18.4; F, 35.4; RI: M, 12.0; F, 21.0; RII: M, 8.3: F, 21.1

—

M, 0.6; F, 3.5

M, 8.06; F, 20.8 M, 1.3; F, 4.9 —

M, 7.0; F, 18.2 —

prevalence BY GENDER (%)

*From Garrigues V, Galvez C, Ortiz V, et al. Prevalence of constipation: Agreement among several criteria and evaluation of the diagnostic accuracy of qualifying symptoms and self-reported definition in a population-based survey in Spain. Am J Epidemiol 2004; 159:520-6. D, dyschezia; F, female; FC, functional constipation; M, male; OD, outlet delay; RI, Rome I criteria; RII, Rome II criteria; SR, self-report.

Mailed questionnaire

Mailed questionnaire

Face-to-face interview with questionnaire Mailed questionnaire

Face-to-face interview Questionnaire administered in person Face-to-face interview Face-to-face interview Mailed questionnaire

SURVEY METHOD

United States17

Olmsted County, Minn

16

Olmsted County, Minn15

East Bristol, UK14

United States11 United States12 Olmsted County, Minn13

United States3 Chapel Hill, NC1

LOCATION OF STUDY (REFERENCE)

Table 18-2  Population-Based Studies of the Prevalence of Constipation

Chapter 18  Constipation 261

262

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 18-3  Risk Factors for Constipation Advanced age Female gender Low level of education Low level of physical activity Low socioeconomic status Nonwhite ethnicity Use of certain medications (see Table 18-4)

all,8,9,12,17 studies showing an increase in prevalence with age. Constipation is particularly problematic in nursing home residents, among whom constipation is reported in almost half and 50% to 74% use laxatives on a daily basis.32,33 Similarly, hospitalized older patients appear to be at high risk of developing constipation. A study of patients on a geriatrics ward in the United Kingdom showed that up to 42% had a fecal impaction.34 Older adults also tend to seek medical assistance for con­ stipation more commonly than their younger counterparts. In an analysis of physician visits for constipation in the United States between 1958 and 1986, the frequency was about 1% in persons younger than 60, between 1% and 2% in those 60 to 65, and between 3% and 5% in those older than 65 years.21 Constipation in older adults is most commonly the result of excessive straining and hard stools30 rather than a decrease in stool frequency. In a community sample of 209 people ages 65 to 93 years, the main symptom used to describe constipation was the need to strain at defecation; 3% of men and 2% of women reported that their average bowel fre­ quencies were less than three/week.29 Possible causes for the increased frequency of straining in older adults include decreased food intake, reduced mobility, weakening of abdominal and pelvic wall muscles, chronic illness, psychological factors, and medications, particularly painrelieving drugs.19,32 Constipation is also common in children younger than 4 years.33 For example, in Great Britain, the frequency of a consultation for constipation in general practice was 2% to 3% for children ages 0 to 4, approximately 1% for women ages 15 to 64, 2% to 3% for both genders ages 65 to 74, and 5% to 6% for patients ages 75 years or older. Fecal retention with fecal soiling is a common cause of impaired quality of life and the need for medical attention in childhood.

ETHNICITY

In North America, constipation is reported more commonly by nonwhites than whites. In a survey of 15,014 persons, the frequency of constipation in whites was 12.2%, com­ pared with 17.3% in nonwhites.3 Both groups demonstrate similar age-specific increases in prevalence.7 In developing countries, constipation is less common among the native populations, in whom stool weights are three to four times more than the median of 106 g daily in Britain.26 In rural Africa, constipation appears to be rare.

SOCIOECONOMIC CLASS AND EDUCATION LEVEL

The prevalence of constipation is influenced by socioeco­ nomic status. In population-based surveys, subjects with a lower income status have higher rates of constipation as compared with those who have a higher income.3,6-8 In a survey of approximately 9000 Australians, men and women

of lower socioeconomic status were more likely to report constipation than those of higher socioeconomic status.35 Similarly, persons who have less education tend to have an increased prevalence of constipation as compared with those who have more education.3,8,9,11,16

DIET AND PHYSICAL ACTIVITY

Cross-sectional studies have not linked low intake of fiber with constipation,29,36 yet data suggest that increased con­ sumption of fiber decreases colonic transit time and increases stool weight and frequency.22 An analysis from the Nurses Health Study, which assessed the self-reported bowel habits of 62,036 women between the ages of 36 and 61 years, demonstrated that women who were in the highest quintile of fiber intake and who exercised daily were 68% less likely to report constipation than women who were in the lowest quintile of fiber intake and exercised less than once a week.24 Although other observational studies have supported a protective effect of physical activity on consti­ pation, results from trials designed to test this hypothesis are conflicting. In a trial designed to assess the effect of regular exercise on chronic constipation, symptoms did not improve after a four-week exercise program.37 In healthy sedentary subjects, a nine-week program of progressively increasing exercise had no consistent effect on whole-gut transit time or stool weight.38 Dehydration has been identified as a potential risk factor for constipation. Some but not all observational studies have found an association between slowed intestinal transit time and dehydration.36,39 Although patients with constipa­ tion are advised routinely to increase their intake of fluid, the benefit of increased fluid intake has not been investi­ gated thoroughly.

MEDICATION USE

Persons who use certain medications are at a substantially higher risk of constipation. In a review of 7251 patients with chronic constipation (and nonconstipated controls) from a general practice database, medications that were signifi­ cantly associated with constipation were opioids, diuretics, antidepressants, antihistamines, antispasmodics, anticon­ vulsants, and aluminum antacids (Table 18-4).40 The use of aspirin or other nonsteroidal anti-inflammatory drugs in the older population is associated with a small but significantly increased risk of constipation.14

COLONIC FUNCTION LUMINAL CONTENTS

The main contents of the colonic lumen are food residue, water and electrolytes, bacteria, and gas. Unabsorbed food entering the cecum contains carbohydrates that are resistant to digestion and absorption by the small intestine, such as starches and nonstarch polysaccharides (NSPs). Some of the unabsorbed carbohydrate serves as substrate for bacterial proliferation and fermentation, yielding short-chain fatty acids and gas (see Chapter 16). On average, bacteria repre­ sent approximately 50% of stool weight.41 In an analysis of feces from nine healthy subjects on a metabolically controlled British diet, bacteria constituted 55% of the total solids, and fiber represented approximately 17% of the stool weight.42 A meta-analysis of the effect of wheat bran on colonic function has suggested that bran increases stool weight and decreases mean colonic transit time in healthy volunteers.42 The effect of bran may be the result primarily of increased

Chapter 18  Constipation Table 18-4  Secondary Causes of Constipation Mechanical Obstruction Anal stenosis Colorectal cancer Extrinsic compression Rectocele or sigmoidocele Stricture Medications Antacids Anticholinergic agents (e.g., antiparkinsonian drugs, antipsychotics, antispasmodics, tricyclic antidepressants) Anticonvulsants (e.g., carbamazepine, phenobarbital, phenytoin) Antineoplastic agents (e.g., vinca derivatives) Calcium channel blockers (e.g., verapamil) Diuretics (e.g., furosemide) 5-Hydroxytryptamine3 antagonists (e.g., alosetron) Iron supplements Nonsteroidal anti-inflammatory drugs (e.g., ibuprofen) Mu-opioid agonists (e.g., fentanyl, loperamide, morphine) Metabolic and Endocrinologic Disorders Diabetes mellitus Heavy metal poisoning (e.g., arsenic, lead, mercury) Hypercalcemia Hyperthyroidism Hypokalemia Hypothyroidism Panhypopituitarism Pheochromocytoma Porphyria Pregnancy Neurologic and Myopathic Disorders Amyloidosis Autonomic neuropathy Chagas’ disease Dermatomyositis Intestinal pseudo-obstruction Multiple sclerosis Parkinsonism Progressive systemic sclerosis Shy-Drager syndrome Spinal cord injury Stroke

bulk within the colonic lumen; the increased bulk stimu­ lates propulsive motor activity. The particulate nature of some fibers also may stimulate the colon. For example, ingestion of coarse bran, 10 g twice daily, was shown to reduce colonic transit time by about one third, whereas ingestion of the same quantity of fine bran led to no signifi­ cant decrease.41 Similarly, ingestion of inert plastic particles similar in size to coarse bran increased fecal output by almost three times their own weight and decreased colonic transit time.43

ABSORPTION OF WATER AND SODIUM

The colon avidly absorbs sodium and water (see Chapter 99). Increased water absorption can lead to smaller, harder stools. The colon extracts most of the 1000 to 1500 mL of fluid that crosses the ileocecal valve, and leaves only 100 to 200 mL of fecal water daily. Less reabsorption of electro­ lytes and nutrients takes place in the colon than in the small intestine, and sodium-chloride exchange and short-chain fatty acid transport are the principal mechanisms for stimu­ lating water absorption. Colonic absorptive mechanisms remain intact in patients with constipation. One proposed pathophysiologic mechanism in slow-transit constipation is that the lack of peristaltic movement of contents through the colon allows more time for bacterial degradation of stool

solids and increased NaCl and water absorption, thereby decreasing both stool weight and frequency.44 The volume of stool water and quantity of stool solids seem to be reduced proportionally in constipated persons.45

DIAMETER AND LENGTH

A wide or long colon may lead to a slow colonic transit rate (see Chapter 96). Although only a small fraction of patients with constipation have megacolon or megarectum, most patients with dilatation of the colon or rectum report constipation. Colonic width can be measured on barium enema films. A width of more than 6.5 cm at the pelvic brim is abnormal and has been associated with chronic constipation.46

MOTOR FUNCTION

Colonic muscle has four main functions (see also Chapter 98): (1) delays passage of the luminal contents so as to allow time for the absorption of water; (2) mixes the contents and allows contact with the mucosa; (3) allows the colon to store feces between defecations; and (4) propels the contents toward the anus. Muscle activity is affected by sleep and wakefulness, eating, emotion, the contents of the colon, and drugs. Nervous control is partly intrinsic and partly extrin­ sic by the sympathetic nerves and the parasympathetic sacral outflow. Transit of contents along the colon takes hours or days (longer than transit in other portions of the gastrointestinal tract). In a study of 73 healthy subjects, the mean colonic transit time was 35 hours.47 In another similar study, the mean colonic transit time in healthy volunteers was 34 hours, with an upper limit of normal of 72 hours.48 Scintigraphic studies in constipated subjects have shown that overall transit of colonic contents is slow. In some patients, the rate of movement of contents is approximately normal in the ascending colon and hepatic flexure but delayed in the transverse and left colon. Other patients show slow transit in the right and left sides of the colon.49 Colonic propulsions are of two basic types, low-amplitude propagated contractions (LAPCs) and high-amplitude prop­ agated contractions (HAPCs).50 The frequency and duration of HAPCs are reduced in some patients with constipation. In one study, 14 chronically constipated patients with proved slow transit of intestinal contents and one or fewer bowel movements weekly were compared with 18 healthy subjects. Four of the patients had no peristaltic movement, whereas peristaltic movement was normal in all the healthy subjects during a 24-hour period. Peristaltic movements in other subjects with constipation were fewer in number and shorter in duration, and thus passed for a shorter distance along the colon, as compared with the findings in the healthy controls. All the healthy subjects reported abdominal dis­ comfort or an urge to defecate during peristaltic movements, and two defecated, whereas only four of the 14 subjects with constipation experienced any sensation during such move­ ments, and none defecated.51

INNERVATION AND THE INTERSTITIAL CELLS OF CAJAL

Proximal colonic motility is under the involuntary control of the enteric nervous system, whereas defecation is volun­ tary. Slow-transit constipation may be related to autonomic dysfunction.52,53 Histologic studies have shown abnormal numbers of myenteric plexus neurons involved in excit­ atory or inhibitory control of colonic motility, thereby resulting in decreased amounts of the excitatory transmitter substance P54 and increased amounts of the inhibitory trans­

263

264

Section III  Symptoms, Signs, and Biopsychosocial Issues mitters vasoactive intestinal polypeptide (VIP) or nitric oxide (NO).55 Interstitial cells of Cajal (ICCs) are the intestinal pace­ maker cells and play an important role in regulating gastro­ intestinal motility. They facilitate the conduction of electrical current and mediate neural signaling between enteric nerves and muscles. ICCs initiate slow waves throughout the gastrointestinal tract. Confocal images of ICCs in patients with slow-transit constipation show not only reduced numbers but also abnormal morphology of ICCs, with irregular surface markings and a decreased number of dendrites. In patients with slow-transit constipa­ tion, the number of ICCs has been shown to be decreased in the sigmoid colon56 or the entire colon.57,58 Pathologic examination of colectomy specimens of 14 patients with severe intractable constipation has revealed decreased numbers of ICCs and myenteric ganglion cells throughout the colon.59

DEFECATORY FUNCTION

The process of defecation in healthy persons begins with a predefecatory period, during which the frequency and amplitude of propagating sequences (three or more succes­ sive pressure waves) are increased. Stimuli such as waking and meals (gastroileal reflex, also referred to as gastrocolic reflex) can stimulate this process. This predefecatory period is blunted, and may be absent, in patients with slow-transit constipation.50 The gastroileal reflex also is diminished in persons with slow-transit constipation. Stool is often present in the rectum before the urge to defecate arises. The urge to defecate is usually experienced when stool comes into contact with receptors in the upper anal canal. When the urge to defecate is resisted, retrograde movement of stool may occur and transit time increases throughout the colon (see Chapter 98).60 Although the sitting or squatting position seems to facili­ tate defecation, the benefit of squatting has not been studied in patients with constipation. Full flexion of the hips stretches the anal canal in an anteroposterior direction and straightens the anorectal angle, thereby promoting emptying of the rectum.61 Contraction of the diaphragm and abdomi­ nal muscles raises intrapelvic pressure, and the pelvic floor relaxes simultaneously. Striated muscular activity expels rectal contents, with little contribution from colonic or rectal propulsive waves. Coordinated relaxation of the puborectalis muscle, which maintains the anorectal angle, and external anal sphincter at a time when pressure is increasing in the rectum results in expulsion of stool (Fig. 18-1). The length of the colon emptied during spontaneous def­ ecation varies but most commonly extends from the descending colon to the rectum.62 When the propulsive action of smooth muscle is normal, defecation usually requires minimal voluntary effort. If colonic and rectal waves are infrequent or absent, however, the normal urge to defecate may not occur.51

During straining

Puborectalis muscle

Anorectal angle Descent of the pelvic floor

Figure 18-1.  Physiology of defecation. Defecation requires relaxation of the puborectalis muscle with descent of the pelvic floor and straightening of the anorectal angle during straining, as well as relaxation of the internal anal sphincter. (From Lembo A, Camilleri M. Chronic constipation. N Engl J Med 2003; 349:1360-8.)

Human stools may vary in consistency from small hard lumps to liquid. The water content of stool determines con­ sistency. Rapid colonic transit of fecal residue leads to diminished water absorption and (perhaps counterintui­ tively) an increase in the bacterial content of the stool. The Bristol Stool Scale25 is used in the assessment of constipa­ tion and is regarded as the best descriptor of stool form and consistency (Fig. 18-2). Stool consistency appears to be a better predictor of whole-gut transit time than of defecation frequency or stool volume.64

CLASSIFICATION Mechanical small and large bowel obstruction, medications, and systemic illnesses can cause constipation, and these causes of secondary constipation must be excluded, espe­ cially in patients presenting with a new onset of consti­ pation (see Table 18-4). Most often, however, constipation is caused by disordered function of the colon or rectum (functional constipation). Functional constipation can be divided into three broad categories—normal-transit con­ stipation, slow-transit constipation, and defecatory or rectal evacuation disorders (Table 18-5). In a study of more than 1000 patients with functional constipation who were evaluated at the Mayo Clinic, 59% were found to have normal-transit constipation, 25% had defecatory disorders, 13% had slow-transit constipation, and 3% had a com­ bination of a defecatory disorder and slow-transit constipation.65

SIZE AND CONSISTENCY OF STOOL

In a study of normal subjects who were asked to expel single hard spheres of different sizes from the rectal ampulla, the intrarectal pressure and time needed to pass the objects varied inversely with their diameters. Small hard stools are more difficult to pass than large soft stools. When larger stimulated stools were tested, a hard stool took longer to expel than a soft silicone rubber object of approximately the same shape and volume. Similarly, more subjects were able to expel a 50-mL water-filled compressible balloon than a hard 1.8-cm sphere.63

PATHOPHYSIOLOGY NORMAL-TRANSIT CONSTIPATION

In normal-transit constipation, stool travels along the colon at a normal rate.66 Patients with normal-transit constipation may have misperceptions about their bowel frequencies and often exhibit psychosocial distress.67 Some patients have abnormalities of anorectal sensory and motor function indistinguishable from those in patients with slow-transit

Chapter 18  Constipation Table 18-5  Clinical Classification of Functional Constipation CATEGORY

FEATURES

CHARACTERISTIC FINDINGS

Normal-transit constipation

Incomplete evacuation; abdominal pain may be present but not a predominant feature Infrequent stools (e.g., ≤1/wk); lack of urge to defecate; poor response to fiber and laxatives; generalized symptoms, including malaise and fatigue; more prevalent in young women Frequent straining; incomplete evacuation; need for manual maneuvers to facilitate defecation

Normal physiologic test results

Slow-transit constipation

Defecatory disorders (pelvic floor dysfunction, anismus, descending perineum syndrome, rectal prolapse)

Whole gut transit time

Type of stool

Description

Pictorial representation

Long transit (e.g., 100 hours) Type 1

Separate hard lumps, like nuts, hard to pass

Type 2

Sausage shaped but lumpy

Type 3

Like sausage but with cracks on its surface

Type 4

Like sausage or snake, smooth and soft

Type 5

Soft blobs with clear-cut edges (passed easily)

Type 6

Fluffy pieces with ragged edges, a mushy stool

Type 7

Watery, no solid pieces

Abnormal balloon expulsion test and/or rectal manometry

similar to those seen in persons with IBS.71 In patients with more severe symptoms, the pathophysiology includes delayed emptying of the proximal colon and fewer HAPCs after meals. Colonic inertia is a term used to describe the disorder in patients with symptoms at the severe end of the spectrum. In this condition, colonic motor activity fails to increase after a meal,72 ingestion of bisacodyl,73 or adminis­ tration of a cholinesterase inhibitor such as neostigmine.74

DEFECATORY DISORDERS

Entirely liquid

Short transit (e.g., 10 hours) Figure 18-2.  Bristol Stool Form Scale. Common stool forms and their consistency in relation to whole-gut transit time are shown. (From Heaton KW, Radvan J, Cripps H, et al. Defecation frequency and timing, and stool form in the general population: A prospective study. Gut 1992; 33:818-24.)

constipation.68 Whether increased rectal compliance and reduced rectal sensation are effects of chronic constipation or contribute to the failure of the patients to experience an urge to defecate is unclear. Most patients, however, have normal physiologic testing. IBS with constipation differs from normal-transit constipation in that abdominal pain is the predominant symptom in IBS (see Chapter 118).

SLOW-TRANSIT CONSTIPATION

Retention in colon of >20% of radiopaque markers five days after ingestion

Slow-transit constipation is most common in young women and is characterized by infrequent bowel movements (less than one bowel movement/week). Associated symptoms include abdominal pain, bloating, and malaise. Symptoms are often intractable, and conservative measures such as fiber supplements and osmotic laxatives are usually ineffec­ tive.69,70 The onset of symptoms is gradual and usually occurs around the time of puberty. Slow-transit constipa­ tion arises from disordered colonic motor function. Patients who have mild delays in colonic transit have symptoms

Defecatory disorders arise from failure to empty the rectum effectively because of an inability to coordinate the ab­ dominal, rectoanal, and pelvic floor muscles. Many patients with defecatory disorders also have slow-transit constipa­ tion75 Defecatory disorders are also known as anismus, dyssynergia, pelvic floor dyssynergia, spastic pelvic floor syndrome, obstructive defecation, or outlet obstruction. These disorders appear to be acquired and may start in childhood. They may be a learned behavior to avoid some discomfort associated with the passage of large hard stools or pain associated with attempted defecation in the setting of an active anal fissure or inflamed hemorrhoids. Patients with defecatory disorders commonly have inappropriate contraction of the anal sphincter when they bear down (Fig. 18-3). This phenomenon can occur in asymptomatic subjects but is more common among patients who complain of difficult defecation.76 Some patients with a defecatory disorder are unable to raise intrarectal pressure to a level sufficient to expel stool, a disturbance that manifests clinically as failure of the pelvic floor to descend on straining.77 Defecatory disorders are particularly common in older patients with chronic constipation and excessive straining, many of whom do not respond to standard medical treat­ ment.78 Defecatory disorders rarely are associated with structural abnormalities such as rectal intussusception, an obstructing rectocele, megarectum, or excessive perineal descent.79 Patients with defecatory disorders may report infrequent bowel movements, ineffective and excessive straining, and the need for manual disimpaction; however, symptoms, particularly in the case of pelvic floor dysfunction, do not correlate with physiologic findings.80 For a diagnosis of a defecatory disorder, a Rome working group81 has specified the criteria listed in Table 18-6. In patients with this dis­ order, constipation is functional and caused by dysfunction of the pelvic floor muscles as determined by physiologic tests. Pelvic floor dyssynergia is a subset of these patients in which the anal sphincter fails to relax more than 20% of its basal resting pressure during attempted defecation, despite the presence of adequate propulsive forces in the rectum.

265

Section III  Symptoms, Signs, and Biopsychosocial Issues

EMG Pressure cm H2O

Control subject

Cough

Pressure cm H2O EMG

Strain

200 µV

Ext. sphincter

100 Anal canal

50 0 Cough (x 3)

Constipated patient

266

Strain

5 sec

200 µV

Ext. sphincter

150 100 50 0

Anal canal 5 sec

Figure 18-3.  Electromyography (EMG) and pressure tracings during defecation in a normal (control) subject and a constipated patient with a defecatory disorder. In both the control subject and constipated patient, a cough produces a rise in pressure. When a normal subject strains (upper tracing), EMG activity of the external anal sphincter is inhibited and pressure in the anal canal falls. In a constipated patient with a defecatory disorder, EMG activity of the anal sphincter is not inhibited on straining, and pressure within the anal canal increases (lower tracing). This paradoxical contraction has been termed anismus, anal dyssynergia, and spastic perineum. (From Preston DM, Lennard-Jones JE. Anismus in chronic constipation. Dig Dis Sci 1985; 30:413-8.)

Table 18-6  Rome III Criteria for Functional Defecation Disorders81* The patient must satisfy diagnostic criteria for functional constipation (see Table 18-1). During repeated attempts to defecate, the patient must have at least two of the following: Evidence of impaired evacuation, based on balloon expulsion test or imaging Inappropriate contraction of pelvic floor muscles (i.e., anal sphincter or puborectalis) or less than 20% relaxation of basal resting sphincter pressure by manometry, imaging, or EMG Inadequate propulsive forces assessed by manometry or imaging *Criteria fulfilled for the previous three months with symptom onset at least six months prior to diagnosis. EMG, electromyography.

Functional fecal retention (FFR) is the most common def­ ecatory disorder in children. It is a learned behavior that results from withholding defecation, often because of fear of a painful bowel movement.82 The symptoms are common and may result in secondary encopresis (fecal incontinence) because of leakage of liquid stool around a fecal impaction. FFR is the most common cause of encopresis in childhood (see Chapter 17).83

DISORDERS OF THE ANORECTUM AND PELVIC FLOOR RECTOCELE

A rectocele is the bulging or displacement of the rectum through a defect in the anterior rectal wall. In women, the perineal body supports the anterior rectal (posterior vaginal)

wall above the anorectal junction, and a layer of fascia runs from the rectovaginal pouch of Douglas to the perineal body and adheres to the posterior vaginal wall. The anterior rectal wall is unsupported above the level of the perineal body, and the rectovaginal septum can bulge anteriorly to form a rectocele (Fig. 18-4). Rectoceles can arise from damage to the rectovaginal septum or its supporting struc­ tures during vaginal childbirth. These injuries are exacer­ bated by repetitive increases in intra-abdominal pressure and the long-term effects of gravity. Prolapse of other pelvic organs may be present. For example, urinary incontinence, as well as a previous hysterectomy, has been reported to be more common in patients with a rectocele than in patients with difficult defecation but no demonstrable rectocele.84 Studies using defecating proctography (see later) have shown that rectoceles are common in symptomless healthy women and may protrude as much as 4 cm from the line of the anterior rectal wall without causing bowel symptoms, although 2 cm is the generally accepted lower limit of a rectocele that may be regarded as clinically significant.85 Symptomatic patients report the inability to complete fecal evacuation, perineal pain, sensation of local pressure, and appearance of a bulge at the vaginal opening on straining. Women may report the need to use their thumb or fingers to support the posterior vaginal wall to complete defeca­ tion.84 Women also may report the need to use a finger to evacuate the rectum digitally. Defecating proctography can be used to demonstrate a rectocele, measure its size, and determine whether barium becomes trapped within the rectocele. In one study, trap­ ping of barium in rectoceles changed with the degree of rectal emptying and was related to the size of the rectocele86; however, the size of the rectocele or degree of emptying on defecation has not been shown to correlate with the outcome of surgical repair.87,88 Asymptomatic women with rectoceles do not require sur­ gical treatment. Kegel exercises (designed to strengthen the pelvic floor muscles that support the urethra, bladder, uterus, and rectum) and instructions to avoid repetitive increases in intra-abdominal pressure may help prevent progression of the rectocele. Surgery should be considered only for patients in whom contrast is retained during defe­ cography and patients in whom constipation is relieved with digital vaginal pressure to facilitate defecation.89 Surgi­ cal repair can be performed by endorectal, transvaginal, or transperineal approaches. Other types of genital prolapse may also be present, and collaboration between the surgeon and gynecologist may be appropriate. In carefully selected patients surgical repair benefits approximately 75% of patients. In a review of 89 women who underwent a com­ bined transvaginal and transanal rectocele repair for symp­ toms of obstructive defecation, the repair was successful in 71% of patients, as assessed by the absence of symptoms after one year.90 Reduction in the size of the rectocele, as judged by defecating proctography, does not appear to cor­ relate clearly with improvement in symptoms.88

DESCENDING PERINEUM SYNDROME

In the descending perineum syndrome, the pelvic floor descends to a greater extent than normal (1 to 4 cm) when the patient strains during defecation, and rectal expulsion is difficult. The anorectal angle is widened as a result of pelvic floor weakness, and the rectum is more vertical than normal. The perineal body is weak (thereby facilitating for­ mation of a rectocele), and the lax muscular support favors intrarectal mucosal intussusception or rectal prolapse. The pelvic floor may not provide the resistance necessary for extrusion of solid stool through the anal canal. A common

Chapter 18  Constipation

Levator plate Rectum

A

Perineal body

Vagina

Rectovaginal septum

Rectocele

B

Figure 18-4.  Development of a rectocele. A, Normal anatomy of the female pelvis. The levator plate is almost horizontal, supporting the rectum and vagina. The perineal body provides support for the lower posterior vaginal wall; above it lies the rectovaginal septum. B, Weakness of the pelvic floor leads to a more vertical levator plate. The perineal body is attenuated, which favors the formation of a rectocele. The laxity of the pelvic floor also favors rectal mucosal prolapse. (From Loder PB, Phillips RKS. Rectocele and pelvic floor weakness. In: Kamm MA, Lennard-Jones JE, editors. Constipation. Peterfield, England: Wrightson Biomedical; 1994. p 281.)

reason for pelvic floor weakness is trauma or stretching during parturition. In some cases, repeated and prolonged defecation appears to be a damaging factor. Symptoms include constipation, incomplete rectal evacuation, exces­ sive straining and, less commonly, digital rectal evacua­ tion.91 Electrophysiologic studies show partial denervation of the striated muscle and evidence of pudendal nerve damage. Histologic examination of operative specimens of the pelvic floor muscles confirms loss of muscle fibers.

DIMINISHED RECTAL SENSATION

The urge to defecate depends in part on tension within the rectal wall (determined by the tone of the circular muscle of the rectal wall), rate and volume of rectal distention, and size of the rectum. Some patients with constipation appear to feel pain normally as the rectum is distended to the maximal tolerable volume, but they fail to experience an urge to defecate with intermediate volumes.92 In a study of women with severe idiopathic constipation, a higher than normal electrical stimulation current applied to the rectal mucosa was required to elicit pain, thereby suggesting a possible rectal sensory neuropathy.93 Rectal hyposensitivity (RH) is defined as insensitivity of the rectum to balloon distention on anorectal physiologic investigation, although the pathophysiology of RH is not entirely clear. Constipation is the most common presenting symptom of RH. In an investigation of 261 patients with RH, 38% had a history of pelvic surgery, 22% had a history of anal surgery, and 13% had a history of spinal trauma.94

RECTAL PROLAPSE AND SOLITARY RECTAL ULCER SYNDROME

Full-thickness rectal prolapse and solitary rectal ulcer syn­ drome are part of a spectrum of defects that arise from weakening of the pelvic floor. Some patients may complain of many fruitless visits to the bathroom, with prolonged straining in response to a constant desire to defecate. The patient has a sense of incomplete evacuation and may spend an hour or more daily on the toilet. The infrequent passage of small hard stools is common, as are other features of a

functional bowel disorder, such as abdominal pain and distention. Rectal prolapse refers to complete protrusion of the rectum through the anus (see Chapter 125). Occult (asymp­ tomatic) rectal prolapse has been found in 33% of patients with clinically recognized rectoceles and defecatory dys­ function.95 Rectal prolapse can be detected easily on phys­ ical examination by asking the patient to strain as if to defecate. A laparoscopic rectopexy—in which the prolapsed rectum is raised and secured with sutures to the adjacent fascia—is the recommended treatment.96 Solitary rectal ulcer syndrome is a rare disorder character­ ized by erythema or ulceration generally of the anterior rectal wall as a result of chronic straining (see Chapter 115). Mucus and blood may be passed when the patient strains during defecation.97,98 Endoscopic findings may include erythema, hyperemia, mucosal ulceration, and polypoid lesions. Misdiagnosis may occur because of the heteroge­ neous findings and misleading name of the syndrome (an ulcer need not be present). In a study of 98 patients with solitary rectal ulcer syndrome, 26% were initially diag­ nosed incorrectly. In patients with a rectal ulcer or mucosal hyperemia, the most common misdiagnoses were Crohn’s disease and ulcerative colitis. In those with a polypoid lesion, the most common misdiagnosis was a neoplastic polyp.99 Histology of full-thickness specimens of the lesion reveals extension of the muscularis mucosa between crypts and disorganization of the muscularis propria. Defecogra­ phy, transrectal ultrasonography, and anorectal manometry are helpful in the diagnosis. Varying degrees of rectal prolapse exist in association with solitary rectal ulcer syndrome. Rectal prolapse and paradoxical contraction of the puborectalis muscle can lead to rectal trauma because of the high pressures generated within the rectum. In addition, rectal mucosal blood flow is reduced.100 Medical treatment may be difficult, and a single optimal therapy does not exist. The patient should be advised to resist the urge to strain. Bulk laxatives and dietary fiber may be of some benefit.101 Surgery may be required; rectopexy is performed most commonly. Of patients who undergo

267

268

Section III  Symptoms, Signs, and Biopsychosocial Issues surgery for solitary rectal ulcer syndrome with rectal pro­ lapse, 55% to 60% report long-term satisfaction, although a colostomy is eventually required in approximately one third of patients.102 Repair of a rectal prolapse may aggravate constipation. Biofeedback appears to be a promising mode of therapy for patients with solitary rectal ulcer syndrome.103

SYSTEMIC DISORDERS HYPOTHYROIDISM

Constipation is the most common gastrointestinal com­ plaint in patients with hypothyroidism. The pathologic effects are caused by an alteration of intestinal motor func­ tion and possible infiltration of the intestine by myxedema­ tous tissue. The basic electrical rhythm that generates peristaltic waves in the duodenum decreases in hypothy­ roidism, and small bowel transit time is increased.104 Myxedema megacolon is rare but can result from myx­ edematous infiltration of the muscle layers of the colon. Symptoms include abdominal pain, flatulence, and constipation.105

DIABETES MELLITUS

The mean colonic transit time is longer in diabetics than in healthy controls. In one study, the mean total colonic transit time in 28 diabetic patients (34.9 ± 29.6 hours; mean ± SD) was significantly longer than that in 28 healthy subjects (20.4 ± 15.6 hours; P < 0.05).106 Among the 28 diabetic patients, 9 of 28 (32%) met the Rome II criteria for consti­pation and 14 of 28 (50%) had cardiovascular autonomic neuropathy. The mean colonic transit times in diabetic patients with and without cardiovascular auto­ nomic neuropathy were similar. By contrast, a previous study reported that asymptomatic diabetic patients with cardiovascular autonomic neuropathy had significantly longer whole-gut transit times (although still within the range of normal) than a control group without evidence of neuropathy.107 In another study, diabetic patients with mild constipation demonstrated delayed colonic myoelec­ trical and motor responses after ingestion of a standard meal, whereas diabetics with severe constipation had no increases in these responses after food. Neostigmine increased colonic motor activity in all diabetic patients, suggesting that the defect was neural rather than muscular (see Chapter 35).108

HYPERCALCEMIA

Constipation is a common symptom of hypercalcemia resulting from hyperparathyroidism.109 It also may be a manifestation of hypercalcemia caused by other conditions, such as sarcoidosis or malignancy involving bone (see Chapter 35).

NERVOUS SYSTEM DISEASE LOSS OF CONSCIOUS CONTROL

A decrease in or complete loss of bodily perception as a result of cerebral disability or dementia may lead to defeca­ tory failure, possibly because of inattention.

PARKINSON’S DISEASE

Constipation occurs frequently in patients with Parkinson’s disease (PD). In a study of 12 patients with PD compared with normal controls, slow colonic transit, decreased phasic

rectal contractions, weak abdominal wall muscle contrac­ tion, and paradoxical anal sphincter contraction on defeca­ tion were all features in patients with PD and frequent constipation.110 Loss of dopamine-containing neurons in the central nervous system is the underlying defect in PD; a defect in dopaminergic neurons in the enteric nervous system also may be present. Histopathologic studies of the myenteric plexuses of the ascending colon in 11 patients with PD and constipation revealed that in 9 patients, the number of dopamine-positive neurons was one tenth or less the number in control subjects. Dopamine concentrations in the muscularis externa were significantly lower in patients with PD than in controls (P < 0.01).111 Another possible contributor to constipation is the inabil­ ity of some patients with PD to relax the striated muscles of the pelvic floor on defecation. This finding is a local mani­ festation of the extrapyramidal motor disorder that affects skeletal muscle. Preliminary observations suggest that injec­ tion of botulinum toxin into the puborectalis muscle is a potential therapy for this type of outlet dysfunction consti­ pation in patients with PD.112,113

MULTIPLE SCLEROSIS

Constipation is common among patients with multiple sclerosis (MS). In an unselected group of 280 patients with MS, the frequency of constipation (defined as diminished bowel frequency, digitation to facilitate defecation, or the use of laxatives) was approximately 43%. Almost 25% of the subjects passed fewer than three stools/week, and 18% used a laxative more than once a week. Constipation cor­ related with the duration of MS but preceded the diagnosis of MS in 45% of subjects. Constipation did not correlate with immobility or the use of medications.114 In another questionnaire study of 221 patients with MS, the frequency of constipation was as high as 54%.115 Constipation in patients with MS can be multifactorial and related to a reduction in postprandial colonic motor activity, limited physical activity, and medications with constipating side effects. Patients with advanced MS and constipation have evi­ dence of a visceral neuropathy. In a group of patients with advanced MS and severe constipation, all had evidence of disease in the lumbosacral spinal cord and decreased com­ pliance of the colon. Motor and electrophysiologic measure­ ments have shown that the usual increase in colonic motor activity after meals is absent. Among less severely affected patients, slow colonic transit and manometric evidence of pelvic floor muscular and anal sphincter dysfunction have been demonstrated. Patients may have fecal inconti­ nence.116,117 Therapy with biofeedback has been reported to relieve constipation and fecal incontinence, although in a study of 13 patients with MS who underwent biofeedback for either constipation or incontinence, only 38% improved (see Chapter 17).118

SPINAL CORD LESIONS Lesions Above the Sacral Segments

Spinal cord lesions or injury above the sacral segments lead to an upper motor neuron disorder, with severe constipa­ tion. The resulting delay in colonic transit affects the recto­ sigmoid colon primarily.119,120 In a study of patients with severe thoracic spinal cord injury, colonic compliance was abnormal, with a rapid rise in colonic pressure on instilla­ tion of relatively small volumes of fluid. Motor activity after meals did not increase, but the colonic response to neostigmine was normal, thereby suggesting absence of myopathy.

Chapter 18  Constipation Studies of anorectal function in patients with severe trau­ matic spinal cord injury have shown that rectal sensation to distention is abolished, although a dull pelvic sensation is experienced by some patients at maximum levels of rectal balloon distention. Anal relaxation on rectal disten­ tion is exaggerated and occurs at a lower balloon volume than in normal subjects. Distention of the rectum leads to a linear increase in rectal pressure, without the plateau at inter­mediate values seen in normal subjects, and ends in high-pressure rectal contractions after a relatively small volume (100 mL) has been instilled into the balloon. As expected, the rectal pressure generated by straining is lower in patients than in control subjects and is less with higher than lower spinal cord lesions. Patients demonstrate a loss of conscious external anal sphincter control, and the sphincter does not relax on straining, suggesting that in normal subjects, descending inhibitory pathways are present.121 These findings explain why some patients with spinal cord lesions experience not only constipation, but also sudden uncontrollable rectal expulsion with incon­ tinence. Other patients cannot empty the rectum in response to laxatives or enemas, possibly because of failure of the external anal sphincter to relax, and they may require manual evacuation. Electrical stimulation of anterior sacral nerve roots S2, S3, and S4 via electrodes implanted for urinary control in para­ plegic patients leads to a rise in pressure within the sigmoid colon and rectum and contraction of the external anal sphincter. Contraction of the rectum and relaxation of the internal anal sphincter persist for a short time after the stimulus ceases. By appropriate adjustment of the stimulus in one study, it was possible for 5 of 12 paraplegic patients to evacuate feces completely and for most of the others to increase the frequency of defecation and reduce the time spent emptying the rectum.122 In another series, left-sided colonic transit time decreased with regular sacral nerve stimulation.123

Lesions of the Sacral Cord, Conus Medullaris, Cauda Equina, and Nervi Erigentes (S2 to S4)

Neural integration of anal sphincter control and rectosig­ moid propulsion occurs in the sacral segments of the spinal cord. The motor neurons that supply the striated sphincter muscles are grouped in Onuf’s nucleus at the level of S2. There is evidence that efferent parasympathetic nerves that arise in the sacral segments enter the colon at the region of the rectosigmoid junction and extend distally in the inter­ muscular plane to reach the level of the internal anal sphincter and proximally to the midcolon via the ascending colonic nerves, which retain the structure of peripheral nerves (see Chapter 98).124 Damage to sacral segments of the spinal cord or to efferent nerves leads to severe constipation. Fluoroscopic studies show a loss of progression of contractions in the left colon. When the colon is filled with fluid, the intraluminal pres­ sure generated is lower than normal, in contrast with the situation after higher lesions of the spinal cord. The distal colon and rectum may dilate, and feces may accumulate in the distal colon. Spasticity of the anal canal can occur. Loss of sensation of the perineal skin may extend to the anal canal, and rectal sensation may be diminished. Rectal wall tone depends on the level of the spinal lesion. In a study of 25 patients with spinal cord injury, rectal tone was signifi­ cantly higher than normal in patients with acute and chronic supraconal lesions but significantly lower than normal in patients with acute and chronic conal or cauda equina lesions.125

STRUCTURAL DISORDERS OF THE COLON, RECTUM, ANUS, AND PELVIC FLOOR OBSTRUCTION

Anal atresia in infancy, anal stenosis later in life, or obstruction of the colon may manifest as constipation. Obstruction of the small intestine generally manifests as abdominal pain and distention, but constipation and inability to pass flatus also may be features (see Chapters 96 and 119).

DISORDERS OF SMOOTH MUSCLE Myopathy Affecting Colonic Muscle

Congenital or acquired myopathy of the colon usually mani­ fests as pseudo-obstruction. The colon is hypotonic and inert (see Chapter 120).

Hereditary Internal Anal Sphincter Myopathy

Hereditary internal anal sphincter myopathy is a rare condi­ tion characterized by constipation with difficulty in rectal expulsion and episodes of severe proctalgia fugax, defined as the sudden onset of brief episodes of pain in the anorectal region.126-128 Three affected families have been reported. The mode of inheritance appears to be autosomal dominant with incomplete penetrance. In symptomatic persons, the inter­ nal anal sphincter muscle is thickened, and resting anal pressure is increased greatly. In two of the described patients, treatment with a calcium channel blocker improved pain but had no effect on constipation. In another family, two patients were treated by internal anal sphincter strip myectomy; one showed marked improvement and one had improvement in the constipation but only slight improve­ ment in the pain. Examination of the muscle strips showed myopathic changes with polyglucosan bodies (glucose poly­ mers) in the smooth muscle fibers and increased endomysial fibrosis.

Progressive Systemic Sclerosis

Progressive systemic sclerosis (scleroderma) may lead to constipation. In patients with progressive systemic sclerosis and constipation, 9 of 10 had no increase in colonic motor activity after ingestion of a 1000-kcal meal. Histologic examination of colonic specimens from these subjects revealed smooth muscle atrophy of the colonic wall (see Chapter 35).129

Muscular Dystrophies

Muscular dystrophies usually are regarded as disorders of striated muscle, but visceral smooth muscle also may be abnormal. In myotonic muscular dystrophy, a condition in which skeletal muscle fails to relax normally, megacolon may be found, and abnormal function of the anal sphincter is demonstrable.130 Cases associated with intestinal pseudoobstruction have been reported (see Chapter 120).131

DISORDERS OF ENTERIC NERVES Congenital Aganglionosis or Hypoganglionosis

Congenital absence or reduction in the number of ganglia in the colon leads to functional colonic obstruction with proximal dilatation, as seen in Hirschsprung’s disease and related conditions (see Chapter 96). In Hirschsprung’s disease, ganglion cells in the distal colon are absent because of an arrest in the caudal migration of neural crest cells in the intestine during embryonic development. Although most patients present during early childhood, often with

269

270

Section III  Symptoms, Signs, and Biopsychosocial Issues delayed passage of meconium, some patients with a rela­ tively short segment of involved colon present later in life.132 Typically, the colon narrows at the area that lacks ganglion cells, and the bowel proximal to the narrowing is usually dilated. Two genetic defects have been identified in patients with Hirschsprung’s disease—a mutation in the RET (rearranged during transfection) proto-oncogene, which is involved in the development of neural crest cells, and a mutation in the gene that encodes the en­ dothelin B receptor, which affects intracellular calcium levels.133,134 Hypoganglionosis is reported when small, sparse myen­ teric ganglia are seen. Neuronal counts can be made on full-thickness tissue specimens and compared with pub­ lished reference values obtained from autopsy material. Establishing the diagnosis of hypoganglionosis is not easy, because of variations in the normal density of neurons.135 Quantitative declines in the number of neurons in the enteric nervous system also are seen in patients with severe slow-transit constipation and characterized morphologi­ cally as oligoneuronal hypoganglionosis.136

Congenital Hyperganglionosis (Intestinal Neuronal Dysplasia)

Congenital hyperganglionosis, or intestinal neuronal dys­ plasia, is a developmental defect characterized by hyperpla­ sia of the submucosal nerve plexus. Clinical manifestations of the disease are similar to those seen in Hirschsprung’s disease and include young age of onset and symptoms of intestinal obstruction (see Chapter 96). In contrast to func­ tional constipation, affected children do not have symptoms of soiling or evidence of a fecaloma.137 A multicenter study of interobserver variation in the histologic interpretation of findings in children with constipation caused by abnormali­ ties of the enteric nervous system has shown complete agreement in the diagnosis of Hirschsprung’s disease but accord in only 14% of children with colonic motility disor­ ders other than aganglionosis. Some of the clinical features and histologic changes previously associated with congeni­ tal hyperganglionosis may be age-related and evolve to normal as children age.135 A diagnosis of congenital hyper­ ganglionosis can be made on the basis of hyperganglionosis of the submucous plexus with giant ganglia and at least one of the following features in rectal biopsy specimens: (1) ectopic ganglia; (2) increased acetylcholinesterase (AChE) activity in the lamina propria; and (3) increased AChE nerve fibers around the submucosal blood vessels. Most patients with congenital hyperganglionosis respond to conservative treatment, including laxatives. Internal anal sphincter myectomy may be performed if conservative management fails.138

Acquired Neuropathies

Chagas’ disease, which results from infection with Trypanosoma cruzi, is the only known infectious neuropathy. The reason for neuronal degeneration in this disorder is unclear but may have an immune basis.139 Patients present with progressively worsening symptoms of constipation and abdominal distention resulting from a segmental mega­ colon that may be complicated by sigmoid volvulus (see Chapter 109). Paraneoplastic visceral neuropathy may be associated with malignant tumors outside the gastrointestinal tract, particularly small cell carcinoma of the lung and carcinoid tumors. Pathologic examination of the affected intestine reveals neuronal degeneration or myenteric plexus inflam­

mation.140 An antibody against a component of myenteric neurons has been identified in some patients with this disorder (see Chapter 120).141 Disruption of the ICCs has been associated with a case of small cell lung carcinoma– related paraneoplastic colonic motility disorder.142

Neuropathies of Unknown Cause

Severe acute neuropathies that present mainly with obstruc­ tive symptoms and not principally with constipation have been described. As noted earlier, neuropathic features affecting the colon may occur in some patients with severe idiopathic constipation.

MEDICATIONS Constipation may be a side effect of a drug or preparation taken long term. Drugs commonly implicated are listed in Table 18-4. Common offenders include opioids used for chronic pain, anticholinergic agents including antispas­ modics, calcium supplements, some tricyclic antidepres­ sants, phenothiazines used as long-term neuroleptics, and antimuscarinic drugs used for parkinsonism.

PSYCHOLOGICAL DISORDERS Constipation may be a symptom of a psychiatric disorder or a side effect of its treatment (see Chapter 21). Healthy men who are socially outgoing, energetic, and optimis­ tic—and not anxious—and who described themselves in more favorable terms than others have heavier stools than men without these personality characteristics.143 Psy­ chological factors associated with a prolonged colonic transit time in constipated patients include a highly depressed mood state and frequent control of anger.144 In one study, women with constipation had higher soma­ tization and anxiety scores than healthy controls, and the psycholo­gical scores correlated inversely with rectal mucosal blood flow (used as an index of innervation of the distal colon).145 In a study that assessed psycholo­ gical characteristics of older persons with constipation, a delayed colonic transit time was related significantly to symptoms of somatization, obsessive-compulsiveness, depression, and anxiety.36 In a study of 28 consecutive female patients undergoing psychological assessment for intractable constipation, 60% had evidence of a current affective disorder. One third reported distorted attitudes toward food. Patients with slow-transit constipation reported more psychosocial distress on rating scales than those with normal-transit constipation.146

DEPRESSION

For some patients, constipation can be a somatic manifesta­ tion of an affective disorder. In a study of patients with depression, 27% said that constipation developed or became worse at the onset of the depression.147 Constipation can occur in the absence of other typical features of severe depression, such as anorexia or psychomotor retardation with physical inactivity. Psychological factors are likely to influence intestinal function via autonomic efferent neural pathways.145 In an analysis of 4 million discharge records of U.S. military veterans, major depression was associated with constipation, and schizophrenia was associated with both constipation and megacolon.148

Chapter 18  Constipation EATING DISORDERS

Patients with anorexia nervosa or bulimia often complain of constipation, and a prolonged whole-gut transit time has been demonstrated in patients with these disorders.149 Colonic transit time returns to normal in most patients with anorexia nervosa once they are consuming a balanced diet and gaining weight for at least three weeks.150 Pelvic floor dysfunction is found in some patients with an eating disorder and does not improve with weight gain and a balanced diet.151 Anorexia nervosa should be considered as a possible diag­ nosis in a young underweight woman who presents with constipation. Patients with an eating disorder often resort to the regular use of laxatives as treatment for constipation or to facilitate weight loss or relieve the presumed conse­ quences of binge eating. Treatment of such patients is directed at the underlying eating disorder (see Chapter 8).

DENIED BOWEL MOVEMENTS

Patients may deny or fail to report defecation when solid inert markers have been demonstrated to disappear from the abdomen by radiologic examination, proving that elimina­ tion has occurred. Such patients who deny that defecation has occurred despite evidence to the contrary need skilled psychiatric help.

CLINICAL ASSESSMENT HISTORY

It is important to determine exactly what the patient means when he or she reports constipation. A detailed history that includes the duration of symptoms, frequency of bowel movements, and associated symptoms such as abdominal discomfort and distention should be obtained. The history should include an assessment of stool consistency, stool size, and degree of straining during defecation. The pres­ ence of warning symptoms or signs, such as unintentional weight loss, rectal bleeding, change in the caliber of the stool, severe abdominal pain, and family history of colon cancer, should be elicited. A long duration of symptoms that have been refractory to conservative measures is suggestive of a functional colorectal disorder. By contrast, the new onset of constipation may indicate a structural disease. Physicians should always evaluate the patient for a structural disease in this situation. A dietary history should be obtained. The amount of daily fiber and fluid consumed should be assessed. Many patients tend to skip breakfast,152 and this practice may exacerbate constipation, because the postprandial increase in colonic motility is greatest after breakfast.72,153,154 Although caffein­ ated coffee (150 mg of caffeine) stimulates colonic motility, the ingestion of a meal has a greater effect.155 A patient’s past medical history must be reviewed. Obstet­ ric and surgical histories are particularly important. Neuro­ logic disorders also may explain some cases of constipation. A carefully taken drug history, including the use of overthe-counter laxatives and herbal medications, and their fre­ quency of intake, is important. A detailed social history may provide useful information as to why the patient has sought help for constipation at this point in time; potentially relevant behavioral back­ ground information also may be obtained. In patients with IBS, the frequency of a history of sexual abuse is increased as compared with healthy controls.156 In a survey of 120 patients with dyssynergia, 22% reported a history of sexual abuse, and 32% reported a history of physical abuse. Bowel

dysfunction adversely affected sexual life in 56% and social life in 76% of patients.157 The physician should be alert to manifestations of depression, such as insomnia, lack of energy, loss of interest in life, loss of confidence, and a sense of hopelessness. A history of physical or sexual abuse may not emerge during the initial visit, but if the physician evinces no surprise at whatever is revealed, indicates that distressing events are common in patients with intestinal symptoms, and maintains a sensitive, encouraging attitude, the full story often gradually emerges during subsequent visits, provided that there is privacy, confidentiality, and adequate time (see Chapters 21 and 118).

PHYSICAL EXAMINATION

The patient’s general appearance or voice may point to a clinical diagnosis of hypothyroidism, parkinsonism, or depression. The general physical examination should exclude major central nervous system disorders, especially spinal lesions. If spinal disease is suspected, the sacral der­ matomes should be examined for loss of sensation. The abdomen should be examined for distention, hard feces in a palpable colon, or an inflammatory or neoplastic mass. If the abdomen appears distended, a hand should be passed under the lumbar spine while the patient is lying supine to exclude anterior arching of the lumbar spine as a cause of postural bloating. The rectal examination is paramount in evaluating a patient with constipation. Placing the patient in the left lateral position is most convenient for performing a thor­ ough rectal examination. Painful perianal conditions and rectal mucosal disease should be excluded, and defecatory function should be evaluated. The perineum should be observed both at rest and after the patient strains as if to have a bowel movement. Normally, the perineum descends between 1 and 4 cm during straining. Descent of the perineum with the patient in the left lateral position below the plane of the ischial tuberosities (i.e., >4 cm) usually suggests excessive perineal descent. A lack of descent may indicate the inability to relax the pelvic floor muscles during defecation, whereas excessive perineal descent may indi­ cate descending perineum syndrome. Patients with descend­ ing perineum syndrome strain excessively and achieve only incomplete evacuation because of lack of straightening of the anorectal angle. Excessive laxity or descent of the perineum usually results from previous childbirth or exces­ sive straining. Eventually, excessive descent of the perineum may result in injury to the sacral nerves from stretching, a reduction in rectal sensation, and ultimately incontinence resulting from denervation.91 Rectal prolapse may be detected when the patient is asked to strain. The perianal area should be examined for scars, fistulas, fissures, and external hemorrhoids. A digital rectal exami­ nation should be performed to evaluate the patient for the presence of a fecal impaction, anal stricture, and rectal mass. A patulous anal sphincter may suggest prior trauma to the anal sphincter or a neurologic disorder that impairs sphincter function. Other important functions that should be assessed during the digital examination are summarized in Table 18-7. Specifically, the inability to insert the exam­ ining finger into the anal canal may suggest an elevated anal sphincter pressure, and tenderness on palpation of the pelvic floor as it traverses the posterior aspect of the rectum may suggest pelvic floor spasm. The degree of descent of the perineum during attempts to strain and expel the exam­ ining finger provides another way of assessing the degree of perineal descent. A thorough history and physical examina­ tion can exclude most secondary causes of constipation (see Table 18-4).

271

272

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 18-7  Clinical Clues to an Evacuation Disorder History Prolonged straining to expel stool Assumption of unusual postures on toilet to facilitate stool expulsion Support of perineum, digitation of rectum, or application of pressure to the posterior vaginal wall to facilitate rectal emptying Inability to expel enema fluid Constipation after subtotal colectomy for constipation Rectal Examination (with patient in left lateral position) Inspection Anus “pulled” forward during attempts to simulate strain during defecation Anal verge descends 4 cm (or beyond ischial tuberosities) during attempts to simulate straining at defecation Perineum balloons down during straining; rectal mucosa partially prolapses through anal canal Palpation High anal sphincter tone at rest precludes easy entry of examining finger (in absence of a painful perianal condition such as an anal fissure) Anal sphincter pressure during voluntary squeeze only minimally higher than anal tone at rest Perineum and examining finger descend 4 cm during simulated straining at defecation Puborectalis muscle tender to palpation through rectal wall posteriorly, or palpation reproduces pain Palpable mucosal prolapse during straining “Defect” in anterior wall of the rectum, suggestive of rectocele Anorectal Manometry and Balloon Expulsion (with patient in left lateral position) Average resting anal sphincter tone >80 cm water (>59 mm Hg) Average anal sphincter squeeze pressure >240 cm water (>177 mm Hg) Failure of balloon expulsion from rectum despite addition of 200-g weight to the balloon

DIAGNOSTIC TESTS The high prevalence of bowel symptoms in the population implies that the symptoms are only a nuisance for most people and do not signify serious disease. Therefore, an investigation is not necessary for most patients who com­ plain of one or more of these symptoms, especially adoles­ cents and young adults. Investigations may be indicated for one of two reasons: (1) to exclude a systemic illness or structural disorder of the gastrointestinal tract as a cause of constipation; or (2) to elucidate the underlying pathophysi­ ologic process when the symptoms are unresponsive to simple treatment.

TESTS TO EXCLUDE SYSTEMIC DISEASE

Determination of the hemoglobin level, erythrocyte sedi­ mentation rate, and biochemical screening test levels, including thyroid function, serum calcium, glucose, and other appropriate investigations, are indicated if the clinical picture suggests that the symptoms may result from an inflammatory, neoplastic, metabolic, or other systemic disorder.

TESTS TO EXCLUDE STRUCTURAL DISEASE OF THE INTESTINE

A barium enema study reveals the width and length of the colon and excludes an obstructing lesion severe enough to cause constipation. When fecal impaction is present, a limited enema study with a water-soluble contrast agent

outlines the colon and fecal mass without aggravating the condition. A barium examination of the small bowel is indicated only if obstruction or pseudo-obstruction involv­ ing the small bowel is suspected (see Chapters 119 and 120). Endoscopy allows direct visualization of the colonic mucosa. The yield of colonoscopy in the absence of “alarm” symptoms in patients with chronic constipation is low and is comparable with that for asymptomatic patients who undergo colonoscopy for colon cancer screening. Therefore, a colonoscopy is recommended only when there has been a recent change in bowel habits, blood in stools, or other alarming symptoms (e.g., weight loss, fever).158,159 All adults older than 50 years who present with constipation should undergo a colonoscopy, flexible sigmoidoscopy and barium enema, or computed tomographic colonography to screen for colorectal cancer, as widely recommended. A flexible sigmoidoscopy is probably sufficient for the evaluation of constipation in patients younger than 50 years without alarm symptoms (e.g., weight loss, recent onset of severe constipation, rectal bleeding) or a family history of colon cancer.

PHYSIOLOGIC MEASUREMENTS

Physiologic testing is unnecessary for most patients with constipation and is reserved for patients with refractory symptoms who do not have an identifiable secondary cause of constipation or in whom a trial of a high-fiber diet and laxatives has not been effective. An American Gastroentero­ logical Association Technical Review on Anorectal Testing Techniques160 has recommended the following investiga­ tions in patients with refractory constipation: symptom diaries to establish a diagnosis of constipation and monitor the efficacy of treatment; colonic transit study to confirm the patient’s complaint of constipation and assess colonic motility for slow transit and regional delay; anorectal manometry to exclude Hirschsprung’s disease and to com­ plement other tests of pelvic floor dysfunction; and surface electromyography (EMG) to evaluate anal sphincter func­ tion and facilitate biofeedback training. Tests of possible value include the following: defecation proctography to document the patient’s inability to defecate; balloon expul­ sion test to document the inability to defecate; and rectal sensory testing to help distinguish functional from neuro­ logic disorders as a cause of constipation.

Measurement of Colonic Transit Time

Radiopaque Markers The normal colonic transit time is less than 72 hours. Measurement of colonic transit time is performed only when objective evidence of slow transit is needed to confirm a patient’s history or as a prelude to surgical treatment. Colonic transit time is measured by performing abdominal radiography 120 hours after the patient has ingested radi­ opaque markers in a gelatin capsule (Fig. 18-5). Before the study, patients should be maintained on a high-fiber diet and should avoid laxatives, enemas, or medications that may affect bowel function. Retention of more than 20% of the markers at 120 hours is indicative of prolonged colonic transit. Because the markers are eliminated only with def­ ecation, the process of measuring colonic transit is discon­ tinuous, and the result of a transit measurement should be regarded with caution, taking recent defecation into account. If the markers are retained exclusively in the sigmoid colon and rectum, the patient may have a defecatory disorder. The presence of markers throughout the colon, however, does not exclude the possibility of a defecatory disorder because delayed colonic transit can result from a defecatory disor­ der. Measurements of transit through different segments of

Chapter 18  Constipation during childbirth or an episiotomy.161 Paradoxical anal sphincter contraction is common in patients with a recto­ cele, suggesting that straining and attempts at emptying against a contracted pelvic floor may facilitate development of a rectocele. The limitations of defecography include vari­ ability among radiologists in interpreting studies, inhibition of normal rectal emptying because of patient embarrass­ ment, and differences in texture between barium paste and stool. Confirmatory studies are needed before a decision about management can be made on the basis of the radio­ graphic findings alone. Importantly, identified anatomic abnormalities are not always functionally relevant. For example, a rectocele is only relevant if it fills preferentially (i.e., instead of the rectal ampulla) and fails to empty after simulated defecation. Magnetic resonance defecography may offer advantages over standard barium defecography162 but is not yet widely available.

Figure 18-5.  Colonic transit study; abdominal film. This constipated patient had ingested 20 inert ring markers 120 hours previously and 20 cube-shaped markers 72 hours previously. Most of the markers are still present, indicating slow whole-gut transit.

the colon are of doubtful value in planning treatment, except for megarectum, in which all the markers move rapidly to the rectum and are retained there. Wireless Motility Capsule The wireless pH and pressure recording capsule (SmartPill, Buffalo, NY) is a novel ambulatory technique for assessing colonic transit without radiation. Colonic transit measure­ ments with the wireless capsule technique have been shown to correlate well with the radiopaque marker test and also allow an assessment of gastric and small bowel transit.161 If surgical treatment for severe constipation is being con­ sidered (see later), studies of gastric emptying, small bowel transit, and segmental colonic transit times are valuable for confirming slow transit and correlating abnormalities with therapeutic outcome. In particular, scintigraphic studies of gastrointestinal transit are indicated.50 Generally, abnormal gastric or small bowel motility precludes surgical treatment of constipation.

Tests to Assess the Physiology of Defecation

Defecography Defecography is performed by instilling thickened barium into the rectum. With the patient sitting on a radiolucent commode, films or videos are taken during fluoroscopy with the patient resting, deferring defecation, and straining to defecate. This procedure evaluates the rate and comple­ teness of rectal emptying, anorectal angle, and amount of perineal descent. In addition, defecography can identify structural abnormalities, such as a large rectocele, internal mucosal prolapse, or intussusception. A rectocele repre­ sents a herniation, usually of the anterior rectal wall into the lumen of the vagina, and usually results from trauma

Balloon Expulsion Test When the rectum is distended with a balloon, the internal anal sphincter relaxes. The inability to evacuate a 50- to 60-mL inflated balloon in the rectum163 while sitting on the toilet for two minutes, with the addition of 200 g of weight to the end of the balloon,93 suggests a defecatory disorder. The balloon expulsion test is an effective and useful screen­ ing tool for identifying patients with a defecatory disorder who do not have pelvic floor dyssynergia. In one study of 359 patients with constipation, the balloon expulsion test was abnormal in 21 of 24 patients with pelvic floor dys­ synergia and an additional 12 of 106 patients without pelvic floor dyssynergia. (The diagnosis of pelvic floor dyssynergia was confirmed by manometric and defecographic findings according to the Rome II criteria.164) Anorectal Manometry Anorectal manometry can provide useful information about patients with severe constipation by assessing the resting and maximum squeeze pressure of the anal sphincters, pres­ ence or absence of relaxation of the anal sphincter during balloon distention of the rectum (rectoanal inhibitory reflex), rectal sensation, and ability of the anal sphincter to relax during straining.93,158,165 Patients with a defecatory dis­ order commonly have inappropriate contraction of the anal sphincter when they bear down. The absence of the recto­ anal inhibitory reflex raises the possibility of Hirschsprung’s disease. A high resting anal pressure suggests the presence of an anal fissure or anismus, the paradoxical contraction of the external anal sphincter in response to straining or pressure within the anal canal. Rectal hyposensitivity suggests a neurologic disorder; however, the volume of rectal content needed to induce rectal urgency also may be increased in patients with fecal retention, and the results of rectal sensitivity testing need to be interpreted with caution. Electromyographic Testing of Striated Muscle Activity EMG studies of the external anal sphincter and puborectalis muscles using concentric needle or surface electrode record­ ings generally are not essential and are rarely indicated. An exception is the use of EMG in patients with suspected spinal cord or cauda equina lesions, in whom bilateral or unilateral dysfunction of the external anal sphincter can be demonstrated. Rectal Sensitivity and Sensation Testing Rectal sensitivity to distention can be measured by intro­ ducing successive volumes of air into a rectal balloon and recording the volume at which the stimulus is first per­

273

274

Section III  Symptoms, Signs, and Biopsychosocial Issues ceived, the volume that produces an urge to defecate, and the volume above which further addition of air can no longer be tolerated owing to discomfort. These measure­ ments are not of value in the routine investigation of con­ stipation but are of research interest. The threshold current needed to elicit sensation when the rectal mucosa is stimu­ lated electrically by a current passed between bipolar elec­ trodes can be used as a test of sensory nerve function, but the test has not been established for general use.93

TREATMENT Initial treatment of constipation is based on nonpharmaco­ logic interventions. If these measures fail, then pharmaco­ logic agents may be used. Figure 18-6 provides an algorithm for the evaluation and treatment of patients with severe constipation. If a defecatory disorder is present, initial treat­ ment should include biofeedback; up to 75% of patients with disordered evacuation respond to biofeedback, and many do not respond well to fiber supplementation or oral laxatives. Otherwise, the initial treatment should include increased fluid, exercise, and intake of fiber, either through changes in diet or use of commercial fiber supplements.

Treat secondary causes of constipation (see Table 18-4)

Patients who do not improve with fiber should be given an osmotic laxative, such as milk of magnesia or polyethy­ lene glycol. The dose of the osmotic laxative should be adjusted until soft stools are attained. Stimulant agents, such as bisacodyl or senna derivatives, should be reserved for patients who do not respond to fiber or osmotic laxatives.

GENERAL MEASURES Reassurance

Some people are raised in childhood to believe that a daily bowel movement is essential for health or derive this opinion from advertisements, and they worry if their bowel habit is irregular or less frequent. They can be helped by being told that an irregular bowel habit and other defecatory symptoms are common in the healthy general population and that their symptoms are not harmful. Such reassurance may be all that they need. Patients who are concerned that their symptoms may indicate disease may be helped by appropriate investigation to relieve their fears.

Lifestyle Changes

The need to set aside an unhurried and, if possible, regular time for defecation and always to respond to a defecatory

History and physical examination

Medication history

Stop/change medication(s)

Supplement diet with 20 g fiber/d No response Colonic transit study

Normal transit

Slow transit

Fiber (>20 g/d), osmotic laxative, stimulant laxative

Assess for defecatory disorder (e.g., anorectal manometry, balloon expulsion test)

No response Assess for defecatory disorder (e.g., anorectal manometry, balloon expulsion test) Normal Treat as for irritable bowel syndrome

Abnormal Abnormal

Normal Defecography

Evacuation disorder

Rectal anatomic defect

Slow-transit constipation

Repair of prolapse Osmotic laxative, stimulant laxative, or rectocele Biofeedback, physical prokinetic agent; therapy, consultation rarely colectomy with psychologist and/or dietitian Figure 18-6.  Algorithm for the evaluation and treatment of severe constipation.

Chapter 18  Constipation urge should be stressed. If patients experience difficulty in expulsion of stool, they should be advised to place a support approximately 6 inches in height under their feet when sitting on a toilet seat so that the hips are flexed toward a squatting posture. For persons with an inactive lifestyle, activity should be encouraged. The use of constipating drugs should be avoided.

Psychological Support

Constipation may be aggravated by stress or may be a mani­ festation of emotional disturbance (e.g., previous sexual abuse; see Chapter 21). For such patients, an assessment of the person’s circumstances, personality, and background and supportive advice may help more than any physical measures of treatment. Behavioral treatment (see later) offers a physical approach with a psychological component and is often acceptable and beneficial. Psychological treat­ ment is needed only when it would be indicated in any circumstance, not specifically for constipation.

Fluid Intake

Dehydration or salt depletion is likely to lead to increased salt and water absorption by the colon, leading in turn to the passage of small hard stools. Although dehydration is generally accepted as a risk factor for constipation, unless a person is clinically dehydrated, no data support the notion that increasing fluid intake improves constipation.166

Dietary Changes and Fiber Supplementation

After studying the dietary and stool patterns of rural Afri­ cans in the early 1970s, Burkitt and colleagues speculated that a deficiency in dietary fiber was contributing to consti­ pation and other colonic diseases in Western societies.167 Since then, studies have shown that when nonconstipated persons increase their intake of dietary fiber, stool weight increases in proportion to their baseline stool weight and frequency of defecation and correlates with a decrease in colonic transit time.168 Every gram of wheat fiber ingested yields approximately 2.7 g of stool expelled. It follows that when an increased intake of dietary fiber leads to an increase in stool weight in constipated subjects who pass small stools, the resulting stool weight may still be lower than normal. For this reason, the therapeutic results of a high-

fiber diet are often disappointing as a treatment for consti­ pation. In a study of 10 constipated women who took a supplement of wheat bran, 20 g/day, average daily stool weight increased from approximately 30 to 60 g/day, with only half of patients achieving a normal average stool weight. Bowel frequency increased from a mean of two to three bowel movements weekly.169 In a controlled, crossover trial, 24 patients took 20 g of bran or placebo daily for four weeks. Although bran was more effective than placebo in improving bowel frequency and oroanal transit rate, the occurrence and severity of constipation experienced by the patients did not differ between the two treatment periods.170 This result probably reflects the observation that patients complain mainly of difficulty in defecation, rather than a decreased frequency of bowel movements. In a series of constipated patients, about half were reported to have gained some benefit from a bran supplement of 20 g daily.171 Dietary fiber appears to be effective in relieving mild to moderate43 but not severe constipation,69 especially if severe constipation is associated with slow colonic transit, evacu­ ation disorders, or medications. Although dietary modifica­ tion may not succeed, all constipated subjects should be advised initially to increase their dietary fiber intake as the simplest, most physiologic, and cheapest form of treatment. Patients should be encouraged to take about 25 g of NSPs daily by eating whole-wheat bread, unrefined cereals, plenty of fruit and vegetables and, if necessary, a supplement of raw bran, either in breakfast cereals or with cooked foods. Specific dietary counseling often is needed to achieve a satisfactory increase in dietary fiber. Because of side effects, adherence with fiber supplemen­ tation is poor, especially during the first several weeks of therapy. Side effects include abdominal distention, bloat­ ing, flatulence, and poor taste. Most controlled studies of the effect of fiber have shown that the minimum supplemen­ tation needed to consistently alter bowel function or colonic transit time significantly is 12 g/day. To improve adherence, patients should be instructed to increase their dietary fiber intake gradually over several weeks to approximately 20 to 25 g/day. If results of therapy are disappointing, commer­ cially packaged fiber supplements should be tried (Table 18-8). Fiber and bulking agents are concentrated forms of

Table 18-8  Commercial Fiber Products AGENT

starting DAILY DOSe (g)

Methylcellulose

4-6

Psyllium

4-6

Polycarbophil

4-6

Guar gum

3-6

COMMENTS Semisynthetic cellulose fiber that is relatively resistant to colonic bacterial degradation and tends to cause less bloating and flatus than psyllium Made from ground seed husk of the ispaghula plant; forms a gel when mixed with water, so an ample amount of water should be taken with psyllium to avoid intestinal obstruction; undergoes bacterial degradation, which may contribute to side effects of bloating and flatus; allergic reactions such as anaphylaxis and asthma have been reported but are rare Synthetic fiber made of polymer of acrylic acid, which is resistant to bacterial degradation Soluble fiber extracted from seeds of the leguminous shrub Cyamopsis tetragonoloba

275

276

Section III  Symptoms, Signs, and Biopsychosocial Issues NSPs based on wheat, plant seed mucilage (ispaghula), plant gums (sterculia), or synthetic methylcellulose deriva­ tives (methylcellulose, carboxymethylcellulose; see later). Some patients, particularly women with markedly delayed colonic transit, find that fiber aggravates abdominal distention. Bran also may be unhelpful in young people with megacolon and in older subjects, in whom it may lead to fecal incontinence. For these patients, a reduction in fiber intake may relieve symptoms.

solution. Guar gum is approved for use in a number of foods and cosmetics and as a supplement. When used in high doses, guar gum has been reported to cause intestinal obstruction.

SPECIFIC THERAPEUTIC AGENTS Commercial Fiber Products

Poorly Absorbed Ions Magnesium, sulfate, and phosphate ions are poorly absorbed by the gut and thereby create a hyperosmolar intraluminal environment. Their primary mode of action appears to be osmotic, but they may have other possible effects with unclear consequences, such as increasing prostaglandin concentrations in the stool.176 In mildly constipated patients, regular use of magnesium hydroxide is a useful and safe laxative. Stool weight increases by 7.3 g for each additional millimole of soluble magnesium excreted.177 Standard doses of magnesium hydroxide (see Table 18-9) contain 40 to 80 mmol of magnesium ion and typically produce a bowel movement within 6 hours. Magnesium sulfate is a more potent laxative that tends to produce a large volume of liquid stool. Patients may complain about this compound because it often leads to abdominal distention and the sudden passage of a liquid foul-smelling stool. Use of mag­ nesium in older adults is limited by adverse effects such as flatulence, abdominal cramps, and magnesium toxicity. A small percentage of magnesium is actively absorbed in the small intestine; the remainder draws water into the intestine along an osmotic gradient.178 Hypermagnesemia can occur in patients with renal failure and in children. Hypermagnesemia-induced paralytic ileus is a rare compli­ cation,179 and hypermagnesemia with coma has occurred in a normal 6-week-old infant given 16 2-mL doses of milk of magnesia.180 Severe toxicity with coma also has occurred in a chronically constipated child given an enema containing 32.5 mg of magnesium sulfate.181 Patients with renal insuf­ ficiency or cardiac dysfunction can experience electrolyte and volume overload from the absorption of magnesium or phosphorus. Even patients who are otherwise healthy can experience these complications, in addition to dehydration, as a result of excessive use. Phosphate can be absorbed by the small intestine, and a substantial dose must be ingested to produce an osmotic laxative effect. One commercial preparation, Fleet Phospho Soda, contains 48 g (400 mmol) of monobasic sodium phosphate and 18 g (130 mmol) of dibasic sodium phosphate/100 mL, resulting in a hypertonic solution. Hyperphosphatemia can occur, especially in patients with renal insufficiency. In addition, a rare but serious form of acute kidney injury has been associated with sodium phos­ phate solution used before colonoscopy, even in patients with normal baseline renal function. Risk factors include hypertension, advanced age, volume depletion, and use of angiotensin-converting enzyme inhibitors or nonsteroidal anti-inflammatory drugs.182,183 The preparation is no longer available over the counter in the United States.

Methylcellulose Methylcellulose is a semisynthetic NSP of varying chain length and degree of methylation. Methylation reduces bac­ terial degradation in the colon. One study of constipated patients with an average daily fecal weight of only 35 g showed an increase in fecal solids with 1, 2, and 4 g of methylcellulose/day, but fecal water increased only with the 4-g dose. Bowel frequency in this group of patients increased from an average of two to four stools weekly, but the patients did not report marked improvement in consis­ tency or ease of passage of stools (see Table 18-8).172 Ispaghula (Psyllium) Ispaghula is derived from the husks of an Asian plant, has high water-binding capacity, is fermented in the colon to a moderate extent, and increases bacterial cell mass. It is available as effervescent suspensions, granules, and a powder. The suspensions, which are popular, need to be drunk quickly before the husk absorbs water. The granules may be stirred briskly in a half-glass of water and swallowed at once; carbonated water may be preferred. Some people like to swallow the solid granules and then drink a glass of water. Ispaghula (3.4 g as Metamucil) has been shown to increase fecal bulk to the same extent as methylcellulose 1 to 4 g daily in constipated subjects. Although both stool dry and wet weights increased, the total weekly weights remained less than those of a healthy control group without treatment. In an observational study, 149 patients were treated with psyllium in the form of Plantago ovata seeds, 15 to 30 g daily, for a period of at least 6 weeks. The response to treat­ ment was poor among patients with slow colonic transit or a disorder of defecation, whereas 85% of patients without abnormal physiologic testing results improved or became symptom-free. Nevertheless, the authors recommend that a trial of dietary fiber be undertaken before diagnostic testing is performed.69 Ispaghula taken by mouth can cause an acute allergic immunoglobulin E–mediated response, with facial swell­ ing, urticaria, tightness in the throat, cough, and asthma.173 Workers who inhale the compound during manufacture or preparation can have a similar reaction174 (see Table 18-8). Calcium Polycarbophil Calcium polycarbophil is a hydrophilic polyacrylic resin that is resistant to bacterial degradation and thus may be less likely to cause gas and bloating. In patients with IBS, calcium polycarbophil appears to improve global symptoms and ease of stool passage175 but not abdominal pain (see Table 18-8). Guar Gum Guar gum is a natural high molecular weight polysaccharide extracted from the seed of the leguminous shrub Cyamopsis tetragonoloba. It hydrates rapidly to form a highly viscous

Other Laxatives

The main groups of laxatives other than fiber are osmotic agents and stimulatory laxatives; stool softeners and emol­ lients are additional therapeutic agents (see later) (Tables 18-9 and 18-10).

Poorly Absorbed Sugars Lactulose.  Lactulose is a nonabsorbable synthetic disaccha­ ride that consists of galactose and fructose linked by a bond resistant to lactase. Therefore, lactulose is not absorbed by the small intestine but undergoes fermentation in the colon to yield short-chain fatty acids, hydrogen, and carbon dioxide, with consequent lowering of the fecal pH. When

Chapter 18  Constipation Table 18-9  Laxatives Commonly Used for Constipation TYPE OF LAXATIVE

GENERIC NAME(S)

DOSE

COMMENTS

Osmotic Laxatives Poorly Absorbed Ions Magnesium

Magnesium hydroxide

15-30 mL once or twice daily

Sulfate

Magnesium citrate Magnesium sulfate Sodium sulfate

150-300 mL every day 15 g every day 5-10 g every day

Hypermagnesemia can occur in patients with renal failure and in children.

Phosphate

Sodium phosphate

0.5-10 mL with 12 oz of water

Lactulose Sorbitol Mannitol

15-30 mL once or twice daily 15-30 mL once or twice daily 15-30 mL once or twice daily

Polyethylene glycol electrolyte

17-34 g once or twice daily

Cascara sagrada Senna

325 mg (or 5 mL) at bedtime 1-2 7.5-mg tablets daily

Castor oil Bisacodyl Phenolphthalein

15-30 mL at bedtime 5-10 mg at bedtime 30-200 mg at bedtime

Sodium picosulfate Docusate sodium Mineral oil

5-15 mg at bedtime 100 mg twice daily 5-15 mL at bedtime

Enemas, Suppositories

Phosphate enema Mineral oil retention enema Tap water enema Soapsuds enema Glycerin suppository Bisacodyl suppository

120 mL 100 mL 500 mL 1500 mL 60 g 10 mg

Chloride Channel Activator

Lubiprostone

8-24 µg twice daily

Poorly Absorbed Sugars Disaccharides Sugar alcohols

Polyethylene glycol

Stimulant Laxatives Anthraquinones

Ricinoleic acid Diphenylmethane Derivatives

Stool Softeners Emollients

normal subjects take lactulose 20 g (30 mL) daily, none of the sugar is detectable in the stool. In larger doses, some of the sugar passes though the colon unchanged and acts as an osmotic laxative. The recommended dose of lactulose for adults is 15 to 30 mL once or twice daily. The time to onset of action is longer than that for other osmotic laxatives, and two or three days are required for lactulose to achieve an effect. Some patients report that lactulose is effective initially but then loses its effect, perhaps because the intestinal flora are altered in response to the medication.184 Adverse effects related to lactulose include abdominal distention or dis­ comfort, presumably as a result of colonic gas production. Cases of lactulose-induced megacolon have been reported. In a group of young, chronically constipated volunteers who reported fewer than three stools a week, lactulose

Sulfate is generally not used by itself as a laxative agent. Hyperphosphatemia can occur, especially in patients with renal failure. Gas and bloating are common side effects. Sorbitol is commonly used as a sweetener in sugar-free products. In older adults, sorbitol has an effect similar to that of lactulose but has a lower cost. Tends to cause less bloating and cramps than other agents; tasteless and odorless, can be mixed with noncarbonated beverages. Typically used to prepare colon for diagnostic examinations and surgery; also available as powder without electrolytes for regular use (MiraLax) Cause apoptosis of colonic epithelial cells phagocytosed by macrophages; result in lipofuscin-like pigmented condition known as pseudomelanosis coli; no definitive association established between anthraquinones and colon cancer or myenteric nerve damage (cathartic colon) Cramping is common. Has effects in small intestine and colon Removed from U.S. market because of teratogenicity in animals Likely has effects only on colon Efficacy in constipation not well established. Long-term use can cause malabsorption of fat-soluble vitamins, anal seepage, and lipoid pneumonia in patients predisposed to aspiration of liquids. Serious damage to rectal mucosa can result from extravasation of enema solution into the submucosa; hypertonic phosphate enemas and large-volume water or soapsuds enemas can lead to hyperphosphatemia and other electrolyte abnormalities if enema is retained; soapsuds enemas can cause colitis. Increases secretion in the intestines

increased bowel frequency and percentage of stool moisture and softened the stools when compared with a control syrup that contained only sucrose. The effectiveness of lactulose was dose-dependent.185 The effect of lactulose in older patients has been studied in two double-blind, placebo-controlled trials. In one trial, only about half of patients were found to be truly consti­ pated, and, among these patients, lactulose was effective in 80%, as compared with 33% of those who received placebo (glucose) (P < 0.01).186 The second trial was conducted in a nursing home over 8 to 12 weeks in 42 older patients with constipation.187 The initial dose of lactulose was 30 mL/day, and the dose was reduced temporarily or permanently to 15 mL, depending on bowel frequency. Lactulose showed an advantage over placebo (a 50% glucose syrup) by increas­ ing the mean number of bowel movements each day and

277

278

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 18-10  Grade of Evidence for the Use of Laxatives According to the American College of Gastroenterology Task Force on Chronic Constipation LAXATIVE Bulking agents   Psyllium   Calcium polycarbophil   Bran Stool softeners Lubricants Osmotic laxatives   PEG   Lactulose   Milk of magnesia Stimulant laxatives Prokinetic agent   Tegaserod‡ Chloride channel activator   Lubiprostone

GRADE of evidence* B B †

B C A A †

B A §

*Grade A: Based on two or more randomized controlled trials (RCTs) with adequate sample sizes and appropriate methodology. Grade B: Based on evidence from a single RCT of high quality or conflicting results from high-quality RCTs or two or more RCTs of lesser quality. Grade C: Based on noncontrolled trials or case reports. † Insufficient data. ‡ Removed from the U.S. market. § Not yet graded. PEG, polyethylene glycol. Data from Brandt LJ, Prather CM, Quigley EM, et al. Systematic review on the management of chronic constipation in North America. Am J Gastroenterol 2005; 100(Suppl 1):S5-21.

markedly reducing episodes of fecal impaction (P < 0.015) and the need for enemas. Sorbitol.  Sorbitol is used widely in the food industry as an artificial sweetener. Ingestion of as little as 5 g causes a rise in breath hydrogen, and 20 g produces diarrhea in about half of normal subjects.188 Sorbitol is as effective as lactulose and less expensive. A randomized, doubleblind, crossover trial of lactulose, 20 g/day, and sorbitol, 21 g/day, in ambulant older men with chronic constipation showed no difference between the two compounds with regard to frequency or normality of bowel movements or patient preference.189 The frequency of side effects was similar except for nausea, which was more common with lactulose. Mannitol is another sugar alcohol that can be used as a laxative. Polyethylene Glycol Polyethylene glycol (PEG) has emerged as a safe and effec­ tive treatment for chronic constipation. It is an isosmotic laxative that is metabolically inert and able to bind water molecules, thereby increasing intraluminal water retention.190 PEG is not metabolized by colonic bacteria. Solutions containing PEG and electrolytes typically are administered orally to prepare the colon for diagnostic examinations or surgery. Ingestion of PEG leads to an increase in stool volume and softer stools, which may become liquid depending on the volume of PEG consumed. PEG is excreted unchanged in the feces. Electrolytes are added to PEG solutions used for colonic lavage before co­ lonoscopy to avoid the potential adverse effects associated with drinking large volumes of a fluid, such as dehydration and electrolyte imbalance. PEG (with electrolytes) is also available as a powder that is mixed in smaller doses with water for regular use to treat constipation.

Several studies have demonstrated the efficacy of PEG in the treatment of chronic constipation. In a trial in which 70 ambulatory outpatients were treated for four weeks with a PEG-electrolyte solution, 250 mL once or twice daily, at the end of the four weeks, bowel frequency had increased to normal, hard stools were uncommon, and straining on def­ ecation was experienced by fewer than 20% of patients, compared with 80% before treatment. The patients were then randomized to continue PEG or a placebo for 20 weeks in a dose of one or two packets daily, as determined to give the best result. In every parameter examined, the active treatment gave significantly improved results over placebo without adverse clinical or laboratory events. At the end of follow-up, complete remission of constipation was reported by 77% and 20% of patients randomized to PEG and placebo, respectively. The dropout rate of 46% in the placebo group, mostly secondary to treatment failure, was notable.191 In another randomized multicenter trial that compared standard and maximum doses of two PEG formulations of different molecular weights, PEG 3350 and PEG 4000, in 266 outpatients, most patients had their first stool within one day of initiating PEG treatment, and stool consistency improved in both treatment groups. The lowest dose of PEG produced the most normal stool consistency, whereas higher doses produced more liquid stools.192 Low-dose PEG appears to be more effective than lactulose in the treatment of chronic constipation.193,194 A study of 307 patients with chronic constipation who were randomized to 17 g of PEG or placebo for six months showed continued benefit of PEG compared with placebo and no electrolyte abnormalities or intestinal malabsorption.194 PEG is approved by the FDA for children, and a doseranging trial in children with constipation, ages six months to 15 years, has suggested that PEG is a potentially useful treatment, provided that the dose is adjusted to the child’s age.195 PEG solutions may be useful for the short-term treat­ ment of fecal impaction. In one study,196 16 severely ill patients, ages 26 to 87 years, who, despite treatment with various laxatives, had not had a bowel movement in the hospital for 5 to 23 days, were treated with PEG. All had a fecal impaction on clinical examination. They were advised to drink 1 L of a PEG-electrolyte solution, taken as two portions of 500 mL, each over 4 to 6 hours. The regimen was repeated on a second and third day, if neces­ sary. The full dose was taken by 12 patients on the first day, and the remainder took at least one half of the recommended dose; only 8 patients needed treatment on the second day and 2 patients on the third day. The treatment was highly effective and, after the last dose, most patients were passing moderate or large volumes of soft stool, with resolution of impaction. No adverse side effects, apart from abdominal rumbling, occurred, and only 1 patient, who was paraplegic, experienced fecal incontinence. Successful treatment with PEG has been described in outpatients with refractory constipation, older adults (with administration of PEG by mouth or by a nasogastric tube), and children with fecal impaction,197 although children have had difficulty drinking the large volume of fluid. The most common adverse effects of PEG include abdomi­ nal bloating and cramps.190 The medication is generally well tolerated, but cases of fulminant pulmonary edema have been reported after administration of PEG solution by naso­ gastric tube, with one fatality.198,199 In each case, the patient had emesis, suggesting aspiration of PEG. PEG also may delay gastric emptying.200

Chapter 18  Constipation Stimulant Laxatives Stimulant laxatives increase intestinal motility and intesti­ nal secretion. They begin working within hours and often are associated with abdominal cramps. Stimulant laxatives include anthraquinones (e.g., cascara, aloe, senna) and diphenylmethanes (e.g., bisacodyl, sodium picosulfate, phe­ nolphthalein). Castor oil is used less commonly because of its side effect profile. The effect of stimulant laxatives is dose-dependent. Low doses prevent absorption of water and sodium, whereas high doses stimulate secretion of sodium, followed by water, into the colonic lumen. Stimulant laxatives sometimes are abused, especially in patients with an eating disorder,201 even though at high doses they have only a modest effect on calorie absorption. Although a cathartic colon (i.e., a colon with reduced motil­ ity) has been attributed to prolonged use of stimulant laxa­ tives, no animal or human data support this effect. Rather, cathartic colon, as seen on a barium enema examination, is probably a primary motility disorder. Overall, stimulant laxatives are well tolerated if used in doses that produce normal, soft, formed stools. They act rapidly and are particularly suitable for use in a single dose for temporary constipation. Most clinicians are cautious about recommending indefinite daily dosing of stimulant laxatives for chronic constipation. Large doses produce abdominal cramping and liquid stools. Stimulant laxatives vary widely in clinical effectiveness, and some patients with severe constipation are not helped by stimulant laxatives. Anthraquinones.  Anthraquinones, such as cascara, senna, aloe, and frangula, are produced by a variety of plants. The compounds are inactive glycosides that when ingested, pass unabsorbed and unchanged down the small intestine and are hydrolyzed by colonic bacterial glycosi­ dases to yield active molecules. These active metabolites increase the transport of electrolytes into the colonic lumen and stimulate myenteric plexuses to increase intestinal motility. The anthraquinones typically induce defecation six to eight hours after oral dosing. Anthraquinones cause apoptosis of colonic epithelial cells, which then are phagocytosed by macrophages and appear as a lipofuscin-like pigment that darkens the colonic mucosa, a condition termed pseudomelanosis coli202 (see Chapter 124 and Figures 124-7 and 124-8). Whether anthra­ quinone laxatives given over the long term cause adverse functional or structural changes in the intestine is contro­ versial. Animal studies have shown neither damage to the myenteric plexus after long-term administration of senno­ sides203 nor a functional defect in motility.204 A case-control study in which multiple colonic mucosal biopsy specimens were examined by electron microscopy showed no dif­ ferences in the submucosal plexuses between patients taking an anthraquinone laxative regularly for one year and those not taking one.205 An association between use of anthraquinones and colon cancer or myenteric nerve damage and the development of cathartic colon has not been established.206 Senna has been shown in controlled trials to soften stools207 and to increase the frequency and wet and dry weights of stool. The formulations available for clinical use vary from crude vegetable preparations to purified and stan­ dardized extracts to a synthetic compound. Castor Oil.  Castor oil comes from the castor bean. After oral ingestion, it is hydrolyzed by lipase in the small intes­ tine to ricinoleic acid, which inhibits intestinal water absorption and stimulates intestinal motor function by dam­ aging mucosal cells and releasing neurotransmitters.206 Cramping is a common side effect.

Diphenylmethane Derivatives.  Diphenylmethane com­ pounds include bisacodyl, sodium picosulfate, and phenol­ phthalein. After oral ingestion, bisacodyl and sodium picosulfate are hydrolyzed to the same active metabolite, but the mode of hydrolysis differs. Bisacodyl is hydrolyzed by intestinal enzymes and thus can act in the small and large intestines. Sodium picosulfate is hydrolyzed by colonic bacteria. Like anthraquinones, the action of sodium picosulfate is confined to the colon, and its activity is unpre­ dictable because its activation depends on the bacterial flora. The effects of bisacodyl, and presumably sodium picosul­ fate, on the colon are similar to those of the anthraquinone laxatives. When applied to the colonic mucosa, bisacodyl induces an almost immediate, powerful, propulsive motor activity in healthy and constipated subjects, although the effect is sometimes reduced in the latter.208 The drugs also stimulate colonic secretion. Like the anthraquinone laxatives, bisacodyl leads to apop­ tosis of colonic epithelial cells, the remnants of which accu­ mulate in phagocytic macrophages, but these cellular remnants are not pigmented.209 Aside from these changes, bisacodyl does not appear to cause adverse effects with long-term use.210 Bisacodyl is a useful and predictable laxative, especially suitable for single-dose use in patients with temporary constipation. Its possible effect on the small bowel is a disadvantage, in contrast to anthraquinones and sodium picosulfate. Long-term use of bisacodyl or related agents is sometimes necessary for patients with chronic severe constipation. In the doses used, liquid stools and cramps tend to result, and it is difficult to adjust the dose to produce soft, formed stools. Phenolphthalein inhibits water absorption in the small intestine and colon by effects on eicosanoids and the Na+/K+ATPase pump present on the surface of enterocytes. The drug undergoes enterohepatic circulation, which may prolong its effects. It has been removed from the U.S. market because it is teratogenic in animals.

Stool Softeners and Emollients

Docusate Sodium Although the detergent dioctyl sodium sulfosuccinate (docusate sodium) is available as a stool softener, further study of its efficacy is needed. The compound stimulates fluid secretion by the small and large intestines but does not increase the volume of ileostomy output or the weight of stools in normal subjects.211,212 A double-blind, crossover trial has shown benefit in 5 of 15 older constipated subjects, as judged by patients and their caregivers, and a significant increase in bowel frequency.213 In a multicenter, doubleblind, randomized trial in adults, docusate sodium was less effective than psyllium for the treatment of chronic idio­ pathic constipation.214 Mineral Oils Mineral oils alter the stool by undergoing emulsification into the stool mass and providing lubrication for the passage of stool. Long-term use can cause intestinal malabsorption of fat-soluble vitamins, anal seepage, and lipoid pneumonia in patients predisposed to aspiration of liquids.

Enemas and Suppositories

Compounds may be introduced into the rectum to stimulate contraction by distention or chemical action, soften hard stools, or both. Serious damage to the rectal mucosa can result from extravasation of the enema solution into the submucosal plane. The anterior rectal mucosa is the site

279

280

Section III  Symptoms, Signs, and Biopsychosocial Issues most vulnerable to trauma from the tip of a catheter intro­ duced through the backward-angulated anal canal (see Chapter 125). The enema nozzle should be directed poste­ riorly after the anal canal has been passed. Phosphate Enemas Hypertonic sodium phosphate enemas are often effective. They cause distention and stimulation of the rectum. A histologic study in normal subjects showed that a single hypertonic phosphate enema caused disruption of the surface epithelium in 17 of 21 biopsy specimens. Scanning electron microscopy showed patchy denudation of the surface epithelium, with exposure of the lamina propria and absence of goblet cells. The proctoscopic appearance of the mucosa was abnormal in every case but returned to normal within one week.215 Therefore, superficially damaged mucosa appears to heal rapidly. Phosphate enemas are used widely, but no conclusive evidence supports their use. Phosphate enemas, if given to a patient who cannot evac­ uate it promptly, can lead to dangerous hyperphosphatemia and hypocalcemic tetany; one patient, age 91 years, died after a single phosphate enema,216 and coma developed in an adult who was given six phosphate enemas at hourly intervals without evacuation.217 Severe hyperphosphatemia, hypocalcemia, and seizure have been reported in a 4-yearold child with normal renal function after retention of two phosphate enemas.218 The use of phosphate enemas in children 3 years and younger is not recommended.219,220 Saline, Tap Water, and Soapsuds Enemas Saline, tap water, or soapsuds enemas can be effective mainly by distending the rectum and softening feces. Stool evacuation typically occurs two to five minutes following administration. A saline enema does no damage to the rectal mucosa and may be effective.215 Water enemas and soapsuds enemas also may be used, but with large volumes, danger­ ous water intoxication can occur if the enema is retained. Large-volume water or soapsuds enemas also can lead to hyperphosphatemia and other electrolyte disturbances if the enema is retained. Soapsuds enemas can cause rectal mucosal damage and necrosis. Stimulant Suppositories and Enemas Glycerin can be administered as a suppository and is often clinically effective. The rectum is stimulated by an osmotic effect. The effect of glycerin, if any, on the rectal mucosa is unknown. Bisacodyl, 10 mg, is available as a suppository that appears to act topically by stimulating enteric neurons.182 In normal subjects, a single bisacodyl suppository or an enema containing 19 mg of bisacodyl in 100 or 200 mL of water produced marked changes in 23 of 25 rectal mucosal biopsy specimens. The epithelium of the surface and within the crypt was altered; with use of the enema, the surface epithelium was absent.215 The regular use of bisacodyl sup­ positories thus appears unwise. Oxyphenisatin (Veripaque), which is no longer available in the United States, is a stimu­ lant enema that was used in the past mainly before diagnos­ tic procedures. When given by mouth, this compound led to some cases of chronic hepatitis.

Chloride Channel Activator

Lubiprostone is a novel bicyclic fatty acid that activates the chloride 2 channel, thereby increasing intestinal fluid secre­ tion and transit221 without altering serum electrolyte levels. In two phase III randomized, placebo-controlled trials, lubiprostone, 24 µg twice daily, increased the number of spontaneous bowel movements in patients with chronic constipation as defined by the Rome II criteria. Lubipros­

tone also significantly decreased straining, improved stool consistency, and reduced overall severity of symptoms. The frequency of spontaneous bowel movements increased in men and women, as well as older patients, who took the drug. A rebound effect after withdrawal of the drug was not evident.222 The most common side effects were nausea, headache, and diarrhea. Lubiprostone, 24 µg twice daily, was approved by the FDA in 2006 for the treatment of men and women with chronic constipation and in 2008, in a dose of 8 µg twice daily, for women who have IBS with constipation.

Prokinetic Agents

Prokinetic agents induce increased contractility in a segment of the gastrointestinal tract. Stimulation of the 5hydroxytryptamine4 (5-HT4) receptor on afferent nerves in the wall of the gastrointestinal tract induces peristaltic con­ traction of the intestine. Several 5-HT4 agonists have been tested for the treatment of constipation. Cisapride, a benzo­ diazepine, has had variable results in treating constipa­ tion.223 Potentially lethal cardiac dysrhythmias led to its withdrawal from the commercial market in the United States in July 2000, although it still remains available through a limited access program. Newer 5-HT4 agonists such as prucalopride and TD-5108 appear promising as future treatments for chronic constipation. Tegaserod Tegaserod, a partial 5-HT4 agonist, is an aminoguanidine indole derivative of serotonin that is structurally different from cisapride. Because of cardiovascular safety concerns, tegaserod was withdrawn from the market in April 2007. The frequency of cardiovascular events in previous clinical trials was 13 in 13,614 (0.11%) compared with 1 in 7031 (0.01%) in control subjects. The cardiovascular events reported were myocardial infarction (n = 3), sudden cardiac death (n = 1), unstable angina (n = 6), and stroke (n = 3). The decision of the U.S. Food and Drug Administration (FDA) to withdraw the drug has been the subject of debate.224 In a randomized, double-blind, placebo-controlled trial, 1348 subjects with chronic constipation were randomized to receive tegaserod 2 mg twice daily, tegaserod 6 mg twice daily, or placebo for 12 weeks. Response was defined as an increase in the number of complete spontaneous bowel movements (CSBMs) of one or more/week as compared with baseline. During the first 4 weeks of treatment, response rates were 41.4%, 43.2%, and 25.1% for tegaserod 2 mg twice daily, tegaserod 6 mg twice daily, and placebo recipi­ ents, respectively. The effect was maintained over 12 weeks. The median time to first CSBM was significantly shorter in patients treated with tegaserod 6 mg twice daily (73 ± 45 hours) and 2 mg twice daily (117 ± 66 hours) than in those treated with placebo (229 ± 123 hours). Tegaserod also improved the frequency and consistency of stools and reduced straining. No rebound effect was seen after with­ drawal of tegaserod. Diarrhea was more common with tega­ serod 2 and 6 mg twice daily (4.5% and 7.3%, respectively) than with placebo (3.8%). Most cases of diarrhea occurred within the first week of therapy and lasted a median of 2.0 days. Treatment with tegaserod did not result in any cases of electrolyte imbalance.225 Tegaserod had also been used in women with constipation-predominant IBS (see Chapter 118).226 Prucalopride Prucalopride, a full 5-HT4 agonist, is a benzofuran deriva­ tive that induces strong contractions in the proximal colon in dogs and accelerates colonic transit in healthy humans

Chapter 18  Constipation and in patients with functional constipation.227 Three large, 12-week, randomized, placebo-controlled phase III trials of similar design that evaluated the efficacy and safety of prucalopride 2 mg or 4 mg once daily versus placebo in patients with chronic constipation have been published.228-230 Patients enrolled in these studies were required to have at least two CSBMs/week, in combination with straining, a sensation of incomplete evacuation, or hard stools, at least 25% of the time. In one of these studies, the percentage of patients achieving more than three CSBMs/week was 30.9% for those receiving prucalopride 2 mg and 28.4% for those receiving prucalopride 4 mg, compared with 12.0% in the placebo group (P < 0.001 for both comparisons). All other secondary efficacy endpoints, including patients’ satisfac­ tion with their bowel function and treatment and their per­ ception of the severity of their constipation symptoms, were improved significantly at week 12 with the use of 2 or 4 mg of prucalopride as compared with placebo. When the results of the phase III studies (N = 1924) were combined, the per­ centage of patients with an average of at least three CSBMs/ week over the 12-week treatment period was 23.6%, 24.7%, and 11.3% for prucalopride 2 mg, prucalopride 4 mg, and placebo, respectively (P < 0.005). The most frequent adverse effects were headaches, nausea, and diarrhea. No cardio­ vascular side effects were observed, nor were any electro­ cardiographic abnormalities reported. TD-5108 TD-5108 is also a full 5-HT4 agonist. In a four-week phase II trial, 401 patients with chronic constipation (less than three CSBMs/week during a two-week baseline period) were randomized to receive TD-5108, 15, 30, or 50 mg or placebo, once daily. The percentages of patients achieving three or more CSBMs for all four weeks were 27%, 19%, 21%, and 5%, respectively (P = 0.0006 for all TD-5108 groups com­ pared with placebo).231

Peripheral Mu-Opioid Antagonists

Peripherally acting opioid antagonists have been shown to reverse opioid-induced bowel dysfunction without revers­ ing analgesia or precipitating central nervous system with­ drawal signs. Preclinical data suggest that these agents may be effective in the treatment of constipation. Methylnaltrex­ one is a peripherally acting mu-opioid receptor antagonist that was approved by the FDA in 2008 for the treatment of opioid-induced constipation in patients with a late-stage, advanced illness who are receiving an opioid on a continu­ ous basis to relieve pain. Approval was based on two phase III trials. One of the studies involved 133 patients with a life expectancy of less than six months and fewer than three bowel movements in the week prior to treatment or no bowel movements for more than two days. Patients were randomized to receive methylnaltrexone, 0.15 mg/kg sub­ cutaneously, or placebo every other day for two weeks, followed by a three-month open-label treatment period. In this study, 47% of patients reported having a bowel move­ ment within four hours of starting methylnaltrexone com­ pared with 15% of those who received placebo (P < 0.001). Methylnaltrexone did not appear to precipitate opioid with­ drawal symptoms or affect central analgesia.232 Results of long-term studies are awaited. Alvimopan is another mu-opioid receptor antagonist approved by the FDA to accelerate bowel recovery follow­ ing surgery. Alvimopam also appears to improve symptoms of opioid-induced constipation. Patients who received alvimopam, 0.5 mg or 1 mg twice daily or 1 mg once daily, had significantly more spontaneous bowel movements than those receiving placebo.233 Other symptoms of constipation

also improved. The most common side effects were abdomi­ nal pain, nausea, and diarrhea.

Other Agents

Colchicine, a drug used for gout, and misoprostol, a prosta­ glandin analog, have been used to treat patients with severe chronic constipation. Treatment of chronic constipation with colchicine has been studied in a randomized, placebocontrolled, double-blind crossover trial in which colchicine increased the frequency of bowel movements as compared with placebo; however, abdominal pain was greater during administration of colchicine than placebo.234 Data for misoprostol are limited, and side effects of the drug are common.235 Cholinergic Agents Cholinergic agents also have been used to treat constipation. Bethanechol, a cholinergic agonist, appears to benefit patients in whom constipation results from therapy with tricyclic antidepressants; however, data to support its use in patients with other causes of constipation are limited. A single intravenous dose of neostigmine, a cholinesterase inhibitor, has been shown to be remarkably effective in decompressing the colon in patients with acute colonic pseudo-obstruction236 (see Chapter 120), but controlled studies of this class of drugs have not been completed in patients with normal-transit or slow-transit constipation. Side effects, such as bradycardia, increased salivation, vomiting, and abdominal cramping, are common. Botulinum Toxin Clostridium botulinum toxin type A (Botox), a potent neu­ rotoxin that inhibits presynaptic release of acetylcholine, has been injected intramuscularly into the puborectalis muscle to treat defecatory disorders. Preliminary data suggest that botulinum toxin may be effective for treating patients with defecatory disorders in which spastic pelvic floor dysfunction causes outlet delay,237 including those who also have Parkinson’s disease.112,113 One study showed that 19 of 24 patients reported improvement in symptoms and physiologic measurements of pelvic floor function at two months.238 Controlled trials have not yet been per­ formed, however, and this approach is not recommended in lieu of biofeedback, for which clinical experience is greater (see later). Newer Agents A newer approach to treating constipation involves using neurotrophins, a multigene family of proteins that includes nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), and neurotrophin-3 (NT-3).239,240 These factors promote growth of subpopulations of sensory neurons and modulate synaptic transmission at developing neuromuscular junctions in Xenopus nerve muscle cul­ tures.241 A specific factor, R-metHuNT-3, has been shown to increase stool frequency and facilitate passage of stool when administered to constipated patients. In healthy persons, R-metHuNT-3 administered subcutaneously has been shown to accelerate gastric, small bowel, and colonic transit. The effects on stool frequency are observed within three days of the start of treatment and last up to five days after cessation of treatment. R-metHuNT-3 has been well toler­ ated, although half of patients treated in the two studies experienced injection site reactions or paresthesias, pre­ sumably by stimulating noncholinergic excitation and sup­ pressing nitrergic inhibition. Linaclotide is a novel, investigational, minimally absorbed guanylate cyclase C agonist that has been shown to reduce

281

282

Section III  Symptoms, Signs, and Biopsychosocial Issues visceral pain and promote intestinal secretion and colonic transit in animal models. In health volunteers, linaclotide was safe and well tolerated. In women with IBS with constipation, oral linaclotide, 1000 µg daily, significantly accelerated ascending colon transit and improved bowel function.241 In a four-week phase IIb study, 307 patients with chronic constipation with less than three spontaneous bowel movements/week were randomized to placebo or linaclotide, 75, 150, 300, or 600 µg. Within 24 hours, 55% of patients who received 300 µg and 59% of patients who received all the linaclotide doses had a spontaneous bowel movement, compared with 37% of those who received placebo. By week four, the mean number of spontaneous bowel movements/week was 5.5 in those who received 300 µg of linaclotide, compared with 3.9 for those who received placebo. Linaclotide was well tolerated; diarrhea was the only dose-dependent adverse effect. No evidence of rebound constipation was noted in the post-treatment period.242

OTHER FORMS OF THERAPY Defecation Training

Defecation training typically involves three to five treatment sessions, each lasting at least 30 minutes. During these ses­ sions, the normal defecation process is taught, and miscon­ ceptions are dispelled. Patients are encouraged to give a detailed description of their bowel symptoms, prompted by a sympathetic listener who is familiar with the full range of problems experienced by those with defecatory dysfunc­ tion. This process is in itself therapeutic because it enables patients to discuss symptoms that otherwise might be regarded as a private burden. Recommendations regarding the proper amount of fiber intake are often given. For patients with infrequent defecation, the importance of developing a regular bowel habit and not ignoring a call to defecate is emphasized. For those who spend excessive time in the bathroom because of ineffective straining, a regimen of less frequent visits to the bathroom and more effective defecation is recommended. The optimum posture for def­ ecation, including the benefit of raising the feet above floor level when using a Western-type toilet, is described. Patients are encouraged to practice what they are taught; that they may be able to help themselves often gives patients new self-confidence. At each visit, patients are encouraged to reduce any dependence on laxatives, enemas, and sup­ positories. Progress is praised at successive treatment sessions.

Anorectal Biofeedback

Anorectal biofeedback typically follows defecation training. During anorectal biofeedback, patients receive visual or auditory feedback, or both, on the functioning of their anal sphincter and pelvic floor muscles. Biofeedback can be used to train patients to relax their pelvic floor muscles during straining and to coordinate this relaxation with abdominal maneuvers to enhance entry of stool into the rectum. Bio­ feedback can be performed with an electromyographic or anorectal manometry catheter. Simulated evacuation with a balloon or silicone-filled artificial stool is commonly taught to patients to emphasize normal coordination of successful defecation.243 Patient education and rapport between the therapist and the patient are integral components of successful biofeedback.244 Patients typically complete from six sessions in six weeks to three sessions/day for 10 successive days. A systematic review of biofeedback studies performed up to 1993 revealed an overall success rate of 67%, although

controlled studies were lacking.245 Biofeedback may be less effective for patients with descending perineum syndrome than for those with spastic pelvic floor disorders.91 In a review of 38 biofeedback studies, psychological factors were found to influence the response to biofeedback.246 More recently, several controlled trials have supported the efficacy of biofeedback.247-249 Patients with pelvic floor dys­ synergia who failed fiber, 20 g/day, plus enemas or sup­ positories were randomized to five weekly biofeedback sessions (n = 54) or PEG, 14.6 to 29.2 g/day, plus five weekly counseling sessions on preventing constipation (n = 55). At six months, major improvement was reported by 80% of patients who underwent biofeedback compared with 22% of the laxative-treated patients (P < 0.001). The benefits of biofeedback were sustained at 12 and 24 months and pro­ duced greater reductions in straining, sensations of incom­ plete evacuation and anorectal blockage, use of enemas and suppositories, and abdominal pain (all P < 0.01). Stool fre­ quency increased in both groups. All biofeedback-treated patients reporting major improvement were able to relax the pelvic floor and defecate a 50-mL balloon at 6 and 12 months.248 In another controlled trial, 77 patients with dys­ synergic defecation were randomized to biofeedback, sham therapy, or standard therapy for three months. Patients who received biofeedback were significantly more likely to correct dyssynergia, improve the defecation index, decrease the balloon expulsion time, increase the number of CSBMs/ week, and decrease the use of digital maneuvers. Global bowel satisfaction was also higher for patients treated with biofeedback than for those in the other groups.249 Originally, biofeedback training was intensive and initi­ ated during admission to the hospital,250 but subsequent experience has shown that training as an outpatient is sat­ isfactory. A small comparative trial has shown no difference in outcome with or without use of an intrarectal balloon or with home training.251 Results are similar when training is conducted with or without access to a visual display of muscular activity. In the absence of a visual display, the instructor gives continuous information and encouragement to the patient and assesses the effect of instruction by observing how the patient strains and by sensing the effec­ tiveness of straining through gentle tension on a rectal balloon. Many patients do not complete defecation training. Of those who do complete the training, most continue to report improvement in symptoms up to two years after completion of training.250,251 Symptoms reported to improve with defe­ cation training include bowel frequency, straining, abdomi­ nal pain, bloating, and need for laxatives.252 Physiologic measurements before and after treatment have shown that training results in appropriate relaxation of the puborectalis and external anal sphincter muscles,253-255 increase in intra­ rectal pressure,75 a widened rectoanal angle on straining during defecation, an increased rate of rectal emptying, an increased rate of colonic transit, and increased rectal mucosal blood flow. Role of Physiologic and Anatomic Investigation Most published series have restricted defecation training and anorectal biofeedback to patients with a defecatory disorder (i.e., paradoxical contraction of the pelvic floor muscles). At one center, however, such training appeared to benefit a high proportion of unselected patients with idiopathic constipation, regardless of the results of investi­ gation of colonic transit or pelvic floor dysfunction, includ­ ing patients with slow colonic transit.254,256 In another series, the results of treatment did not depend on the presence or absence of a rectocele, intussusception, or perineal

Chapter 18  Constipation descent.252 Other investigators, however, have shown that patients who fail to respond to defecation training and biofeedback have a greater degree of perineal descent than those who respond.91 Defecation training has benefited some patients in whom constipation developed after hysterectomy257 and some patients with solitary rectal ulcer syndrome.258

Complementary and Alternative Medical Therapies

Many complementary and alternative therapies are used by patients with constipation,259 but clinical studies are limited and generally of poor quality (see Chapter 127). A system­ atic review of approximately 90 trials of acupuncture for the treatment of chronic constipation identified in the Chinese Biomedical Database is in progress.260

Sacral Nerve Stimulation

Uncontrolled data suggest that sacral nerve stimulation may be helpful for patients with severe constipation.261 Further investigation is in progress.

Surgery

The goal of surgical treatment for patients with severe constipation is to increase bowel frequency and ease of defecation; a possible additional benefit is relief of abdomi­ nal pain and distention. Procedures may be divided into three groups—partial or total colectomy, construction of a stoma, and anorectal operations undertaken to improve defecatory function. Colectomy Colectomy for constipation produces variable results. A review of 32 published studies of surgery for chronic con­ stipation has found considerable variability in rates of patient satisfaction (39% to 100%).262 The most common complications following surgery are small bowel obstruc­ tion, diarrhea, and incontinence; however, diarrhea and incontinence tend to improve after the first year following surgery. Selection of Patients for Colectomy.  Preoperative psy­ chological assessment is essential, because poor results are common among patients who are psychologically dis­ turbed.263 Because the aim of surgery is to increase bowel frequency, slow colonic transit must be demonstrated by an objective method. Also, defecatory function must be assessed, inasmuch as the inability to expel stool from the rectum may be a major factor in causing symptoms. Finally, a generalized intestinal dysmotility or pseudo-obstruction syndrome should be excluded, as much as possible, by appropriate radiographic study of the small intestine and, when available, studies of gastric emptying and small bowel transit. Series in which these steps have been taken to select a homogeneous group of patients have shown the best results, although longer follow-up is awaited. For example, at one center, only 74 of 1009 patients referred for possible surgi­ cal treatment of chronic constipation underwent surgery. Measurement of intestinal transit and tests of pelvic floor function revealed that 597 patients had no quantifiable abnormality and that 249 patients had pelvic floor dysfunc­ tion without slow colonic transit. Colectomy with an ileorectal anastomosis was performed in 52 patients with demonstrable slow colonic transit and normal defecatory function. The operation also was performed in 22 patients with slow colonic transit and pelvic floor dysfunction after the latter had been treated by a training program. Of the 74 patients treated surgically, 97% were satisfied with the result, and 90% had a good or improved quality of life after

a mean follow-up of 56 months. There was no operative mortality, but 7 patients had a subsequent episode of small bowel obstruction.41 Type of Operation.  Several series have shown that the results of colectomy with cecorectal or ileosigmoid anasto­ mosis are inferior to those for a subtotal colectomy with an ileorectal anastomosis.264 Occasional reports have described proctocolectomy with ileoanal anastomosis and construc­ tion of an ileal pouch, usually following failure of colec­ tomy and ileorectal anastomosis.265 In one patient, ileorectal anastomosis failed because the rectum had a larger than normal capacity.266 Laparoscopic subtotal colectomy appears to be as effective as an open approach.267,268 Construction of a Stoma A colostomy is occasionally performed for slow-transit constipation because it is reversible and the results of colectomy are uncertain. Most patients report subjective improvement after a colostomy performed as a primary procedure for slow-transit constipation or for neurologic disease.89 Many patients, however, continue to require laxa­ tives or regular colonic irrigation. An ileostomy occasionally is performed after failure of colectomy and ileorectal anastomosis for slow-transit con­ stipation, either because constipation persists or because severe diarrhea and incontinence occur. Patients who do not benefit from colectomy with ileorectal anastomosis are likely to be those with a generalized disorder of gut motility or those with a psychological disturbance. Creation of a continent catheterizable appendicostomy, through which antegrade enemas can be administered, can sometimes benefit patients with paraplegia and severe con­ stipation and incontinence. Such a procedure can decrease the time and medication needed for bowel care; most of the experience is in children.269 Operations for Defecatory Disorders The stapled transanal rectal resection (STARR) procedure has been used with some success, particularly for patients who also have a rectocele and intussusception.270-272 Pubo­ rectalis or internal anal sphincter muscle division is un­successful in patients with slow-transit constipation.273 Procedures to correct a rectocele should be considered only for patients who have evidence of retained contrast during defecating proctography or in women in whom constipation is relieved with digital vaginal pressure.89

KEY REFERENCES

Bharucha AE, Wald A, Enck P, Rao S. Functional anorectal disorders. Gastroenterology 2006; 130:1510-18. (Ref 81.) Camilleri M, Bharucha AE, Ueno R, et al. Effect of a selective chloride channel activator, lubiprostone, on gastrointestinal transit, gastric sensory, and motor functions in healthy volunteers. Am J Physiol Gastrointest Liver Physiol 2006; 290:G942-7. (Ref 221.) Camilleri M, Kerstens R, Rykx A, Vandeplassche L. A placebocontrolled trial of prucalopride for severe chronic constipation. N Engl J Med 2008; 358:2344-54. (Ref 228.) Chiarioni G, Whitehead WE, Pezza V, et al. Biofeedback is superior to laxatives for normal transit constipation due to pelvic floor dyssyn­ ergia. Gastroenterology 2006; 130:657-64. (Ref 248.) Dipalma JA, Cleveland MV, McGowan J, Herrera JL. A randomized, multicenter, placebo-controlled trial of polyethylene glycol laxative for chronic treatment of chronic constipation. Am J Gastroenterol 2007; 102:1436-41. (Ref 194.) Fletcher JG, Busse RF, Riederer SJ, et al. Magnetic resonance imaging of anatomic and dynamic defects of the pelvic floor in defecatory disorders. Am J Gastroenterol 2003; 98:399-411. (Ref 162.) Higgins PD, Johanson JF. Epidemiology of constipation in North America: A systematic review. Am J Gastroenterol 2004; 99:750-9. (Ref 10.)

283

284

Section III  Symptoms, Signs, and Biopsychosocial Issues Holzer B, Rosen HR, Novi G, et al. Sacral nerve stimulation in patients with severe constipation. Dis Colon Rectum 2008; 51:524-9. (Ref 261.) Lembo A, Camilleri M. Chronic constipation. N Engl J Med 2003; 349:1360-8. (Ref 2.) Locke GR 3rd, Pemberton JH, Phillips SF. AGA technical review on constipation. American Gastroenterological Association. Gastroenter­ ology 2000; 119:1766-78. (Ref 22.) Longstreth GF, Thompson WG, Chey WD, et al. Functional bowel dis­ orders. Gastroenterology 2006; 130:1480-91. (Ref 6.) Markowitz GS, Stokes MB, Radhakrishnan J, D’Agati VD. Acute phos­ phate nephropathy following oral sodium phosphate bowel purga­ tive: An underrecognized cause of chronic renal failure. J Am Soc Nephrol 2005; 16:3389-96. (Ref 183.)

Rao SS, Seaton K, Miller M, et al. Randomized controlled trial of bio­ feedback, sham feedback, and standard therapy for dyssynergic def­ ecation. Clin Gastroenterol Hepatol 2007; 5:331-8. (Ref 249.) Steinman TI, Samir AE, Cornell LD. Case records of the Massachusetts General Hospital. Case 27-2008. A 64-year-old man with abdominal pain, nausea, and an elevated level of serum creatinine. N Engl J Med 2008; 359:951-60. (Ref 182.) Thomas J, Karver S, Cooney GA, et al. Methylnaltrexone for opioidinduced constipation in advanced illness. N Engl J Med 2008; 358:2332-43. (Ref 232.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

19 Gastrointestinal Bleeding Thomas J. Savides and Dennis M. Jensen

CHAPTER OUTLINE Initial Assessment and Management of Acute Gastrointestinal Bleeding  285 History  285 Physical Examination  286 Laboratory Studies  288 Clinical Determination of the Bleeding Site  288 Hospitalization  288 Resuscitation  288 Initial Medical Therapy  288 Endoscopy  289 Endoscopic Hemostasis  291 Radiologic Imaging  292 Surgery  293 Upper Gastrointestinal Bleeding  293 Epidemiology  293 Risk Factors and Risk Stratification  293

The annual rate of hospitalization for any type of gastro­ intestinal (GI) hemorrhage in the United States is estimated to be 350 hospital admissions/100,000 population, with more than 1,000,000 hospitalizations annually.1 Approxi­ mately 50% of admissions for GI bleeding are for upper GI (UGI) bleeding (from the esophagus, stomach, duodenum), 40% are for lower GI (LGI) bleeding (from the colon and anorectum), and 10% are for obscure bleeding (from the small intestine). This chapter focuses on overt GI bleeding that may be severe and prompts the patient to seek medical attention. The source of most GI bleeds can be suspected by the clini­ cal symptoms and physical examination and confirmed by upper or lower endoscopy. Initial management focuses on medical resuscitation, followed by endoscopic diagnosis and interventions to stop acute bleeding and prevent recurrent bleeding. Pharmacologic therapy is playing an increasingly important role in the management of UGI bleeding from peptic ulcers and varices. Optimal patient outcomes depend on successful medical resuscitation, precise endoscopic diagnosis, and appropriate use of thera­ peutic endoscopy. Severe gastrointestinal bleeding is defined as documented gastrointestinal bleeding (i.e., hematemesis, melena, hema­ tochezia, or positive nasogastric lavage) accompanied by shock or orthostatic hypotension, a decrease in the hema­ tocrit value by at least 6% (or a decrease in the hemoglobin level of at least 2 g/dL), or transfusion of at least two units of packed red blood cells. Most patients with severe gastro­ intestinal bleeding are admitted to the hospital for resusci­ tation and treatment. Overt bleeding implies visible signs of blood loss from the GI tract. Hematemesis is defined as vomiting of blood, which is indicative of bleeding from the esophagus, stomach, or duodenum. Hematemesis includes vomiting of bright red blood, which suggests

Upper Endoscopic Technique  293 Peptic Ulcer  294 Other Causes  303 Varices  306 Lower Gastrointestinal Bleeding  308 Risk Factors and Risk Stratification  309 Mortality  309 Diagnostic and Therapeutic Approach  309 Causes and Management  311 Obscure Overt Gastrointestinal Bleeding  315 Causes and Diagnostic and Therapeutic Approaches  315 Diagnostic Tests  318 Obscure Occult Gastrointestinal Bleeding and Iron Deficiency Bleeding  320 Fecal Occult Blood  320 Iron Deficiency Anemia  321

recent or ongoing bleeding, and dark material (coffeeground emesis), which suggests bleeding that stopped some time ago. Melena is defined as black tarry stool and results from degradation of blood to hematin or other hemochromes by intestinal bacteria. Melena can signify bleeding that originates from UGI, small bowel, or proximal colonic source. Melena generally occurs when 50 to 100 mL or more of blood is delivered into the GI tract (usually the upper tract), with passage of characteristic stool occurring several hours after the bleeding event.2,3 Hematochezia refers to bright red blood per rectum, and suggests active UGI or small bowel bleeding, or distal colonic or anorectal bleeding. Occult gastrointestinal bleeding refers to subacute bleeding that is not clinically visible. Obscure gastrointes­ tinal bleeding is bleeding from a site that is not apparent after routine endoscopic evaluation with esophagogastro­ duodenoscopy (upper endoscopy) and colonoscopy, and possibly small bowel radiography. An algorithm for the initial management of acute, severe UGI bleeding is shown in Figure 19-1.

INITIAL ASSESSMENT AND MANAGEMENT OF ACUTE GASTROINTESTINAL BLEEDING HISTORY

Initial assessment of the patient with acute GI bleeding includes medical history taking, obtaining vital signs, per­ forming a physical examination, including a rectal examina­ tion, and nasogastric lavage. During history taking, patients should be questioned about risk factors and historical fea­ tures that help identify diagnostic possibilities for the bleed­ ing source (Table 19-1). Bleeding from a peptic ulcer should be suspected in patients with a history of an ulcer or those

285

286

Section III  Symptoms, Signs, and Biopsychosocial Issues Severe upper gastrointestinal bleeding

History and physical examination Onset in hospital, syncope, shock, comorbidities, hematochezia Admission to intensive care unit

Hemodynamic resuscitation (ongoing)

Type and crossmatch, complete blood count, chemistry panel, liver biochemical tests, coagulation tests (transfusions as indicated)

Electrocardiogram and chest x-ray

Nasogastric (or orogastric) tube and gastric lavage

Gastroenterology consultation

Proton pump inhibitor may be started before endoscopy if peptic ulcer is suspected

If patient is known or suspected to have chronic liver disease, consider beginning octreotide (bolus and infusion) Figure 19-1.  Algorithm for the initial management of severe upper gastrointestinal (UGI) bleeding. The steps in the algorithm may take place simultaneously or in varying orders and in the emergency department depending on the clinical situation.

taking daily aspirin or other nonsteroidal anti-inflammatory drugs (NSAIDs). Patients who have known or suspected liver disease or who are alcoholics should be suspected of bleeding related to portal hypertension. Patients with heavy alcohol intake, a feeding or chronic nasogastric tube, or a history of gastroesophageal reflux disease are at risk of erosive esophagitis. Patients who have had prior surgical repair of an abdominal aortic aneurysm should be consid­ ered to have a fistula from the graft to the duodenum until proven otherwise. Patients on an anticoagulant such as warfarin should be evaluated for the possibility of excessive anticoagulation. Prior radiation to the abdomen should raise the possibility of radiation enteritis or colitis. Weight loss suggests possible malignancy. Abdominal pain may suggest malignancy, inflammatory bowel disease, or ische­ mic colitis. A change in stool caliber suggests colon cancer or a colonic stricture. Chest pain or syncope suggests pos­ sible cardiovascular complications related to blood loss.

PHYSICAL EXAMINATION

On initial evaluation, physical examination should focus on the patient’s vital signs, with attention to signs of hypovole­ mia such as hypotension, tachycardia, and orthostasis. The abdomen should be examined for surgical scars, tenderness, and masses. Signs of chronic liver disease include spider

Upper endoscopy (generally within 6–12 hours of arrival)

Specific endoscopic treatment (see Figure 19-2); for ulcer, begin proton pump inhibitor if not already started

angiomata, palmar erythema, gynecomastia, ascites, spleno­ megaly, caput medusae, and Dupuytren’s contracture. The skin, lips, and buccal mucosa should be examined for telangiectasias, which are suggestive of hereditary hemor­ rhagic telangiectasia (HHT), or Osler-Weber-Rendu disease. Pigmented lip lesions may suggest Peutz-Jeghers syndrome. Purpuric skin lesions may suggest Henoch-Schönlein purpura. Acanthosis nigricans may suggest underlying malignancy, especially gastric cancer. The patient’s feces should be observed to identify melena or maroon and red stool; however, the subjective description of stool color varies greatly among patients and physicians.4 Nasogastric or orogastric tube placement to aspirate and visually characterize gastric contents can be useful to deter­ mine the presence or absence of large amounts of red blood, coffee-ground material, or nonbloody fluid. Occult blood testing of a nasogastric tube aspirate is not useful, however, because trauma from the nasogastric tube may cause suffi­ cient, although scant, bleeding to cause a false-positive result. Patients who have coffee-ground emesis or fresh bloody emesis that is witnessed do not require placement of a nasogastric tube for diagnostic purposes but may need a nasogastric tube to help clear the gastric blood for better endoscopic visualization and to minimize the risk of aspiration.

Chapter 19  Gastrointestinal Bleeding Table 19-1  Suspected Source of Gastrointestinal Bleeding as Suggested by a Patient’s History SUSPECTED SOURCE OF BLEEDING

PATIENT HISTORY

Nasopharynx

History of nasopharyngeal radiation Recurrent epistaxis Prior nasopharyngeal malignancy Hemoptysis Gastroesophageal reflux disease Heartburn Heavy alcohol use Odynophagia Pill ingestion Traumatic nasogastric tube placement Dysphagia Weight loss Alcohol binge Vomiting Large hiatal hernia Cirrhosis Chronic liver disease Heavy alcohol use Chronic kidney disease Epigastric discomfort Frequent aspirin or nonsteroidal anti-inflammatory drug use History of peptic ulcer disease Early satiation Weight loss Prior severe acute unexplained bleeding Prior abdominal aortic aneurysm surgical repair with synthetic graft Recent endoscopic sphincterotomy Recent liver biopsy or cholangiography Pancreatitis, pseudocyst Recent pancreatography Hereditary nonpolyposis colorectal cancer History of intra-abdominal metastatic cancer Intermittent small intestinal obstruction Recurrent unexplained gastrointestinal bleeding Weight loss Unexplained gastrointestinal bleeding since childhood Use of aspirin or other nonsteroidal anti-inflammatory drug Frequent nosebleeds Hereditary hemorrhagic telangiectasia (Osler-Weber-Rendu disease) Age > 60 years Hematochezia without abdominal pain History of diverticulosis Change in bowel habits Personal or family history colon neoplasia Subacute bleeding Weight loss Cardiovascular disease Hematochezia with abdominal pain or discomfort Family history of IBD Bloody diarrhea History of ulcerative colitis Family history of IBD History of Crohn’s disease Chronic abdominal discomfort Hematochezia with anal pain Dripping blood with bowel movements Hematochezia with normal bowel movements Recent colonoscopy with polypectomy Use of anticoagulants or antiplatelet drugs Age > 70 yr Cardiovascular disease Recurrent bleeding of variable severity Prior intestinal surgical anastomosis History of abdominal radiation therapy

Lungs Esophageal ulceration

Esophageal cancer Mallory-Weiss tear Cameron’s erosions Esophageal or gastric varices or portal hypertensive gastropathy Gastric angiodysplasia Peptic ulcer Gastric cancer Primary aortoenteric fistula Secondary aortoenteric fistula Ampulla of Vater Bile ducts Pancreatic ducts Small intestine malignancy

Meckel’s diverticulum Small intestine or colon ulcerations Small intestine telangiectasias Small intestine angiodysplasia Colonic diverticulosis Colonic neoplasia

Ischemic colitis Ulcerative colitis Crohn’s disease Anal fissure Hemorrhoids Postpolypectomy ulcer Colonic or small intestinal angioectasias Anastomotic ulceration Radiation enteritis or proctitis IBD, inflammatory bowel disease.

287

288

Section III  Symptoms, Signs, and Biopsychosocial Issues LABORATORY STUDIES

Blood from the patient with acute GI bleeding should be sent for standard hematology, chemistry, liver biochemical, and coagulation studies and for typing and crossmatching for packed red blood cells. The hematocrit value immedi­ ately after the onset of bleeding may not reflect blood loss accurately because over 24 to 72 hours there is equilibration of red blood cells in the vascular space with extravascular fluid and hemodilution resulting from intravenous admin­ istration of saline.5 The mean corpuscular volume (MCV) is an important indicator of the chronicity of blood loss; an MCV lower than 80 fL suggests chronic GI blood loss and iron deficiency, which can be confirmed by the finding of low blood iron, total iron-binding capacity (TIBC), and ferritin levels. A low MCV and negative fecal occult blood test result raise the possibility of celiac disease. A high MCV (>100 fL) suggests chronic liver disease or folate or vitamin B12 deficiency. An elevated white blood cell count may occur in more than half of patients with UGI bleeding and has been associated with greater severity of bleeding.6 A low platelet count can contribute to the severity of bleed­ ing and suggests chronic liver disease or a hematologic disorder. The blood urea nitrogen (BUN) and serum creatinine levels can help assess the patient for hemoconcentration (elevated levels) or chronic kidney disease, which may lead to chronic anemia because of decreased erythropoietin production. In patients with UGI bleeding, the BUN level typically increases to a greater extent than the serum crea­ tinine level because of increased intestinal absorption of urea after the breakdown of blood proteins by intestinal bacteria.7 The prothrombin time (PT) and international normalized ratio (INR) assess whether a patient has impairment of the extrinsic coagulation pathway. Values can be elevated in chronic liver disease or with the use of warfarin. Liver biochemical test levels may indicate the presence of acute or chronic liver disease; a low serum albumin level suggests possible chronic liver disease, malnutrition, or protein loss via the intestine or kidney.

CLINICAL DETERMINATION OF THE BLEEDING SITE

Presentation with hematemesis, coffee-ground emesis, or nasogastric lavage with return of a large amount of blood or coffee-ground emesis indicates an UGI source of bleeding. A small amount of coffee-ground material or pink-tinged fluid that clears easily may represent mucosal trauma from the nasogastric tube rather than active bleeding from an UGI source. A clear (nonbloody) nasogastric aspirate does not necessarily indicate a more distal GI source bleeding, because 16% of patients with actively bleeding UGI lesions have a clear nasogastric aspirate.8 The presence of bile in the nasogastric aspirate makes UGI bleeding unlikely but can be seen with an intermittently bleeding UGI source. Melena generally indicates an UGI source but can be seen with small intestinal or proximal colonic bleeding. Hema­ tochezia generally implies a colonic or anorectal source of bleeding unless the patient is hypotensive, which could indicate a severe, brisk UGI bleed with rapid transit of blood through the GI tract.4 Maroon-colored stool can be seen with an actively bleeding UGI source or a small intestinal or proximal colonic source.

HOSPITALIZATION

On the basis of the patient’s initial history, physical exami­ nation, and laboratory test results, the location of bleeding

(upper or lower), suspected bleeding lesion, and severity of bleeding can be predicted. Patients with severe GI bleeding require hospitalization, whereas those who present with only mild acute bleeding (self-limited hematochezia or infrequent melena) and who are hemodynamically stable (not suspected to be volume depleted), have normal blood test results, and can be relied on to return to the hospital if symptoms recur may be candidates for semiurgent outpa­ tient endoscopy rather than direct admission to the hospi­ tal.9,10 On the other hand, patients should be hospitalized in an intensive care unit if they have large amounts of red blood in the nasogastric tube or per rectum, have unstable vital signs, or have had severe acute blood loss that may exacerbate other underlying medical conditions. Patients who have had an acute GI bleed but are hemodynamically stable can be admitted to a monitored bed (step-down unit) or standard hospital bed, depending on their clinical condition. Several small studies have suggested that urgent endos­ copy performed in the emergency department in patients with suspected UGI bleeds can help determine optimal hos­ pital placement; however, widespread implementation of this practice is unlikely.11,12

RESUSCITATION

Resuscitation efforts should be initiated at the same time as initial assessment in the emergency department and continue during the patient’s hospitalization. At least one large-bore (14- or 16-gauge) catheter should be placed intra­ venously, and two should be placed when the patient has ongoing bleeding. Normal saline is infused as fast as needed to keep the patient’s systolic blood pressure higher than 100 mm Hg and pulse lower than 100/min. Patients are transfused with packed red blood cells, platelets, and freshfrozen plasma as necessary to keep the hematocrit value higher than 24%, platelet count higher than 50,000/mm3, and prothrombin time less than 15 seconds, respectively. A GI endoscopist should be consulted as soon as possible to expedite the patient’s assessment and determine the optimal timing of endoscopy. In hospitals with a liver transplanta­ tion program, the transplantation hepatology service should also be notified if the patient is known to have advanced liver disease and is a potential transplant candidate. The patient’s vital signs should be monitored frequently, as appropriate to the level of hospitalization. Laboratorydetermined hematocrit values (not fingerstick hematocrit values, which are less reliable) should be obtained every four to eight hours until the hematocrit value is stable. In patients with active bleeding, an indwelling urinary catheter should be placed to monitor the patient’s urine output. Endotracheal intubation should be considered in patients with active ongoing hematemesis or with altered mental status to prevent aspiration pneumonia. Patients who are older than 60 years, have chest pain, or have a history of cardiac disease should be evaluated for myocardial infarc­ tion with electrocardiography and serial troponin measure­ ments. A chest x-ray should also be considered.

INITIAL MEDICAL THERAPY

Administration of a proton pump inhibitor (PPI) is useful for reducing rebleeding rates in patients with peptic ulcer disease (see later). Starting a PPI in the emergency depart­ ment or intensive care unit (ICU) before endoscopy is per­ formed in patients with severe UGI bleeding has become a common practice but is still controversial.13 Several clinical studies and meta-analyses have shown that infusion of a high-dose PPI before endoscopy accelerates the resolution

Chapter 19  Gastrointestinal Bleeding of endoscopic stigmata of bleeding in ulcers (see later) and reduces the need for endoscopic therapy but does not result in improved clinical outcomes in the transfusion require­ ment, rebleeding rate, need for surgery, or death rate.14-17 Patients with a strong suspicion of portal hypertension and variceal bleeding should be started empirically on intrave­ nous octreotide (bolus followed by infusion [see later and Chapter 90]), which can reduce the risk of rebleeding to a rate similar to that associated with endoscopic therapy (Fig. 19-2; also see Fig. 19-1).18,19

ENDOSCOPY

GI endoscopy will identify the bleeding site and permit therapeutic hemostasis in most patients with GI bleeding. Endoscopy should be done only when it is safe to do so and

when the information obtained from the procedure will influence patient care. Ideally, the patient should be hemo­ dynamically stable, with a heart rate of less than 100 beats/ min and a systolic blood pressure higher than 100 mm Hg. Respiratory insufficiency, altered mental status, or ongoing hematemesis indicates the need for endotracheal intubation before emergency upper endoscopy to stabilize the patient and protect the airway. Coagulopathy and thrombocytope­ nia should be corrected with transfusions prior to endos­ copy. Proper medical resuscitation will not only allow safer endoscopy, but also ensure a better diagnostic examination for lesions, such as varices, that are volume dependent and will allow more effective hemostasis because of the correc­ tion of coagulopathy (Figs. 19-3 and 19-4; also see Figs. 19-1 and 19-2).

Upper endoscopy

Major stigmata (active bleeding, NBVV, or clot)

Oozing

Flat pigmented spot or clean-based ulcer

Combination endoscopic hemostasis (e.g., epinephrine injection and multipolar electrocoagulation)

Hemoclip or thermal hemostasis

Oral PPI and early discharge

Oral PPI twice daily High-dose PPI (IV bolus plus infusion for 72 hr), followed by oral PPI

Figure 19-2.  Algorithm for the endoscopic and medical management of severe ulcer hemorrhage following hemodynamic stabilization. IV, intravenous; NBVV, nonbleeding visible vessel; PPI, proton pump inhibitor; UGIB, upper gastrointestinal bleed.

Severe hematochezia Ongoing hemodynamic resuscitation History, physical examination, nasogastric tube Consult gastroenterologist + surgeon Oral or nasogastric tube colonic purge

Anoscopy Colonoscopy (or flexible sigmoidoscopy)

Source identified (see Figure 19-4)

No source identified

Upper endoscopy or push enteroscopy

Source identified: Treat appropriately (see Figures 19-1 and 19-2)

Source identified: Arteriographic embolization or surgery

No source identified: RBC scintigraphy Angiography

No source identified: Consider repeat endoscopic studies, capsule endoscopy, balloon enteroscopy, or surgery

Figure 19-3.  Algorithm for the management of severe hematochezia. RBC, red blood cell.

289

290

Section III  Symptoms, Signs, and Biopsychosocial Issues Severe hematochezia Ongoing hemodynamic resuscitation History, physical examination, nasogastric tube History of cirrhosis, ulcers, melena, or hematemesis Upper endoscopy and/or push enteroscopy Source identified: Treat

No source identifed

History of hemorrhoids, pelvic or abdominal radiation, colitis, diarrhea

No identifiable risk factors, painless hematochezia

Anoscopy and flexible sigmoidoscopy Source identified: Treat

No source identified

Colonic purge and urgent colonoscopy

No source identified: Push enteroscopy

Source identified: Treat

Source identified: Treat

No source identified: Capsule endoscopy or RBC scintigraphy or angiography

Source identified: Treat (may require balloon enteroscopy)

No source identified: Balloon enteroscopy or surgery

Figure 19-4.  Algorithm for the management of severe hematochezia modified according to the patient’s history. RBC, red blood cell.

Patients with active hemorrhage (i.e., a high-volume bloody gastric lavage or ongoing hematochezia) should undergo emergency endoscopy soon after medical resuscita­ tion. Where emergency endoscopy should be performed in the hospital depends somewhat on local circumstances, but, in general, emergency endoscopy is best performed once the patient has reached an ICU bed, rather than in the emer­ gency department, because resources (personnel, medica­ tions, space) are more readily available in the ICU. Patients suspected of having cirrhosis or an aortoenteric fistula or who rebleed in the hospital should undergo emergent endoscopy, usually within six hours of admission or rebleeding. Patients who are hemodynamically stable without evidence of ongoing bleeding can undergo urgent endoscopy (within 12 hours), often in the GI endoscopy unit rather than the ICU. Middle-of-the-night endoscopy should be avoided, except for the most severely bleeding or highrisk patients, because well-trained endoscopy nurses, endo­ scopic equipment, and surgical backup may not be available at night. In the rare patient with massive bleeding and refractory hypotension, endoscopy can be performed in the operating room, with the immediate availability of surgical management, if necessary. In patients with severe UGI bleeding, gastric lavage with a large (34-Fr) orogastric tube should be performed to evacu­ ate blood and clots from the stomach to prevent aspiration and allow good endoscopic visualization. Special lavage systems can help remove blood rapidly. The intravenous

administration of erythromycin (a gastric prokinetic agent) 30 to 90 minutes before upper endoscopy to induce gastric contraction and push blood from the stomach into the small intestine helps endoscopic visualization.20,21 Therapeutic single- or double-channel endoscopes with large-diameter suction channels should be used to allow quick removal of fresh blood from the GI tract during endoscopy. Addition­ ally, a water pump can be used to irrigate target lesions through an accessory channel and dilute blood for suction­ ing, both of which facilitate visualization. Using iced saline lavage to prevent or decrease UGI bleeding is of no particu­ lar value and may impair coagulation and cause hypother­ mia. Gastric lavage with lukewarm tap water is as safe as lavage with sterile saline and much less expensive. In patients with severe hematochezia and suspected active colonic bleeding, urgent colonoscopy can be under­ taken after a rapid purge (see Figs. 19-3 and 19-4).22,23 Patients should receive 4 to 8 L of polyethylene glycol purge orally or via a nasogastric tube over four to six hours until the rectal effluent is clear of stool, blood, and clots. Addi­ tional polyethylene glycol bowel purge may be required in some patients, particularly those with active bleeding, severe constipation, or the onset of hematochezia in the hospital. Metoclopramide, 10 mg, may be given intrave­ nously before the purge and repeated every four to six hours to facilitate gastric emptying and reduce nausea. In patients with severe or ongoing active hematochezia, urgent co­ lonoscopy should be performed within 12 hours, but only

Chapter 19  Gastrointestinal Bleeding Severe overt obscure gastrointestinal bleeding Hematochezia

Melena

Urgent colonoscopy after colonic purge

Upper endoscopy and/or push enteroscopy

Source identified: Treat

No source identified

Source identified: Treat

Source identified: Treat

No source identified: Colonoscopy with examination of terminal ileum

No source identified: Capsule endoscopy

Source identified No source identified: Balloon endoscopy In proximal small intestine

In distal small intestine

Deep enteroscopy*

Retrograde ileoscopy (via balloon enteroscopy or colonoscopy)

Source identified: Treat or laparotomy and intraoperative enteroscopy

No source identified: Supportive care

Figure 19-5.  Algorithm for the management of severe overt obscure gastrointestinal bleeding. *Deep enteroscopy includes double-balloon enteroscopy, single-balloon enteroscopy, and spiral enteroscopy.

after thorough cleansing of the colon. Patients with mild or moderate self-limited hematochezia should undergo co­ lonoscopy within 24 hours of admission, and a colonic purge is also recommended in this situation to cleanse the colon thoroughly. Patients with maroon stool in whom there is pretest uncertainty about the bleeding source should be considered for an urgent polyethylene bowel preparation as well. Co­ lonoscopy immediately after push enteroscopy (see later), while the patient is still sedated, will expedite a patient’s care if push enteroscopy does not provide a diagnosis (and is also indicated for colon cancer screening in patients older than 50 years; Fig. 19-5). Wireless small bowel capsule endoscopy (or capsule endoscopy; see later) can be useful in patients with overt GI bleeding who have normal push enteroscopy and colonos­ copy results and in whom a small bowel source of bleeding is suspected.24 Capsule endoscopy has the advantages of directly visualizing the small intestine to identify potential sources or active bleeding. Disadvantages are that the pro­ cedure takes eight hours to complete and additional time to download and review the images, does not permit therapeu­ tic hemostasis, and is difficult to perform in inpatients because of limited availability of staff trained to place the capsule. A follow-up endoscopic procedure, such as singleor double-balloon enteroscopy or retrograde ileoscopy, may be indicated for definitive diagnosis and treatment if a focal bleeding site is found on capsule endoscopy. Complications related to emergency endoscopy and endo­ scopic hemostasis may occur in up to 1% of patients,

depending on the type of endoscopy and treatment per­ formed.25,26 The most common complications include GI tract perforation, aspiration pneumonia, induced hemor­ rhage, an adverse medication reaction, hypotension, and hypoxia (see Chapter 40).

ENDOSCOPIC HEMOSTASIS

Thermal contact probes have been the mainstay of endo­ scopic hemostasis since the 1970s. These probes come in diameters of 7 and 10 Fr and in lengths that can fit through panendoscopes, enteroscopes, or colonoscopes. Contact probes can physically tamponade a blood vessel to stop bleeding and interrupt underlying blood flow, and thermal energy is then applied to seal the underlying vessel (coap­ tive coagulation). The most commonly used probe is a mul­ tipolar electrocoagulation (MPEC) probe, also referred to as a bipolar electrocoagulation probe, with which heat is created by current flowing between intertwined electrodes on the tip of the probe. Animal studies in which MPEC probes were used to stop bleeding in mesenteric vessels have shown that optimal coagulation occurs with lowpower settings (12 to 16 W) applied for a moderate amount of time (8 to 10 seconds), with moderate pressure on the bleeding site.27 Heater probes can provide a predetermined amount of joules of energy, which does not vary with tissue resistance. Animal studies have shown that heater probes can effectively coagulate arteries up to 2 mm in diameter, a diameter considerably larger than most secondary or ter­ tiary branches of arteries (usually 1 mm) found in resected bleeding human peptic ulcers.28,29 The main risk of using a

291

292

Section III  Symptoms, Signs, and Biopsychosocial Issues thermal probe is perforation with excessive application of coagulation or pressure, especially in acute or nonfibrotic lesions. Thermal probes can also cause a coagulation injury that can make lesions larger and deeper and may induce delayed bleeding in patients with a coagulopathy. Injection therapy is performed most commonly with the use of a sclerotherapy needle to inject epinephrine, diluted to a concentration of 1 : 10,000 or 1 : 20,000, submucosally into or around the bleeding site or stigma of hemorrhage (see later). The advantages of this technique are its wide availability, relatively low cost, and safety in patients with a coagulopathy. Additionally, it is associated with a lower risk of perforation and thermal burn damage than the thermal techniques. The disadvantage of epinephrine injec­ tion is that it is not as effective for definitive hemostasis as thermal coagulation, hemoclipping (see later), or combina­ tion therapy.30,31 Injection therapy can also be performed with a sclerosant, such as ethanolamine or alcohol, but these agents are associated with increased tissue damage and other risks. Endoscopic hemoclips (or clips) have been available since 1974, and have become popular as technical improvements have been introduced.32 Hemoclips serve to apply mechani­ cal pressure to a bleeding site, as is done with surgical clips or sutures. Endoscopic hemoclips differ from surgical clips, however, in that they do not have as much compressive strength, and the currently available clips do not close com­ pletely but leave a small space between the prongs. Animal studies have shown that the first-generation hemoclips could not stop bleeding in vessels larger than a diameter of 1 mm.33 Subsequent hemoclips have been larger and stronger and have had a grasp and release mechanism that improves endoscopic deployment and hemostasis. By not causing significant thermal damage, hemoclips are espe­ cially useful for patients with malnutrition or coagulopa­ thy.34 Nevertheless, hemoclips can also be difficult to deploy depending on the location of the bleeding site, the degree of fibrosis of the underlying lesion, and limitations to endo­ scopic access. Band ligation is a technique in which mucosal (with or without submucosal) tissue is suctioned into a cap placed on the end of the endoscope, and a rubber band is rolled off the cap and over the lesion to compress its base. This tech­ nique is widely used for the treatment of esophageal varices (see Chapter 90) and occasionally can be used for other bleeding lesions. An advantage of band ligation is that it is relatively easy to perform; however, a disadvantage is that sufficient mucosa must be suctioned into the cap for ligation to be successful. Depending on the manufacturer, some band ligation devices can only fit on diagnostic endoscopes, and switching from a larger therapeutic endoscope to a smaller diagnostic endoscope during a case is timeconsuming and inefficient.

RADIOLOGIC IMAGING

Angiography may be used to diagnose and treat severe bleeding, especially when the cause cannot be determined by upper and lower endoscopy. Angiography generally is diagnostic of extravasation into the intestinal lumen only when the arterial bleeding rate is at least 0.5 mL/min.35 The sensitivity of mesenteric angiography is 30% to 50% (with higher sensitivity rates for active GI bleeding than for recur­ rent acute or chronic occult bleeding), and the specificity is 100%.36 An advantage of angiography is that it permits therapeutic intra-arterial infusion of vasopressin or trans­ catheter embolization for hemostasis if active bleeding is detected, without the need for bowel cleansing. Neverthe­ less, the rate of major complications, including hematoma

formation, femoral artery thrombosis, contrast dye reac­ tions, acute kidney injury, intestinal ischemia, and transient ischemic attacks, is 3%.37 Another disadvantage of angiog­ raphy is that it usually does not identify the specific cause of bleeding, only its location. Radionuclide imaging is occasionally helpful for patients with unexplained GI bleeding, although it is used less fre­ quently now than in the past because of the widespread use of endoscopy and lack of availability of nuclear medicine services for emergencies, particularly at night and on week­ ends. Radionuclide imaging can be performed relatively quickly and may help localize the general area of bleeding and thereby guide subsequent endoscopy, angiography, or surgery. The technique involves injecting a radiolabeled substance intravenously into the patient’s bloodstream and then performing serial scintigraphy to detect focal collec­ tions of radiolabeled material. Radionuclide imaging has been reported to detect bleeding at a rate of 0.04 mL/min.38 The two tracers used for radionuclide imaging for bleeding (bleeding scans) are technetium sulfur colloid and techne­ tium pertechnetate–labeled autologous red blood cells. Technetium sulfur colloid is cleared rapidly from the blood­ stream and is therefore useful for identifying acute, active bleeding. Technetium pertechnetate–labeled red blood cells remain in the circulation for up to 24 hours and there­ fore can be used for repeated scanning in patients with intermittent bleeding. A comparative study of patients with suspected GI bleeding has found technetium per­ technetate–labeled red blood cell scans to be more sensitive, specific, and accurate than technetium sulfur colloid scans.39 The overall rate of a diagnostic radionuclide scan is approximately 45%, with a 78% accuracy rate in the local­ ization of the true bleeding site.40,41 Up to 25% of bleeding scans suggest a site of bleeding that proves to be incor­ rect.41-43 The rate of true-positive scans is higher for active bleeding with hemodynamic instability than for less severe bleeding.44 The most common reason for a false-positive result is rapid transit of luminal blood, so that labeled blood is detected in the colon even though it originated from a more proximal site in the GI tract. Caution is recommended in using the results of delayed scans to localize and target lesions for resective surgery.45 Technetium pertechnetate scintigraphy can identify ectopic gastric mucosa in a Meckel’s diverticulum. This diagnosis should be considered in a pediatric or young adult patient with unexplained GI bleeding. The positive predictive value, negative predictive value, and overall accuracy of a so-called Meckel’s scan has been reported to be higher than 90% in young patients.46,47 In patients older than 25 years, however, Meckel’s scans are much less sensitive (80 — 60 yr Bleeding onset in hospital Comorbid medical illness Shock or orthostatic hypotension Fresh blood in nasogastric tube Coagulopathy Multiple transfusions required Higher lesser curve gastric ulcer (adjacent to left gastric artery) Posterior duodenal bulb ulcer (adjacent to gastroduodenal artery) Endoscopic finding of arterial bleeding or visible vessel

No diagnosis or >1 type of lesion No diagnosis or >1 type of lesion 59%

Other

33% 39%

7% 22%

Pathogenesis

Peptic ulcers are most commonly caused by a decrease in mucosal defense mechanisms attributable to aspirin or other NSAIDs, Helicobacter pylori infection, or both.75,76 In one large multicenter study of patients with severe peptic ulcer bleeding, 57% of those with bleeding from a gastric ulcer (n = 2057) took aspirin or other NSAID, and 45% were infected with H. pylori, whereas 53% of those with a bleed­ ing duodenal ulcer (n = 2033) took aspirin or other NSAID, or both, and 50% were infected with H. pylori.77 Of the patients with a bleeding peptic ulcer in this study, 10% had no obvious cause for the ulcer (H. pylori–negative, no aspirin or other NSAID use, no cancer, no gastrinoma).

11%

Ulcers

17%

12% Gastroesophageal varices

Other 1983–1992 (n = 945)

2000–2008 (n = 300)

Figure 19-6.  The frequencies of major causes of severe upper gastrointestinal bleeding during two time periods in patients seen at the University of California, Los Angeles, Center for Ulcer Research Education. (All differences between the two time periods are statistically significant; P < 0.05.) Note that in the more recent period, the overall number of cases of severe upper gastrointestinal bleeding and the percentage of cases caused by peptic ulcer have declined.

Chapter 19  Gastrointestinal Bleeding H. pylori infection is common, with a prevalence of over 80% of the population in many developing countries and 20% to 50% in industrialized countries.78 H. pylori gastritis most commonly involves the antrum and pre­ disposes patients to duodenal ulcers, whereas gastric body– predominant gastritis is associated with gastric ulcers. The lifetime risk of peptic ulcer disease from H. pylori infec­ tion ranges from 3% in the United States to 25% in Japan (see Chapter 50). NSAIDs are the most widely used medication in the United States, with 11% of the adult population using NSAIDs on a daily basis.79 NSAIDs, including aspirin, pre­ dominantly cause ulceration by inhibiting cyclooxygenasemediated prostaglandin synthesis and thereby impairing mucosal protection, rather than causing direct topical injury.76 Gastroduodenal ulcers are found at endoscopy in 15% to 45% of patients who take NSAIDs regularly.80,81 Gastric ulcers are approximately four times as common as duodenal ulcers in patients who take NSAIDs.82 In a large study of patients with UGI hemorrhage and NSAIDassociated ulcers, however, gastric and duodenal ulcers occurred with equal frequencies.77

Histopathology

In a landmark study by Swain and colleagues, the patho­ logic examination of 27 surgically resected bleeding gastric ulcers with endoscopically visible vessels revealed an un­ derlying artery in 96% of specimens.28 Approximately 50% of the vessels protruded above the surface of the ulcer, whereas the other 50% had clot in continuity with a breach in the vessel wall. The bleeding arteries had a mean di­ ameter of 0.7 mm, with a range of 0.1 to 1.18 mm.

Endoscopic Risk Stratification

Endoscopy not only detects a peptic ulcer, but also can be used to evaluate the ulcer for stigmata associated with an increased risk of rebleeding. The Forrest classification is used to categorize findings during endoscopic evaluation of bleeding peptic ulcers, as follows: active spurting bleeding (Forrest IA); oozing bleeding (Forrest IB); pigmented protu­ berance or nonbleeding visible vessel (NBVV; Forrest IIA); adherent clot (Forrest IIB); flat pigmented spot (Forrest IIC); and clean-based ulcer (Forrest III).83 Overall interobserver agreement among experts for classifying these stigmata of recent bleeding is only fair to moderate, with poor agree­ ment for NBVVs.84,85 Endoscopic stigmata of recent hemorrhage from an ulcer are shown in Figure 19-7, and the risk of rebleeding asso­ ciated with each stigma is shown in Figure 19-8. Patients at high risk of rebleeding without treatment are those with active arterial bleeding (90%), an NBVV (50%), or an adherent clot (33%).86,87 These patients benefit from endoscopic hemostasis (see later). An endoscopically identified NBVV that has a translucent (pearl or whitish) color has a higher risk of rebleeding than a darkly colored pigmented pro­tuberance (clot), because the translucent stigma likely represents the arterial wall.88,89 A multi­ variate analysis of predictors of persistent or recurrent bleeding in patients with nonvariceal UGI bleeding is shown in Table 19-5. Patients with major stigmata of ulcer hemorrhage (spurting, NBVV, or adherent clot) benefit most from endoscopic hemostasis, whereas those with a flat spot or clean ulcer base do not. Patients with oozing bleeding and no other stigma (e.g., a clot or NBVV) have an intermediate risk of rebleeding and may benefit from

*

A

C

B

D

Figure 19-7.  Endoscopic stigmata of recent peptic ulcer bleeding. A, Active bleeding with spurting. B, Visible vessel (arrow) with an adjacent clot. C, An adherent clot. D, Slight oozing of blood after washing in the center of an ulcer without a clot or visible vessel.

295

Section III  Symptoms, Signs, and Biopsychosocial Issues

Risk (%)

296

100 90 80 70 60 50 40 30 20 10 0

were treated with an intravenous histamine H2 antagonist and cessation of aspirin and other NSAIDs.88,90,91 Natural history studies of untreated NBVVs have found that these lesions resolve over four days and adherent clots tend to resolve over two days.92

90%

50% 33% 10%

7%

3%

Clean Active Non- Adherent Oozing Flat ulcer bleeding bleeding clot spot base visible vessel Figure 19-8.  Rebleeding rates without endoscopic therapy or administration of a proton pump inhibitor in patients with ulcers demonstrating various stigmata of recent hemorrhage. (From the University of California, Los Angeles, Center for Ulcer Research Education database, unpublished.)

Doppler Probe Ultrasound Portable Doppler ultrasound probes can be passed through the working channel of an endoscope and applied to an ulcer to determine if blood flow is present beneath a stigma in the ulcer base.93,94 The presence of a blood flow signal correlates with the risk of rebleeding before and after endo­ scopic therapy. Conflicting results have been reported, however, as to whether use of Doppler ultrasound improves the outcome of endoscopic hemostasis in patients with acute peptic ulcer bleeding.95,96 A decision-analysis study has found that Doppler ultrasound is the preferred costminimizing strategy over conventional endoscopic therapy alone in patients with acute peptic ulcer bleeding,97 but the area remains one of active investigation.

Endoscopic Hemostasis

Table 19-5  Independent Risk Factors for Persistent or Recurrent Gastrointestinal Tract Bleeding

RISK FACTOR Clinical Factors Health status (ASA class 1 vs. 2-5) Comorbid illness Shock (systolic blood pressure < 100 mm Hg) Erratic mental status Ongoing bleeding Age ≥ 70 yr Age > 65 yr Transfusion requirement Presentation of Bleeding Hematemesis Red blood on rectal examination Melena Laboratory Factors Coagulopathy Initial hemoglobin ≤ 10 g/dL Endoscopic Factors Ulcer location on superior wall of duodenum Ulcer location on posterior wall of duodenum Active bleeding High-risk stigmata Ulcer size ≥ 2 cm Ulcer location high on lesser curve Diagnosis of gastric or duodenal ulcer Clot over ulcer

range of ODDS RATIOs FOR INCREASED RISK 1.94-7.63 1.6-7.63 1.2-3.65 3.21 3.14 2.23 1.3 NA 1.2-5.7 3.76 1.6 1.96 0.8-2.99 13.9 9.2 2.5-6.48 1.91-4.81 2.29-3.54 2.79 2.7 1.72-1.9

ASA, American Society of Anesthesiologists; NA, not applicable. Data from Barkun A, Bardou M, Marshall JK. Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med 2003; 139:843-57.

endoscopic hemostasis but not from high-dose PPI infu­ sion (see later). The risk of rebleeding from a peptic ulcer decreases sig­ nificantly 72 hours after the initial episode of bleeding. This conclusion is based on studies in which serial endoscopies were performed and only active bleeding was treated endo­ scopically, with all other stigmata observed. All patients

Active Bleeding and Nonbleeding Visible Vessels Many well-conducted randomized controlled trials, metaanalyses, and consensus conferences have concluded that endoscopic hemostasis with epinephrine injection or coap­ tive thermal probe therapy significantly decreases the rates of ulcer rebleeding, urgent surgery, and mortality in patients with high-risk stigmata, such as active bleeding and NBVVs.98-101 The rebleeding rates for peptic ulcers with various endoscopic stigmata are shown in Figure 19-8. These rebleeding rates are based on studies performed before the widespread use of high-dose infusions of PPIs and predominantly used injection therapy, MPEC therapy, or a combination of injection and thermal probe therapy. In general, for the lesions at highest risk, including those with active bleeding (90% risk of ongoing bleeding) or NBVVs (50% risk of ongoing bleeding), endoscopic hemo­ stasis alone decreases the rebleeding rate to approximately 15% to 30% (Table 19-6). The adjunctive administration of a high-dose intravenous PPI (e.g., pantoprazole, 80-mg bolus and 8 mg/hr for 72 hours) decreases this rate even further, as discussed in the next section. Intravenous formulations of pantoprazole, lansoprazole, and esomeprazole are avail­ able in the United States. The most commonly used treatment for ulcer bleeding worldwide is epinephrine injection therapy, because it is widely available, easy to perform, safe, and inexpensive. Therapy with epinephrine alone seems to be more effective when used in high doses (13 to 20 mL) than in low doses (5 to 10 mL).102 Injection of epinephrine results in a fivefold increase in circulating plasma epinephrine levels but rarely is thought to cause clinically significant cardiovascular events.103 Although epinephrine injection alone is effective com­ pared with placebo, numerous studies and meta-analyses have shown that the addition of a thermal or mechanical hemostatic modality further decreases the rates of rebleed­ ing, surgery, and mortality significantly.30,104,105 Several studies have suggested that the only benefit to adding epi­ nephrine injection to thermal probe therapy is in patients with active bleeding and that no benefit is seen in patients with NBVVs.106,107 Mechanical endoscopic clips have not been studied as well as injection and thermal probe techniques but seem to be more effective than epinephrine injection alone and have shown mixed results when compared with thermal probe therapy.108-111 In a meta-analysis of outcomes for ulcer

Chapter 19  Gastrointestinal Bleeding Table 19-6  Endoscopic Stigmata of Recent Ulcer Hemorrhage ENDOSCOPIC APPEARANCE

FREQUENCY (%)

RISK OF REBLEEDING (%)

12 22 10 14 10 32

90 50 33 10 7 3

Active arterial bleeding Visible vessel Adherent clot Oozing without stigmata Flat spot Clean ulcer base

RISK OF REBLEEDING AFTER ENDOSCOPIC HEMOSTASIS (%)* 15-30 15-30 0-5 0-5 NA NA

*Reduction in bleeding risk is without the administration of a proton pump inhibitor. NA, not applicable.

Table 19-7  Endoscopic Technical Parameters for Using Multipolar Electrocoagulation in the Treatment of Bleeding Lesions* Peptic Ulcer

Epinephrine Injection Probe size|| Pressure¶ Power setting (W)** Pulse duration (sec) Endpoint

ACTIVE BLEEDING

NONBLEEDING VISIBLE VESSEL

ADHERENT CLOT

MALLORYWEISS TEAR

DIEULAFOY’S LESION

GASTRIC ANGIOECTASIA

COLON DIVERTICULum with VISIBLE VESSEL

Yes† Large Firm 12-15 8-10 Bleeding stops

No Large Firm 12-15 8-10 Flat vessel

Yes‡ Large Firm 12-15 8-10 Flat stigma

Maybe Large or small Moderate 10-15 4 Bleeding stops

Yes Large Firm 10-15 8-10 Flat vessel

No Large Light 10-15 2 White

Maybe§ Large or small Light 10-15 2 Flat vessel

COLON ANGIOECTASIA No Large or small Light 10-15 2 White

*These guidelines from UCLA CURE have been derived from experimental and randomized endoscopic studies. Power, pressure, and duration settings must be reduced for small, acute, or deep bleeding lesions. Epinephrine (1 : 20,000) injected in 1-mL aliquots into each of 4 quadrants should be used to control bleeding initially, followed by coagulation. ‡ Epinephrine (1 : 20,000) injected in 1-mL aliquots into each of 4 quadrants should be injected around clot initially, followed by piecemeal snare resection and treatment of underlying stigmata. § Colonic diverticulum with active bleeding can be treated with epinephrine (1 : 20,000) injected into the neck or base. If a visible vessel is seen at the neck, it can be treated with multipolar electrocoagulation. || Large probe is 10 Fr (3.2-mm diameter) and fits through a 3.8-mm endoscope channel. Small probe is 7 Fr (2.4 mm) and fits through a 2.8-mm endoscope channel. ¶ Pressure is the tamponade pressure exerted en face or tangentially via the contact probe directly on the lesion. **Power setting using BICAP II generator. Power settings are general guidelines and may vary based on the generator used. CURE, Center for Ulcer Research Education; UCLA, University of California, Los Angeles; W, watts. †

hemorrhage, application of hemoclips was shown to be superior to epinephrine injection alone but comparable to thermocoagulation.31 Hemoclips have the advantage of being able to be used in patients with severe coagulopathy without the risk of inducing bleeding. The disadvantage of hemoclips is that they can be difficult to deploy, depend­ ing on the position of the endoscope in approaching an ulcer. Adherent Clots An adherent clot is generally defined as a blood clot over an ulcer that is resistant to several minutes of vigorous target jet water irrigation. The rebleeding rate for ulcers with an adherent clot with medical therapy alone is 8% to 35%, with most large studies reporting rebleeding rates of 30% to 35%.112-115 Randomized controlled studies have shown that endoscopic treatment of adherent clots can decrease the rebleeding rate to less than 5% (see Table 19-6). A metaanalysis has found that endoscopic therapy is superior to medical therapy for preventing recurrent bleeding from peptic ulcers with an adherent clot, but no differences in the need for surgery, duration of hospitalization, number of transfusions, or mortality rate are observed.116 These studies were performed prior to the widespread use of PPIs, which also decrease rates of rebleeding.

Clean-Based Ulcers Patients with clean-based ulcers at endoscopy after target irrigation have a rebleeding rate of less than 5%. Laine and colleagues have found no difference in outcomes between patients who immediately resumed eating and those who waited several days before they resumed eating after an UGI bleed.117 Longstreth and Feitelberg showed that selected low-risk patients with clinically mild UGI bleeds and clean-based ulcers can be discharged safely to home with a significant savings in cost.9,10

Techniques for Endoscopic Hemostasis

Active Bleeding The technique used at the University of California, Los Angeles (UCLA) Center for Ulcer Research and Education (CURE) for actively spurting ulcer bleeding is to inject 0.5to 1.0-mL aliquots of epinephrine (1 : 20,000) via a sclero­ therapy needle into four quadrants of the ulcer within 1 to 2 mm of the bleeding site (Table 19-7). When combination therapy is performed, coagulation is performed with a large 10-Fr multipolar probe. After epinephrine injection, the thermal probe is placed directly on the bleeding site to tamponade the site and stop the bleeding, and coagulation is applied with long (10-second) pulses and firm pressure at a low (12- to 15-W) power setting (Fig. 19-9). The probe

297

298

Section III  Symptoms, Signs, and Biopsychosocial Issues

A

B

C

D

E

F

Figure 19-9.  Actively bleeding gastric ulcer treated with a combination of epinephrine injection, multipolar electrocoagulation, and hemoclip placement. A, Clot with oozing of blood is seen. B, After injection of epinephrine, the oozing has subsided; the edge of the ulcer is seen inferior to the clot. C, Multipolar electrocoagulation is applied with a probe. D, The appearance of the ulcer after electrocoagulation; some oozing is noted at the 7 o’clock position at the edge of the crater. E, A single hemoclip has been applied; bleeding has ceased entirely. F, A second hemoclip has been applied.

is then removed slowly from the ulcer (sometimes with gentle irrigation to prevent pulling coagulated tissue), and thermal coagulation is repeated as needed to stop bleeding and flatten any underlying visible vessel. Epinephrine injec­ tion can be repeated if rebleeding persists. With successful endoscopic hemostasis, the rebleeding rate can be decreased to 30% with monotherapy and 15% with combination therapy (see Table 19-6). Alternatively, injection of epi­ nephrine followed by hemoclip placement directly across the actively bleeding site is also effective, although some investigators recommend that clips be placed prior to injec­ tion of epinephrine to allow placement of the clip directly on the vessel rather than on a submucosal epinephrinefilled cushion. Nonbleeding Visible Vessel In contrast to active arterial bleeding, no significant dif­ ference in results between thermal therapy alone and com­ bination thermal and epinephrine injection therapy is seen with NBVVs. We use the same technique as that used to stop active bleeding; visible vessels are flattened using a large probe, firm pressure, and a low power setting (Fig. 19-10). Hemoclipping can also be effective for preventing rebleeding from an NBVV if the clip is placed across the NBVV and a high-dose PPI is administered intravenously

for 72 hours (Fig. 19-11).77,118 With successful endoscopic hemostasis, the rebleeding rate can be reduced to 30% with injection alone and 10% to 15% with thermal coagulation, hemoclipping, or combination therapy (see Table 19-6). Adherent Clot Our current recommendations for treating an adherent clot on an ulcer are first to inject epinephrine (1 : 20,000) in 1-mL increments in four quadrants around the pedicle of the clot and then use a rotatable cold snare to guillotine the clot piecemeal, without pulling it off the base, until an underlying stigma of hemorrhage is identified in the ulcer base or a 3-mm or smaller clot pedicle is left. Coagulation or hemoclipping is performed if active bleeding, a visible vessel, or residual pedicle is seen. Figure 19-12 shows an example of combination treatment for an adherent clot. The combination technique decreases the rebleeding rate from up to 35% (with medical therapy alone) to 5%. Adherent clots are considered a high-risk stigma, and administration of a high-dose PPI is recommended after endoscopic hemostasis.115,116 Oozing of Blood from an Ulcer without Other Stigmata Minor bleeding from the edge or base of an ulcer (without other stigmata) that continues despite water irrigation and

Chapter 19  Gastrointestinal Bleeding

A

B

C

D

Figure 19-10.  A, Epinephrine injection and multipolar electrocoagulation for hemostasis of a chronic gastric ulcer (thick arrow) with a nonbleeding visible vessel (thin arrow). B, The nonbleeding visible vessel is injected with epinephrine, after which blanching and swelling of the surrounding mucosa occur. C, A multipolar electrocoagulation probe is applied with firm pressure and coagulation. D, After completion of treatment, the visible vessel has been coagulated and flattened.

A

B

C

Figure 19-11.  A, A gastric ulcer with a nonbleeding visible vessel (arrow) treated by endoscopy with epinephrine injection (B) and hemoclip placement (C).

observation suggests the need for endoscopic treatment. The rebleeding rate for ulcers with persistent oozing treated medically varies from 10% (UCLA CURE) to 27% (Hong Kong). Monotherapy with probes or epinephrine injection reduces the rebleeding rate to less than 5%. In patients with oozing, the bleeding arteries may be small and the outcomes better than those in patients with active

arterial bleeding.119 Patients with oozing and no other stigmata of hemorrhage (e.g., a clot or NBVV) can be treated effectively with epinephrine injection alone and have an added benefit from combination therapy. After successful endoscopic hemostasis, patients with oozing and no other stigmata do not benefit from administration of a high-dose PPI.

299

300

Section III  Symptoms, Signs, and Biopsychosocial Issues

A

B

C

D

E

F

G

H

Figure 19-12.  A, Endoscopic treatment of a duodenal ulcer with an adherent clot. B, The clot was injected with epinephrine, followed by piecemeal snare polypectomy to trim away the clot (C-E), after which an underlying vessel was revealed (F; arrow). G, H, Two endoscopic hemoclips are placed across the visible vessel.

Clean-Based Ulcers Patients with clean-based ulcers at endoscopy have a rebleeding rate of less than 5% and therefore do not require endoscopic therapy. If the patient has a clean-based gastric ulcer, biopsies from the ulcer edge should be con­ sidered to exclude underlying malignancy as well as to assess for H. pylori infection (see Chapter 54). These patients can be fed after the endoscopy and treated with oral acid suppression medication; they do not require continued hospitalization unless indicated for other medical problems.

Assessing for Helicobacter pylori Infection

In a patient with a bleeding gastric or duodenal ulcer, endo­ scopic mucosal biopsies of the normal appearing antrum and midbody greater curvature should be obtained to assess for the presence of H. pylori infection. Biopsies can be obtained safely after successful endoscopic hemostasis; however, bleeding reduces the sensitivity of rapid urease testing (see Chapter 50).

Pharmacologic Therapy

Acid Suppression Medication In vitro studies have shown that a luminal gastric pH higher than 6.8 is required for normal clotting function (platelet aggregation and fibrin formation) and that a pH less than 5.4 almost abolishes platelet aggregation and plasma coagula­ tion.120 Platelet aggregates lyse at an acidic pH, an effect that is enhanced by the presence of pepsin. Therefore, reducing the risk of acute bleeding and rebleeding from a peptic ulcer is theoretically possible by maintaining a gastric pH higher than 6. Intravenous H2 receptor antagonists can raise the intragastric pH acutely, but tolerance to these agents devel­ ops rapidly and the pH usually returns to 3 to 5 within 24 hours. Several studies have shown that in normal subjects, administration of an intravenous PPI can consistently keep gastric pH higher than 4 (and often 6) over a 72-hour infu­ sion in contrast to an intravenous H2 receptor antago­ nist.121,122 Trials of intravenous H2 receptor antagonists for the prevention of recurrent ulcer bleeding have shown no definite benefit.123,124

Chapter 19  Gastrointestinal Bleeding Several studies of PPIs have shown that these agents are effective in reducing rebleeding rates from peptic ulcer. In a study from India, patients with endoscopic high-risk stig­ mata of peptic ulcer bleeding (active bleeding, NBVV, clot, oozing) who did not undergo endoscopic hemostasis were randomized to omeprazole, 40 mg orally twice daily, or placebo. The rebleeding rate in the omeprazole-treated group was 11% compared with 36% in the placebo-treated group (P < 0.001).125 Another study from the same investiga­ tors showed that omeprazole, 40 mg orally twice daily for five days, decreased the rebleeding rate after endoscopic hemostasis with injection therapy for ulcers with active bleeding, an NBVV, or a clot from 21% in the placebotreated group to 7% in the oral omeprazole-treated group (P = 0.02).126 In a study from Hong Kong, patients who had undergone successful endoscopic hemostasis for active bleeding or an NBVV were randomized to high-dose intra­ venous omeprazole, 80-mg bolus, followed by 8 mg/hour or placebo. The 30-day rebleeding rate was 6.7% in the omeprazole-treated group compared with 22.5% in the placebo-treated group (P < 0.05).127 The same investigators from Hong Kong found that the 30-day rebleeding rate in patients with an adherent clot or NBVV who received intra­ venous omeprazole alone was 12% compared with 1% in those who received intravenous omeprazole and underwent endoscopic hemostasis (P < 0.05).128 Another study from Hong Kong found that starting intravenous omeprazole before upper endoscopy in patients with UGI bleeding resulted in a decrease in the number of high-risk stigmata found and the need for endoscopic therapy, but no differ­ ence in clinical outcomes such as the number of units trans­ fused, frequency of recurrent bleeding, or rates of surgery and death.129 Systematic and Cochrane reviews of the clinical effective­ ness and cost-effectiveness of PPIs in acute UGI bleeding by Leontiadis and colleagues have found that PPI treatment initiated after endoscopic diagnosis of peptic ulcer bleeding significantly reduces the rates of rebleeding and surgery compared with placebo or H2 receptor blockers and that the benefit is more pronounced in Asian than in non-Asian studies.130-132 PPI treatment was associated with decreased mortality in the Asian studies as well as in patients with high-risk endoscopic stigmata. The initiation of PPI treat­ ment prior to endoscopy significantly reduced the pro­ portion of patients with stigmata of recent hemorrhage at index endoscopy compared with placebo or H2 receptor blockers but did not reduce the rate of mortality, rebleeding, or surgery. Some caution is advised in generalizing the results of PPI trials in Asian patients with peptic ulcer hemorrhage to heterogeneous non-Asian populations. The Asian patients are generally more responsive than heterogeneous popula­ tions or whites to PPIs.133 Asian patients have a smaller average parietal cell mass, are slower metabolizers of PPIs, and often have H. pylori infection, all of which increase the effectiveness of PPIs. These factors may explain the lower mortality rates in Asians compared with non-Asians in meta-analyses of PPI trials for peptic ulcer hemorrhage. A number of issues related to PPIs and UGI bleeding are still unresolved. Whether a PPI should be given before or after endoscopy is uncertain. Although some small random­ ized studies have not shown pre-endoscopy administration of a PPI to improve clinical outcomes (although the number of high-risk stigmata that require treatment is reduced), most modeling studies have suggested that pre-endoscopy administration of a PPI is cost-effective.14,16,17,129,132 Intrave­ nous administration of a PPI by high-dose continuous drip

or intermittent bolus infusion is also a matter of contro­ versy, but most data favor continuous drip, with small comparative studies suggesting that continuous infusion decreases the rate of rebleeding and need for surgery com­ pared with intermittent dosing.134 Whether administration of an oral PPI is as effective as intravenous administration is unclear, although studies have shown that high-dose oral PPI administration (e.g., omeprazole, 40 mg twice daily) reduces rebleeding to rates that would be expected from endoscopic hemostasis. In addition, studies have shown that the increase in intragastric pH with high-dose oral PPI administration is almost identical (although delayed by one hour) to that with intravenous PPI administration.125,135 Whether intravenous administration of a PPI alone is suf­ ficient therapy (without endoscopic hemostasis) in patients with recent UGI bleeding and some stigmata of hemorrhage, such as an NBVV, oozing, or clot, is controversial. In an Asian study, Sung and colleagues reported that the 30-day rebleeding rate with intravenous PPI administration alone (12%) is similar to that in previous studies of endoscopic hemostasis, although they also found that the rebleeding rate with a combination of endoscopic therapy and an intra­ venous PPI is even lower (1%).136 Finally, because almost all the major studies of PPIs in acute peptic ulcer bleeding have been conducted in Asian populations, who have a greater pharmacodynamic response to PPIs than non-Asian populations, studies in non-Asian populations are needed to confirm the Asian data. One large international study has confirmed the benefit of high-dose intravenous PPI admin­ istration in high-risk patients with active arterial bleeding, a NBVV, or an adherent clot in a study of a predominantly white population.137 Somatostatin and Octreotide A meta-analysis has suggested that intravenous administra­ tion of somatostatin or its long-acting form octreotide decreases the risk of rebleeding from peptic ulcers when compared with placebo or an H2 receptor blocker.138 The proposed mechanisms of action include reductions in splanchnic and gastroduodenal mucosal blood flow, decreases in gastrointestinal motility, inhibition of gastric acid secretion, inhibition of pepsin secretion, and gastric mucosal cytoprotective effects. These drugs have not been studied, however, in the era of endoscopic therapy or PPI use and therefore cannot be considered for routine use.139 Somatostatin or octreotide can be considered in patients with severe ongoing bleeding who are not responsive to endoscopic therapy, an intravenous PPI, or both, and are not surgical candidates, although their effectiveness in these patients is uncertain. Intravenous octreotide may also be useful in patients with portal hypertension and peptic ulcer hemorrhage as an adjunct to endoscopic hemostasis and a PPI (see Chapter 90).

Second-Look Endoscopy

Routine repeat, or second-look, endoscopy 24 hours after initial endoscopic hemostasis, with additional endoscopic hemostasis if persistent high-risk endoscopic stigmata are found, has been proposed as a way to improve patient out­ comes. The results of four prospective randomized trials have yielded conflicting results, with no benefit in the majority of studies.140-143 Therefore, routine second-look endoscopy is not recommended for most patients with peptic ulcer bleeding.100 The exception to this recommenda­ tion is for patients who had an incomplete initial endo­ scopic examination because of excessive blood that obscured the view or technical problems with hemostasis. Patients

301

302

Section III  Symptoms, Signs, and Biopsychosocial Issues with clinically significant rebleeding also should undergo a second endoscopy. For patients who are treated with epinephrine injection alone, second-look endoscopy and repeat treatment are considered a routine part of the man­ agement.119 Endoscopic retreatment with administration of more epinephrine for major stigmata of hemorrhage or oozing is required in 20% to 25% of patients with peptic ulcer hemorrhage.

Rebleeding after Endoscopic Treatment

The risk of rebleeding from peptic ulcers that started bleeding in the outpatient setting and required endoscopic hemostasis is greatest in the first 72 hours after diagnosis and treatment. Such patients, therefore, should be kept on a PPI in high doses for at least 72 hours after endoscopic hemostasis, after which they can be switched to a standard dose. Before the widespread use of intravenous PPIs, the rebleeding rate after endoscopic hemostasis of actively bleeding ulcers or those with a NBVV was as high as 30%; now, with the use of PPIs, the rate is less than 10%. The difference between ulcer hemorrhage that starts in the outpatient setting and hemorrhage that starts in the inpatient setting is substantial (Table 19-8). Because the time to rebleeding can be much longer for inpatient (than outpatient) ulcer hemorrhage and the risk of rebleeding is high, combination endoscopic hemostasis and high-dose intravenous PPI administration for at least 72 hours ought to be considered. Further studies are warranted in this highrisk group to define the optimal management. If rebleeding from a peptic ulcer is severe, an urgent repeat endoscopy (rather than surgery) should be performed. A large, well-designed, randomized trial from Hong Kong has found that when endoscopic hemostasis is repeated in patients with hemodynamically significant rebleeding after initial endoscopic hemostasis, 73% of patients achieve sus­ tained hemostasis and do not require surgery.144 The overall mortality rate was the same in both groups, but the rate of complications was significantly higher in the surgical group. Factors that predicted failure of endoscopic retreatment included an ulcer size of at least 2 cm and hypotension on initial presentation.

Angiography and Surgery

Patients with recurrent bleeding despite two sessions of endoscopic hemostasis should be considered for angio­ graphic embolization or surgical therapy. Interventional

Table 19-8  Comparison of Outpatient and Inpatient Onset of Peptic Ulcer Bleeding* Onset PARAMETER Frequency (%) American Society of Anesthesiologists comorbidity score† Time to rebleeding (%)   ≤72 hr   4-7 days   8-30 days   >30 days

OUTPATIENT

INPATIENT

80-90 ≤3

10-20 >3

70-80 10-15 1-5 0

40-50 15-20 15-20 5-10

*Data from the UCLA CURE database. † 1 point signifies a healthy person; 5 points signifies high likelihood of mortality within 24 hr. CURE, Center for Ulcer Research Education; UCLA, University of California, Los Angeles

angiography with embolization has become widely avail­ able. Several retrospective series have reported no signifi­ cant difference between angiography with embolization and surgery in rates of rebleeding and mortality, despite the older age of and more serious medical problems in patients treated by angiography than in those treated by surgery.145,146 These studies suggest that angiography can be considered after failure of endoscopic therapy. If embolization therapy does not control the bleeding, surgery remains an option. Acute surgical intervention is indicated for patients who have exsanguinating bleeding and those who cannot be medically resuscitated. Surgery should also be considered if the endoscopist does not feel comfortable treating a large or pulsating visible vessel (e.g., one in a deep, posterior duodenal ulcer that may represent the gastroduodenal artery). Another indication for surgery is a locally confined bleeding malignant ulcerated mass. The ultimate choice between interventional angiography and surgery often depends on local availability and exper­ tise, as well as the patient’s medical suitability to undergo surgery safely.

Immediate Postendoscopic Management

High-Risk Endoscopic Stigmata Patients who have undergone endoscopic hemostasis for active arterial bleeding, an NBVV, or an adherent clot should be observed in the hospital for 72 hours while they receive high-dose intravenous infusions of a PPI. After suc­ cessful endoscopic treatment and recovery from moderate sedation, the patient can be started on a liquid diet, with subsequent advancement of the diet. Ideally, NSAIDs or warfarin should be withheld for as long as safely possible to help allow the ulcer to heal. For patients with severe atherosclerotic cardiovascular disease who require aspirin, however, a dose of 81 mg/day should be started within seven days to prevent stroke and myocardial infarction. Intermediate-Risk Patients Patients with oozing from an ulcer and no other stigmata (e.g., spurting, NBVV, or clot), severe comorbidity, or shock on presentation should undergo endoscopic hemostasis. Ini­ tiation of an oral PPI and observation in the hospital for 24 to 48 hours after endoscopic hemostasis are recommended. Such patients do not benefit from a high-dose intravenous PPI after successful endoscopic hemostasis. Low-Risk Endoscopic Stigmata Patients with a clean-based ulcer or flat spot in the ulcer base can generally resume a normal diet immediately, begin an oral PPI once daily, and be discharged from the emer­ gency department or hospital when stable.117 Several studies have shown that patients at low risk for rebleeding on the basis of these endoscopic findings and a stable clinical status can avoid hospitalization entirely or be discharged early.9,10,64,147 Generally these patients are young and hemo­ dynamically stable, have no severe coexisting medical ill­ nesses, a hemoglobin level higher than 10 mg/dL, and normal coagulation parameters, had the onset of ulcer bleed­ ing outside the hospital, have good social support systems at home in case rebleeding occurs, and have a clean-based ulcer or ulcer with a flat spot on initial endoscopy. Aspirin and Clopidogrel For patients with severe atherosclerotic cardiovascular disease, aspirin and clopidogrel may need to be restarted soon after hemostasis is achieved to prevent myocardial infarction or stroke. The patient’s cardiologist or neurologist should be consulted in these high-risk cases.

Chapter 19  Gastrointestinal Bleeding Prevention of Recurrent Ulcer Bleeding

Helicobacter pylori Infection All patients with peptic ulcer bleeding should be tested for H. pylori infection and, if the result is positive, should receive antibiotic therapy in standard fashion (see Chapter 50).78 One caveat is that bleeding can lead to a false-negative rapid urease test result, and the patient may need to undergo an alternative method of testing for H. pylori in this setting. Antibiotic therapy does not need to be started urgently and can be initiated on an outpatient basis when the patient has resumed a normal diet. Patients who are H. pylori–positive and who will need long-term PPI treatment because of the concomitant need for aspirin or other NSAID do not neces­ sarily need to be treated for H. pylori infection, because recurrent ulceration will be prevented by the PPI. In patients who are found to have an H. pylori–induced ulcer, confir­ mation of the eradication of H. pylori after treatment is recommended. Aspirin, Other Nonsteroidal Anti-inflammatory Drugs, and Clopidogrel Ideally, patients with ulcer bleeding caused by aspirin or another NSAID should stop the drug. If the patient is also positive for H. pylori, the organism should be eradicated with antibiotics (see Chapter 50).148 In patients with a history of ulcer bleeding who are H. pylori–positive and need to continue taking low-dose aspirin (81 mg daily), eradication of H. pylori alone results in ulcer rebleeding rates similar to those associated with daily PPI therapy (if H. pylori is not eradicated).149 By con­ trast, in patients with a history of ulcer bleeding who are H. pylori–positive and need to continue full-dose NSAID therapy, eradication of H. pylori alone leads to a signifi­ cantly higher rebleeding rate than use of a daily PPI in conjunction with the NSAID. In patients with ulcer bleeding who do not have H. pylori infection but who need to con­ tinue daily aspirin, co-therapy with a daily PPI significantly reduces the rebleeding rate compared with placebo in com­ bination with aspirin.150 Patients who require an antiplatelet medication and have a history of ulcer bleeding will have less chance of recurrent bleeding if they take aspirin 81 mg and a PPI daily com­ pared with clopidogrel alone.151 Patients who require an NSAID after an ulcer bleed may be considered for a selective cyclooxygenase-2 (COX-2) inhibitor. Selective COX-2 inhibitors cause fewer ulcers than nonselective NSAIDs but are associated with a greater rate of cardiovascular complications. Because selective COX-2 inhibitors result in rebleeding rates similar to those associated with NSAID and PPI cotherapy, their use may not be worth the cardiovascular risk.152

Repeat Endoscopy to Confirm Gastric Ulcer Healing

Repeat upper endoscopy should be considered in patients with a gastric ulcer after 6 to 10 weeks of acid suppressive therapy to confirm healing of the ulcer and absence of malig­ nancy (see Chapter 54). In areas of the world in which the population is at intermediate risk for gastric cancer, 2% to 4% of repeat upper endoscopies to confirm ulcer healing have been reported to disclose gastric cancer.153-155 Some experts have suggested that when the index endoscopy with biopsies is negative for malignancy and the ulcer appears benign endoscopically, a follow-up endoscopy is unneces­ sary.156 A small retrospective study has found that when gastric cancer is detected on repeat endoscopy to evaluate gastric ulcer healing, survival is no better than that for patients who did not undergo the recommended follow-up endoscopy.153

OTHER CAUSES Esophagitis

Patients with severe erosive esophagitis can present with hematemesis or melena. A multivariate analysis from a center in France, in which 8% of all UGI bleeding was caused by erosive esophagitis, found that independent risk factors for bleeding esophagitis were grade 3 or 4 (moderate to severe) esophagitis by the Savary-Miller grading system (see Chapter 43), cirrhosis, a poor performance status, and anticoagulant therapy.157 A history of heartburn was obtained in only 38% of patients. Severe bleeding from gastroesopha­ geal reflux–induced esophagitis is treated medically with a PPI (see Chapter 43). Upper endoscopy is critical to diag­ nosing severe erosive esophagitis, but endoscopic therapy generally has no role unless a focal ulcer with a stigma of recent hemorrhage is found. These patients should be treated with a daily PPI for 8 to 12 weeks and undergo repeat endoscopy to exclude underlying Barrett’s esophagus (see Chapter 44). Patients can sometimes present with mild UGI bleeding from esophagitis not related to gastroesophageal reflux disease, such as infections (e.g., Candida, herpes simplex virus, cytomegalovirus) or pill-induced esophagitis. Endos­ copy with biopsies and brushings is critical for making these diagnoses and determining the appropriate pharma­ cologic therapy (see Chapter 45).

Ulcer Hemorrhage in Hospitalized Patients

Hemorrhage from an ulcer or erosions in hospitalized patients typically falls into two categories. The classic cause is stress-related mucosal injury (SRMI, or stress ulcers), which is characterized by diffuse bleeding from erosions and superficial ulcers. The second category is inpatient ulcers, which are large, focal, chronic-appearing ulcers that are painless and present with severe inpatient UGI hemorrhage manifested by hematochezia, melena, or bloody emesis. On emergency endoscopy, focal inpatient ulcers often are actively bleeding or demonstrate a visible vessel or adherent clot and are marked by high rebleeding rates, despite combination endoscopic therapy, and delayed healing on a high-dose PPI. SRMI occurs in the UGI tract of severely ill inpatients in an ICU and is likely caused by a combination of decreased mucosal protection and mucosal ischemia. SRMI usually occurs in the stomach but can also be seen in the duode­ num, esophagus, and even rectum. Diffuse oozing is common, and patients have a poor prognosis and high rebleeding rate, often related to impaired wound healing and multiple organ failure. Bleeding from SRMI is now uncommon, with a frequency of approximately 1.5% patients in an ICU. The two main risk factors are severe coagulopathy and mechanical ventila­ tion for longer than 48 hours.158 The frequency of clinically significant GI bleeding with either or both of these risk factors is 3.7% compared with 0.1% when neither risk factor is present. Other proposed risk factors include a history of UGI bleeding, sepsis, an ICU admission longer than seven days, occult GI bleeding for more than five days, and treatment with high-dose glucocorticoids. ICU patients with risk factors for bleeding are the main target groups for pharmacologic prevention of bleeding SRMI. Therapy with an H2 receptor antagonist has been shown to decrease the rate of clinically significant bleeding in ICU patients at high risk of SRMI.159 Well-designed and adequately powered studies that compare H2 receptor block­ ers and PPIs are few in number, but one large multicenter study found that prophylactic treatment with oral omepra­ zole or intravenous cimetidine results in similar bleeding

303

304

Section III  Symptoms, Signs, and Biopsychosocial Issues rates but that omeprazole is more effective than cimetidine in maintaining the luminal gastric pH above 4.160 A potential harmful effect of gastric acid suppression to prevent stress ulcers is that a decrease in gastric pH may allow proli­ feration of bacteria in the stomach and the potential for aspiration and ventilator-associated pneumonia; however, randomized trials in which acid suppression (with an H2 receptor blocker or antacids) and sucralfate (which does not lower gastric pH) were compared have not shown convincingly that lowering gastric pH increases the risk of pneumonia.161,162 Generally, if a patient with SRMI or an inpatient ulcer is supported hemodynamically and medically, the lesion will heal as the patient’s overall medical status improves. Because SRMI is diffuse, endoscopic therapy is generally not feasible. By contrast, focal inpatient ulcer hemorrhage often requires endoscopic hemostasis for severe hemor­ rhage; however, rebleeding rates are higher and healing rates are slower than those in patients in whom bleeding starts before hospitalization (see Table 19-8).163,164 A study in which epinephrine injection plus hemoclip placement was compared with epinephrine injection plus MPEC in a cohort of patients who had a high frequency of in-hospital ulcers found a significantly lower rebleeding rate in the group that underwent injection and hemoclip placement.118 Figure 19-9 illustrates combination treatment using injec­ tion therapy, bipolar probe coagulation, and endoscopic hemoclipping for an inpatient ulcer hemorrhage.

Dieulafoy’s Lesion

A Dieulafoy’s lesion is a large (1- to 3-mm) submucosal artery that protrudes through the mucosa, is not associated with a peptic ulcer, and can cause massive bleeding. It usually is located in the gastric fundus, within 6 cm of the gastroesophageal junction, although lesions in the duode­ num, small intestine, and colon have been reported. The cause is unknown and congenital and acquired (related to mucosal atrophy or an arteriolar aneurysm) causes are thought to occur. Dieulafoy’s lesion can be difficult to identify at endoscopy because of the intermittent nature of the bleeding; the over­ lying mucosa may appear normal if the lesion is not bleed­ ing. An NBVV or adherent clot without an ulcer may be seen on endoscopy. If a massive UGI bleed seems to be emanating from the stomach, a careful inspection of the proximal stomach should be carried out to look for a protuberance that might be a Dieulafoy’s lesion. Endoscopic Doppler ultrasound has been used to help identify a Dieulafoy’s lesion that is not visualized on endoscopy.165 Because of the difficulty of identifying the bleeding site, we recommend that if a Dieulafoy’s lesion is found and treated, the site be marked with submucosal injection of ink to tattoo the area in case of rebleeding and the need for retreatment. Endoscopic hemostasis of a Dieulafoy’s lesion can be performed with injection therapy, a thermal probe, or clip device or by band ligation.165-169,170 Large case series have reported an initial hemostasis rate of approximately 90%, with the need for surgery in 4% to 16% of cases.171 Rebleed­ ing after successful hemostasis appears to be rare. Although all the endoscopic hemostasis techniques seem to be effec­ tive, perforation and delayed rebleeding have been reported after band ligation (see Chapter 36).

Mallory-Weiss Tears

Mallory-Weiss tears are mucosal or submucosal lacerations that occur at the gastroesophageal junction and usually extend distally into a hiatal hernia (Fig. 19-13). Patients generally present with hematemesis or coffee-ground emesis

Figure 19-13.  Endoscopic appearance of a Mallory-Weiss tear with mild oozing. Note that the tear starts at the gastroesophageal junction (large arrow) and extends distally into the hiatal hernia (small arrow).

and typically have a history of recent nonbloody vomiting followed by hematemesis, although some patients do not recall any vomiting. The tear is thought to result from increased intra-abdominal pressure, possibly in combina­ tion with a shearing effect caused by negative intrathoracic pressure above the diaphragm, which is often related to vomiting in patients with a history of alcohol abuse. Mallory-Weiss tears have been reported in patients who vomit while taking a bowel purge before colonoscopy.171 Endoscopy usually reveals a single tear that begins at the gastroesophageal junction and extends several millimeters distally into a hiatal hernia sac. Occasionally, more than one tear is seen. A retroflexed view in the stomach may provide better visualization than the forward viewing position. The bleeding stigmata of Mallory-Weiss tears can include a clean base, oozing, or active spurting. Usually, the bleed­ ing is self-limited and mild, but occasionally it can be severe. Superficial (mucosal) Mallory-Weiss tears can start healing within hours and can heal completely within 48 hours. Although approximately 50% of patients hospitalized with UGI bleeding from a Mallory-Weiss tear receive blood transfusions, the tear manifests as mild, self-limited hematemesis in most patients, who do not seek medical care.172 The rebleeding rate among patients hospitalized for a Mallory-Weiss tear is approximately 10%; risk factors for rebleeding include shock at presentation and active bleed­ ing at endoscopy.173 Because of the risk of continued and recurrent bleeding, patients with active bleeding from a Mallory-Weiss tear should undergo endoscopic therapy, which can be performed successfully with epinephrine injection, MPEC, hemoclip placement, or band ligation. Randomized trials that compared MPEC and medical therapy with an H2 receptor antagonist have found that endoscopic therapy reduces the rates of rebleeding, blood transfusions, and emergency surgery.174 Our current endoscopic technique for treating actively bleeding Mallory-Weiss tears in patients without portal hypertension or esophageal varices is to apply endoscopic hemoclips to stop the bleeding and close the tear. If hemo­ clips are not available, MPEC at a low power setting and with light pressure for one to two seconds is recommended. The management of patients with esophageal varices caused by portal hypertension who also have a Mallory-Weiss tear should be targeted toward the esophageal varices, with

Chapter 19  Gastrointestinal Bleeding

Figure 19-14.  Endoscopic appearance of Cameron’s lesions. Note that these linear ulcerations (arrows) are located at the distal end of a hiatal hernia.

esophageal band ligation or variceal sclerotherapy (see later and Chapter 90). Patients with a Mallory-Weiss tear are often treated with antiemetics if they have nausea or vomiting and a PPI to accelerate mucosal healing. The PPI may allow better natural hemostasis by raising the gastric pH to improve coagulation and possibly speed the healing of the tear, but this approach has not been well studied and most MalloryWeiss tears heal within days anyway. Therefore, long-term treatment with a PPI is generally not needed.

Cameron’s Lesions

Cameron’s lesions are linear erosions or ulcerations in the proximal stomach at the end of a large hiatal hernia, near the diaphragmatic pinch (Fig. 19-14).175 Cameron’s lesions are thought to be caused by mechanical trauma and local ischemia as the hernia moves against the diaphragm and only secondarily by acid and pepsin. They can be a source of acute UGI bleeding but more commonly may present as slow GI bleeding and iron deficiency anemia. Cameron’s lesions are a common cause of obscure GI bleeding (see later) and not uncommonly are missed by an unsuspecting endoscopist. Endoscopic management has been reported.176 The long-term medical management is usually with iron supplements and an oral PPI (see Chapter 36).177,178 Occasionally surgical repair of the hiatal hernia may be needed.

Upper Gastrointestinal Malignancy

Malignancy accounts for 1% of severe UGI bleeds. The tumors are usually large, ulcerated masses in the esophagus, stomach, or duodenum. Endoscopic hemostasis with MPEC, laser, injection therapy, or hemoclips can temporarily control acute bleeding in most patients and allow time to determine the appropriate long-term management.179,180 Patients with an ulcerated subepithelial mass (which usually is a gastrointestinal stromal tumor or leiomyoma) should undergo surgical resection of the mass to prevent rebleeding and to prevent the risk of metastasis. Angiogra­ phy with embolization should be considered for patients with severe UGI bleeding caused by malignancy who do not respond to endoscopic therapy. External beam radiation can

Figure 19-15.  Endoscopic appearance of gastric antral vascular ectasia (GAVE), or watermelon stomach. The pattern seen in this view is considered classic, with rows of ectatic mucosal blood vessels emanating from the pylorus.

provide palliative hemostasis for patients with bleeding from advanced gastric or duodenal cancer (see Chapter 54).

Gastric Antral Vascular Ectasia

Gastric antral vascular ectasia (GAVE), also described as watermelon stomach, is characterized by rows or stripes of ectatic mucosal blood vessels that emanate from the pylorus and extend proximally into the antrum (Fig. 19-15). The cause is uncertain, and the lesion may represent a response to mucosal trauma from contraction waves in the antrum. GAVE has been associated with cirrhosis and scleroderma (see Chapters 35 and 90). Patients with GAVE who do not have portal hypertension demonstrate linear arrays of angio­ mas (classic GAVE), whereas those with portal hypertension have more diffuse antral angiomas.181 The diffuse type of antral angiomas and occasionally classic GAVE are some­ times mistaken for gastritis by an unsuspecting endoscopist. Such cases are a common cause of obscure GI bleeding in referral centers (see later).48 Patients usually present with iron deficiency anemia or melena, with a mildly decreased hematocrit value sugges­ tive of a slow UGI bleed. GAVE is most commonly reported in older women181 and also seems to be more common in patients with end-stage renal disease. Endoscopic hemostasis with thermal heat modalities such as laser, MPEC, or argon plasma coagulation has been used successfully. Endoscopic hemostasis and ablation with thermal modalities can result in good palliation with an increase in the hematocrit value and a decrease in the need for blood transfusions and hospitalization.181,182 Usually, several sessions, approximately four to eight weeks apart, are needed to achieve eradication of the lesions and a reduction in bleeding from the antral ectasias. Endoscopic therapy with argon plasma coagulation has been shown to be equally (80%) effective in cirrhotic and noncirrhotic patients with GAVE.183 Pilot studies have demonstrated that mucosal band ligation, radiofrequency ablation, and cryo­ therapy can also lead to eradication of GAVE in selected patients.184-186

305

306

Section III  Symptoms, Signs, and Biopsychosocial Issues Placement of a transjugular intrahepatic portosystemic shunt (TIPS) in patients with portal hypertension and cir­ rhosis does not decrease bleeding from GAVE or diffuse antral angiomas. Patients who have ongoing severe chronic bleeding from GAVE rarely may require surgical antrectomy to control symptoms (see Chapters 36 and 90).187

Portal Hypertensive Gastropathy

Portal hypertensive gastropathy (PHG) is caused by increased portal venous pressure and severe mucosal hyperemia that results in ectatic blood vessels in the proximal gastric body and cardia and oozing of blood. Less severe grades of PHG appear as a mosaic or snakeskin pattern and are not asso­ ciated with bleeding.188 Usually, patients with severe PHG present with chronic blood loss, but they occasionally can present with acute bleeding. Severe PHG with diffuse bleeding is treated by measures that decrease portal pressure, usually with β-adrenergic receptor blockers or possibly with placement of a TIPS or surgical portacaval shunt. Endoscopic management has no role unless an obvious focal bleeding site is identified. The best treatment is liver transplantation (see Chapters 36 and 90).

Hemobilia

Hemobilia may occur in patients who have experienced liver trauma, undergone a liver biopsy, manipulation of the hepatobiliary system, as occurs with endoscopic retrograde cholangiopancreatography (ERCP), percutaneous transhe­ patic cholangiography, or TIPS, or have hepatocellular carcinoma or a biliary parasitic infection.189 Patients may present with a combination of GI bleeding and elevated liver biochemical test levels. The diagnosis can be confirmed by using a side-viewing duodenoscope to identify bleeding from the ampulla (Fig. 19-16). Ongoing or recurrent bleed­ ing is treated with arterial embolization via arteriography.

Hemosuccus Pancreaticus

Hemosuccus pancreaticus is a rare form of UGI bleeding that occurs most commonly in patients with acute pancreatitis,

chronic pancreatitis, pancreatic pseudocyst, or pancreatic cancer or after ERCP with pancreatic duct manipulation. It can also result from rupture of a splenic artery aneurysm into the pancreatic duct.190 CT can demonstrate pancreatic pathology if previously unsuspected. Endoscopy with a side-viewing duodenoscope reveals blood coming out of the ampulla. Management of severe hemorrhage is usually with angiographic embolization or surgery.

Postsphincterotomy Bleeding

Bleeding following endoscopic sphincterotomy occurs in approximately 2% of patients (see Chapter 40).191 Potential risk factors include coagulopathy, use of anticoagulants, portal hypertension, renal failure, and the type and length of sphincterotomy. Successful hemostasis of postsphinc­ terotomy bleeding is usually achieved with endoscopic methods such as injection of epinephrine, hemoclips, or MPEC (see Chapter 40).

Aortoenteric Fistula

Bleeding from an aortoenteric fistula is usually acute and massive, with a high mortality rate.192 A primary aortoen­ teric fistula is a communication between the native ab­ dominal aorta (usually an atherosclerotic abdominal aortic aneurysm) and, most commonly, the third portion of the duodenum.193 Often, a self-limited herald bleed occurs hours to months before a more severe, exsanguinating bleed. Occasionally, the diagnosis of an aortoenteric fistula is sus­ pected by a history of an abdominal aortic aneurysm or by palpation of a pulsatile abdominal mass. The diagnosis can be difficult to make on endoscopy in the absence of active bleeding. Demonstration of an aortic aneurysm on ab­ dominal CT scan (with intravenous contrast) suggests the diagnosis of a fistula.49 Secondary aortoenteric fistulas usually occur between the small intestine and an infected abdominal aortic surgical graft. The fistula typically occurs between the third portion of the duodenum and the proximal end of the graft but may occur elsewhere in the GI tract. The fistula usually forms between three and five years after graft placement. Patients often experience a herald bleed that is mild and self-limited, and occasionally intermittent, before massive bleeding occurs.194 A secondary fistula can also occur between the third part of the duodenum and an endovascular stent, in which case the fistula occurs as a result of pressure from the stent against the duodenum, infection of the stent, or pos­ sibly expansion of the native aneurysm.195 Patients with an acute UGI bleed and a history of an aortic aneurysm repair should undergo urgent CT with intrave­ nous contrast, push enteroscopy to evaluate the third portion of the duodenum for compression or blood, as well as to exclude other bleeding sources, and a vascular surgery consultation. CT may show inflammation around the graft and may demonstrate the fistula. Surgical treatment is required to remove the infected graft. Therapeutic endos­ copy plays no role in the management of bleeding from an aortoenteric fistula (see Chapter 36).

VARICES

Figure 19-16.  Endoscopic appearance of the ampulla of Vater and hemobilia. Note the fresh red blood exuding from the ampulla of a patient who earlier that day had undergone a percutaneous liver biopsy.

Variceal hemorrhage is an important cause of UGI bleeding and is discussed in more detail in Chapter 90. Esophageal variceal bleeding related to portal hypertension is the second most common cause of severe UGI bleeding (after peptic ulcer disease). The acute mortality rate with each bleed is approximately 30%, and the long-term survival rate is less than 40% after one year with medical management alone.196 Despite advances in medical therapy, endoscopic hemostasis, and portosystemic shunt procedures, overall

Chapter 19  Gastrointestinal Bleeding long-term survival rates have not improved for patients with variceal bleeding. Liver transplantation, however, can improve survival in selected patients. Survival in nontrans­ planted patients with variceal bleeding is heavily influ­ enced by the severity of underlying liver disease, with poorer survival rates for patients with Child (or Child-Pugh) class C cirrhosis than for those with Child class A or B cirrhosis (see Chapter 90). Bleeding gastric varices are a difficult therapeutic problem because, in contrast to bleeding esophageal varices, most available nonsurgical treatments are ineffective, except when isolated gastric varices are found without accompany­ ing esophageal varices, as occurs with splenic vein throm­ bosis and often in association with pancreatitis or pancreatic cancer. The diagnosis of splenic vein thrombosis can be made with Doppler ultrasound, magnetic resonance imaging, or angiography. Bleeding from gastric varices caused by splenic vein thrombosis is treated by splenectomy.

Medical Management of Acute Variceal Bleeding

Somatostatin and its long-acting analog octreotide cause selective splanchnic vasoconstriction and lower portal pressure, without causing the cardiac complications seen with vasopressin (even in combination with nitroglycerin). Studies have shown mixed results as to whether soma­ tostatin is more effective than placebo in managing vari­ ceal bleeding, but it seems to be at least as effective as vasopressin and much safer. A meta-analysis has shown that vasoactive drugs (e.g., octreotide, somatostatin, terli­ pressin [a long-acting vasopressin analog]) are as effective as sclerotherapy for controlling variceal bleeding and cause fewer adverse events.19 No studies have shown a survival benefit to vasopressin or somatostatin in patients with variceal bleeding. Given the potential ability of octreotide to control acute variceal hemorrhage, its low toxicity, and its availability in the United States (unlike somatostatin and terli­pressin), octreotide appears to be the pharmacologic drug of choice as an adjunct to endoscopic therapy for the treatment of variceal hemorrhage. The dose of octreotide for acute variceal hemorrhage is a 50-µg bolus followed by a continuous infusion of 50  µg/hour for up to five days. Patients with a prolonged prothrombin time that does not correct with fresh frozen plasma may benefit from infusion of human recombinant factor VIIa. In one uncontrolled trial, a single 80-µg/kg dose of recombinant factor VIIa norma­ lized the prothrombin time in all 10 patients within 30 minutes, with immediate control of bleeding in all patients.197 In a large, randomized, placebo-controlled study, administration of recombinant factor VII in addition to endoscopic hemostasis decreased rebleeding rates in patients with Child class B and C cirrhosis who had bled from varices.198 Because recombinant factor VIIa is expen­ sive, its use should be reserved for patients with severe ongoing bleeding and irreversible coagulopathy, pending the results of additional clinical and cost-effectiveness studies. Up to 20% of cirrhotic patients who are hospitalized with GI bleeding have a bacterial infection at the time of admis­ sion to the hospital, and infection develops during the hos­ pitalization in up to 50%. Meta-analyses have suggested that administration of an antibiotic to cirrhotic patients with variceal bleeding is associated with a decrease in the rates of mortality and bacterial infections.199,200 The optimal type and duration of antibiotic is unknown. The most commonly prescribed antibiotics are fluoroquinolones, including oral norfloxacin, 400 mg twice daily, intravenous ciprofloxacin, 400 mg every 12 hours, intravenous levofloxacin, 500 mg

every 24 hours, and intravenous ceftriaxone, 1 g every 24 hours, administered for seven days.

Balloon Tamponade

Balloon tamponade of varices is seldom used now to control variceal bleeding; it may be used to stabilize a patient with massive bleeding prior to definitive therapy. Varices lie in the esophageal and gastric submucosa and are amenable to physical tamponade. Three types of tamponade balloons are available. The Sengstaken-Blakemore tube has gastric and esophageal balloons, with a single aspirating port in the stomach. The Minnesota tube also has gastric and esopha­ geal balloons and has aspiration ports in the esophagus and stomach. The Linton-Nicholas tube has a single large gastric balloon and aspiration ports in the stomach and esophagus. Most reports suggest that balloon tamponade provides initial control of bleeding in 85% to 98% of cases, but vari­ ceal rebleeding recurs soon after the balloon is deflated in 21% to 60% of patients.201 The major problem with tampon­ ade balloons is a 30% rate of serious complications, such as aspiration pneumonia, esophageal rupture, and airway obstruction. Patients should be intubated before placement of a tamponade balloon to minimize the risk of pulmonary complications. Clinical studies have not shown a significant difference in efficacy between vasopressin administration and balloon tamponade.

Endoscopic Sclerotherapy

Endoscopic variceal sclerotherapy involves injecting a sclerosant into or adjacent to esophageal varices. The most commonly used sclerosants are ethanolamine oleate, sodium tetradecyl sulfate, sodium morrhuate, and ethanol. Cyanoacrylate is a glue that when injected into esophageal or gastric varices, effectively stops bleeding, but it is difficult to use and is not approved by the U.S. Food and Drug Administration. Various techniques are used; their common goal is to achieve initial hemostasis and perform sclerotherapy on a weekly basis until all varices are obliterated. Esophageal varices are much more amenable than gastric varices to eradication with endoscopic therapy. Prospective randomized trials have shown mixed results but suggest improved immediate hemostasis and a reduc­ tion in acute rebleeding with sclerotherapy compared with medical therapy alone for bleeding esophageal varices.202-205 Hemostasis can be achieved in 85% to 95% of cases, with a rebleeding rate of 25% to 30%.206 Complications of endo­ scopic variceal sclerotherapy include esophageal ulcers, which can bleed or perforate, esophageal strictures, medi­ astinitis, pleural effusions, aspiration pneumonia, acute respiratory distress syndrome, chest pain, fever, and bacte­ remia and account in part for the use of esophageal variceal band ligation as the preferred endoscopic therapy for vari­ ceal bleeding.

Endoscopic Band Ligation

The technique of endoscopic band ligation is similar to that used for band ligation of internal hemorrhoids. A rubber band is placed over a varix, which subsequently undergoes thrombosis, sloughing, and fibrosis. Prospective, random­ ized, controlled trials have shown that endoscopic band ligation is as effective as sclerotherapy in achieving initial hemostasis and reducing the rate of rebleeding from esopha­ geal varices. Acute hemostasis generally can be achieved in 80% to 85% of cases, with a rebleeding rate of 25% to 30%. Subsequent studies have shown that band ligation is associ­ ated with fewer local complications, especially esophageal strictures, and requires fewer endoscopic treatment sessions

307

308

Section III  Symptoms, Signs, and Biopsychosocial Issues than sclerotherapy.206 A meta-analysis has reported that variceal band ligation reduces the rates of rebleeding, overall mortality, and death from bleeding compared with sclero­ therapy.207 Band ligation, however, may be more technically difficult to perform than sclerotherapy during active vari­ ceal bleeding. Devices used for band ligation allow up to 10 bands to be placed, without the need to remove the endo­ scope to reload the banding device. The strategy is to control active bleeding and place two bands on each esophageal variceal column, one distally near the gastroesophageal junction and another 4 to 6 cm proximally.

Transjugular Intrahepatic Portosystemic Shunt

Placement of a transjugular intrahepatic portosystemic shunt (TIPS) is an interventional radiologic procedure in which an expandable metal stent is placed via percutaneous insertion between the hepatic and portal veins, thereby cre­ ating an intrahepatic portosystemic shunt. TIPS is effective for the short-term control of bleeding gastroesophageal varices.208,209 Initially envisioned as a bridge to liver trans­ plantation, it has been used with increased frequency in nontransplantation situations. Randomized trials that have compared TIPS with endoscopic sclero­therapy suggest that TIPS is more effective for the long-term prevention of rebleeding.210 The main problems with TIPS are a rate of shunt occlusion of up to 80% (less with polytetrafluoro-eth­ ylene-coated stents) within one year and the development of new or worsening hepatic ence­phalopathy in approximately 20% of patients.211 TIPS does not prolong the survival of patients with variceal bleeding compared with endoscopic treatment. In the management of acute variceal bleeding, TIPS is reserved for patients who fail endoscopic treatment.

Portosystemic Shunt Surgery

A variety of portosystemic shunt operations can be per­ formed to reduce portal venous pressure. When compared with sclerotherapy, surgical shunts decrease the rebleeding rate significantly but do not improve survival.206,212-215 Some groups have suggested that survival can be improved by using a combination of endoscopic sclerotherapy and surgi­ cal portosystemic shunt rescue for those who rebleed despite sclerotherapy. Surgical shunts may be associated with hepatic encephalopathy and can make future liver trans­ plantation technically more difficult, but they have an advantage over endoscopic variceal therapy in reducing portal hypertension and treating gastric variceal bleeding. Surgical shunts are considered for selected patients who have failed endoscopic therapy and who are not expected to become candidates for liver transplantation (see Chapters 90 and 95).

LOWER GASTROINTESTINAL BLEEDING LGI bleeding generally signifies bleeding from the colon or anorectum. The annual incidence of LGI bleeding is approx­ imately 20 cases/100,000 population, with an increased risk in older adults.216 The rate of hospitalization for LGI bleed­ ing is lower than that for UGI bleeding. Most patients are older than 70 years. Patients usually present with painless hematochezia and a decrease in their hematocrit value but without orthostasis. If orthostasis is associated with hema­ tochezia, a briskly bleeding UGI source should be excluded (see earlier). Severe painless hematochezia results from a foregut source in approximately 15% of patients.217 The sites of origin within the GI tract of severe hematochezia at UCLA CURE are shown in Figure 19-17.

Small intestine 5% (n = 30) No source identified 3% (n = 18)

Colon 75% (n = 486)

UGI tract 17% (n = 113) Figure 19-17.  The frequencies of the sources of severe hematochezia in patients seen at the University of California, Los Angeles, Center for Ulcer Research Education. Note that in most cases (75%), severe hematochezia is from the colon, 17% is from an upper gastrointestinal (UGI) (esoph­agus, stomach, or duodenum) source, and 5% is from a small intestinal source.

Table 19-9  Causes of Severe Hematochezia (%) Study LESION Diverticulosis Colon cancer or polyps Colitis Ischemic colitis IBD Noninfectious colitis Infectious colitis Angioectasia Postpolypectomy Rectal ulcer Hemorrhoids Anorectal source (unspecified) Radiation colitis Other Unknown

reference 218

reference 219

UCLA CURE

30 18

33 21

30 6

17 NP NP NP

17 7 4 5

21 12 9 0

NP 7 6 NP NP 4

1 6 NP 1 20 3

0 3 8 6 14 0

0 8 16

0.5 3 0

3 6 0

CURE, Center for Ulcer Research Education; IBD, inflammatory bowel disease; NP, not provided; UCLA, University of California, Los Angeles.

Patients with LGI bleeding initially should be resuscitated medically. After they have been stabilized, they generally should undergo colonoscopy after a polyethylene glycol purge.22 In early reports, urgent colonoscopy resulted in a diagnosis in approximately 70% of cases.40,218 In more recent reports, however, the combination of urgent colonoscopy and, if necessary, push enteroscopy, anoscopy, and capsule endoscopy has resulted in a diagnosis in 95% of cases (see Fig. 19-4).217 The most common causes of LGI bleeding are shown in Table 19-9. Diverticulosis is generally the most common cause of acute LGI bleeding, occurring in approximately 30% of cases. Colonic polyps or cancer, colitis, and anorec­ tal disorders each account for approximately 20% of cases.219 In most cases, acute LGI bleeding will stop spontane­ ously, thereby allowing nonurgent diagnosis and treatment. For patients with ongoing or recurrent hematochezia, urgent diagnosis and treatment are required to control the bleeding. In a large series of patients at the UCLA Medical Center and Wadsworth Veterans Administration Hospital, 64% of patients with severe hematochezia required a therapeutic intervention to control continued bleeding or rebleeding22; 39% underwent endoscopic hemostasis, 1% underwent angiographic embolization, and 24% underwent surgery.

Chapter 19  Gastrointestinal Bleeding Table 19-10  Clinical Predictors of Severe Acute Lower Gastrointestinal Bleeding*

TOTAL risk POINTS† 0 1-3 ≥4

FREQUENCY (%)

RISK OF SEVERE BLEEDING (%)

NEED FOR SURGERY

MORTALITY rate (%)

HOSPITAL DAYS

mean NUMBER OF UNITS TRANSFUSED (PACKED RED BLOOD CELLS)

6 75 19

6 43 79

0 1.5 7.7

0 2.9 9.6

2.8 3.1 4.6

0 1 3

*Severe lower gastrointestinal bleeding is defined as continued bleeding within the first 24 hr of hospitalization (transfusion of two or more units of packed red blood cells and/or hematocrit value drop of 20% or more) and/or recurrent bleeding after 24 hr of stability (need for additional transfusions, further hematocrit value decrease of 20% or more, or readmission to the hospital for lower gastrointestinal bleed within 1 wk of discharge). † Risk factors (1 point each): aspirin use; more than two comorbid illnesses; heart rate ≥ 100 beats/min; nontender abdominal examination; rectal bleeding within the first 4 hr of evaluation; syncope; systolic blood pressure ≤ 115 mm Hg. Data from Strate LL, Saltzman JR, Ookubo R, et al. Validation of a clinical prediction rule for severe acute lower intestinal bleeding. Am J Gastroenterol 2005; 100:1821-7.

RISK FACTORS AND RISK STRATIFICATION

Nonselective NSAIDs increase the risk of LGI bleeding com­ pared with placebo.220,221 The main risk factors for NSAIDassociated LGI bleeding appears to be an age of 65 years or older and prior history of LGI bleeding.222 It is uncertain whether the use of long-term selective COX-2 inhibitors is associated with a lower risk of LGI bleeding than nonselec­ tive COX-2 inhibitors. Table 19-10 shows clinical factors that are predictive of severe LGI bleeding (defined as continued bleeding within the first 24 hours of hospitalization, with a transfusion requirement of at least two units of packed red blood cells or a decrease in the hematocrit value of 20% or more) or recurrent bleeding after 24 hours of stability (defined as the need for additional transfusions, a further decrease in the hematocrit value of at least 20%, or readmission for LGI bleeding within one week of discharge). Predictive factors include tachycardia, hypotension, syncope, a nontender abdomen, rectal bleeding on presentation, aspirin use, and more than two comorbid illnesses.223,224 Such a prognostic scoring system could identify patients at the highest risk for severe LGI bleeding, who account for 19% of patients with LGI bleeding and who might benefit most from urgent co­ lonoscopy, although this benefit remains unproved. A single-institution case series of 94 patients admitted for LGI bleeding has identified similar risk factors.225 This study found that 39% of all cases of LGI bleeding requiring hos­ pitalization were severe, as defined by the passage of red blood after the patient left the emergency department, with hypotension or tachycardia, or the need for a transfusion of more than two units of packed red blood cells during the hospitalization. Independent risk factors for severe LGI bleeding were an initial hematocrit value of 35% or lower, abnormal vital signs (a systolic blood pressure less than 100 mm Hg or a heart rate higher than 100 beats/min) on admission, and gross blood on initial rectal examination. Artificial neural networks also have been used to develop prediction models for severe LGI bleeding.226,227 The main problem with the use of such networks from a clinical point of view is that a large number of variables need to be entered into a computer program for analysis, thus limiting wide­ spread routine clinical use.

MORTALITY

A large United States database study of 227,000 patients with a discharge diagnosis of LGI bleeding in 2002 reported an overall mortality rate of 3.9%.219 Multivariate analysis has found that independent predictors of in-hospital mortal­ ity are age older than 70 years, intestinal ischemia, at least two comorbid illnesses, bleeding during a hospitalization

for an unrelated condition, coagulopathy, hypovolemia, transfusion of packed red blood cells, and male gender. Colorectal polyps and hemorrhoids were associated with a lower mortality risk. The low risk of death from LGI bleed­ ing identified in this study is consistent with data from smaller series such as those from Kaiser San Diego (2.4%) and the University of California, San Francisco (3.2%).216,225 The Kaiser study also found an increased risk of death with in-hospital LGI bleeding.

DIAGNOSTIC AND THERAPEUTIC APPROACH

Patients with hematochezia should undergo the same careful history taking, physical examination, and laboratory testing described earlier for the general approach to the patient with acute GI bleeding (see Table 19-1). The history should focus specifically on identifying sources of LGI bleeding. Diverticular bleeding should be suspected in patients with painless severe acute hematochezia and a history of diverticulosis. Ischemic colitis may be suspected in those with painful acute hematochezia with mild abdom­ inal discomfort. A recent polypectomy suggests a postpol­ ypectomy bleed. A history of recent antibiotic use or inflammatory bowel disease and bloody diarrhea suggests colitis. The acute onset of bright red blood that drips from the anus suggests a hemorrhoidal bleed. Prior radiation sug­ gests radiation proctitis. Patients should be medically resuscitated. Because LGI bleeding is generally less severe than UGI bleeding, blood transfusions may not be needed. Most patients should undergo initial evaluation with colonoscopy after bowel preparation, although in selected cases anoscopy or flexible sigmoidoscopy without any bowel cleansing or after an enema may be performed. Other diagnostic tests, including radionuclide bleeding scans or angiography, may be used in selected cases or when colonoscopy fails to detect a source of bleeding.

Anoscopy

Anoscopy can be useful for patients in whom actively bleed­ ing internal hemorrhoids or other anorectal disorders (e.g., fissures, fistulas, proctitis) are suspected and allows imme­ diate treatment with rubber band ligation (see Chapter 125). Most patients, however, especially if older than 50 years, will also require colonoscopy, at least electively, to evaluate the remainder of the colon.

Flexible Sigmoidoscopy

Flexible sigmoidoscopy can evaluate the rectum and left side of the colon for a bleeding site and can be performed without a standard colonoscopy bowel preparation.

309

310

Section III  Symptoms, Signs, and Biopsychosocial Issues Although not adequate for evaluation of the anal canal, flex­ ible sigmoidoscopy alone will result in a diagnosis in approximately 9% of cases.228 If the distal colon can be adequately cleansed with enemas, an urgent flexible sig­ moidoscopy can be useful for patients suspected of having a solitary rectal ulcer, ulcerative colitis, radiation proctitis, ischemic colitis, postpolypectomy bleeding (in the rectosig­ moid), or internal hemorrhoids (see Chapters 39, 112, 115, 122, 124, and 125). Therapeutic hemostasis can be provided with injection therapy, hemoclip placement, band ligation, and MPEC. Monopolar electrocautery (e.g., argon plasma coagulation, snare polypectomy, or hot biopsy forceps) should not be used if a bowel preparation has not been administered to avoid the risk of ignited flammable colonic gas (see Chapter 16).

patient is having hematochezia with abdominal pain. One study from France reported that CT accurately identified 17 of 19 LGI bleeding sites, including diverticula, tumors, angi­ omas, and varices.231 Multidetector CT has been shown to be more accurate than technetium-tagged red blood cell scanning in patients with LGI bleeding.232 CT colonography is being used increasingly to screen persons for colonic polyps and cancer and may be of some benefit in patients with LGI bleeding. CT colonography detects large polyps (>1 cm) or cancers with a sensitivity rate of 90%.233 Faster multidetector scanners also allow CT angiography to be performed, as well as evaluation of the small bowel. This capability could allow detection of masses and vascular lesions and is a potential advantage of CT angiography over other radiologic imaging techniques.

Nuclear Scintigraphy

Colonoscopy

Nuclear scintigraphy involves injecting a radiolabeled sub­ stance into the patient’s bloodstream and performing serial scintigraphy to detect focal collections of radiolabeled mate­ rial (see earlier). This technique has been reported to detect bleeding at a rate as low as 0.04 mL/min,38 with an overall positive diagnostic rate of approximately 45% and a 78% accuracy rate for localizing the true bleeding site.40 The disadvantages of radionuclide bleeding scans are that delayed scans may be misleading, and determining the specific cause of bleeding often depends on endoscopy or surgery. False-positive results are most likely to occur when transit of luminal blood is rapid, so that radiolabeled blood is detected in the colon, even though it originated in the UGI tract. Radionuclide scanning may be helpful in cases of obscure GI bleeding (see later) or prior to angiography to help localize a lesion, particularly if an early scan (e.g., 30 minutes to four hours after injection of the radiolabeled material) is positive for red blood cell extravasation.

Angiography

Angiography is most likely to detect a site of bleeding when the rate of arterial bleeding is at least 0.5 mL/min.35 The diagnostic yield depends on patient selection, the timing of the procedure, and the skill of the angiographer, with posi­ tive results in 12% to 69% of cases. An advantage of angi­ ography is that embolization can be performed to control some bleeding lesions. Major complications, however, occur in 3% of cases and include bowel ischemia, hema­ toma formation, femoral artery thrombosis, contrast dye reactions, acute kidney injury, and transient ischemic attacks.37 Other disadvantages of angiography are the absence of active bleeding in most patients at the time of angiography, expense of the test, and inability to determine the specific lesion responsible for bleeding in many cases. A small retrospective case series of 11 patients with colonic bleeding who underwent angiographic emboliza­ tion reported that the bleeding ceased in 10, mesenteric ischemia developed in 7, and 6 died.229 Another study of 65 patients with acute LGI bleeding who did not undergo colonoscopy as a first diagnostic step found that diagnostic angiography provided little additional clinical information because the bleeding stopped spontaneously in most patients. Moreover, angiography did not help guide sub­ sequent surgery and was associated with a complication rate of 11%.230

Computed Tomography and Computed Tomography Colonography

Multidetector CT can identify abnormalities in the colon that could be a source of bleeding, such as diverticulosis, colitis, masses, and varices. CT is often performed if the

Urgent colonoscopy following a rapid bowel purge has been shown to be safe, provide important diagnostic information, and allow therapeutic intervention.22,217 Patients usually ingest 4 to 8 L of polyethylene glycol solution orally or via a nasogastric tube over four to six hours until the rectal effluent is clear of stool, blood, and clots. Metoclopramide, 10 mg, may be given intravenously before the purge and repeated every three to four hours to facilitate gastric empty­ ing and reduce nausea. Sodium phosphate bowel prepara­ tions probably should be avoided in patients with suspected LGI bleeding because of potential risks of the high phos­ phate and sodium loads. Urgent colonoscopy for LGI bleeding generally is per­ formed 6 to 36 hours after the patient is admitted to the hospital. Because most bleeding stops spontaneously, colo­ noscopy often is performed semielectively on the day after initial hospitalization to allow the patient to receive blood transfusions and the bowel preparation on the first day of hospitalization. The overall rate of detecting a presumed or definite cause of LGI bleeding by colonoscopy ranges from 48% to 90%, with an average of 68%, based on a review of 13 studies.40 The problem with interpreting these data, however, is that making a definite diagnosis of the cause of the bleeding is often not possible unless a bleeding stigma such as active bleeding, a visible vessel, an adherent clot, mucosal friabil­ ity or ulceration, or the presence of fresh blood limited to a specific segment of the colon is seen. The optimal time for performing urgent bowel prepara­ tion and colonoscopy is unknown. Theoretically, the sooner endoscopy is performed, the higher the likelihood of finding a lesion, such as a bleeding diverticulum or polyp stalk, that might be amenable to endoscopic hemostasis. A retrospec­ tive study from the Mayo Clinic, however, has suggested that in patients with diverticular bleeding, the timing of endoscopy (0 to 12 hours, 12 to 24 hours, or more than 24 hours after admission) is not associated significantly with the finding of active bleeding or other stigmata that would prompt colonoscopic hemostasis.234 Early colonos­ copy (soon after admission) has been associated with a shorter length of hospitalization, principally because of improved diagnostic yield rather than therapeutic interven­ tions.235 A consensus on a single approach to patients with severe hematochezia has not been reached, and the approach used depends on local resources and expertise. In large centers, the approach detailed in Figure 19-4 is recom­ mended. With use of an urgent endoscopic approach for diagnosis and treatment, the diagnostic yield of definitive and presumptive bleeding sites is more than 90%, and the estimated direct costs are significantly less than the costs associated with an elective evaluation.23

Chapter 19  Gastrointestinal Bleeding Barium Enema

Emergency barium enema has no role in patients with LGI bleeding. This test is rarely diagnostic because it cannot demonstrate vascular lesions and may be misleading if only diverticula are seen. It fails to detect 50% of polyps larger than 10 mm. In addition, the barium contrast liquid can make urgent colonoscopy more difficult by impairing visu­ alization.236 Subsequent colonoscopy is needed for any sus­ picious lesions seen on barium enema or for lesions that require therapy.

Role of Surgery

Surgical management is rarely needed in patients with LGI bleeding because most bleeding is self-limited or easily managed with medical or endoscopic therapy. The main indications for surgery are malignancy, diffuse bleeding that fails to cease with medical therapy (as in ischemic or ulcer­ ative colitis), and recurrent bleeding from a diverticulum. Therefore, most stable patients can be managed on a medical service rather than a surgical service.

CAUSES AND MANAGEMENT

Visualizing active bleeding during colonoscopy is not always possible. Unlike the case with angiography or nuclear red blood cell scanning, however, colonoscopy permits identification of stigmata of recent hemorrhage (visible vessels, clots, or spots) and provides information on the location of the lesion and on risk stratification. The earlier the colonoscopy is carried out, the higher the chance of detecting an actively bleeding lesion or stigmata of recent hemorrhage. A definite diagnosis of a bleeding lesion can usually be made if active bleeding, a visible vessel, or a clot is seen. A presumptive diagnosis of the cause of bleeding can be made if a lesion that is a potential cause of bleeding is seen and no other possible sources are identified by anos­ copy, full colonoscopy with intubation of the terminal ileum, and, in some cases, push enteroscopy.23,217

Diverticulosis

Colonic diverticula are herniations of colonic mucosa and submucosa through the muscular layers of the colon (see Chapter 117). Histopathologically, diverticula in the colon are actually pseudodiverticula, because they do not contain all layers of the colonic wall. Diverticula form when colonic tissue is pushed out by intraluminal pressure at points of entry of the small arteries (vasa recta), where they penetrate the circular muscle layer of the colonic wall. The entry points of the vasa recta are areas of relative weakness through which the mucosa and submucosa can herniate when intraluminal pressure is increased. Diverticula vary in diameter from a few milliliters to several centimeters and are located most commonly in the left colon. Most colonic diverticula are asymptomatic and remain uncomplicated. Bleeding may occur from vessels at the neck or base of a diverticulum.237 In our experience with definitive diverticu­ lar hemorrhage (see later), bleeding was from the base in 52% and from the neck in 48% of diverticula.217 Diverticula are common in Western countries, with a fre­ quency of 50% in older adults.238 By contrast, diverticula are found in fewer than 1% of continental African and Asian populations.239 It has been hypothesized that the regional differences in prevalence rates can be explained by the low amount of dietary fiber in Western diets. Presum­ ably, a low-fiber diet results in less stool content, longer fecal transit times, increased colonic muscle contraction and, ultimately, increased intraluminal pressure, which results in the formation of propulsion diverticula. Further­ more, diverticula increase in frequency with advanced age,

possibly because of weakening of the colonic wall and muscle tone. Diverticular bleeding develops in an estimated 3% to 5% of patients with diverticulosis.240 Although most diverticula are in the left colon, several series have sug­ gested that diverticula in the right colon are more likely to bleed.237,240-242 Two thirds of definitive diverticular bleeds (with stigmata of hemorrhage) emanate from the region of the splenic flexure of the colon or proximally.217 Diverticular hemorrhage should be classified carefully based on findings at colonoscopy, angiography, or surgery,23 particularly in the case of older patients with severe hema­ tochezia who are likely to have colonic diverticulosis. Definitive diverticular hemorrhage is diagnosed when stig­ mata of recent hemorrhage (e.g., active bleeding, visible vessel, adherent clot) are seen on colonoscopy or active bleeding is demonstrated on angiography or nuclear red blood cell scanning, with later confirmation of a diverticu­ lum in that location as the source of bleeding by colonos­ copy or surgery. Presumptive diverticular hemorrhage is diagnosed when colonoscopy reveals diverticulosis without stigmata and no other significant lesions are seen in the colon and by anoscopy, terminal ileum examination, and push enteroscopy. The term incidental diverticulosis is used when another lesion is identified as the cause of hema­ tochezia and colonic diverticulosis is evident. In a large, prospective cohort study in which the management algo­ rithm shown in Figure 19-4 was used in our institutions to classify patients with hematochezia, colonic diverticulosis was incidental in 52%, presumptive diverticular hemor­ rhage occurred in 31%, and definitive diverticular hemor­ rhage was established in 17% of cases.217 Patients with diverticular bleeding typically are older, have been taking aspirin or other NSAIDs, and present with painless hematochezia.243 In at least 75% of patients with diverticular bleeding, the bleeding stops spontaneously, and these patients require transfusion of fewer than four units of packed red blood cells. In one surgical series, surgi­ cal segmental colonic resection was needed in 60% of patients, most of whom had had continued bleeding despite trans­fusion of four units of blood.241 Patients who under­ went resection for a bleeding diverticulum had a rebleeding rate of 4%. Among patients who stopped bleeding spontane­ ously, the rebleeding rate from colonic diverticulosis has been reported to range from 25% to 38% over the next four years, with most patients having mild rebleeding.216,241 These data, however, are not based on colonoscopic docu­ mentation of diverticular bleeding, and the actual rate of rebleeding appears to be lower. In a large, prospective, cohort study of patients with documented colonic diverticu­ lar hemorrhage (definitive or presumptive) by our group, the overall rate of rebleeding was 18% in four years—9% from recurrent diverticular hemorrhage and 9% from other GI sources.217 Occasionally, urgent colonoscopy reveals a stigma of recent bleeding, such as active bleeding, a visible vessel, a clot, or blood limited to one segment of the colon. As noted, earlier colonoscopy for LGI bleeding is likely to result in a greater frequency of finding stigmata of recent diverticular bleeding, although a small case series study from the Mayo Clinic did not find any difference in the rate of detection of these stigmata whether colonoscopy was performed between 0 and 12 hours, 12 and 24 hours, or more than 24 hours from the time of hospital admission.234 Stratifying the risk of diverticular rebleeding by applying the same endoscopic stigmata used in high-risk peptic ulcer bleeding (active bleeding, visible vessel, and clot) has been attempted, but the natural history associated with each of these stigmata is unknown (Fig. 19-18). The pigmented pro­

311

312

Section III  Symptoms, Signs, and Biopsychosocial Issues

A

B

C

Figure 19-18.  Endoscopic stigmata of recent colonic diverticular bleeding. A, Active bleeding (arrow). B, Adherent clot (arrow). C, Nonbleeding visible vessel (arrow).

tuberance found on the edge of some diverticula is usually clot at the edge of a ruptured blood vessel on histopathol­ ogy.244 The UCLA CURE group found that among 17 patients with stigmata of recent diverticular hemorrhage (active bleeding in 6, visible vessel in 4, and adherent clot in 7), the rates of rebleeding (53%) and emergency surgery (35%) were high.23 Endoscopic Hemostasis Colonoscopic hemostasis of actively bleeding diverticula has been reported using MPEC, epinephrine injection, hemoclips, fibrin glue, or combinations of epinephrine and MPEC or hemoclips.23,244-249 If fresh red blood is seen in a focal segment of colon, that segment should be irrigated vigorously with water to remove the blood and detect the underlying bleeding site. If the bleeding is coming from the edge of a diverticulum or a pigmented protuberance is seen on the edge, a sclerotherapy needle can be used to inject epinephrine (diluted 1 : 20,000 in saline) in 1-mL aliquots submucosally into four quadrants around the bleeding site. Subsequently, MPEC at a low power setting (10 to 15 W) and light pressure can be carried out for a one-second pulse duration to cauterize the diverticular edge and stop bleeding or flatten the visible vessel, or hemoclips can be applied. A nonbleeding adherent clot can be injected with 1 : 20,000 epinephrine into four quadrants, 1 mL/quadrant, after which the clot can be removed piecemeal by guillotining it with a cold polyp snare until it extends 3 mm above the diverticulum. The underlying stigma is treated with MPEC or hemoclips (see earlier). After endoscopic hemostasis of a bleeding diverticulum is completed, a permanent submucosal tattoo should be placed around the lesion to allow identification of the site in case colonoscopy or surgery is repeated for recurrent bleeding. After colonoscopic hemostasis, patients are told to avoid aspirin and other NSAIDs and to take a daily fiber supplement on a long-term basis. In 2000, Jensen and the UCLA CURE group published their results on urgent colonoscopy for the diagnosis and treatment of severe diverticular hemorrhage23 and reported that 20% of patients with severe hematochezia had endo­ scopic stigmata, suggesting a definitive diverticular bleed. This group of patients, who underwent colonoscopic hemostasis, had a rebleeding rate of 0% and an emergency hemicolectomy rate of 0% compared with 53% and 35%, respectively, in a historical control group of patients who had high-risk stigmata but did not undergo colonoscopic hemostasis. No rebleeding had occurred after three years of follow-up in the patients who underwent colonoscopic

hemostasis. Other studies also have suggested that endo­ scopic hemostasis is effective for immediate diverticular bleeding, although subsequent rebleeding and the need for surgery have been reported in some cases.250 Angiography and Surgery Angiographic embolization can be performed in selected cases of diverticular bleeding, but with a risk of bowel infarction, contrast reactions, and acute kidney injury. One study found that routine angiography prior to surgical resection is not helpful in reducing the overall risk of complications.230 Surgical resection for diverticular bleeding is rarely needed and is reserved for recurrent bleeding. The decision to operate is best guided by colonoscopic, angiographic, or nuclear medicine studies that demonstrate the likely segment of colon from which the bleeding is emanating and by the presence of medical comorbidities. Diverticular bleeding is usually mild, and the risks of surgical com­ plications are increased in older patients. Blind subtotal colectomy, often performed in the past when a definite bleeding site could not be identified, should be avoided, if possible.

Colitis

The term colitis refers to any form of inflammation of the colon. Severe LGI bleeding may be caused by ischemic colitis, inflammatory bowel disease, or possibly infectious colitis. Ischemic colitis can present as painless or painful hema­ tochezia with mild left-sided abdominal discomfort. The painless subtype usually results from mucosal hypoxia, and is thought to be caused by hypoperfusion of the intramural vessels of the intestinal wall, rather than by large-vessel occlusion, which is often painful and clinically more severe and has worse outcomes. The incidence of ischemic colitis is estimated to be 4.5 to 44 cases/100,000 person-years.251 Most cases do not have a recognizable cause. Risk factors associated with ischemic colitis have been reported to include older age, shock, cardiovascular surgery, congestive heart failure, chronic obstructive pulmonary disease, ileos­ tomy, colon cancer, abdominal surgery, irritable bowel syn­ drome, constipation, laxative use, oral contraceptive use, and use of an H2 receptor antagonist.251-254 The superior mesenteric artery supplies blood to the right colon (cecum, ascending colon, hepatic flexure, proximal transverse colon, and midtransverse colon) whereas the inferior mesenteric artery supplies blood to the left colon (distal transverse colon, splenic flexure, descending colon, sigmoid colon,

Chapter 19  Gastrointestinal Bleeding and rectum). The colon has an abundant blood supply, but the watershed area between the superior and inferior mes­ enteric arteries has the fewest collateral vessels and is at most risk for ischemia. The colon normally receives 10% to 35% of cardiac output, and ischemia can occur if blood flow decreases by more than 50%. Although ischemia is most likely to occur in the watershed area of the splenic flexure, it can occur anywhere in the colon.255 The diagnosis of ischemia is usually made by colonos­ copy and is suspected by the demonstration of thumbprint­ ing on plain film radiographs or colonic wall thickening on CT scans. The colonoscopic appearance of the mucosa includes erythema, friability, and exudate. Mucosal biopsy specimens may suggest ischemic changes but generally are used to exclude infectious colitis or Crohn’s disease. Ische­ mic colitis generally resolves in a few days and generally does not require colonoscopic hemostasis or antibiotic therapy. In the UCLA CURE experience, approximately 10% of patients with ischemic colitis and severe hemato­ chezia had a focal ulcer with a major stigma of hemorrhage on urgent colonoscopy. The recommended treatment in these cases is epinephrine injection and hemoclipping (see Chapter 114). In a large retrospective series from Kaiser, no episodes of rebleeding from ischemic colitis occurred over a four-year follow-up period.216 On the other hand, inpa­ tients in whom ischemic colitis develops or those with large-vessel mesenteric ischemia usually have worse out­ comes, including higher rates of rebleeding, perforation, need for surgery, and death. Inflammatory bowel disease that involves the colon can rarely cause severe acute LGI bleeding. In a case series from the Mayo Clinic, most of these patients had Crohn’s disease, and most were successfully treated medically.256 Three of the 31 patients in this series underwent endoscopic therapy with epinephrine injection alone or with MPEC for an adherent clot or an oozing ulcer. These 3 patients had no rebleeding, but 23% of the other patients had rebleeding at a median of three days (range, 1 to 75 days) after the initial bleed. Thirty-nine percent of the patients with severe bleeding eventually required surgery (see Chapters 111 and 112). Infectious colitis should be excluded in any patient with severe LGI bleeding and colitis. LGI bleeding can occur with infection caused by Campylobacter jejuni, Salmonella, Shigella, enterohemorrhagic Escherichia coli (O157:H7), cytomegalovirus, or Clostridium difficile. Significant blood loss is rare except in patients with severe coagulopathy. Diagnosis is made by stool cultures and flexible sigmoidos­ copy or colonoscopy. Treatment is with medical manage­ ment; the use of antibiotics depends on the causative organism. Endoscopic management generally has no role in infectious colitis (see Chapter 107).

Postpolypectomy Bleeding

Bleeding occurs after approximately 1% of colonoscopic polypectomies. The bleeding occurs most commonly five to seven days after polypectomy but can occur from 1 to 14 days after the procedure; it generally is self-limited and mild to moderate, with 50% to 75% of patients requiring blood transfusions.257-260 Reported risk factors for postpoly­ pectomy bleeding include a large polyp size (more than 2 cm), thick stalk, sessile type, location in the right colon, use of warfarin or heparin, and use of aspirin or other NSAID. In most cases of delayed postpolypectomy bleeding, an ulceration at the site of the polypectomy is found on colonoscopy at the time of bleeding. In patients with severe bleeding, stigmata of recent hemorrhage may be found in the ulceration.261 Figure 19-19 illustrates a postpolypectomy

Figure 19-19.  Endoscopic appearance of post-polypectomy bleeding in the colon. Bleeding occurred seven days after a snare polypectomy for a large pedunculated polyp. Note the nonbleeding visible vessel (arrow) in the ulcerated polypectomy site.

ulcer with evidence of recent hemorrhage. Endoscopic man­ agement techniques for delayed postpolypectomy ulcer bleeding depend on the stigma found and are similar to those used for peptic ulcer hemorrhage, including epineph­ rine injection, thermal coagulation, hemoclip placement, and combination therapy.

Colon Polyps and Cancer

Patients with colon polyps and cancer can present with acute hematochezia. Often, these patients have a microcytic iron deficiency anemia consistent with slow GI blood loss (see later) before more overt bleeding occurs. At colonos­ copy, epinephrine can be injected into the lesion to slow active bleeding, and hemoclips can be applied to treat stig­ mata of hemorrhage on ulcerated lesions that cannot be resected endoscopically. When possible, colon polyps can be removed to stop bleeding. Surgical resection is usually required to prevent rebleeding from a large ulcerated sessile lesion (see Chapters 122 and 123).

Radiation Proctitis

Radiation proctitis usually causes mild chronic hematoche­ zia but occasionally can cause acute severe LGI bleeding. Ionizing radiation can cause acute and chronic damage to the normal colon and rectum when used to treat gyneco­ logic, prostatic, bladder, or rectal tumors. Acute self-limited diarrhea, tenesmus, abdominal cramping, and rarely bleed­ ing will develop for a few weeks in approximately 75% of patients who have received a radiation dose of 4000 cGy. Chronic radiation effects occur 6 to 18 months after comple­ tion of treatment and manifest as bright red blood with bowel movements. Bowel injury resulting from chronic radiation is related to vascular damage, with subsequent mucosal ischemia, thickening, and ulceration. Much of this damage is thought to result from chronic hypoxic ischemia and oxidative stress. Flexible sigmoidoscopy or colonoscopy reveals telangiec­ tasias, friability, and sometimes ulceration in the rectum (Fig. 19-20). At times, active bleeding is noted, and usually numerous telangiectasias are seen. Internal hemorrhoids are often seen as well and frequently are considered in the dif­ ferential diagnosis of the rectal bleeding. Treatment initially focuses on avoidance of aspirin and other NSAIDs, consumption of a high-fiber diet, and iron

313

314

Section III  Symptoms, Signs, and Biopsychosocial Issues medical therapy, those with severe bleeding are likely to require some form of endoscopic or surgical treatment (see Chapter 125).

Anal Fissures

Patients with an anal fissure usually present with extremely painful bowel movements but can present with hematoche­ zia. Generally, the hematochezia is mild and is noticed with wiping; rarely, hematochezia is moderate to severe. Treat­ ment focuses on resolving the anal fissure, rather than using specific hemostasis techniques. A topical calcium channel blocker (such as 2% topical diltiazem cream) along with fiber supplementation, stool softeners, and sitz baths will heal most anal fissures (see Chapter 125).

Rectal Varices

Figure 19-20.  Endoscopic appearance of radiation proctitis. Note the diffuse oozing and telangiectasias.

supplementation if the patient is anemic. Medical therapy with topical or oral 5-aminosalicylic acid (mesalamine), sucralfate, or glucocorticoids can be tried but is not gener­ ally effective.262 Thermal therapy can be successful, includ­ ing repeated treatments with MPEC or argon plasma coagulation.263 Topical formalin applied directly to the rectal mucosa can reduce bleeding,264 as can the use of hyperbaric oxygen.265 Antioxidant vitamins, such as vita­ mins E and C, also have been reported to decrease bleeding from chronic radiation proctitis (see Chapter 39).266

Colonic Angioectasias

Colonic bleeding from angioectasias, an important cause of LGI bleeding, is discussed in the section on small bowel and obscure bleeding. When angioectasias are the cause of bleeding in the colon, they are often multiple, making endo­ scopic hemostasis a challenge (see Chapter 36).

Internal Hemorrhoids

Hemorrhoidal bleeding is characterized by bright red blood per rectum that can coat the outside of the stool, may drip into the toilet bowel, can be seen on tissue after wiping, and often appears as a large amount of fresh blood in the toilet. Usually, bleeding is mild, intermittent, and self-limited but occasionally, severe, transfusion-requiring bleeding may occur from hemorrhoids.267 In a large study of patients with hematochezia discharged from the hospital, 20% were thought to have had bleeding from hemorrhoids.219 In the UCLA CURE series of patients hospitalized for severe hema­ tochezia (see earlier), internal hemorrhoids were the second most common cause (see Table 19-9).217 Hemorrhoids were documented by urgent anoscopy and colonoscopy after a colonic preparation. The diagnosis can be made with anoscopy, sigmoidoscopy, or colonoscopy, especially if per­ formed while bleeding is ongoing. The treatment of hemorrhoids usually starts with medical therapy consisting of fiber supplementation, stool softeners, lubricant rectal suppositories (with or without glucocor­ ticoids), and warm sitz baths. Anoscopic therapy can also be used and includes injection sclerotherapy, rubber band ligation, cryosurgery, infrared photocoagulation, MPEC, and direct current electrocoagulation. Although most patients with mild hemorrhoidal bleeding respond to

Ectopic varices may develop in the rectal mucosa between the superior hemorrhoidal veins (portal circulation) and middle and inferior hemorrhoidal veins (systemic circula­ tion) in patients with portal hypertension. On sigmoidos­ copy, rectal varices are seen during retroflexion as vascular structures located several centimeters above the dentate line and extending into the rectum; they are distinct from inter­ nal hemorrhoids. The frequency of rectal varices increases with the degree of portal hypertension. Approximately 60% of patients with a history of bleeding esophageal varices have rectal varices. The treatment of rectal varices is similar to that for esophageal varices, with sclerotherapy, band liga­ tion, or a portosystemic shunt (see Chapter 90).268-270

Rectal Dieulafoy’s Lesions

Dieulafoy’s lesions are large submucosal arteries without overlying mucosal ulceration that can cause massive bleed­ ing. They can occur anywhere in the GI tract, although usually in the stomach (see earlier). Bleeding Dieulafoy’s lesions in the rectum, which have been treated successfully with endoscopic hemostasis, have been described in several reports.271,272

Rectal Ulcers

Several case series have described seriously ill hospitalized patients with the sudden onset of painless severe hemato­ chezia from a solitary or multiple rectal ulcers located 3 to 10 cm above the dentate line. In one series of 19 cases from Taiwan, 2.7% of patients evaluated for severe hematochezia were diagnosed with acute hemorrhagic rectal ulcer syn­ drome.273 The patients had a mean age of 71 years and had been hospitalized for other medical problems from 3 to 14 days (average 7.5 days) prior to the onset of bleeding. All developed hypotension and required transfer to an ICU and blood transfusions. Colonoscopy revealed an equal number of cases of multiple and solitary ulcers located 1 to 7 cm from the dentate line; most of the ulcers were large (more than 1 cm) and circumferential or geographic in appear­ ance. The patients were treated with combinations of thermal coagulation, injection therapy, and suture ligation and had a mortality rate of 26% because of multiorgan failure. The pathology of the lesions revealed necrosis sug­ gestive of mucosal ischemia, as seen with gastric stress ulcers. This entity appears to be a different disease from solitary rectal ulcer syndrome, colitis cystica profunda, infectious ulcers, radiation ulcer, NSAID ulcers, or constipation-induced stercoral ulcer and can be considered a type of stress ulcer of the rectum, similar to that seen in the duodenum, in extremely ill hospitalized patients. Solitary or multiple painless rectal ulcers were the third most common cause of severe hematochezia in the UCLA CURE study (see Table 19-9). In contrast to solitary rectal

Chapter 19  Gastrointestinal Bleeding Table 19-11  Causes of Obscure Gastrointestinal Bleeding

Figure 19-21.  Endoscopic appearance of bleeding from a solitary rectal ulcer with a visible vessel (arrow) seen on a retroflexed view.

ulcer syndrome (see Chapter 124), they occur in older patients with severe constipation, ICU patients, and persons who are bedridden and, on colonoscopy, are chronicappearing, large, and single or multiple. They often have stigmata of recent hemorrhage and can be treated endo­ scopically (Fig. 19-21).274 Patients with inpatient hemato­ chezia from a rectal ulcer have a higher rate of rebleeding than those who present from home (see Chapters 115 and 124).

OBSCURE OVERT GASTROINTESTINAL BLEEDING Obscure GI bleeding is commonly defined as GI bleeding of uncertain cause after a nondiagnostic upper endoscopy, colonoscopy, and barium small bowel follow-through.275 Obscure GI bleeding may have an overt or occult presenta­ tion. Overt obscure GI bleeding refers to patients who have visible acute GI bleeding (e.g., melena, maroon stool, hema­ tochezia) and a nondiagnostic upper endoscopy, colonos­ copy, and small bowel series. Occult obscure GI bleeding refers to patients with a positive fecal occult blood test result, usually in association with unexplained iron defi­ ciency anemia. In most large series, a diagnostic lesion is not found on upper endoscopy and colonoscopy in 5% of hospitalized patients with overt GI bleeding, and in 75% of these patients a bleeding site is found in the small intestine. In patients with obscure GI bleeding, the following pos­ sibilities exist: (1) the lesion was within reach of a standard endoscope and colonoscope but not recognized as the bleed­ ing site (e.g., Cameron’s lesions, angioectasias, or internal hemorrhoids); (2) the lesion was within reach of the endo­ scope and colonoscope but was difficult to visualize (e.g., a blood clot obscured visualization of the lesion, varices became inapparent in a hypovolemic patient, or a lesion was hidden behind a mucosal fold) or present intermittently (e.g., Dieulafoy’s lesion, angioectasias); or (3) the lesion was in the small intestine beyond the reach of standard endo­ scopes (e.g., neoplasm, angioectasias, diverticulum). In several series, approximately 50% of patients referred to a tertiary medical center for evaluation of obscure bleeding were found to have a lesion within reach of standard endo­ scopes (i.e., a missed lesion or difficult to see lesion that

Upper Gastrointestinal Tract* Cameron’s lesions Dieulafoy’s lesions Gastric antral vascular ectasia Small Intestine Angioectasias Aortoenteric fistula Dieulafoy’s lesion Diverticulosis Meckel’s diverticulum Neoplasm Pancreatic or biliary disease Ulceration Colon Angioectasias Diverticulosis Hemorrhoids Varices *After exclusion of common causes of upper gastrointestinal bleeding.

accounted for the bleeding). Causes of obscure GI bleeding are shown in Table 19-11.276 In a patient with recurrent severe unexplained hemato­ chezia, without hypotension, a colonic source should be suspected and a repeat colonoscopy is warranted (by the same or a different endoscopist). Colonic lesions that can bleed profusely and then stop, such as diverticulosis or hemorrhoids, should be considered. In patients with recur­ rent severe melena, upper endoscopy to re-examine the esophagus, stomach, and duodenum, as well as the proxi­ mal jejunum, for a missed or unrecognized lesion should be considered. If an upper GI endoscopic examination is repeated, push enteroscopy is recommended (see later). Once it is certain that a bleeding lesion in the UGI or LGI tract was not missed, the evaluation should focus on the small intestine. In the past, the principal imaging modality of the small intestine was barium radiography, but this technique has been limited by the length, mobility, and motility of the small bowel and by overlying loops of bowel. Because small bowel bleeding is often intermittent, nuclear medicine bleeding scanning or angiography has limited value in the diagnostic evaluation. Since the late 1990s, diagnostic options for evaluating the small intestine have expanded greatly and have been revolutionized by the development of new small bowel imaging techniques, including wireless video capsule endoscopy, deep enteros­ copy, and CT enterography, which now allow greater visu­ alization and more therapeutic options than in the past.

CAUSES AND DIAGNOSTIC AND THERAPEUTIC APPROACHES

A number of lesions can cause obscure bleeding (see Table 19-11). The cause of bleeding is more likely to be a tumor, Meckel’s diverticulum, or Crohn’s disease in patients younger than 40 years and angioectasias or an NSAIDinduced ulcer in those ages 40 years and older.

Angioectasia

A variety of vascular lesions may cause bleeding from the GI tract (see Chapter 36). Angioectasias, sometimes referred to as angiodysplasias, are aberrant blood vessels found throughout the GI tract that develop with advancing age. They are distinct from arteriovenous malformations, which are congenital, and angiomas, which are neoplastic. Telan­ giectasia is the lesion that results from dilatation of the

315

316

Section III  Symptoms, Signs, and Biopsychosocial Issues terminal aspect of a vessel. Any of the vascular lesions may cause overt or obscure GI bleeding in adults, particularly in older adults and those who take antiplatelet and anticoagu­ lant drugs. Acquired vascular lesions (angioectasias and telangiectasias) occur in association with various disorders, such as chronic kidney disease, cirrhosis, rheumatoid dis­ orders, and severe heart disease.48 Although angioectasias may present as overt bleeding, they often manifest as occult bleeding or iron deficiency anemia. The most common loca­ tions are the colon and small intestine. The histopathology of angioectasias in the colon is char­ acterized by ectatic, dilated submucosal veins.277,278 A pro­ posed mechanism for their formation in the colon is that partial, intermittent, low-grade obstruction of submucosal veins during muscular contraction and distention of the cecum results in dilatation and tortuosity of the submucosal veins. Over time, the increased pressure also results in dila­ tation of the venules, capillaries, and arteries of the mucosal vasculature. Finally, precapillary sphincters can become incompetent, thereby causing arteriovenous communica­ tions to develop and possibly resulting in local mucosal ischemia. Because angioectasias can occur elsewhere in the GI tract, other mechanisms are postulated, including a response to mucosal irritation or local ischemia, as occurs after radiation. Most angioectasias occur in patients older than 60 years and can involve any segment of the GI tract. Usually, the lesions are multiple in a given segment of intestine. Approx­ imately 20% (and probably more) of patients have angioec­ tasias in at least two sections of the GI tract.279,280 In studies of asymptomatic persons who underwent co­ lonoscopy, angioectasias were found in 1% to 3%.281,282 In these individuals, the angioectasias were mostly in the right colon, with the following distribution: cecum, 37%; ascend­ ing colon, 17%; transverse colon, 7%; descending colon, 7%; sigmoid colon, 18%; and rectum, 14%. Among asymp­ tomatic persons found incidentally to have colonic angioec­ tasias, no bleeding occurred during a three-year follow up. Several conditions appear to be associated with an increased frequency of angioectasias. Patients with chronic kidney disease and uremia have an increased rate of intes­ tinal angioectasias. A study of patients with and without chronic kidney disease who had obscure GI bleeding found angioectasias as the presumptive source in 47% compared with 18% of those without kidney disease.283 The increased risk of bleeding from angioectasias in patients with chronic kidney disease may be associated with uremia-induced platelet dysfunction. Von Willebrand’s disease (congenital or acquired) also has been associated with bleeding angioectasias. Von Wil­ lebrand’s factor is needed for effective platelet aggregation. A small case series of 10 patients with bleeding GI angioectasias and von Willebrand’s disease suggested an association between the two disorders.284 Subsequently, a well-controlled prospective study found that almost all patients with bleeding GI (upper GI and colonic) angioecta­ sias, as opposed to nonbleeding angioectasias or bleeding diver­ticulosis, had acquired von Willebrand’s disease asso­ ciated with selective loss of the largest multimeric forms of von Willebrand’s factor, as well as with aortic stenosis.285 Because the large von Willebrand multimers promote primary hemostasis in a microcirculation characterized by high shear forces, as occurs in angioectasias, the loss of the large multimers may explain why bleeding occurs in some patients with angioectasias. Aortic stenosis has been associated with GI bleeding from angioectasias (Heyde’s syndrome).286 This association is controversial because both conditions are common and an

association may not imply cause and effect.287 Nevertheless, aortic stenosis has been shown to be associated with an acquired form of von Willebrand’s disease in 67% to 92% of patients because of mechanical disruption of von Willebrand proteins during passage through the stenotic aortic valve; the acquired von Willebrand’s disease, in turn, increases the risk of bleeding from angioectasias.288,289 Several series have reported the cessation of bleeding from angioectasias after aortic valve replacement, even though the angioectasias persisted, an observation consistent with the hypothesis that bleeding was the result of the damaged von Willebrand factors that normalized after aortic valve replacement.290 Because many older persons with bleeding from intestinal angioectasias have cardiovascular disease, but not severe aortic stenosis, other cardiovascular disorders such as mild to moderate aortic stenosis, aortic sclerosis, hypertrophic cardiomyopathy, and peripheral vascular disease may result in sufficiently high shear rates to disrupt von Willebrand factors and contribute to bleeding angioectasias.290 On endoscopy, angioectasias appear as 2- to 10-mm red lesions, with arborizing ectatic blood vessels that emanate from a central vessel (Fig. 19-22). Application of pressure on an angioectasia with an endoscopic probe may cause the lesion to blanch. One study has suggested that sedation of a patient with a narcotic during endoscopy can make visualization of angioectasias difficult because of transient hypotension that leads to decreased filling or causes vasoconstriction and that reversal with naloxone, a narcotic antagonist, can make the angioectasia more prominent.291 In practice, however, this maneuver is unlikely to be useful clinically and might make the patient more uncomfortable. Endoscopic treatment of angioectasias can be performed with various modalities, including injection therapy with epinephrine, thermal probe coagulation, argon plasma coagulation, and band ligation. Assessing efficacy can be difficult, given the heterogeneity of patients affected and intermittent nature of the blood loss. One series of 16 patients with transfusion-requiring angioectasias found no difference in the frequency of continued bleeding (50%), whether treatment was with surgery, endoscopic therapy, or blood transfusions alone, presumably because of the diffuse locations of the angioectasias.292 In another study of 33 patients with iron deficiency anemia and small bowel angioectasias seen on push enteroscopy, no changes in clinical or endoscopic findings were found in most patients one year after endoscopic therapy.293 By contrast, in another study of patients with GI bleeding suspected from small bowel angioectasias, treatment with electroco­ agulation led to a significant decrease (but not elimination of the need for) blood transfusions compared with observa­ tion alone.294 In a pilot study of double-balloon enteroscopy, endoscopic treatment was performed in approximately one half of patients with angioectasias, and rebleeding rates during follow-up were similar in the treated and nontreated patients.295 In a small case series, hormonal therapy with estrogen was suggested to have a benefit in controlling bleeding from telangiectasias in patients with chronic kidney disease.296 Case reports have also suggested that estrogen decreases bleeding in patients with hereditary hemorrhagic telan­ giectasia (Osler-Weber-Rendu disease; see later) and von Willebrand’s disease. A multicenter, randomized, controlled trial involving 72 patients, however, found no difference between an estrogen-progesterone combination and placebo in the rates of rebleeding, which were 39% and 46%, respectively.297 Therefore, routine use of hor­

Chapter 19  Gastrointestinal Bleeding

A

B

Figure 19-22.  Endoscopic appearance of jejunal angioectasia before (left) and after (right) multipolar probe electrocoagulation.

mones for managing bleeding from angioectasias cannot be recommended. Most patients with intermittently bleeding GI angioec­ tasias require medical treatment in addition to endoscopic hemostasis. Use of medications that can exacerbate chronic low-level bleeding, in particular aspirin, other NSAIDs, warfarin, and clopidogrel, should be avoided or at least minimized. Many patients can be managed with chronic administration of iron (orally or intravenously) and occasionally may need erythropoietin injections as well to maintain adequate blood counts, despite ongoing bleeding.

Hereditary Hemorrhagic Telangiectasia

HHT, also known as Osler-Weber-Rendu disease, is a hered­ itary condition characterized by diffuse telangiectasias and large arteriovenous malformations. The most striking clinical feature is telangiectasias on the lips, oral mucosa, and fingertips. Additionally, up to one third of patients have pulmonary, hepatic, or cerebral arteriovenous malfor­ mations (AVMs). Patients generally present with recurrent severe nosebleeds, GI bleeding, and iron deficiency anemia. Usually. the epistaxis, rather than GI bleeding, causes the more profound blood loss and anemia. HHT can be lifethreatening because of embolic strokes or brain abscesses related to the pulmonary and cerebral AVMs. Symptoms of HHT generally develop in childhood or early adulthood. HHT is inherited as an autosomal dominant trait, with varying phenotypic expression. Mutations occur in at least four genes (ENG [encodes endoglin], ALK-1 [encodes activin receptor-like kinase 1], MADH4, and HHT3) that encode proteins needed to maintain the integrity of the vascular endothelium; defects in these proteins allow the formation of AVMs. The diagnosis of HHT is based on four criteria: (1) spon­ taneous and recurrent epistaxis; (2) multiple mucocutane­ ous telangiectasias; (3) visceral AVMs (GI, pulmonary, brain, liver); and (4) a first-degree relative with HHT.298 Genetic testing to detect mutations in the ENG, ALK-1, or MAHD4 genes may be helpful in selected cases. Patients suspected of having HHT should be screened for cerebral and pulmonary AVMs, and family members of the patient should consider genetic testing.

Telangiectasias can occur anywhere in the small intestine in patients with HHT. In a case series in which capsule endoscopy was performed in 32 patients with and 48 patients without HHT who were being evaluated for small bowel bleeding, small bowel telangiectasias were found in 81% of patients with HHT compared with 29% of those without HHT.299 The telangiectasias were evenly distributed throughout the small bowel, but all actively bleeding lesions were found in the duodenum or proximal jejunum and within reach of a standard push enteroscope. The detec­ tion of five or more telangiectasias had a sensitivity of 75% and a positive predictive value of 86% for a diagnosis of HHT. The treatment of HHT is generally focused on the control of acute bleeding (epistaxis and GI bleeding), prevention of rebleeding, and treatment of anemia (with iron supple­ ments). Patients with GI bleeding should undergo endos­ copy (or push enteroscopy) and colonoscopy to look for any GI tract lesions that may be bleeding. Focal GI tract bleeding can be treated with endoscopic coagulation. Hormonal therapy also has been reported as a treatment for small bowel bleeding in HHT.300 Patients who have symp­ tomatic or large cerebral or pulmonary AVMs should be considered for radiologic embolization of these lesions (see Chapter 36).

Blue Rubber Bleb Nevus Syndrome

Blue rubber bleb nevus syndrome is a rare syndrome characterized by venous malformations in the skin, soft tissues, and GI tract. Bleeding usually occurs in childhood and continues into adulthood and results in chronic iron deficiency requiring iron replacement and transfusions. On endoscopy, lesions appear as large protuberant polypoid venous blebs; they can occur anywhere in the GI tract, but especially in the small bowel and colon, and can be treated by endoscopic band ligation or surgical resection (see Chapter 36).301,302

Meckel’s Diverticulum

A Meckel’s diverticulum is a congenital blind intestinal pouch that results from incomplete obliteration of the vitel­ line duct during gestation.303 Characteristic features of Meckel’s diverticula have been described by the “rule of

317

318

Section III  Symptoms, Signs, and Biopsychosocial Issues twos”: They occur in 2% of the population, are found within 2 feet of the ileocecal valve, are 2 inches long, result in a complication in 2% of cases, have two types of ectopic tissue (gastric and pancreatic) within the diverticulum, present clinically most commonly at age 2 (with intestinal obstruction), and have a male-to-female ratio of more than 2 : 1. The most common complications of Meckel’s diver­ ticula are bleeding, obstruction, and diverticulitis, which can occur in children or adults. Histopathologic evaluation of bleeding diverticula reveal ectopic gastric mucosa, which can lead to acid secretion and ulceration, in 75% of patients. The diagnostic test for a Meckel’s diverticulum is a 99mTcpertechnetate scan (Meckel’s scan), because technetium pertechnetate has an affinity for gastric mucosa. Meckel’s scans have a high specificity (almost 100%) and positive predictive value but can be negative in the 25% to 50% of patients in whom the diverticulum does not contain ectopic gastric mucosa.304 The accuracy of the Meckel’s scan can be improved with administration of an H2 receptor antagonist for 24 to 48 hours before the test. Meckel’s diverticula also have been diagnosed by capsule endoscopy and doubleballoon enteroscopy (via an oral or rectal approach; see Chapter 96).

Nonsteroidal Anti-Inflammatory Drug-Induced Small Intestinal Erosions and Ulcers

Mucosal erosions or ulcers that can be seen on capsule endoscopy develop in 25% to 55% of patients who take full-dose nonselective NSAIDs.305-309 Patients who take selective COX-2 inhibitors have lower rates of mucosal ulcers on capsule endoscopy (see Chapter 115).

Small Intestinal Neoplasms

Tumors of the small intestine comprise only 5% to 7% of all GI tract neoplasms but are the most common cause of obscure GI bleeding in patients younger than 50 years.310 The most common small intestine neoplasms are adenomas (usually duodenal), adenocarcinomas (Fig. 19-23), carcinoid tumors (usually ileal), gastrointestinal stromal tumors, lym­ phomas, hamartomatosis polyps (Peutz-Jeghers syndrome), and juvenile polyps (see Chapters 29 to 31, 121, and 122).

Small Intestinal Diverticula

The duodenum is the most common site of small intestinal diverticula. In one large series,311 79% of small intestinal

Figure 19-23.  Ileal adenocarcinoma detected on deep balloon enteroscopy in a patient with a history of hereditary nonpolyposis colorectal cancer who had obscure overt gastrointestinal bleeding. The lesion was initially visualized on a video capsule endoscopy study.

diverticula occurred in the duodenum, 18% were in the jejunum or ileum, and only 3% were in all three segments— duodenum, jejunum, and ileum. Duodenal diverticula are noted in up to 20% of the population, with an increasing frequency with age.311-314 They usually are located along the medial wall of the second part of the duodenum within 1 to 2 cm of the ampulla of Vater. Bleeding from a duodenal diverticulum appears to be rare. Several reports have described bleeding from a duodenal diverticulum that was managed endoscopically.314,315 Jejunal and ileal diverticula occur in 1% to 2% of the population, are most commonly associated with a motility disorder and small intestinal bac­ terial overgrowth, and only rarely have been associated with bleeding (see Chapter 23).

Dieulafoy’s Lesion of the Small Intestine

Several reports have described Dieulafoy’s lesions of the duodenum, jejunum, and ileum.316 Most affected persons are younger than 40 years, in contrast to those with gastric Dieulafoy’s lesions, who tend to be older. The lesions are often challenging to find and in the past were detected by angiography and intraoperative endoscopy. Currently, capsule endoscopy can localize these lesions, which can be treated via a single- or double-balloon enteroscope.

DIAGNOSTIC TESTS Radiologic Studies

Barium small bowel follow-through has long been used to evaluate the small intestine because it is relatively easy to perform and readily available, but it has a low yield for determining the cause of obscure GI bleeding because of its limited ability to distend the bowel and visualize mucosal surfaces. Barium enteroclysis involves placement of a naso­ enteric catheter to infuse contrast at a variable rate and to insufflate air to produce an air contrast barium radiograph. Barium enteroclysis is more accurate than a barium small bowel follow-through, with reported rates of positive find­ ings of 10% to 20% in patients with obscure bleeding.317 Conversely, enteroclysis is more difficult for the patient to tolerate, requires more involvement by the radiologist to perform, and is not widely available. A major limitation of barium studies is the inability to visualize mucosal angioec­ tasias, which are a common cause of small bowel bleeding (see earlier). Barium studies are not recommended for patients with acute bleeding because the barium can make urgent endoscopy, colonoscopy, or angiography more diffi­ cult to perform because of residual contrast in the GI tract. CT of the abdomen has the advantage of imaging extralu­ minal structures as well as mucosal and intramural lesions in the small bowel. High-quality abdominal CT (with and without oral contrast) can show thickening of the small bowel, suggestive of Crohn’s disease or malignancy. Standard CT is less accurate than barium enteroclysis for the diagnosis of low-grade bowel obstruction, mucosal ulcerations, and fistulas. CT enteroclysis using a multidetec­ tor scanner provides better views of the small intestine than standard CT. Because placement of a nasoduodenal tube is usually required, patients sometimes receive moderate sedation for CT enteroclysis.318 CT enterography with a high volume of an oral contrast agent to distend the small bowel may have a diagnostic yield similar to that for CT entero­ clysis, without the need for a nasoduodenal tube. Magnetic resonance imaging (MRI) enteroclysis and enterography have also been described, but preliminary studies suggest that results to date are inferior to those with a multidetector CT. MRI techniques have the advantage of not exposing the patient to radiation.

Chapter 19  Gastrointestinal Bleeding Nuclear medicine studies and angiography can be used to evaluate obscure GI bleeding. A Meckel’s (99mTcpertechnetate) scan is useful for the diagnostic evaluation of a Meckel’s diverticulum, as discussed earlier. Radionu­ clide scanning with technetium-labeled red blood cells has limited benefit because of its poor ability to localize the bleeding site in the small bowel. Angiography can be useful for patients with active, acute, small bowel bleeding because of the possibility of therapeutic embolization. Small case series have described provocative angiography, in which heparin or another anticoagulant is administered to provoke GI bleeding that has been intermittent. The technique increases the yield of detecting a bleeding lesion but at the risk of causing a life-threatening complication.319

Endoscopy

Push Enteroscopy Push enteroscopy can be performed with a colonoscope (160 to 180 cm in length) or dedicated push enteroscope (220 to 250 cm in length). These endoscopes can be used to evaluate the esophagus, stomach, duodenum, and proximal jejunum approximately 50 to 150 cm beyond the ligament of Treitz. Insertion is often limited by looping of the endo­ scope in the stomach. Push enteroscopy identifies a poten­ tial bleeding site in approximately 50% of patients, and approximately 50% of the lesions found are within reach of a standard upper endoscope, suggesting that the lesion was missed or unrecognized on the initial examination.276,320 The overall diagnostic yield of push enteroscopy is approxi­ mately 40%, with a range of 3% to 80% in various studies; the most commonly detected lesions are angioectasias.275 In our UCLA CURE hemostasis experience in patients with recurrent, overt, severe obscure GI bleeding manifesting as melena, the diagnostic yield has been 80%.48 The lesions were categorized as those missed by upper endoscopy, those in the duodenum (first to fourth portion), and those in the jejunum; most lesions were in reach of a push enteroscope. Focal lesions were treated endoscopically, biopsied, or tat­ tooed. Patients in whom a diagnosis was not made by push enteroscopy underwent further studies (see Fig. 19-5). Intraoperative Endoscopy and Surgical Exploration Surgical exploration of the small intestine can be performed when other studies are nondiagnostic. At surgery, the small bowel should be palpated (“running the bowel”) to detect mass lesions. In general, a standard exploratory laparotomy or laparoscopy is performed first to lyse any adhesions and to look for obvious tumors, a Meckel’s diverticulum, or large vascular lesions. The small bowel is usually extracted through the abdominal incision to allow the surgeon to assist with advancement of an endoscope within the lumen of the GI tract, which allows mucosal visualization as well as transillumination. Any endoscope can be used (panen­ doscope, pediatric colonoscope, or push enteroscope), depending on the route of access. The endoscope can be passed transorally for a natural orifice luminal examination or via an enterotomy with use of a sterile endoscope. Because air insufflation will distend the entire small intes­ tine and thereby make laparoscopic or open visualization difficult, the surgeon should pinch the intestine, manually or with an atraumatic clamp, distal to the tip of the endo­ scope, to trap enough air to permit visualization. Addition­ ally, insufflation of the bowel with carbon dioxide rather than room air allows faster diffusion of air out of the bowel. The surgeon helps advance the endoscope by pleating the small bowel over the endoscope. Any lesion identified can be addressed surgically or endoscopically, depending on the nature of the lesion. Most series report

complete enteroscopy of the entire small bowel in 50% to 75% of cases.321,322 The diagnostic yield of intraoperative enteroscopy ranges from 58% to 88%, but rebleeding after intraoperative enteroscopy has also been reported in 13% to 60% of patients.275 The moderate performance character­ istics, as well as risks of surgical exploration, limit this procedure as a diagnostic tool, but in selected patients, combined endoscopic and surgical evaluation can be useful and definitive. Capsule Endoscopy With capsule endoscopy, the patient ingests a pill camera that transmits images of the small intestine over the course of approximately eight hours. In patients with severe, recurrent GI bleeding, this technique can identify a transi­ tion point at which fresh blood appears in the small bowel and thereby possibly detect a potential bleeding site. Capsule endoscopy does not permit the application of therapy and can only localize a lesion in the small bowel on the basis of the time of passage down the small intestine, as determined by sensors on the abdomen and telemetry. The information can be useful, however, in directing subsequent therapeutic procedures such as deep (balloon) enteroscopy, angiography, or surgery. Although capsule endoscopy may occasionally detect gastric, duodenal, or colonic lesions, it is not a substitute for upper endoscopy and colonoscopy. Compared with small bowel barium studies, capsule endoscopy has significantly improved detection rates for small bowel lesions (67% vs. 8%) and findings that influ­ ence clinical management (42% vs. 6%).323,324 A small series has found capsule endoscopy to be superior to CT entero­ clysis for the diagnosis of obscure GI bleeding because of its ability to identify angioectasias.325 An evaluation of published studies that have compared push enteroscopy with capsule endoscopy in patients with obscure bleeding (79% overt, 21% occult) has found the average rate of positive findings to be 23% for push enter­ oscopy and 63% for capsule endoscopy.275 A similar result was found in a meta-analysis of published trials and abstracts; the diagnostic yield for push enteroscopy was 28% and that for capsule endoscopy was 63%.324 A random­ ized trial that compared push enteroscopy with capsule endoscopy as a first-line approach to obscure GI bleeding reported identification of a bleeding source in 24% of the push enteroscopy examinations and 50% of the capsule studies (P = 0.02).326 In this study, capsule endoscopy missed lesions in 8% of patients; all the missed lesions were within reach of a standard upper endoscope. Capsule endoscopy was compared with intraoperative endoscopy in one study of 47 patients who underwent both procedures, primarily for obscure overt GI bleeding.327 Using intraoperative endoscopy as the gold standard, capsule endoscopy had a sensitivity of 95%, specificity of 75%, positive predictive value of 95%, and negative predictive value of 85%. Most of the bleeding lesions were angioectasias. Several studies have found that the diagnostic yield of capsule endoscopy increases in the setting of ongoing or recent (less than two weeks) overt GI bleeding or severe chronic GI bleeding (hemoglobin < 10 g/dL, iron deficiency anemia, or more than one overt bleeding episode).327-329 In a study from Greece of 34 patients who had active, mild to moderate overt GI bleeding, and negative upper endoscopy and colonoscopy results, and who underwent an urgent capsule endoscopy study while still in the hospital, the diagnostic rate was 92%, as defined by the identification of a bleeding lesion (18 angioectasias, 3 ulcers, 2 tumors) or

319

320

Section III  Symptoms, Signs, and Biopsychosocial Issues the segment of intestine with bleeding (11 patients).24 By contrast, the same group from Greece found that the diag­ nostic yield of capsule endoscopy in patients with obscure occult bleeding and iron deficiency anemia was 57% (angio­ ectasias in 24%, multiple jejunal or ileal ulcers in 12%, multiple erosions in 8%, solitary ulcers in 6%, polyps in 4%, and other tumors in 4%).330 Deep Enteroscopy of the Jejunum and Ileum Specially designed, ultraflexible, 200-cm-long enteroscopes are used in conjunction with an overtube to advance the endoscope by pleating the small intestine over it. The avail­ able systems include a double-balloon endoscope (with a balloon on the tip of the endoscope and another balloon on the overtube), a single-balloon system (a balloon on the overtube only), and a spiral overtube (no balloon used). All enteroscopes work by pleating the small intestine over the endoscope. These enteroscopes can be inserted orally (ante­ grade) and advanced into the proximal to midileum or inserted rectally (retrograde) and advanced to the distal to midileum. Rarely, a complete enteroscopy of the small intestine to the cecum can be performed via the antegrade approach. Deep enteroscopy allows not only visualization but also interventions such as biopsy, hemostasis, and tat­ tooing of lesions. The endoscopes used for deep enteros­ copy have standard working channels that allow the passage of accessories such as biopsy forceps, MPEC thermal probes, endoscopic hemoclips, and injection needles that fit through a standard colonoscope. The risks of deep enteroscopy are similar to those for push enteroscopy or colonoscopy, with the additional small risk of pancreatitis. The preponderance of available data is derived from studies using the doubleballoon enteroscopy system. A compilation of 12 case series of double-balloon enter­ oscopy for obscure bleeding in 723 patients found an overall diagnostic yield of 65% (Table 19-12).275 Comparative studies of capsule endoscopy and double-balloon enteros­ copy have revealed a slightly higher diagnostic yield for capsule endoscopy. The agreement between the two approaches in one large multicenter study of 115 patients was 74% for angioectasias, 96% for ulcers, 94% for polyps, and 96% for other large tumors.331 Another comparative study found that for patients with obscure bleeding, the agreement between the two was 92%, but the yield in a given segment of intestine in patients with polyposis was only 33% for capsule endoscopy compared with 67% for double-balloon enteroscopy; however, capsule endoscopy may detect polyps beyond the reach of the double-balloon enteroscope.332

Table 19-12  Small Intestinal Lesions Found in 488 Patients during Double-Balloon Enteroscopy for Obscure Gastrointestinal Bleeding LESION None Angioectasias Ulcerations Malignancy Other

FREQUENCY, % (RANGE) 40 31 13 8 6

(0-57) (6-55) (2-35) (3-26) (2-22)

Data from Raju GS, Gerson L, Das A, Lewis B. American Gastroenterological Association (AGA) Institute technical review on obscure gastrointestinal bleeding. Gastroenterology 2007; 133:1697-717.

Overall Approach to the Patient with Overt Obscure Gastrointestinal Bleeding

For patients with unexplained overt GI bleeding and nega­ tive upper endoscopy and colonoscopy results, capsule endoscopy is generally recommended as the next step. If capsule endoscopy reveals a lesion in the proximal jejunum, push enteroscopy can be performed. If a lesion is found in the mid–small intestine, deep enteroscopy or surgery may be considered, depending on the nature of the lesion. A lesion in the terminal ileum may prompt deep enteroscopy via the colonic route. If no lesion is detected on capsule endoscopy, but a high suspicion for a lesion remains, capsule endoscopy should be repeated or deep enteroscopy performed. With the increased avail­ ability of deep enteroscopy equipment, deep enteroscopy could become the preferred initial diagnostic step before capsule endoscopy. Modeling studies have suggested that this approach might be a cost-effective strategy,333,334 but the question ideally should be addressed in a randomized study. The UCLA CURE group’s algorithm for the management of patients who have had unexplained, severe, overt GI bleeding, with a history of melena and the need for blood transfusions, is shown in Figure 19-5. For such patients, the diagnostic yield is over 80%.48

OBSCURE OCCULT GASTROINTESTINAL BLEEDING AND IRON DEFICIENCY BLEEDING FECAL OCCULT BLOOD

Occult GI bleeding is usually detected with a routine fecal occult blood test (FOBT), with no visible blood in the stool and with or without iron deficiency. Normal fecal blood loss is 0.5 to 1.5 mL/day.335 Many FOBTs are available for detect­ ing increased amounts blood in the stool and are described in detail in Chapter 123. The approach to the patient with a positive FOBT result depends on why the test was obtained. If the FOBT was obtained for colon cancer screening in a patient older than 50 years, the patient should undergo colonoscopy and pos­ sibly upper endoscopy. Upper endoscopy should also be considered in a patient with a positive FOBT result who does not have iron deficiency anemia; this recommendation is based on the results of a study of 248 patients with fecal occult blood in whom more lesions were found in the UGI tract by upper endoscopy (mostly esophagitis, gastritis, and ulcers) than in the colon by colonoscopy (mostly large adenomas and cancer).336 If the FOBT was performed for iron deficiency anemia, the patient should be evaluated with upper endoscopy and colonoscopy. If both examina­ tions are negative, the small bowel should be imaged, as described earlier, with capsule endoscopy, possibly fol­ lowed by deep enteroscopy if a lesion is detected on capsule endoscopy. Although colon cancer screening with FOBT generally is based on six samples of spontaneously passed stool, a posi­ tive FOBT result, not uncommonly, may be found when stool is obtained during digital examination of the rectum. Although a digital rectal examination could potentially cause trauma to the anal canal, several studies have found no increase in the false-positive rate of FOBTs when stool is obtained from a digital examination.337,338 Therefore, a positive FOBT result should be approached in the same manner regardless of the method by which the stool sample was obtained. Additionally, a single negative FOBT result

Chapter 19  Gastrointestinal Bleeding on a digital rectal examination is not considered adequate colon cancer screening and does not reduce a patient’s chances of having advanced neoplasia.339

IRON DEFICIENCY ANEMIA

Iron deficiency anemia is common, with a frequency of 2% to 5% in adult men and postmenopausal women.340 Iron deficiency anemia represents 4% to 13% of all referrals for outpatient gastroenterology consultation.341 The approach to iron deficiency anemia depends on the patient’s gender and the result of an FOBT. Young women with iron deficiency anemia should be considered to have menstrual blood loss as the cause of anemia and, depending on clinical circumstances, may not need a GI evaluation. By contrast, men and postmenopausal women with iron defi­ ciency anemia should always be evaluated for a GI cause of iron deficiency. Iron deficiency anemia should be considered in patients with a low mean corpuscular volume (MCV) and anemia. In iron deficiency anemia, the serum iron concentration is decreased, and the level of transferrin (TIBC) is increased. A transferrin saturation index (serum iron divided by TIBC) lower than 15% is a sensitive indicator of iron deficiency anemia. A serum ferritin level lower than 15 ng/mL has a sensitivity of 59% and specificity of 99% for iron defi­ ciency, whereas a cutoff ferritin level of 41 ng/mL has a sensitivity and specificity of 98%.342 A bone marrow aspi­ rate can provide information about body stores of iron but is rarely necessary. Iron deficiency can result from overt or occult blood loss (from GI tract luminal lesions, menses, epistaxis, pulmonary lesions, or urinary tract lesions), intestinal iron malabsorp­ tion (as in celiac disease or gastric atrophy or after gastric bypass surgery), treatment with erythropoietin (because of excess iron requirements), and red blood cell destruction (hemolysis). The GI evaluation of a patient with iron deficiency should focus on endoscopy (upper and lower) to detect treatable lesions, especially malignancies. Recognizing iron malab­ sorption from the GI tract as a cause of iron deficiency is especially important. The duodenum is the site of iron absorption in the small intestine. Most dietary iron is in the ferric form, but only the ferrous form of iron can be absorbed by the duodenum. Ascorbic acid at a low pH is required to release nonheme iron and convert it to the ferrous form for absorption in the small intestine.343 Several studies have shown that 20% to 30% of patients with iron deficiency anemia have gastric atrophy and therefore do not produce an acid milieu that facilitates iron absorption.340,344,345 Iron deficiency anemia has also been associated with H. pylori infection.346 These studies suggest that gastric biopsies should be obtained during upper endoscopy in patients with unexplained iron deficiency anemia (see Chapters 50, 51, and 100). Celiac disease commonly manifests as iron deficiency anemia, primarily because of iron malabsorption resulting from blunted duodenal villi. Patients with celiac disease have been reported to have higher rates of positive FOBT results than healthy controls, but subsequent studies in which radiolabeled red cells were used did not find a true increase in blood loss.347,348 Nevertheless, it is possible that the cause of iron deficiency anemia in patients with celiac disease is multifactorial. Any patient who is evaluated for iron deficiency anemia and undergoes upper endoscopy should have duodenal biopsy samples obtained to look for celiac disease (see Chapter 104).

Patients who have undergone Roux-en-Y gastric bypass surgery are at high risk of iron malabsorption because of bypass of the duodenum, where most iron is absorbed. These patients can present with severe unexplained iron deficiency without occult blood in the stool. They often have extremely low body stores of iron and require intrave­ nous iron supplementation. The differential diagnosis of iron deficiency anemia includes anemia of chronic disease and thalassemia. In anemia of chronic disease, both the serum iron level and TIBC are low, with a normal serum ferritin level. Patients with thalassemia have a family history of anemia, spleno­ megaly, target cells on peripheral blood smear, and normal serum ferritin levels. Patients with unexplained iron deficiency anemia should undergo upper endoscopy and colonoscopy to rule out a GI tract lesion that may cause chronic blood loss. In a prospec­ tive study of 100 patients with iron deficiency anemia, GI tract lesions were found in 62 patients, with 36 having lesions in the UGI tract (mostly ulcers), 25 in the colon (mostly cancer), and 1 in both the UGI tract and colon.349 In patients with unexplained iron deficiency anemia who undergo upper endoscopy, duodenal biopsy specimens should be obtained to rule out celiac disease as a cause of iron malabsorption. Gastric biopsy samples also should be obtained to rule out gastropathy and H. pylori infection. Depending on the severity of iron deficiency anemia, even without a positive FOBT result, evaluation of the small intestine for a bleeding lesion, as discussed earlier, should be considered. If a specific cause of anemia is not identified, patients should be advised to avoid antiplatelet and antico­ agulant drugs and take supplemental iron.

KEY REFERENCES

Barkun A, Bardou M, Marshall JK. Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med 2003; 139:843-57. (Ref 100.) Gralnek IM, Barkun AN, Bardou M. Management of acute bleeding from a peptic ulcer. N Engl J Med 2008; 359:928-37. (Ref 139.) Jensen DM, Kovacs TO, Jutabha R, et al. Randomized trial of medical or endoscopic therapy to prevent recurrent ulcer hemorrhage in patients with adherent clots. Gastroenterology 2002; 123:407-13. (Ref 115.) Jensen DM, Machicado GA. Diagnosis and treatment of severe hemato­ chezia. The role of urgent colonoscopy after purge. Gastroenterology 1988; 95:1569-74. (Ref 22.) Jensen DM, Machicado GA, Jutabha R, Kovacs TO. Urgent colonoscopy for the diagnosis and treatment of severe diverticular hemorrhage. N Engl J Med 2000; 342:78-82. (Ref 23.) Khuroo MS, Yattoo GN, Javid G, et al. A comparison of omeprazole and placebo for bleeding peptic ulcer. N Engl J Med 1997; 336:1054-8. (Ref 125.) Laine L. Multipolar electrocoagulation in the treatment of active upper gastrointestinal tract hemorrhage. A prospective controlled trial. N Engl J Med 1987; 316:1613-17. (Ref 174.) Lau JY, Leung WK, Wu JC, et al. Omeprazole before endoscopy in patients with gastrointestinal bleeding. N Engl J Med 2007; 356:163140. (Ref 129.) Lau JY, Sung JJ, Lam YH, et al. Endoscopic retreatment compared with surgery in patients with recurrent bleeding after initial endo­ scopic control of bleeding ulcers. N Engl J Med 1999; 340:751-6. (Ref 144.) Lau JY, Sung JJ, Lee KK, et al. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med 2000; 343:310-16. (Ref 127.) Raju GS, Gerson L, Das A, Lewis B. American Gastroenterological Asso­ ciation (AGA) Institute technical review on obscure gastrointestinal bleeding. Gastroenterology 2007; 133:1697-717. (Ref 275.) Rockey DC, Koch J, Cello JP, et al. Relative frequency of upper gastro­ intestinal and colonic lesions in patients with positive fecal occultblood tests. N Engl J Med 1998; 339:153-9. (Ref 336.)

321

322

Section III  Symptoms, Signs, and Biopsychosocial Issues Rockey DC, Cello JP. Evaluation of the gastrointestinal tract in patients with iron-deficiency anemia. N Engl J Med 1993; 329:1691-5. (Ref 349.) Strate LL, Ayanian JZ, Kotler G, Syngal S. Risk factors for mortality in lower intestinal bleeding. Clin Gastroenterol Hepatol 2008; 6:100410. (Ref 219.)

Sung JJ, Chan FK, Lau JY, et al. The effect of endoscopic therapy in patients receiving omeprazole for bleeding ulcers with nonbleeding visible vessels or adherent clots: A randomized comparison. Ann Intern Med 2003; 139:237-43. (Ref 136.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

20 Jaundice Steven D. Lidofsky

CHAPTER OUTLINE Bilirubin Metabolism and Measurement  323 Metabolism  323 Measurement  324 Differential Diagnosis  325 Isolated Disorders of Bilirubin Metabolism  325 Liver Disease  326 Obstruction of the Bile Ducts  329 Diagnostic Approach to Jaundice  330 History and Physical Examination  330

Jaundice (icterus), is a condition characterized by yellow discoloration of the skin, conjunctivae, and mucous membranes as a result of widespread tissue deposition of the pigmented metabolite bilirubin. Although jaundice is generally associated with liver and biliary tract disease, it has many causes. It is thus not surprising that the diagnosis and management of jaundice have challenged clinicians for centuries. Attempts to classify icteric syndromes appeared as early as the treatises of Hippocrates. By the time of Osler, distinctions were already made between biliary tract obstruction and nonobstructive causes of jaundice. In the latter part of the twentieth century, elucidation of the molecular mechanisms that underlie bilirubin metabolism, as well as the development of more sophisticated biochemical and imaging techniques, made it possible to pinpoint the cause of jaundice in most cases. Despite the impressive array of tools available today, it should be emphasized that to minimize risk to the patient, an effective approach to jaundice requires selection of diagnostic and therapeutic modalities on the basis of a careful assessment of the likelihood of possible underlying diseases. This chapter covers four major areas: (1) bilirubin metabolism; (2) differential diagnosis of jaundice; (3) role of the history, physical examination, and routine biochemical tests in narrowing the differential diagnosis and the usefulness of selected laboratory and hepatobiliary imaging studies; and (4) therapeutic approaches to the management of jaundice.

BILIRUBIN METABOLISM AND MEASUREMENT METABOLISM

Bilirubin, a hydrophobic and potentially toxic compound, is a tetrapyrrole that is an end product of heme degradation. Bilirubin metabolism has been reviewed in depth elsewhere1,2 and is summarized briefly in Figure 20-1. Each day, a healthy adult produces approximately 4 mg/kg of bilirubin (i.e., almost 0.5 mmol in a 70-kg person). Most

Initial Laboratory Studies  331 Overall Approach  332 Imaging Studies  332 Other Studies  334 Therapeutic Approaches  334 Biliary Obstruction  334 Other Conditions  335

bilirubin (70% to 80%) is derived from degradation of hemoglobin from senescent erythrocytes, and a minor component arises from premature destruction of newly formed erythrocytes in the bone marrow or circulation (i.e., ineffective erythropoiesis). Most of the remaining 20% to 30% is formed from breakdown of hemoproteins, such as catalase and cytochrome (CYP family) oxidases, in hepatocytes. Although nonhemoglobin heme-containing proteins are also present in extrahepatic tissues, their mass is so small or their turnover rate so slow (as for myoglobin) that their overall contribution to bilirubin production is minimal. The breakdown of heme to bilirubin occurs by a two-step process. First, heme is converted to biliverdin by heme oxygenase, which functions predominantly as an integral membrane protein of the smooth endoplasmic reticulum. Second, biliverdin is converted rapidly to bilirubin by the cytosolic protein biliverdin reductase. Catabolism of erythrocyte-derived hemoglobin to bilirubin takes place prim­arily in reticuloendothelial cells in the spleen, liver, and bone marrow. By contrast, free hemoglobin, haptoglobinbound hemoglobin, and methemalbumin are catabolized to bilirubin predominantly in hepatocytes. Bilirubin circulates in plasma tightly, but noncovalently, bound to albumin. Excretion of bilirubin requires con­ version to water-soluble conjugates by hepatocytes and subsequent secretion into bile. Bilirubin metabolism and elimination is a multistep process for which several inherited disorders have been identified (see later). Bilirubin is taken up across the sinusoidal (basolateral) membrane of hepatocytes by a carrier-mediated mechanism. The uptake of bilirubin is inhibited competitively by certain organic anions such as sulfobromophthalein (BSP) and indocyanine. Bilirubin uptake has been suggested to be mediated by a liver-specific sinusoidal organic anion transport protein, (OATP1B1, SLC21A6), but this is not entirely certain.3,4 After uptake, bilirubin is directed by cytosolic binding proteins (e.g., glutathione S-transferase B, fatty acid binding protein) to the endoplasmic reticulum, where it is conjugated with uridine diphosphate (UDP)–glucuronic acid by the enzyme bilirubin UDP–glucuronyl transferase (B-UGT).

323

324

Section III  Symptoms, Signs, and Biopsychosocial Issues jugated by bacterial enzymes in the terminal ileum and colon and converted to colorless tetrapyrroles called urobilinogens. Up to 20% of urobilinogens are resorbed and ultimately excreted in bile and urine.

Hemoglobin Other hemoproteins Br Alb-Br Plasma Br

BrG

Urine MRP3?

OATP?

B-UGT BrG

Hepatocytes

MRP2

BrG Bile Figure 20-1.  Overview of bilirubin formation, metabolism, and transport. Heme from hemoglobin and other hemoproteins is converted to biliverdin and then to bilirubin (Br), predominantly in reticuloendothelial cells in the bone marrow and spleen. Br is released into plasma (in its unconjugated form), where it is tightly but reversibly bound to albumin (Alb). Br is then taken up at the sinusoidal membrane of hepatocytes, possibly via a member of the organic anion transporter (OATP) family. Br is conjugated, via the activity of bilirubin UDP-glucuronyl transferase (B-UGT), to form bilirubin mono- and di-glucuronides (BrG). Biliary secretion of BrG occurs at the canalicular membrane by the multispecific organic anion transporter MRP2. Under physiological conditions, the vast majority of BrG is eliminated in bile. Small amounts of BrG are transported at the sinusoidal membrane back into plasma, possibly via the multispecific organic anion transporter MRP3. Plasma BrG enters the renal circulation, where it undergoes glomerular filtration and elimination into urine. Thus, under normal conditions, at least 95% of bilirubin in plasma is present in the unconjugated form. If abnormally high concentrations of BrG are retained over a prolonged period, BrG-Alb complexes, which do not dissociate and cannot undergo glomerular filtration, are formed. MRP, multidrug resistanceassociated protein.

Conjugation converts hydrophobic bilirubin into a watersoluble form suitable for excretion. Conjugated bilirubin is then directed primarily toward the canalicular (apical) membrane, where it is transported into the bile canaliculus by an adenosine triphosphate (ATP)-dependent export pump. The responsible protein, multidrug resistance– associated protein-2 (MRP2, ABCC2), appears to function as a multispecific transporter of various organic anions (including BSP, glutathione, and conjugated bile salts).5 Small amounts of bilirubin glucuronides are secreted across the sinusoidal membrane via a pathway postulated to be mediated by a distinct multispecific organic ion export pump, MRP3 (ABCC3)6; conjugated bilirubin in plasma undergoes renal elimination (see Fig. 20-1). This pathway may be upregulated in disorders characterized by cholestasis (impaired bile flow). With prolonged cholestasis (or a metabolic disorder of conjugated hyperbilirubinemia; see later), an increasing proportion of conjugated bilirubin in plasma becomes covalently bound to albumin, and this covalently bound bilirubin cannot be excreted into urine. Approximately 80% of bilirubin in human bile is in the form of diglucuronides. Almost all the rest is in the form of monoglucuronides, and only trace amounts are unconjugated. Resorption of conjugated bilirubin by the gallbladder and intestine is negligible; however, bilirubin can be decon-

MEASUREMENT

The normal bilirubin concentration in the serum of adults is lower than 1 to 1.5 mg/dL. In general, jaundice is not evident until the serum bilirubin concentration exceeds 3 mg/dL. In healthy persons, most bilirubin circulates in its unconjugated form; less than 5% of circulating bilirubin is present in conjugated form. In cholestatic conditions, the proportion of unconjugated bilirubin may increase as a consequence of upregulated MRP3 expression. The importance of accurate measurement of bilirubin is underscored by its incorporation as a critical variable in scoring systems such as the Model for End-stage Liver Disease (MELD), which provide estimates of survival in various acute and chronic liver disorders.7 Serum bilirubin is detected conventionally by the diazo van den Bergh reaction. With this colorimetric method, bilirubin is cleaved by compounds such as diazotized sulfanilic acid to form an azodipyrrole that can be assayed by spectrophotometry. Conjugated bilirubin is cleaved rapidly (directly) by diazo reagents. By contrast, unconjugated bilirubin reacts slowly with diazo reagents because the site of chemical cleavage is rendered inaccessible by internal hydrogen bonding. Therefore, reliable measurement of total bilirubin concentration requires the addition of an accelerator compound, such as ethanol or urea, which disrupts this hydrogen bonding and facilitates the cleavage of unconjugated bilirubin by the diazo reagent. The concentration of the indirect bilirubin fraction is calculated by subtracting the direct bilirubin concentration (measured in the absence of the accelerator compound) from that of the total bilirubin concentration (measured in the presence of the accelerator compound). Although the direct bilirubin concentration is influenced by changes in conjugated bilirubin levels, the two are not equivalent. Similarly, indirect bilirubin is not equivalent to unconjugated bilirubin. In particular, reliance on direct and indirect bilirubin measurements can lead to errors in the diagnosis of isolated disorders of bilirubin metabolism (e.g., suspected Gilbert’s syndrome; see later). Many clinical laboratories have abandoned measurements of direct and indirect bilirubin and instead use automated reflectance spectroscopic assays that provide more accurate estimates of conjugated and unconjugated bilirubin. These assays are useful clinically in the management of physiologic jaundice of the newborn (see later), in which neurotoxicity may result from the passage of unconjugated bilirubin across the blood-brain barrier (kernicterus). In disorders characterized by prolonged cholestasis, however, such assays may underestimate the conjugated bilirubin concentration, because they do not accurately detect albumin-bound conjugated bilirubin (so-called delta bilirubin). Indeed, if an isolated disorder of bilirubin metabolism is suspected, the diagnosis may require more sophisticated chromatographic techniques that precisely measure the concentrations of unconjugated, monoglucuronidated, and diglucuronidated bilirubin, as well as conjugated bilirubin-albumin complexes.2 In practice, these techniques are not widely used. Even with such accurate methods, measurements of con­jugated and unconjugated bilirubin will not distinguish hepatic disorders from biliary obstruction. Therefore, in most cases, these tests are of limited use.

Chapter 20  Jaundice DIFFERENTIAL DIAGNOSIS Jaundice can result from an increase in the formation of bilirubin or a decrease in the hepatobiliary clearance of bilirubin. From a practical standpoint, conditions that produce jaundice can be classified under the broad categories of isolated disorders of bilirubin metabolism, liver disease, and obstruction of the bile ducts (Table 20-1).

ISOLATED DISORDERS OF BILIRUBIN METABOLISM Unconjugated Hyperbilirubinemia

Three basic mechanisms can lead to isolated unconjugated hyperbilirubinemia: (1) increased bilirubin production; (2) decreased hepatocellular uptake of unconjugated bilirubin; Table 20-1  Differential Diagnosis of Jaundice and Hyperbilirubinemia Isolated Disorders of Bilirubin Metabolism Unconjugated Hyperbilirubinemia Increased bilirubin production (e.g., hemolysis, ineffective erythropoiesis, blood transfusion, resorption of hematomas) Decreased hepatocellular uptake (e.g., drugs such as rifampin, Gilbert’s syndrome [secondary mechanism]) Decreased conjugation (e.g., Gilbert’s syndrome, Crigler-Najjar syndrome, physiologic jaundice of the newborn, drugs such as indinavir, atazanavir) Conjugated or Mixed Hyperbilirubinemia Dubin-Johnson syndrome Rotor’s syndrome Liver Disease Hepatocellular Dysfunction Acute or subacute hepatocellular injury (e.g., viral hepatitis, hepatotoxins [such as ethanol, acetaminophen, Amanita phalloides]; drugs such as isoniazid, phenytoin; ischemia such as in hypotension, vascular outflow obstruction; metabolic disorders such as Wilson disease, Reye’s syndrome; pregnancy-related as in acute fatty liver of pregnancy, pre-eclampsia) Chronic hepatocellular disease (e.g., viral hepatitis; hepatotoxins such as ethanol, vinyl chloride, vitamin A; autoimmune hepatitis; metabolic disorder such as hemochromatosis, Wilson disease, nonalcoholic fatty liver disease, α1-antitrypsin deficiency; celiac disease) Hepatic Disorders with Prominent Cholestasis Diffuse infiltrative disorders (e.g., granulomatous diseases such as mycobacterial infections, sarcoidosis, lymphoma, Wegener’s granulomatosis; amyloidosis; malignancy) Cholangiocyte injury (e.g., primary biliary cirrhosis; graft-versus-host disease; drugs such as erythromycin, trimethoprimsulfamethoxazole; cystic fibrosis) Miscellaneous conditions (e.g., benign recurrent intrahepatic cholestasis; drugs such as estrogens, anabolic steroids; total parenteral nutrition; bacterial infections; paraneoplastic syndromes; intrahepatic cholestasis of pregnancy; benign postoperative cholestasis) Obstruction of the Bile Ducts Choledocholithiasis Diseases of the Bile Ducts Inflammation, infection (e.g., primary sclerosing cholangitis, AIDS cholangiopathy, hepatic arterial chemotherapy, postsurgical strictures) Neoplasms (e.g., cholangiocarcinoma) Extrinsic Compression Neoplasms (e.g., pancreatic carcinoma, metastatic lymphadenopathy, hepatocellular carcinoma, ampullary adenoma, lymphoma) Pancreatitis Vascular enlargement (e.g., aneurysm, cavernous transformation of the portal vein [portal cavernoma]) AIDS, acquired immunodeficiency syndrome.

and (3) decreased bilirubin conjugation. In each of the resulting conditions, liver function is otherwise normal, and the results of standard biochemical liver tests other than the serum bilirubin concentration are normal. Increased Bilirubin Production Processes that can generate excessive bilirubin production include hemolysis, ineffective erythropoiesis, and resorption of a hematoma.2 Jaundice may thus complicate the clinical course of patients with hemolytic anemias, megaloblastic anemia from folate or vitamin B12 deficiency, iron deficiency anemia, sideroblastic anemia, and polycythemia vera. With these disorders, bilirubin concentration does not generally exceed 4 to 5 mg/dL. Jaundice can follow massive blood transfusions, because the foreshortened lifespan of transfused erythrocytes leads to excessive hemoglobin release. Hyperbilirubinemia resulting from resorption of hematomas and blood transfusions also may develop in patients who have experienced major trauma.8 Decreased Bilirubin Uptake A decrease in hepatocellular uptake of bilirubin can be seen with certain drugs. For example, the antituberculosis agent rifampin has been shown to inhibit bilirubin uptake by hepatocytes competitively and may produce jaundice by inhibiting the transport protein OATP1B1 (SLC21A6); similar effects may be produced by the immunosuppressive drug cyclosporine A.9,10 Decreased bilirubin uptake also may contribute to phenotypic expression of the hereditary disorder Gilbert’s syndrome, in which the predominant abnormality is impaired bilirubin conjugation resulting from reduced B-UGT activity.11 Decreased Bilirubin Conjugation Three autosomally inherited disorders of unconjugated hyperbilirubinemia are attributable to impaired bilirubin conjugation (Table 20-2). The most common of these dis­ orders is Gilbert’s syndrome, which has a prevalence of approximately 10% in white populations. The disorder is entirely benign and rarely produces clinical jaundice. Serum bilirubin levels may rise two- to threefold with fasting or dehydration but are generally below 4 mg/dL. Patients with Gilbert’s syndrome typically present during or after adolescence, when isolated hyperbilirubinemia is detected as an incidental finding on routine multiphasic biochemical screening. The molecular basis of Gilbert’s syndrome has been linked to a reduction in transcription of the B-UGT gene UGT1A1 as a result of mutations in the promoter region and, less commonly, in the coding region.1 Mutations in the coding region of UGT1A1 appear to be responsible for Crigler-Najjar syndrome.12 In type I CriglerNajjar syndrome, B-UGT activity is absent, and many patients die of kernicterus in the neonatal period (see Table 20-2). Phototherapy (see later) is required to prevent kernicterus, and liver transplantation can be lifesaving. Persons with type II Crigler-Najjar syndrome have markedly reduced B-UGT activity, with serum bilirubin levels between those of patients with Gilbert’s syndrome and those with type I Crigler-Najjar syndrome (see Table 20-2). In contrast to patients with type I Crigler-Najjar syndrome, those with type II Crigler-Najjar syndrome are not ill during the neonatal period and may not be diagnosed until early childhood. Although the degree of jaundice can wax and wane, most patients with type II Crigler-Najjar syndrome experience a fall in serum bilirubin levels to 2 to 5 mg/dL with phenobarbital, an agonist for the constitutive androstane receptor CAR, which increases expression of UGT1A1

325

326

Section III  Symptoms, Signs, and Biopsychosocial Issues Table 20-2  Hereditary Disorders of Bilirubin Metabolism and Transport Syndrome CRIGLER-NAJJAR TYPE I

CRIGLER-NAJJAR TYPE II

6%-12% UGT1A1 ↓Bilirubin conjugation

Very rare UGT1A1 No bilirubin conjugation

Uncommon UGT1A1 ↓↓Bilirubin conjugation

≤3 in absence of fasting or hemolysis, almost all unconjugated Usually normal, occasional ↑lipofuscin

Usually >20 (range, 17-50), all unconjugated Normal

Usually 500 mL/day) Moderate output (200-500 mL/day) Low output (95%) of GISTs, but KIT is not expressed by true smooth muscle tumors of the GI tract nor by stromal tumors at other anatomic locations, such as endometrial stromal tumors. Although the origin of the neoplastic cells of GISTs remains a matter of active investigation, some data suggest that GISTs originate from CD34-positive stem cells residing within the wall of the gut, which can then differentiate incompletely toward the ICC phenotype.27-29 GISTs characteristically exhibit expression of CD117 by immunohistochemical assays (>95% of lesions but, importantly, not 100%, because there are true KIT-negative GIST cases).30 Levels of expression of CD117 (KIT) are generally diffuse and strong in the spindle cell GIST subtype (Fig. 30-1). In contrast, in the epithelioid subtype, CD117 expression is typically focal and weakly positive in a dot-like pattern (Fig. 30-2). As noted, CD34 expression is neither sensitive nor specific for GIST because this antigen can also

Figure 30-2.  High-power view of an epithelioid gastrointestinal stromal tumor (GIST) showing epithelioid cytomorphology, fibrillary cytoplasm, and a lack of mitotic activity (Hematoxylin and eosin, ×200). (Courtesy of Dr. Brian P. Rubin, Cleveland, Ohio.)

be noted in desmoid tumors, and approximately 30% to 40% of GIST lesions are negative for CD34.21,26 True leiomyosarcomas express two smooth muscle markers, smooth muscle actin (SMA) and desmin, but fail to express CD117. Schwannomas are usually positive for the neural antigen S100 but are also negative for CD117. Normal mast cells and ICCs in the surrounding stromal tissues serve as ideal positive internal controls because these normal cells strongly express CD117. Activating mutations in the KIT gene were identified in five of six cases of human GISTs originally analyzed by Hirota and colleagues,20 with evidence that the mutations resulted in uncontrolled, ligand-independent activation of the KIT kinase. Genetically engineered cells harboring the mutant overactive KIT proteins were tumorigenic in nude mice, serving as proof of concept that the malignant phenotype was directly induced by the aberrant signaling pathways associated with uncontrolled KIT kinase activation. The oncogenic potential of mutant, uncontrollably active KIT in the pathogenesis of GISTs in humans has also been

463

464

Section IV  Topics Involving Multiple Organs supported by the identification of familial syndromes with an autosomal dominant inheritance pattern and an abnormally high incidence of GISTs, usually occurring as multiple foci within any affected individual (see later).21,31,32 Genetic analysis of such kindreds reveals that they harbor germline-activating KIT mutations, similar to the mutations that were first described in sporadic cases of GISTs. With analyses using optimal tumor specimens and sophisticated technology, it has become clear that KIT mutations can be detected in more than 90% of GIST cells.33-35 Constitutive activation of the KIT kinase enzymatic function has been reported to characterize every GIST sample analyzed by immunoblotting technique, even in cases in which there are no detectable mutations in the KIT gene. The mechanisms whereby nonmutated (wild-type) KIT is maintained in an uncontrollably phosphorylated state are poorly understood and are likely to be fertile ground for future research, with therapeutic potential. Importantly, the vast majority of GIST cells at initial presentation demonstrate only a single site of mutation in the KIT gene; complex genetic changes in untreated GISTs at initial presentation are vanishingly rare. Gain of function mutations have been identified most commonly in exon 11 of KIT (up to 70% of cases), an exon that encodes the intracellular juxtamembrane domain of the KIT protein. Mutations in the KIT gene locus have also been described in other regions including (in decreasing order of prevalence) exon 9 (the KIT extracellular domain), exon 13 (kinase domain), and exon 17 (kinase domain).33-35 Structural biology studies have revealed the mechanism whereby normal (wild-type) KIT is kept in an autoinhibited con­ formation until ligand binds; mutational changes in conformation interfere with this autoinhibition and lead to a structural basis for the aberrant activation of the KIT kinase function.36,37 Another key advance in the understanding of GISTs has been the recognition that signaling through other uncontrolled kinases in addition to KIT could drive the neoplastic phenotype of GIST cells. Specifically, it is now recognized that approximately 5% of GIST cells are not through activation and aberrant signaling of the KIT receptor, but rather through mutational activation of the structurally related kinase known as the platelet-derived growth factor receptoralpha (PDGFRA).38,39 The definitive diagnostic criteria of uncommon CD117negative lesions that are nevertheless truly GISTs are currently somewhat obscure. GIST lesions can be heterogeneous in the expression of CD117, even within a single mass. It is possible that a needle biopsy could yield cells consistent with a GIST yet be CD117-negative simply by sampling bias alone. However, expert pathologists can also define a rare subset of GISTs ( collagenoma > café-au-lait macules > lipomas CNS tumors—meningiomas, ependymomas, schwannomas Smooth muscle tumors— leiomyomas, leiomyosarcomas Thyroid tumors—adenomas

AVERAGE FREQUENCY (RANGE), % OF ALL PATIENTS 97 (78-100) 81-82 80-100 (microscopic), 0-13 (symptomatic) 54 (20-61) 18 (7-31) 3 (1-6) 1 (1-12) 0-1 5 mEq/hr proposed to distinguish patients with ZES with or without previous gastric acid–reducing surgery, respectively, from patients without ZES. Left upper panel, Prominent gastric folds found on endoscopy in 92% of ZES patients, as compared with a normal subject. Left lower panel, Fasting serum gastrin levels from 309 National Institutes of Health (NIH) ZES patients and 2229 ZES patients from the literature. Fasting gastrin levels are expressed as a multiple (fold) of the upper limit of normal on the horizontal axis and as the cumulative percentage of patients with the indicated gastrin level on the vertical axis. Only 0.3% of NIH patients and 3.4% of all patients in the literature had normal values; 60% of patients had a 5 yr) of persistent symptoms (>50%) Bleeding, perforation, penetration (15%) No helicobacter pylori (PUD: 10%-50%) Family history of PUD or gerd Endocrinopathy (25%) Refractoriness to treatment In Patients with Persistent Diarrhea Malabsorption that is unexplained Abdominal pain (55%) Esophageal disease/symptoms (45%) Weight loss (15%) Lack of response to disease-specific treatments (gluten-free or lactose-free diet, antibacterial treatment) Family history of endocrinopathies or PUD (25%) Secretory diarrhea Signs Prominent gastric folds on UGI endoscopy or X-ray (90%) Multiple peptic ulcers or ulcers in unusual locations (1%-5%) Esophageal stricture caused by PUD (3%-4%) Complicated PUD (gastric outlet obstruction, 10%; perforation, 7%; recurrent bleeding, 5%-10%) GERD, gastroesophageal reflux disease; PUD, peptic ulcer disease; UGI, upper gastrointestinal; ZES, Zollinger-Ellison syndrome. From reference 72.

studies, in which up to 100% of patients presented with or developed a complication of advanced acid peptic disease (bleeding, penetration, esophageal stricture, perforation, obstruction), at present less than 30% of patients develop these complications, even though 71% have a confirmed history of peptic ulcer disease. Patients presenting with duodenal gastrinomas do not different clinically from patients with pancreatic tumors.124 In the 25% of patients with ZES who have MEN-I syndrome, the clinical presentation is similar to that of those with sporadic disease (see Table 32-5).47,48,67,72 Some important clues that should suggest MEN-I are the following: (1) a history of nephrolithiases and/or renal colic is much more frequent in patients with ZES with MEN-I than with sporadic ZES (47% vs. 4%); (2) patients with MEN-I present with ZES at a younger age (34 vs. 43 years); and (3) 72% have a family history of endocrinopathies. In a review of 107 patients with ZES with MEN-I prospectively studied at the NIH and 1007 cases from the literature,48 88% to 94% developed hyperparathyroidism, 31% to 60% pituitary disease, 6% to 30% various carcinoids (gastric, bronchial, thymic), and 6% to 16% other functional PETs, each of which may mask the diagnosis of ZES and be the presenting feature of the syndrome.

DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS

Despite the widespread availability of gastrin radioimmunoassays and the publication of more than 3300 articles on ZES, as well as most physicians’ general awareness of the features of ZES, the diagnosis of ZES continues to be delayed four to six years after onset of symptoms (see Table 32-5).46,48,72 This likely occurs for a number of reasons. ZES is an uncommon cause of peptic disease (1 to 3 new cases/ million/year). Initially, it is frequently clinically indistinguishable from patients with peptic ulcer disease (2300 cases/million/year), and GERD, which can affect 3% to 4%

of the population.28,67,112 In contrast to older studies, the initial symptoms of abdominal pain or GERD symptoms that most ZES patients currently exhibit are generally indistinguishable from those seen in peptic disease. Also, the characteristics of the ulcer itself (single, duodenal in location) do not distinguish it from those typically seen in patients with duodenal ulcer disease. Furthermore, the widespread use of proton pump inhibitors (PPIs) is complicating and delays the diagnosis of ZES. One study139 has analyzed the number of patients referred and diagnosed with ZES in two well-established referral centers (Università La Sapienza, Rome, and the NIH) before and after the introduction of PPIs. The widespread use of PPIs decreased referrals of patients with possible ZES by 62%, decreased by 40% the number of cases of ZES diagnosed, and was associated with a two- to six-fold increase in patients with a false diagnosis of ZES referred to the U.S. center. This false diagnosis occurred because chronic treatment with PPIs causes hypergastrinemia in 80% to 100% of all patients with peptic ulcer disease or GERD and it frequently reaches five times normal levels, a range seen in 60% of patients with ZES. The use of histamine H2 receptor antagonists often suggests the diagnosis of ZES because they often do not control acid hypersecretion in ZES with the conventional doses used to treat routine peptic ulcer disease or GERD. Thus, treatment failure with these drugs leads to a suspicion of the diagnosis of ZES.9 In contrast, PPIs mask the diagnosis of ZES because they control symptoms in most patients with ZES with conventional doses used in the treatment of peptic ulcer disease or GERD, and therefore a treatment failure rarely occurs with PPIs.71,140 There are a number of distinctive clinical and laboratory features that should suggest the diagnosis of ZES in a patient with acid peptic disease or diarrhea (see Table 32-6). A few important points will be briefly discussed. Diarrhea is now infrequent in patients with peptic disease or GERD because higher doses of antacids are rarely used, so its presence should lead to a suspicion of ZES because it occurs in 73% of ZES patients (see Tables 32-5 and 32-6). Diarrhea alone can be the presenting symptom in up to 27% of patients with ZES.28,72,112 MEN-I is a cause of ZES in 20% to 25% of patients and is characterized by endocrine abnormalities in multiple endocrine glands (parathyroid > pancreas > pituitary). Therefore, anyone with a personal or family history of endocrinopathies, or laboratory evidence of endocrinopathies, should be suspect of having ZES.46,48 Helicobacter pylori is an important cause of peptic ulcer disease (present in 80% to 100%); however, it is present in only 10% to 50% of patients with ZES.141 Prominent gastric folds on upper gastrointestinal (UGI) endoscopy should suggest the diagnosis of ZES (see Fig. 32-2). Whereas enlarged gastric folds were reported in only 13% to 30% in older studies of patients with ZES, an NIH prospective study72 involving 261 patients, in which all patients underwent UGI endoscopy, reported this finding in 94% of cases. This endoscopic finding particularly contrasts with the loss of gastric folds reported in many patients with acid hyposecretory disorders, resulting in fasting hypergastrinemia (e.g., atropic gastritis, pernicious anemia; see Chapter 51), and therefore should be an important clue to the possible presence of ZES. The diagnosis of ZES requires the demonstration of acid hypersecretion in the presence of hypergastrinemia (see Fig. 32-2).28,46,65,112,113,124 Therefore, to diagnose ZES, assessments of fasting serum gastrin levels and acid secretion are required. At the NIH, over 99% of 309 ZES patients studied had fasting hypergastrinemia and, in 2229 cases from the literature, 97% had fasting hypergastrinemia.142 Hence, if

Chapter 32  Endocrine Tumors of the Pancreas and Gastrointestinal Tract the fasting serum gastrin level is normal, especially in repeated determinations, the diagnosis of ZES is very unlikely. There are two specific exceptions to this general conclusion. First, in MEN-I patients after parathyroidectomy, correction of hyperparathyroidism may result in normalization of fasting gastrin levels in patients with ZES.67,79,143 Second, after gastrinoma resection, the fasting serum gastrin level may be normalized, even though the patient is not cured.144 In these latter two special cases and the rare case in which ZES is suspected clinically even though the fasting gastrin level is normal, a secretin provocative test should be performed. Because PPIs can elevate fasting serum gastrin levels up to three- to five-fold, a range that overlaps with gastrin levels in 60% of patients with ZES, it is difficult to diagnose ZES when the patient is taking PPIs.9,46,71,139 Therefore, if the fasting gastrin level is elevated while taking a PPI, the fasting gastrin level should be performed after stopping the PPI for at least one week (see Fig. 32-2). However, abruptly stopping PPIs can lead to the rapid development of peptic complications in a small percentage of ZES patients. In one prospective study of acid secretion in 235 ZES patients and review of 984 cases from the literature,113 99% of patients had a fasting gastric pH lower than 2; therefore, gastric pH should be determined when the fasting gastrin is repeated (see Fig. 32-2) to exclude physiologic hypergastrinemia (i.e., caused by hypochlorhydria or achlorhydria). The most common causes of physiological hypergastrinemia include atrophic gastritis, H. pylori infection, pernicious anemia, the use of potent acid suppressant drugs (particularly PPIs), chronic renal failure, and following gastric acid-reducing surgery.28,46,65,71 In patients with fasting serum gastrin levels less than 10-fold increased (i.e., usually 48 hrs) and potency, allow once or twice daily dosing in most patients (>95%) and are now the drugs of choice.140,150 Histamine H2 receptor antagonists are also effective; however, higher than conventional doses are required (mean dosages—cimetidine, 3.6 g/day; ranitidine, 1.2 g/day; famotidine, 0.25 g/day).110 All PPIs have been shown to be effective in ZES (omeprazole, lansoprazole, pantoprazole, esomeprazole, rabeprazole).149-152 PPIs are usually started with a dosage equivalent

503

504

Section IV  Topics Involving Multiple Organs to 60 mg/day of omeprazole, except in patients with complicated ZES–MEN-I, severe GERD, or previous Billroth II resection, for whom an initial dose equivalent to 60 mg twice daily of omeprazole is recommended.138 Sufficient antisecretory drug needs to be given to reduce acid hypersecretion to lower than 10 mEq/hr in patients without previous gastric acid-reducing surgery (5 µg/L in men and >10 µg/L in women) and the demonstration of elevated plasma GH-RF levels. In normal subjects and patients with acromegaly not caused by a GRFoma, plasma GH-RF–IR are within the normal to low-normal range, which in most laboratories is from 50 to 100 pg/mL.221,222 The lowest GH-RF–IR level in a patient with a proven GRFoma reported was 300 pg/mL. It has thus been suggested that a plasma GRF level higher than 300 pg/mL is strongly suggestive of the presence of a GRFoma.9 Plasma insulin-like growth factor I (IGF-I) is also elevated in patients with GRFomas.

TREATMENT

Tumor localization studies should be performed to evaluate the extent of disease (see later). In patients without metastatic disease to the liver, surgical resection of the GRFoma should be carried out. Prior to surgery and in those patients with nonresectable lesions, various agents may be helpful to reduce plasma growth hormone levels. Even though dopamine agonists such as bromocriptine are widely used in patients with classic acromegaly, they are able to reduce plasma GRF levels in only 25% of patients with GRFomas. Octreotide is the agent of choice.31,32,100-102,221,223,224 Treatment in a small number of cases suggests the long-acting forms of somatostatin (octreotide LAR or lanreotide SR) will also be effective. In most cases, but not all, octreotide significantly suppressed or normalized growth hormone and IGF-I levels and, in some cases, this was associated with pituitary shrinkage.9 The suppression of growth hormone secretion by octreotide was mainly caused by suppression at the pituitary level because plasma GH-RF levels never become undetectable. Surgical resection should be directed at the primary tumor, not the pituitary. Surgery has resulted in regression of the GRFoma syndrome in a small number of cases.7 The actual number of patients who achieve longterm cure is unknown.

PPOMAS AND NONFUNCTIONING PANCREATIC ENDOCRINE TUMORS DEFINITION

A PPoma is a tumor, usually of the pancreas, that secretes excessive amounts of pancreatic polypeptide (PP). The clinical symptoms are caused by local effects of the tumor itself, not by the actions of PP.225 Strictly speaking, a nonfunctioning PET is a tumor of the pancreas that has typical histologic features of a PET, is not associated with elevated plasma levels of any known peptide, and whose symptoms are entirely caused by the local effects of the tumor itself. The term nonfunctional PET is most widely used to indicate a PET occurring in a patient in whom there are no clinical symptoms caused by hormone overproduction. However, such nonfunctional PETs frequently secretes peptides. This would include PPomas and PETs secreting neurotensin, HCG subunits, ghrelin, chromogranin, or neuron-specific enolase.

PATHOPHYSIOLOGY AND PATHOLOGY

Except in patients with MEN-I,47,174 NF-PETs are usually large.9,22,225,226 In one series, the mean size was 4 cm.23 They are usually solitary tumors, except in patients with MENI.66,227,228 Sixty percent of sporadic NF-PETs occur in the pancreatic head. The malignancy rate in sporadic NF-PETs varies from 38% to 92% in different series. Histologically, NF-PETs are similar and cannot be differentiated from other PETs, even by immunocytochemistry. In one series of 30 NF-PETs, 50% had insulin-like immunoreactivity (IR), 30% glucagon IR, 43% PP IR, 13% somatostatin IR, and only 13% produced none of these peptides. Elevated plasma levels of chromogranin A and B are found in 69% to 100% of patients with these tumors, neuron-specific enolase in 31%, PP in 50% to 75%, α-HCG in 40%, and β-HCG in 20%. An elevated plasma PP level in a patient with a pancreatic tumor is suggestive of a PET because none of 53 patients with adenocarcinoma of the pancreas had elevated plasma PP levels. Infusions of PP into animals and humans have shown this peptide to have numerous biological effects, including the following: a net secretory effect on water and electrolytes in the small intestine; inhibitory effects on fluid, electrolyte, and enzyme secretion by the pancreas; effects on esophageal, gastric, intestinal, and gallbladder motility; and metabolic effects, such as inhibition of somatostatin or insulin release. In various studies,225 patients with PPomas have been reported to have symptoms attributed to elevated plasma levels of PP, including persistent watery diarrhea, diabetes mellitus, weight loss, decreased gastric acid secretion, peptic disease, flushing, and acute psychosis.9 Furthermore, the plasma PP level is frequently elevated in other symptomatic PETs. However, the symptoms of these patients do not differ from those without elevation of plasma PP levels; thus, it is now generally agreed that plasma elevations of PP are not associated with specific symptoms.35 At present, it is unclear why patients who have elevated plasma levels of PP do not have specific symptoms.

CLINICAL FEATURES AND DIAGNOSIS

Typically, the patient with a NF-PET is 40 to 60 years of age. These tumors occur approximately equally in both genders. The median delay in the time from diagnosis to the first symptoms varies from 0.5 to 2.7 years. In different studies, 36% to 56% of patients with NF-PETs presented with abdominal pain, 27% to 40% with jaundice, 28% to 46% with weight loss, and 8% to 40% with abdominal

Chapter 32  Endocrine Tumors of the Pancreas and Gastrointestinal Tract mass.9,22,23,225,226,229,230 In 16% to 35% of patients, the tumors were found incidentally at surgery and the remaining patients presented with a variety of symptoms caused by the tumor mass. The main diagnostic challenge is differentiating the NF-PET from a nonendocrine pancreatic tumor and also in determining whether the tumor is associated with a symptomatic tumor syndrome (e.g., insulinoma, glucagonoma, gastrinoma). Elevated plasma levels of PP do not establish the diagnosis of a PPoma even when a pancreatic mass is present. Plasma PP levels are reported to be elevated in 22% to 71% of patients with functional PETs in various studies,28,225 as well as in nonpancreatic carcinoid tumors. In one large study,9 plasma PP levels exceeded a value of 1000 pg/mL in 45% of patients with various PETs, with this elevation in 32% of gastrinomas, 21% of insulinomas, 57% of glucagonomas, 74% of VIPomas, 33% of somatostatinomas, and 45% of carcinoid tumors. Furthermore, elevated plasma levels of PP can occur in other situations such as old age, after bowel resection, with alcohol abuse, during certain infections, in chronic noninfective inflammatory disorders, acute diarrhea, chronic renal failure, diabetes, chronic relapsing pancreatitis, hypoglycemia, or after eating. To increase the specificity of an elevated plasma PP level for a pancreatic tumor, an atropine suppression test has been proposed. In one study of 48 patients with elevated plasma PP levels, atropine (1 mg IM) did not suppress the levels in any of the 18 patients with PETs, but did suppress the level by 50% or more in all patients without tumors. Somatostatin receptor scintigraphy (SRS), discussed in detail in the following section on tumor localization, has also been shown to be useful in distinguishing pancreatic adenocarcinoma from a NF-PET.22,100-103,230,231

TREATMENT

With NF-PETs, treatment needs to be directed only at the tumor itself because no hormonal syndrome is present. Tenyear survival is better in patients with smaller tumors of 3 cm or smaller at presentation (80%) than in larger tumors (40%) in patients who were asymptomatic when the tumor was discovered (70% vs. 50% if symptomatic) and if no metastases were present (75% vs. 25% if there were metastases).9,23 Overall survival is 30% to 63% at five years, with a median survival of 72 months.22 In one series, a diagnosis of a NF-PET was never made preoperatively. Of 25 cases in this series, a Whipple pro­ cedure was done in five patients (20%), partial or total pancreatectomy in 25%, and tumor excision in 10%. The remaining patients had a biopsy only. The survival rates were 60% at three years and 44% at five years.9 In eight other studies, curative resection was attempted in 26% to 79% of patients with NF-PETs, with a five-year survival of 44% to 63% and a median survival varying from 2.5 to 4.8 years.226,229 The cure rate of these tumors at present is generally low. In one study,23 75% of patients undergoing resection were alive at the end of the study (mean follow-up, 4.4 years), which was better than the 47% survival in those not undergoing any surgical resection.

ing intestinal motility; and stimulating jejunal and ileal fluid, electrolyte, pancreatic protein, and bicarbonate secretion. Clinical features of patients with a PET and possible neurotensinomas include hypokalemia, weight loss, diabetes mellitus, cyanosis, hypotension, and flushing. The existence of a specific neurotensinoma syndrome has been questioned. In one study of patients with gastrinomas,35 those with or without an elevated neurotensin level did not clinically differ. Patients with PETs with Cushing’s syndrome (ACTHoma) have been reported.10,11,44,232 In various reviews, 4% to 16% of ectopic Cushing’s syndrome cases were caused by a PET.10 Cushing’s syndrome was reported in 19% of patients with gastrinoma and MEN-I; in these patients, the disease was of pituitary origin and was mild. Cushing’s syndrome occurs in 4% to 5% of sporadic gastrinoma cases; in these patients, Cushing’s syndrome was severe, caused by ectopic ACTH production, usually occurred with metastatic tumors, responded poorly to chemotherapy, and was associated with a poor prognosis. In a large prospective study,45 the development of Cushing’s in patients with ZES was shown to be an independent predictor of poor survival, with patients having a mean survival of 1.7 years after its onset. Cushing’s syndrome as the only manifestation of a PET occurs occasionally and may precede any other hormonal syndrome. In every case in one series, ectopic Cushing’s syndrome caused by a PET only occurred in the presence of metastatic disease. Hypercalcemia caused by a PET secreting PTH-rP or by an unknown hypercalcemic substance that mimics the action of PTH and causes hyperparathyroidism has been reported.9,16,233 The tumors are usually large and metastatic to the liver by the time of diagnosis, although in some cases resection of the PET resulted in cure or remission. PETs causing the carcinoid syndrome are rare but have been well described (see Chapter 31). Pancreatic carcinoid tumors are usually large and 67% to 88% are malignant.9,12,183 The carcinoid syndrome is present in 23% to 65% of these patients and in 23% in one review of 156 pancreatic carcinoids in the literature. Even though foregut carcinoids, which include pancreatic PETs,170 may lack Dopa decarboxylase, the enzyme that converts 5-hydroxytryptophan to serotonin (5-hydroxytryptamine), 84% of patients with PETs causing the carcinoid syndrome have increased urinary 5-hydroxyindoleacetic acid (5-HIAA) levels, which can be used for their detection. There is only one case described of a malignant PET secreting renin, which resulted in hypertension, although renin-secreting juxtaglomerular cell tumors, Wilms’ tumors, and ovarian tumors have been described.13 Similarly, there is only one reported case of a malignant PET secreting erythropoietin resulting in polycythemia.15 Ectopic release of erythropoietin has also been reported with pheochromocytomas, renal cell cancers, posterior fossa tumors, hemangioblastomas, and Wilms’ tumors. PETs secreting LH have been described, which can result in masculinization in females and loss of libido in males.9,14 PETs may contain ghrelin IR material and secrete ghrelin234 but, as noted in the historical section earlier, these do not seem to produce a distinct clinical syndrome.

OTHER PANCREATIC ENDOCRINE TUMORS In a few patients with PETs secreting the peptide neurotensin, a neurotensinoma syndrome has been proposed. Neurotensin is a 13–amino acid peptide, originally isolated from bovine brain, that has a number of biological effects, including causing tachycardia, hypotension, and cyanosis; affect-

TUMOR LOCALIZATION It is essential for the correct management of patients with PETs that extent (stage) of the tumor and often the site of the primary tumor be established (see Chapter

515

516

Section IV  Topics Involving Multiple Organs Table 32-10  Sensitivities of Imaging Methods for Localization of Pancreatic Endocrine Tumors (%)* Other Pancreatic Endocrine Tumors IMAGING STUDY Abdominal ultrasonography CT MRI Arteriography Selective venous sampling   Portal venous sampling   Post-intraarterial calcium Somatostatin receptor scintigraphy Endoscopic ultrasonography Intraoperative ultrasonography

Insulinomas

PRIMARY

LIVER METASTASES

30 (0-64) 31 (16-60) 10 (0-25) 60 (41-81)

22 42 27 70

(6-70) (33-100) (21-100) (35-100)

44 (14-76) 70 (35-100) 80 (67-100) 71 (33-86)

80 (63-96) 88 54 (15-56) 81 (57-92) 89 (80-100)

71 (17-94) ND 70 (58-77) 70 (40-100) 91 (80-100)

NA NA 93 (88-100) NA NA

*Values shown are mean values (and range). CT, computed tomography; NA, not applicable; ND, no data available. From references 9, 26, 28-30, 56, 69, 107, 242, and 253.

31).22,30,69,71,110,181,199,235 The ability of standard imaging studies, such as computed tomography (CT), ultrasonography (US), and MRI to localize the PET depends on the tumor size. CT and MRI localize less than 10% of PETs smaller than 1 cm, 30% to 40% of tumors 1 to 3 cm, and more than 50% of PETs larger than 3 cm in diameter.9,26,28,69 PETs are hypervascular tumors and the ability to localize different PETs appear to be influenced more by tumor size and location than PET type. Insulinomas are usually small (4 cm).29,46,58,65,71,73-77,135 Furthermore, insulinomas, nonfunctioning PETs, glucagonomas and VIPomas in adults are almost always located in the pancreas, whereas somato­ statinomas, gastrinomas, and VIPomas in children frequently occur extrapancreatically. Table 32-10 lists the sensitivities of the various localization studies to detect insulinomas, as well as the primary tumor and liver metastases of other PETs. The results with insulinomas are presented separately because they are almost always benign, small (85%)239,240 and, in some studies, is clearly more sensitive than SRS, especially for insulinomas. EUS in MEN-I patients can detect PETs not seen by other modalities and may prove particularly useful for serial assessments to determine changes in the size of small lesions,174,228,250,251 which are frequently not routinely resected (see surgical section). Figure 32-8 shows the ability of EUS to identify an insulinoma in the pancreatic tail. EUS, especially of the pancreas, requires considerable expertise, whereas SRS can be performed in most nuclear medicine departments. Furthermore, EUS6 may yield false-positive results, although the false-positive rate is not clear, especially for tumors outside the pancreas. Cytologic confirmation of a functional PET is rarely needed, but EUS-guided fine-needle aspiration for cytology may sometimes be helpful in distinguishing a nonfunctional PET from another pancreatic tumor.252 EUS also provides information on adjacent lymph nodes, but provides no information on more distant sites of metastases, which may affect the surgical approach.56 Neither SRS nor EUS appear to identify small extrapancreatic PETs, especially duodenal gastrinomas.56,69 Some studies recommend SRS be used in combination with EUS in patients with gastrinomas to improve sensitivity.99 Functional localization of PETs by determining the site of the maximal hormonal gradient using selective venous blood sampling still remains a useful technique in some cases.105,107,108 Originally, this approach involved considerable expertise because transhepatic catheterization of portal venous tributaries was required. Furthermore, procedurerelated complications occurred in 20% of patients. A simplified method has been described.253 By using various secretagogues, such as secretin for gastrinomas or calcium for insulinomas, selective intra-arterial injection during angiography with hepatic venous hormone sampling has helped localize gastrinomas and insulinomas. This latter procedure is easier to perform than portal venous sampling, has fewer complications, and has equal to greater sensitivity and thus has replaced portal venous sampling. During this procedure the secretagogue is injected selectively into various arteries (superior mesenteric, splenic, right and left hepatic, gastroduodenal arteries) and, when the selected artery supplies the area of the PET, there is a sharp increase in the hepatic venous hormone concentration with the secretagogue injection. Figure 32-9 shows the ability of calcium infusion to localize an insulinoma accurately to the pancreatic body, a location that was suspected from the MRI study. In a comparative study in insulinomas, which are frequently smaller than 1 cm in diameter and difficult to localize, the intra-arterial calcium test with selective hepatic venous sampling was positive in 88% of patients, US in 9%, CT in 17%, MR in 43%, selective angiography in 36%, and portal venous sampling in 67%. Other studies support the conclusion that the intra-arterial calcium test with hepatic venous insulin sampling is a highly sensitive method to localize PETs. Furthermore, the intra-arterial calcium test may allow differentiation of the cause of the hypoglycemia and whether it is caused by nesidioblastosis or insulinoma.254 This is becoming an increasingly important distinction with the increased occurrence of hypoglycemia after gastric bypass surgery for obesity, which is primarily caused by nesidioblastosis,92 but on occasion can be caused by an insulinoma.255 Calcium infusion may also increase the release of hormones from VIPomas, PPomas, glucagonomas,

517

518

Section IV  Topics Involving Multiple Organs BONE SCAN

R

SRS

L

R

L

Liver mets

Figure 32-7.  Comparison of the ability of bone scanning (left panel) and somatostatin receptor scintigraphy (SRS; right panel) to localize bone metastases (mets) in a patient with a metastatic pancreatic endocrine tumor (PET). The PET was secreting both glucagon and gastrinoma. This patient had bone metastases in the lumbar spine, left pelvis, and left scapula. The SRS (right panel) demonstrates the metastases in each area (solid arrows), whereas the bone scan (left panel) shows only a questionable metastasis in the left scapula (solid arrow). The primary tumor in the pancreatic tail is shown by the dotted arrow on the SRS. Liver mets are seen on SRS. These results demonstrate the greater sensitivity of SRS in detecting bone metastases in patients with malignant PETs.134

A

B

Figure 32-8.  Endoscopic ultrasonography (EUS) for localization of pancreatic endocrine tumors (PETs). A, The EUS balloon is in the stomach. EUS shows a sonolucent pancreatic endocrine tumor (insulinoma) 2 cm in diameter (three arrows) near the splenic vein (small arrow, labeled V) B, EUS of the pancreatic neck region showing a 0.7-cm pancreatic endocrine tumor (gastrinoma). (Courtesy of Dr. Norman Thompson, Ann Arbor, Mich.)

Chapter 32  Endocrine Tumors of the Pancreas and Gastrointestinal Tract tumors that cannot be identified by other methods68,122,123,125,126 and its use has been shown to increase the disease-free survival rate (see Fig. 32-4). Newer modalities are being used that may increase the sensitivity for detection and localization of primary and metastatic tumors in patients with PETs. An example is the use of hybrid scanning, particularly combining CT and SRS,57,258,259 and the use of scanning, either alone or combined with MRI or CT. SRS alone does not allow exact localization of a lesion within a given region, such as the distinction between a gastrinoma in the duodenum versus the pancreatic head. However, SRS is increasingly being combined with CT (SRS-CT hybrid scanners), which allows better definition of the exact location. Similarly, numerous recent studies, primarily from Europe, have shown that PET scanning for PETs using 11C-5-hydroxytryptophan or 68Galabeled somatostatin analogs may have improved sensitivity over SRS alone or conventional imaging studies (see Chapter 31).241,260-263 In a number of studies,260,264 PET scanning was reported to have greater accuracy and/or sensitivity than SRS alone. At present, PET is not approved for this use in the United States and has limited availability.

T

MRI

INTRA-ARTERIAL CALCIUM GLUCONATE INJECTION 1000

Plasma insulin level (µM/mL)

Splenic artery 800

MANAGEMENT OF METASTATIC PANCREATIC ENDOCRINE TUMORS

600

400

200

Hepatic artery GDA SMA

0 Pre

20

40

60

90

Seconds (postinjection) Figure 32-9.  Localization of an insulinoma by MRI (top) and by intraarterial calcium injection and hepatic venous sampling for insulin concentrations (bottom). Top, On the MRI scan, a probable insulinoma (T) was seen in the pancreatic body, an area supplied by the splenic artery. Bottom, Ca gluconate (10%; 0.025 mEq Ca/kg) was selectively injected into the superior mesenteric, gastroduodenal, common hepatic, and splenic arteries. Venous samples were collected prior to and 30, 40, 60, and 90 seconds postinjection and assayed for insulin concentration. A significant (>50%) increase in hepatic venous insulin concentration occurred at all times after calcium injection into the splenic artery but not after calcium injection into the other vessels. At surgery, an insulinoma was found in the pancreatic body. GDA, gastroduodenal artery; SMA, superior mesenteric artery.

GRFomas, or somatostatinomas, so that a similar approach may be useful with these tumors.9 At surgical exploration a number of procedures are helpful in localizing different PETs. Intraoperative ultrasonography (IOUS) is useful, particularly for localizing intrapancreatic PETs, and will identify some insulinomas not found by other means.256,257 It is essential for the surgeon to perform a Kocher maneuver to palpate the pancreatic head carefully as well as mobilize the pancreatic tail to allow its careful palpation. Lastly, in patients with gastrinomas, a routine duodenotomy is essential to detect small duodenal

The treatment of all metastatic PETs is considered together because in most aspects it is similar for each of the tumors (see Chapter 31). Metastatic PETs are relatively slowgrowing compared with other more common malignant adenocarcinomas.45,135,265,266 The long-term natural history of most functional PETs is not known, because until recently effective treatment for the clinical syndrome with functional PETs was not available, and therefore patients often died of complications of the hormonal excess rather than the tumor. However, this is changing with the recent availability of agents such as long-acting somatostatin analogs. In contrast, with nonfunctional tumors and gastrinomas, for which effective therapy for the gastric hypersecretion has existed for more than 30 years, the natural history of the malignant tumor itself has been assessed. Because of the similar biological behavior of all PETs, the assessment of these latter tumors will likely also provide insights into the natural history of all these less common malignant PETs.28,135 Insights into PET biology and natural history have identified prognostic factors that determine survival.25 An awareness of these factors is essential in planning the type and timing of treatment of advanced disease.

TUMOR BIOLOGY, PROGNOSTIC FACTORS, AND SURVIVAL

With increased ability to control the hormone excess state, the survival of patients with functional PETs is increasingly being determined by the tumor’s biology and natural history of the PET’s growth pattern.25,45,54,265 Information has mainly been obtained from studies of the natural history and prognostic factors determining survival in patients with nonfunctional PETs22 and from patients with gastrinomas because of the ability to control the gastric hypersecretion medically in these patients for a number of years.45 Studies have demonstrated that PETs grow at different rates in different patients.25,45,135,137,266 As noted, in approximately 25% of patients with gastrinomas followed long term, the gastrinoma demonstrates aggressive growth, whereas in the remaining 75% growth is indolent or no

519

520

Section IV  Topics Involving Multiple Organs Table 32-11  Prognostic Factors Associated with Decreased Survival in Patients with Various Pancreatic Endocrine Tumors Female gender Absence of MEN-I syndrome Liver metastases Extent Growth Lymph node metastases Bone metastases Incomplete tumor resection Nonfunctional tumor Ectopic Cushing’s syndrome (gastrinomas) Depth of tumor invasion Tumor size Histologic features High nuclear atypia Poor tumor differentiation High growth indices (Ki-67 index > 2%, PCNA expression) Capsular invasion Vascular or perineural invasion Necrosis Flow cytometric features (i.e., aneuploidy) Laboratory findings Elevated serum chromogranin A (in some studies) Elevated serum gastrin level (gastrinomas) Lack of progesterone receptors Ha-Ras oncogene or p53 overexpression Molecular biological features High HER2/neu gene expression (gastrinomas) High 1q loss of heterozygosity (gastrinomas) Increased EGF or IGF receptor expression (gastrinomas) Chromosomal instability CGH findings (loss = 1p, 3p, 3q, 6q, 9q, 12q; gains = 7q, 17q, 17p, 20q) Gene microarray studies may also have prognostic potential (see text). EGF, epidermal growth factor; IGF, insulin-like growth factor; MEN-I, multiple endocrine neoplasia type I; PCNA, proliferating cell nuclear antigen; PET, pancreatic endocrine tumor. From references 40, 45, 53-56, 100-102, 135, 265, 266, 268, and 314-317.

growth occurs. Similarly, aggressive growth occurred in only 40% of patients with liver metastases and all deaths occurred in this subset of patients. At present, the molecular basis for this difference in growth remains unclear. In different PETs, prognostic factors have been defined (Table 32-11). The most important prognostic factor in all studies was the development of liver metastases.77,267,268 In one large study involving 221 patients with gastrinomas, the 15-year survival for all patients was 90%, without liver metastases was 96%, and with liver metastases was 26%.45 The development of liver metastases, their extent (one lobe, both, diffuse), presence of bone or lymph node metastases, larger primary tumor size, primary location (pancreatic gastrinomas have a worse prognosis than duodenal), various histologic features, laboratory, flow cytometric, and molecular features all have predictive value. Recently, microarray analysis53,59,60,62 has been used to attempt to identify genes that contribute to aggressive growth of PETs. In these studies, a large number of genes have been found to be up-regulated or down-regulated in association with aggressive growth behavior, but none have been shown to be clinically useful for management of patients. Because of the variable growth rate of different PETs and the importance of prognostic factors in patient management, there have been a number of proposed classification systems for PETs and, for the first time, a TMN classification50 (see Chapter 31). Some been shown to have prognostic value.269-272

Two of the most important clinical outcomes are the development of bone metastases and ectopic Cushing’s syndrome, both of which were independent predictors of poor survival, with a mean survival from their onset of less than two years.25,45 Most would agree that treatment directed at metastatic disease that is increasing in size is indicated, as well as treatment to prevent the development of metastases. However, there is currently no agreement about what type of therapy is most appropriate for patients with metastatic disease, when therapy should be started, and even the efficacy of various therapies. Chemotherapy,273-276 debulking (cytoreductive) surgery, with or without chemotherapy,56,176,177 hepatic arterial embolization with or without chemoembolization,277-279 hormonal therapy with longacting somatostatin analogs,100-102,183,280 interferon-α, liver transplantation,183,281-283 and targeted radiotherapy using radiolabeled somatostatin analogs284 have all been reported to be useful in some cases (see Chapter 31).

CHEMOTHERAPY

Most studies of chemotherapy in metastatic PETs have included mixtures of patients with functional PETs, occasionally nonfunctional PETs, and in some cases patients with carcinoid tumors.28,273,274 Results from these studies are limited in a number of ways. There are often small numbers of cases. Some studies have suggested that responsiveness to chemotherapy is equal in the different PETs, but others have suggested there may be important differences. For example, up to 93% of metastatic glucagonomas are reported to respond to dacarbazine (DTIC) including some complete remissions, whereas DTIC produces a low response rate in other PETs. Similarly, up to 90% of VIPomas respond to streptozotocin, whereas only 5% to 40% of metastatic gastrinomas respond to streptozotocin, with no complete responses. The current recommendation is the combination of streptozotocin and doxorubicin.273-275 This recommendation is based on two studies from the Eastern Cooperative Oncology Group (ECOG) published in 1980286 and 1992.285 The study in 1980 demonstrated that streptozotocin plus 5fluorouracil (5-FU) is more effective than streptozotocin alone. The 1992 study demonstrated that streptozotocin plus doxorubicin causes tumor regression in 69% of patients, which is significantly better than the 45% with streptozotocin alone or 30% with chlorozotocin. Furthermore, the patients treated with streptozotocin and doxorubicin had a significantly better survival. Streptozotocin is a glycosamine nitrourea compound originally derived from a Streptomyces species, and in preclinical studies was found to have cytotoxic effects on pancreatic islets. Streptozotocin was found to have clinical effectiveness against a PET in 19689 and since then has been used as the initial agent either alone or in combination with other agents for treating metastatic PETs. In various series, streptozotocin alone produces an objective tumor response in 36% to 62%.286 In contrast, other agents, such as doxorubicin, DTIC, tubercidin, etoposide, and carboplatin, have generally had a lower response rate of 6% to 33%. Streptozotocin causes nausea and vomiting in almost all patients and transient dose-related renal dysfunction, including proteinuria (40% to 50%) and a decrease in creatinine clearance, as well as abnormalities in hepatic function, leukopenia, and thrombocytopenia in 6%. In the 1992 ECOG study, nine patients developed renal failure and seven required dialysis. The nausea and vomiting can now be controlled in almost all patients using 5-HT3 receptor antagonists such as ondansetron. Chlorozotocin is structurally closely related to streptozotocin but causes less nausea and vomiting. When given alone or combined with

Chapter 32  Endocrine Tumors of the Pancreas and Gastrointestinal Tract 5-fluorouracil, it gives similar results to those seen with streptozotocin. Because of the limited effectiveness of single agents, various combinations have been investigated.273,274 The combination of streptozotocin and 5-fluorouracil was found to be more effective than streptozotocin alone. However, in a later study, streptozotocin plus 5-FU was less effective than streptozotocin plus doxorubicin.285 In the 1980 ECOG study,286 the 42 patients treated with streptozotocin had a 36% response rate, with 12% showing a complete response, whereas with streptozotocin plus 5-FU, 63% demonstrated a response, with 33% having a complete response. Response rates in this study between different functional tumors or between functional and nonfunctional tumors did not differ. In more recent prospective studies, the response rate with streptozotocin plus 5-FU was significantly lower in patients with metastatic gastrinomas to the liver. No patient had a complete response and there was no difference in survival for responders and nonresponders. At present, the difference in response rates between the early study and more recent ones remains unexplained. Streptozotocin has been used in combination with other agents such as doxorubicin or tubercidin in small numbers of cases, with response rates reported from 20% to 100%. Streptozotocin combined with 5-FU, with doxorubicin, with both these agents, with 5-FU plus tubercidin, or with doxorubicin and cisplatin has been used in different studies.9,273-275,285,286 Only the combination of streptozotocin and doxorubicin are established as superior to streptozotocin and 5-FU. The combination of etoposide and cisplatin has been evaluated in patients with PETs and carcinoid tumors.273,274,287,288 In one study,287 12 of 18 anaplastic neuroendocrine tumors, 2 of 14 PETs, and 0 of 13 of metastatic carcinoid tumors demonstrated partial to complete regression. In a second study,288 only 1 of 12 patients with a welldifferentiated neuroendocrine tumor responded, whereas 42% of 41 patients with a poorly differentiated tumor showed an objective tumor response. Hematologic toxicity occurred in 60% and there was one treatment-related death.

SURGICAL TREATMENT

Systematic removal of all resectable tumor (debulking or cytoreductive surgery) has been recommended, if possible, for all PETs, including gastrinomas, VIPomas, glucagonomas, and somatostatinomas.128,176,177,181,199,276,289 In various studies in patients with advanced PETs, cytoreductive surgery is reported to result in occasional cures, five-year survivals of 75% to 80% in resected patients, and increased survival over those not undergoing resection.56 Unfortunately, such resection is possible in only a small proportion of patients (5% to 15%). Even though this approach is recommended, it is not clear whether such an approach actually increases survival. This approach may be required in patients with symptomatic PETs in whom octreotide or the use of chemotherapy alone cannot reduce plasma hormone levels sufficiently to control symptoms.

HEPATIC ARTERY EMBOLIZATION AND CHEMOEMBOLIZATION

Hepatic artery embolization with or without postocclusion chemotherapy has been used successfully in patients with metastatic PETs to the liver (see Chapter 31).277-279 Because the liver derives only 20% to 25% of its blood supply from the hepatic artery (the rest comes from the portal vein), and because most PETs are vascular with an arterial supply, hepatic artery embolization has been possible. In some studies, 68% to 100% of treated patients have symptomatic

improvement. Chemotherapy using doxorubicin or other chemotherapeutic agents in iodized oil, combined with gelatin or sponge particles, has been reported to improve symptoms in 66% to 100% of patients and decrease tumor size and/or hormone levels in 37% to 100% of patients. Almost all patients report abdominal pain, nausea, vomiting and fever, usually lasting three to ten days, with severe complications occurring in 10% to 15%, including hepatic failure, acute renal failure, infection, and death. With the availability of SRS, it is now possible to assess the extent of metastatic disease easily. In a patient with diffusely metastatic disease to the liver, with minimal or no bone metastases, in whom hormone symptoms cannot be controlled by octreotide, chemotherapy, or other medical treatments, this hepatic artery therapy should be considered.

RADIOFREQUENCY ABLATION

Radiofrequency thermal ablation (RFA) works by converting RF waves into heat, which results in cellular destruction at temperatures higher than 60°C.279,290,291 RFA can be applied to liver metastases if their number is limited (usually less than five) and if they are not too large (usually 80%), have lasted more than two years and the procedure-related morbidity is low (50% decrease in hormone levels) and 12% had a decrease in tumor size, with a mean duration of 20 months (range, 2 to 96 months). Disease stabilization is seen in 75% to 80% of patients with metastatic neuroendocrine tumors. Studies295 have demonstrated that interferon-a can induce an increase in bcl-2 expression in NETs, which may contribute to the tumoristatic effect by stabilizing cells at the G0 phase of the cell cycle (see Chapter 3). These results suggest that interferon-a, similar to somatostatin analogs, may possibly extend survival by decreasing tumor growth rate. Interferon-α has been used in combination with somatostatin analogs,296-298 including randomized, prospective studies, to compare its antigrowth effect with either agent alone. The results did not support the results of nonrandomized studies, which had shown an enhanced antitumor action of the combination over

521

522

Section IV  Topics Involving Multiple Organs each drug alone. A number of methodologic issues have been raised with these studies, and thus it remains controversial whether this drug combination is better than either agent alone.

LIVER TRANSPLANTATION

Liver transplantation has been carried out in a small number of patients with metastatic PETs.281-283,299 In a review of 103 cases of patients with malignant NETs who underwent liver transplantation (including 48 PETs), the five-year survival rate was 45%. However, the recurrence-free survival rate was low (200/mm3 favors common bacteria and other nonopportunistic diseases; CD4 count 1500 U/L) is now mostly related to zone three necrosis from circulatory causes such as SOS and hypoxic hepatitis, and not to infection. However, acute hepatitis caused by HSV, VZV, adenovirus or HBV can be fatal after HCT.108,157,158 Except for sporadic cases of

Chapter 34  Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation adenovirus hepatitis, infections by these other viruses have become rare because of antiviral prophylaxis and preemptive treatment. Hepatic infections caused by CMV and HCV are seldom severe.114 Despite acyclovir, HHV-6 and HHV-8 reactivation have been associated with the development of fever, rash, and hepatitis in HCT recipients.159 Other non­ infectious causes of acute hepatocellular injury include a hepatitic presentation of GVHD and drug toxicity.155,156 When there is uncertainty about the cause of rising serum ALT, DNA blood tests for herpesviruses, adenovirus, and HBV; transvenous measurement of the wedged hepatic venous pressure gradient; and liver biopsy are indicated (see Fig. 34-5). If acyclovir is not being given, it should be started empirically, particularly if the patient presents with abdominal complaints typical of VZV infection (see Fig. 34-5E).160 Adenovirus hepatitis should be suspected if the patient has concomitant pulmonary, renal, bladder, or intestinal symptoms (see Fig. 34-5F); the most effective treatment is cidofovir when given early in the course of infection.161,162 Fulminant hepatitis B may develop during immune reconstitution in patients at risk, but can be prevented with prophylactic lamivudine or adefovir.108,163 If severe hepatitis B reactivation does occur, usually because a diagnosis of HBV was not made prior to HCT,117 antiviral therapy with the most potent antiviral drug available should be initiated immediately, although progression to fatal liver failure is not uncommon.164 Fulminant hepatitis B has also occurred after discontinuation of prophylactic antiviral therapy; all HBV-infected patients, particularly those with high pretransplant HBV DNA levels, should be monitored following antiviral drug withdrawal.165,166 HCV infections are seldom severe; asymptomatic elevations of serum ALT are commonly seen from days 60 to 120, frequently coinciding with the tapering of immunosuppressive drugs.114 Therapy directed at chronic HCV infection should be considered once the patient has ceased all immunosuppressive drugs and has no evidence of active GVHD.167 About a third of HCV-infected transplant survivors will develop cirrhosis over 20 to 30 years. Fungal and Bacterial Infections Antifungal prophylaxis has significantly reduced the incidence of invasive fungal disease in HCT recipients, particularly in those requiring ongoing immunosuppression for treatment of GVHD.106,168,169 If invasive fungal disease does occur, infection with resistant Candida species or molds is likely. Signs are fever and tender hepatomegaly, with increased serum alkaline phosphatase levels. Highresolution CT or MRI may demonstrate multiple fungal abscesses, and serological tests for fungal antigens may be useful for diagnosis. Antifungal drugs and return of neu­ trophil function after HCT can lead to resolution of previously treatment-refractory mold infection.105 Bacterial liver abscesses are rare in HCT recipients, probably because of the high use of systemic antibiotics; however, latent mycobacterial infection may reactivate within the liver with prolonged immunosuppressive therapy. Disseminated bacille Calmette-Guérin (BCG) infection with liver involvement has been reported. Disseminated clostridial infection and gallbladder infection with gas-producing organisms may lead to air in the liver and biliary system. Gallbladder and Biliary Disease Gallbladder sludge (calcium bilirubinate) is almost universally present in HCT patients. Although sludge is usually asymptomatic, passage through the bile duct may cause epigastric pain, nausea, and elevated serum liver enzymes. Endoscopic papillotomy is rarely indicated. Biliary sludge

may be a cause of acute “acalculous” cholecystitis, acute pancreatitis, and bacterial cholangitis.170,171 Acute cholecystitis is uncommonly seen in HCT recipients and is frequently acalculous. Cholecystitis in this setting may also be due to leukemic relapse with gallbladder involvement or infection by CMV or fungi. Diagnosis is difficult because of the high frequency of gallbladder abnormalities on ultrasonography following HCT. Pericholecystic fluid, gall­ bladder wall necrosis, or localized tenderness suggests cholecystitis. A radionuclide bile excretion study, with morphine infusion to enhance gallbladder filling, can be useful; nonvisualization of the gallbladder suggests cholecystitis. Biliary obstruction is a rare event, caused by a variety of disorders (e.g., lymphoblastic infiltration of the bile duct and gallbladder in EBV-LPD; CMV-related biliary disease; dissecting duodenal hematoma complicating endoscopic biopsy; inspissated biliary sludge; and leukemic relapse [chloroma] in the head of the pancreas).170,171 Therapeutic endoscopic cholangiopancreatog­ raphy may be needed if there is cholangitis or per­sistent obstruction. Malignant Disorders EBV-LPD was commonly seen two to four months after HCT, particularly in recipients of HLA-mismatched T cell– depleted grafts and after potent anti–T cell therapies. Liver involvement occurred in more than 50%, manifested by abnormal serum alkaline phosphatase and massive hepa­ tosplenomegaly. EBV-LPD is now infrequent because of EBV-DNA surveillance and preemptive treatment with rituximab. Idiopathic Hyperammonemia and Coma A syndrome of hyperammonemia and coma has been described in patients who received high dose chemotherapy, including conditioning for HCT.172 Patients present with progressive lethargy, confusion, weakness, incoordination, vomiting and hyperventilation. The diagnosis is confirmed when the plasma ammonia exceeds 200 µmol/L and there is no evidence of liver failure. This syndrome is rare, but is associated with a high mortality. Its pathogenesis involves the unmasking of a latent genetic disorder similar to ornithine transcarbamylase deficiency.173,174

Gastrointestinal Bleeding

Bleeding that does not require transfusion is very common after HCT, particularly when platelet counts are low. Causes include retching-induced trauma to the esophageal or gastric mucosa (Fig. 34-6A), mucosal injury from conditioning therapy, peptic esophagitis, C. difficile colitis, anal fissures and hemorrhoids, and mild acute GVHD. The incidence of severe GI bleeding, particularly in patients with adequate platelet counts, is less than 1% because of effective prophylaxis against viruses, fungi, and acute GVHD.175 Mortality from severe intestinal bleeding, however, remains at 40%.175,176 The most common cause of severe bleeding is refractory acute GVHD, which can result in bleeding from extensive ulceration in the small intestine and cecum (see Fig. 34-4). In some patients with GVHD, bleeding may appear to be coming from specific areas of the mucosa, but when such patients are operated on or come to autopsy, diffuse rather than focal mucosal ulceration is the rule. Ulcers in the stomach or duodenum that develop after HCT are usually caused by acute GVHD or CMV infection, but with preemptive ganciclovir therapy, bleeding CMV ulcers have become rare.175 Gastric ulcerations also may be caused by infection by VZV, bacteria (phlegmonous gastritis), or EBV (lymphoproliferative disease) (see Chapter 51).

549

550

Section IV  Topics Involving Multiple Organs

A

B

C

D

E

Figure 34-6.  Uncommon gastrointestinal problems in hematopoietic cell transplant recipients. A, Esophagus: barium contrast radiograph of an intramural hematoma that occupies one wall of the esophagus from the aortic arch to the lower esophagus. The red line approximates the normal esophageal contour. This hematoma was caused by retching in a thrombocytopenic patient. B, Stomach: diffuse oozing of blood through the mucosa of the gastric antrum in gastric antral vascular ectasia (GAVE). When the blood was washed off, the underlying mucosa was not ulcerated, yet blood reappeared. C, High-power view of an antral biopsy specimen illustrating the features of GAVE: capillary dilation, thrombosis, and fibromuscular hyperplasia (Hematoxylin and eosin). D, Duodenum: linear ulceration with yellow exudates (arrows) caused by Rhizopus infection in a transplant patient receiving immunosuppressive therapy for graft-versus-host disease (GVHD). The surrounding mucosa is abnormal because of GVHD. E, Colon: sigmoid colon in adenovirus colitis showing diffuse mucosal edema, ulceration, and hemorrhage.

Gastric antral vascular ectasia, is also a cause of severe upper intestinal bleeding in HCT recipients who received oral busulfan as part of conditioning therapy.177,178 Diffuse areas of hemorrhage are seen in the gastric antrum and proximal duodenum, but the underlying mucosa is intact (see Fig. 34-6B). Histology is diagnostic, revealing abnormal dilated capillaries, thromboses, and fibromuscular hyperplasia in the lamina propria (see Fig. 34-6C). Endoscopic laser therapy is the treatment of choice to control bleeding, but multiple laser treatments may be required to obliterate ectatic lesions.178 Rare gastroduodenal causes of bleeding post HCT include ulcers caused by molds (see Fig. 34-6D), Dieulafoy lesions, Curling (stress) ulcers, duodenal biopsy sites, adenovirus colitis (see Fig. 34-6E), and C. septicum infection (typhlitis).179 There is no effective therapy for mucosa that is diffusely oozing blood other than raising the platelet count and treating the underlying condition. In GVHD, re-epithelialization of ulcerated intestinal mucosa occurs very slowly (see Fig. 34-4F). Focal bleeding lesions, especially those caused by mucosal infection, can be treated with endoscopic cautery, heater probe, or epinephrine injection provided platelet counts are adequate. Unless the underlying disease process is eliminated, these endoscopic methods will not cure the bleeding problem. Attempts to resect large segments of diffusely bleeding intestine involved with GVHD have not been successful.

Dysphagia

Mucositis, acid-peptic esophagitis, and pill esophagitis are the leading causes of dysphagia. Infections of the esophagus have largely disappeared because of antiviral and antifungal prophylaxis. Desquamation of oropharyngeal epithelium caused by conditioning therapy may lead to pain on initiating a swallow and inability to move a bolus past the cricopharyngeus. Rarely, non-healing esophageal ulcerations, strictures, and dysphagia result from conditioning therapy. The abrupt onset of severe retrosternal pain, hematemesis, and painful swallowing suggests a hematoma in the wall of the esophagus, a result of retching when platelet counts are very low (see Fig. 34-6A).180 Endoscopy is relatively contraindicated because many intramural hematomas represent contained perforations. The course of intramural hematomas is one of slow resolution over one or two weeks. In patients with severe GVHD, esophageal edema, erythema, and a peeling epithelium lead to ulcerations (see Fig. 34-4A).181 Pill esophagitis occurs after ingestion of medications that might be used after HCT, such as phenytoin, foscarnet, captopril, oral bisphosphonates, ascorbic acid, ciprofloxacin, clindamycin, and oral potassium chloride.

Diarrhea (see Table 34-3) Diarrhea caused by mucosal damage from myeloablative conditioning therapy is seldom severe, usually resolving by

Chapter 34  Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation Table 34-3  Causes of Diarrhea after Hematopoietic Cell Transplantation CAUSE

FREQUENCY

DIAGNOSIS

SEVERITY

TREATMENT

Myeloablative conditioning therapy

Common

Exclude infection, hyperacute GVHD

Usually mild, can be severe after some regimens

Self-limited; octreotide useful in severe cases

Acute GVHD

Common after allografts; occurs in ≈10% of autografts

Association with skin and liver GVHD. Exclude infection. Mucosal histology in problematic cases

Ranges from mild to intractable, highvolume diarrhea

Immunosuppressive drugs, usually prednisone initially

Now uncommon

CMV antigen or DNA in blood; viral culture, immunohistology of mucosal biopsy Adenovirus DNA in blood; viral culture, immunohistology of mucosal biopsy Research test only ELISA of stool

Potentially fatal if not detected early

Ganciclovir or foscarnet

Serotype dependent; may be rapidly fatal

Cidofovir

Self-limited Serotype dependent; can be severe Usually fatal when lymphomatous intestinal involvement develops

None None

Viral infections CMV

Adenovirus

Sporadic

Astrovirus, norovirus Rotavirus

Unusual Rare

EBV-LPD

Now rare

EBV DNA in blood; mucosal biopsy

Common

Toxin and antigen in stool

Usually mild to moderate

Sporadic Rare except in endemic areas

Clinical syndrome of typhlitis Stool, blood culture

Potentially fatal Potentially fatal

Oral vancomycin preferred to metronidazole Imipenem, oral vancomycin Based on organism sensitivities

Rare Rare

Stool EIA Stool microscopy, PCR

Can be protracted Often protracted

Metronidazole Recovery of immunity

Rare

Potentially fatal

Rare

Stool microscopy, antigen, DNA; possibly serum EIA Stool microscopy

Potentially fatal

Metronidazole or tinidazole, followed by paromomycin Ivermectin

Common Common Common

Clinical diagnosis Clinical diagnosis Clinical diagnosis

Dose dependent Diet dependent Medication dependent

Reduce dose; switch to IV Mg++ Disaccharide dietary restriction Restore flora

Common Unusual Common

Clinical diagnosis Clinical diagnosis Clinical diagnosis

Dose dependent Dose dependent Dose dependent

Unusual

Clinical diagnosis

Medication dependent

Reduce dose; switch to IV Mg++ Reduce dose; loperamide Reduce dose; substitute mycophenolic acid formulation; loperamide Discontinue

Bacterial infections Clostridium difficile Clostridium septicum Enteric pathogens Parasitic infections Giardia lamblia Cryptosporidium parvum Entamoeba histolytica Stronglyoides stercoralis Osmotic diarrhea Oral magnesium salts CHO malabsorption Antibiotic use Medication related Oral magnesium salts Tacrolimus Mycophenolate mofetil Metoclopramide

Rituximab when detected early; withdrawal of immunosuppressive drugs

CHO, carbohydrate; CMV, cytomegalovirus; DNA, deoxyribonucleic acid; EBV-LPD, Epstein-Barr virus–lymphoproliferative disease; EIA, enzyme immunoassay; GVHD, graft-versus-host disease; IV, intravenous; PCR, polymerase chain reaction.

days 12 to 15. Cytarabine-containing regimens, high-dose melphalan, and multiple alkylating regimens cause more severe, protracted diarrhea. Intravenous octreotide and oral loperamide at (4 mg by mouth every six hours) may be effective for severe diarrhea associated with conditioning therapy. Acute GVHD is the most common cause of diarrhea after day +15.176,182 The onset of diarrhea can be sudden, with daily stool volumes in excess of 2 L in severe cases. The diarrheal fluid is watery and green, with ropy strands of mucoid material that reflect transmucosal protein loss. In an allografted patient with skin and liver abnormalities typical of acute GVHD, this diarrheal syndrome is almost diagnostic of intestinal GVHD, particularly when there is falling serum albumin and negative stool studies for infection. In GVHD, abdominal imaging (CT, positron-emission

tomography [PET]-CT, or focused ultrasonography) may reveal intestinal edema, but this finding does not differen­ tiate acute GVHD from CMV infection.148,183,184 Pneumatosis intestinalis, which may be associated with GVHD or CMV enteritis, may be seen by plain x-ray, CT, or MRI. A definitive diagnosis of GVHD in problematic cases requires mucosal biopsy. In mild cases, the gastroduodenal and rectosigmoid mucosa are grossly normal, but moderately severe GVHD causes diffusely edematous and erythematous mucosa (see Fig. 34-4).134 Severe GVHD may lead to ulcerations and large areas of mucosal sloughing in the stomach, small intestine and colon (see Fig. 34-4F). Even when the endoscopic appearance is normal, biopsies often reveal intestinal crypt cell necrosis and apoptotic bodies diagnostic of acute GVHD (see Fig. 34-4D). In severe cases of GVHD,

551

552

Section IV  Topics Involving Multiple Organs whole crypts are destroyed, then adjacent crypts, and finally whole segments of intestinal mucosa (see Fig. 34-4E and F). Other histologic findings that support the diagnosis of GVHD include pericapillary hemorrhage,185 infiltrating neutrophils,186 or eosinophils.187 The use of capsule endoscopy for diagnosis of GVHD can provide visual inspection of the small intestine that cannot be seen with routine endoscopy, but does not allow biopsy.188 The negative predictive value of a capsule endoscopy examination of the small intestine appears to be high and is thus useful for excluding more severe acute GVHD.188 Bleeding often accompanies diarrhea in patients with mucosal ulceration.175 Successful treatment of acute GVHD with immunosuppressive therapy results in a dramatic reduction in stool volume, with resolution of accompanying symptoms of abdominal pain, nausea, and vomiting. The management of patients whose diarrhea and other symptoms of intestinal GVHD persist after 7 to 14 days of immunosuppressive therapy is unsatisfactory because the rate of failure of secondary therapy is high.189 Overall prognosis in patients with GVHD can be estimated by use of the acute GVHD Activity Index.190 In allograft recipients, infectious causes of diarrhea are far less common than GVHD, accounting for only 10% to 15% of diarrheal episodes.176,182 In countries where intestinal parasitism and bacterial contamination of water are endemic, the spectrum of infections may be wider.98 C. difficile colitis is usually a relatively mild, treatable disease when diagnosed at the onset of diarrhea, but the recent emergence of more virulent strains of C. difficile has changed the natural history of this infection. The inappropriate use of proton pump inhibitors for marginal indications191 increases the risk of C. difficile colitis twofold.192 Other relatively common causes of infectious diarrhea include astrovirus, norovirus, rotavirus, CMV, and adenovirus.99,182,193,194 Some serotypes of adenovirus cause necrotizing enteritis and rapidly fatal multiorgan failure involving the gut, liver, lungs, and kidneys (see Fig. 34-6E).157,158,195 There should be a sense of urgency in identifying adenovirus as a cause of enteritis, as early treatment with cidofovir appears to be effective.161,162 CMV is the only cause of enteritis after HCT that requires an intestinal biopsy for diagnosis.182 Otherwise the negative predictive value of a stool examination (including PCR) for other viruses, bacteria, fungi, and parasites is high. Watery diarrhea secondary to intestinal parasite infection (Cryptosporidium, G. lamblia, and E. histolytica) and mycobacteria infection are rare outside of endemic areas.100,194,196,197 Diarrhea may also result from carbohydrate malabsorption (particularly in patients on antibiotics), oral magnesium salts, tacrolimus (a motilin agonist), metoclopramide, and MMF.198,199 MMF gut toxicity can be addressed by switching to mycophenolic acid.64

Abdominal Pain

It is extremely important to distinguish abdominal pain as an indicator of a rapidly progressive, fatal condition from illnesses with a benign natural history that require only conservative management. The causes of abdominal pain after HCT are listed in Table 34-4. The illnesses that may progress rapidly include intestinal perforation, some infections (e.g., typhlitis caused by C. septicum, adenovirus, and VZV), gallbladder necrosis, liver abscess, and acute GVHD presenting only as abdominal pain.200 Fortunately, these disorders are far less common than intestinal pseudoobstruction, hepatic pain related to SOS, multisystem acute GVHD, and hemorrhagic cystitis. Intestinal perforation may develop in the setting of lysis of a transmural lymphoma or metastatic carcinoma shortly after conditioning therapy, or later on, from CMV ulcers or diverticular perforation. Perforation may present with only mild to moderate abdominal

pain and pneumoperitoneum on plain abdominal x-ray. Dilation of the bowel in the absence of a mechanical obstruction is the most common cause of moderate to severe abdominal pain. Most patients with pseudo-obstruction have an underlying intestinal disease, such as enteritis from conditioning therapy, GVHD, or infection, but frequently the acute presentation is related to increasing use of muopioid medications. Pseudo-obstruction is more frequent among patients with lymphoma, a result of intestinal neuropathy from repeated use of vincristine (see Chapter 120). To allow pain relief without affecting colon motility, methylnaltrexone can be used.77 Alternatively, colon distention may decrease after switching from a mu-opioid to a kappaopioid agonist (for example, to butorphanol). Neostigmine (2 mg intravenously) has been successfully used in patients with acute colonic pseudo-obstruction after HCT.201 In visceral VZV infection, abdominal distention, severe pain, fever, and rising serum ALT may precede cutaneous manifestations by up to 10 days.160 In rare instances, a skin rash never develops. Acyclovir should be started on clinical suspicion while serum is analyzed by PCR for VZV DNA.160 More severe acute intestinal GVHD may present with nausea, anorexia, periumbilical crampy abdominal pain, and diarrhea. The sudden onset of intestinal edema (see Fig. 34-4B and C) can cause a rigid abdomen with rebound tenderness preceding the development of a skin rash or diarrhea. The decision to treat a patient empirically with prednisone when definitive evidence of GVHD is not at hand can be difficult, but when the pretest probability of GVHD is high (e.g., an HLA-mismatched or unrelated donor; engraftment; a nascent skin rash) and that of perforation or infection low, treatment should be started while GVHD is sought by endoscopic mucosal biopsy (see Fig. 34-4). Pancreatitis is an uncommon cause of abdominal pain in HCT patients, but in a study of autopsied patients the prevalence of acute pancreatitis was 28%.202 Symptoms of pancreatitis had been absent in many of the patients who were found to have florid pancreatitis at autopsy, suggesting that symptoms had been masked by immunosuppressive drugs. Patients with low platelet counts or prolongation of blood clotting may rarely bleed into the retroperitoneum, abdominal wall, or intra-abdominal viscera, particularly after duodenal biopsy, causing significant pain. Intestinal infections presenting with significant pain are listed in Table 34-4. Typhlitis (C. septicum infection) occurs in granulocytopenic patients but is not common after HCT. Symptoms include fever, right lower quadrant pain, nausea and vomiting, diarrhea, occult blood in stool, and shock.102 Typhlitis is usually diagnosed using imaging studies; laparotomy is rarely necessary.203 If typhlitis is a possibility, imipenem and oral vancomycin therapy should be started along with coverage for luminal bacteria and fungi.204

Perianal Pain

Perianal pain after HCT can be caused by an anal fissure, a thrombosed external hemorrhoid, cellulitis related to tissue maceration, fistulas, and abscesses. In patients with granulocytopenia, infections in the perineum or perianal spaces are usually polymicrobial, arising either from anal crypts or from tears in the anal canal. After HCT these infections can be difficult to recognize because they may not produce abscesses but rather a spreading cellulitis. Extensive supralevator and intersphincteric abscesses may be present without being apparent on external examination. CT, MRI, or endoscopic ultrasonography can give a clear view of the anatomy involved, particularly if there is pus present.205 When antibiotics covering anaerobic and aerobic bacteria are given to patients with incipient perianal infection, far fewer patients require surgical drainage than in the past.

Chapter 34  Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation Table 34-4  Causes of Acute Abdominal Pain after Hematopoietic Cell Transplantation CAUSE

FREQUENCY

DIAGNOSIS

SEVERITY

TREATMENT

Sinusoidal obstruction syndrome (SOS)

Now uncommon

Tender hepatomegaly, weight gain, jaundice

Potentially fatal

See section on SOS (see Table 34-2)

Intestinal damage from conditioning therapy

Unusual cause of pain

Examination, imaging

Can be protracted after some regimens

None

Colon pseudo-obstruction

Common, particularly in patients with prior vincristine exposure and current mu-opioid therapy. May also occur with GVHD and VZV or CMV infection.

Distention, tympany, abdominal radiographs

When medication related, usually resolves; when a sign of GVHD or viral infection, may be severe

Reduce opioid, anticholinergic drug exposure; rule out treatable underlying causes; consider methylnaltrexone or kappa agonist opioid; neostigmine if persistent and severe

Hemorrhagic cystitis

Common after cyclophosphamide and with viral bladder infection

Suprapubic pain, hematuria, viral cultures (JC/BK virus or adenovirus)

Can be protracted with viral infection

Urologic therapy, antiviral drugs if appropriate

Acute GVHD

Common, particularly with more severe GVHD

Evaluate skin, intestinal symptoms, serum bilirubin level Intestinal imaging (CT, ultrasound); mucosal biopsy

Potentially fatal

Immunosuppressive drug therapy

Biliary pain

Unusual

RUQ/epigastric localization; gallbladder sludge, edema, gas; biliary dilation on ultrasonography

Passage of sludge is usually self-limited; necrotic gallbladder requires surgery

Persistent biliary obstruction requires stent placement; surgery for gallbladder necrosis

Pancreatitis

Unusual

Serum lipase

Usually self-limited, but pancreatic necrosis may occur

Address biliary, infection, and medication causes

Hematomas

Rare; can be seen after duodenal biopsy

Examination, abdominal imaging, endoscopy

Can be protracted

Restoration of platelet counts; intestinal obstruction may require surgery

Intestinal infection

Unusual

Diagnostic and imaging tests for clostridial infection, VZV, CMV, adenovirus, molds

Potentially fatal if not treated (especially C. septicum, viral, or mold infection)

Treat organism identified

Intestinal perforation

Rare

Plain films, CT

Potentially fatal

Surgery; identification of underlying cause (CMV ulcer, intestinal tumor necrosis, diverticula)

Liver abscess/bacterial infection

Rare (usually fungal)

Liver imaging (MRI preferred), examination, serum fungal antigen detection

Potentially fatal if not treated

Appropriate antifungal, antimycobacterial therapy

Intestinal infarction

Rare (usually disseminated Aspergillus infection)

Intestinal imaging, examination, chest film, galactomannan EIA

Uniformly fatal

Antifungal drugs active against Aspergillus

EBV-lymphoproliferative disease

Rare with surveillance for EBV DNA in serum

Abdominal imaging, endoscopy

Usually fatal once tumor masses form

Rituximab when detected early; withdrawal of immunosuppressive drugs

CMV, cytomegalovirus; CT, computed tomography; DNA, deoxyribonucleic acid; EBV, Epstein-Barr virus; EBV-LPD, Epstein-Barr virus–lymphoproliferative disease; EIA, enzyme immunoassay; GVHD, graft-versus-host disease; JC/BK; polyomaviruses; MRI, magnetic resonance imaging; RUQ, right upper quadrant; VZV, varicella-zoster virus.

PROBLEMS IN LONG-TERM TRANSPLANT SURVIVORS Liver Disease Caused by Graft-Versus-Host Disease Long-term survivors of transplant who have hepatic GVHD usually have other evidence of chronic GVHD, a pleomorphic immune disorder characterized by oral and ocular sicca; ulceration in squamous epithelium of the skin, mouth,

esophagus, and vagina; subcutaneous fibrosis; contractures and myositis; immunodeficiency; and other manifestations of immune dysregulation. Liver involvement in these patients, like gut mucosal inflammation, is considered to be a protracted form of acute GVHD because clinical and histologic changes in these organs are identical to those in patients with acute GVHD that occurs in the months

553

554

Section IV  Topics Involving Multiple Organs following HCT.206 Patients with isolated elevations of serum alkaline phosphatase or ALT related to GVHD should be followed closely, but may not require high-dose immunosuppressive therapy. By the time that jaundice develops in patients with chronic GVHD, liver biopsy shows extensive damage to small bile ducts and, in severe cases, ductopenia. In patients receiving no, or tapering doses of, immunosuppression, liver GVHD may also present as an acute hepatitis, with abrupt elevations of serum aminotransferase levels to more than 2000 U/L.155 A similar presentation can be seen after donor lymphocyte infusion (DLI).156 In patients presenting with an acute hepatitis, blood samples for viral DNA or ribonucleic acid (RNA) or liver biopsy are essential in excluding acute viral hepatitis due to a herpesvirus (HSV or VZV) or a hepatitis virus and to make a definitive diagnosis of hepatic GVHD (see Fig. 34-5D and E). A serum autoantibody test for CYP1A2 may prove diagnostically useful in the diagnosis of hepatitic GVHD because this enzyme appears to be a target antigen in GVHD.207 Immunosuppressive drug treatment of chronic GVHD is successful in 50% to 80% of patients with extensive multiorgan disease, but refractory chronic GVHD is a fatal illness. The addition of ursodeoxycholic acid (15 mg/kg/day) may result in biochemical improvement.208 Ductopenic GVHD is potentially reversible if ongoing immunologic destruction of biliary epithelium ceases, but this process may take months before resolution of jaundice.155 Liver transplantation, including living-donor transplantation from the original stem cell donor,209 has been performed for patients with liver failure due to chronic hepatic GVHD, although frequently there are contraindications to this approach.210

Chronic Viral Hepatitis and Cirrhosis

HCV infection in HCT survivors almost always results in chronic hepatitis.114,115 In the first 10 years of HCV infection after HCT, there is little liver-related morbidity.114 However, a third of patients transplanted before the 1990s will develop cirrhosis related to chronic HCV infection over a 20- to 40-year time frame.115 The reasons for more rapid progression of fibrosis after HCT may be related to conco­ mitant liver involvement with GVHD, immunosuppression, and iron overload.211 Iron overload is particularly severe in thalassemic patients who have undergone HCT.212 Patients with chronic HCV should be offered therapy with com­ bination pegylated INF-α plus ribavirin, unless there are contrain­dications.167,213 Pegylated interferons, with their longer half-lives, should be administered with caution because some HCT patients experience rapid falls in platelet and granulocyte counts. INF-α may also activate chronic GVHD. Liver transplantation should be considered in any HCT survivor with incipient liver decompensation; in some cases, the original allogeneic cell donor can be a partial liver donor.214 The prevalence of chronic HBV infection among HCT survivors varies widely depending on the country. Patients who remain viremic or HBsAg positive after HCT are at risk of flares of hepatitis B at times of reduction of immunosuppression, such as during tapering or cessation of treatment for chronic GVHD. All long-term survivors with chronic hepatitis B should be regularly monitored to assess the need for antiviral therapy. Hepatitis B e antigen and antibody status should be determined in all patients, and liver enzymes monitored every 6 to 12 months. The need for antiviral treatment with an oral nucleoside or nucleotide analog (for instance, entecavir, telbivudine, adefovir, tenofovir) is based on the ALT and HBV DNA levels and the severity of hepatic fibrosis.215 These levels may change over time, emphasizing the need to continually reassess patients

who are not on treatment. As in the peritransplant period, patients with viremia or positive HBsAg should receive an antiviral agent whenever they receive immunosuppressive or cytotoxic therapy.216

Hemosiderosis

Iron overload is caused by a combination of multiple red cell transfusions (e.g., for thalassemia) and dyserythropoiesis leading to increased iron transport by the intestine. After HCT, iron accumulation stops and body iron stores fall slowly over time.217 The consequences of extreme iron overload in HCT survivors are primarily those of cardiac, pituitary, and endocrine pancreatic dysfunction. Iron overload may be an important cofactor in liver disease in longterm survivors of HCT and should be part of a screening panel.218,219 In the past, liver biopsy with liver iron determination was required, but increasingly noninvasive methods (e.g., MRI129) are being used to provide assessments of liver iron concentration and distribution. Patients with liver iron content greater than 15,000 µg/g dry weight should be treated aggressively with phlebotomy and chelation; when liver iron content is 7000 to 15,000 µg/g dry weight, phlebotomy is indicated; when liver iron content is less than 7000 µg/g dry weight, treatment is indicated only if there is evidence of liver disease.220 Mobilization of iron from heavily overloaded patients improves cardiac function, normalizes serum ALT levels, and results in improved liver histology.219-222

Hepatic Drug Toxicity

Drug-induced liver injury may be related to drugs in common use by transplant survivors, including antihypertensive drugs, lipid-lowering agents, hypoglycemic agents, NSAIDs, antidepressants, antibiotics, and herbal preparations.223 Some drug reactions may result in chronic liver disease.224

Fungal Liver Infections

Fungal abscesses can recur after apparently successful antifungal therapy when high-dose immunosuppressive drugs are started for GVHD. Oral, nonsterile herbal remedies contaminated by molds may lead to liver abscesses in immunosuppressed HCT survivors.225

Liver Cancer

Compared with the general population, patients who survive more than 10 years post HCT have an eightfold risk of developing a new solid malignancy. The risk of hepatocellular carcinoma is particularly elevated.226 Transplant survivors with risk factors for hepatocellular carcinoma (HCV or HBV infection, obesity, diabetes, low platelet count) should be screened at yearly intervals (see Chapter 94).227,228 Chronic hepatitis C may also be a risk factor for development of lymphoma229 and other lymphoproliferative disorders230 after transplant.

Other Hepatobiliary Disorders

There is a higher than expected incidence of gallstones and stone-related biliary problems after HCT than in an agematched population, probably related to earlier formation of biliary sludge. Chronic cyclosporine dosing also may lead to gallstones and biliary symptoms. Patients who have experienced SOS from either chemotherapy or a myeloablative conditioning regimen may rarely develop hepatic nodularity caused by atrophy of zone three and hypertrophy of zone one hepatocytes, without fibrosis.231 This process (nodular regenerative hyperplasia) is usually clinically silent unless portal hypertension develops, manifested by variceal bleed-

Chapter 34  Gastrointestinal and Hepatic Complications of Solid Organ and Hematopoietic Cell Transplantation ing, ascites, splenomegaly, and thrombocytopenia, but with preserved liver function. Focal nodular hyperplasia (nodules readily seen by liver imaging, with characteristic central scars) has been described as an incidental finding in 12% of a cohort of HCT survivors, up to 14 years after transplant.232 These lesions are likely the result of sinu­ soidal liver injury caused by myeloablative conditioning regimens.

Esophageal Disorders

Some patients with extensive chronic GVHD have eso­ phageal desquamation, webs, submucosal fibrous rings, bullae, and long, narrow strictures in the upper and midesophagus.233-236 Although the most common symptom is dysphagia, some patients present with insidious weight loss, retrosternal pain, and aspiration of gastric contents. The diagnosis is made by barium contrast radiography and endoscopy,236 which should be done with caution because perforations have been reported.233 Tight strictures are difficult to dilate safely, but failure to dilate strictures may lead to progressive esophageal narrowing. Esophageal involvement can be prevented by prompt treatment of chronic GVHD at its early stages. Therapy with proton pump inhibitors should be considered if there is uncontrolled acid reflux. Myasthenia gravis may also complicate chronic GVHD, with dysphagia as its presenting complaint.237 Sporadic cases of fungal and rarely viral esoph­ agitis may occur in patients with chronic GVHD on immunosuppressive and antibiotic therapy. Esophageal strictures may be sequelae of earlier herpesvirus infection or mucositis. Squamous cell carcinoma of the esophagus has been reported in HCT survivors, usually with concomitant chronic GVHD of the oropharynx.238

Gastrointestinal Disorders

The incidence of diarrhea falls sharply after day 100 except in patients who have received allografts following reduced intensity conditioning therapy239 and in those whose acute GVHD has never resolved. Patients with protracted acute GVHD, however, often have symptoms that wax and wane with intensity of immunosuppressive therapy for up to 15 years after HCT, with each exacerbation similar to the presenting signs of GVHD that occurred earlier after HCT (satiety, poor appetite, nausea, episodic diarrhea, and weight loss).240,241 The endoscopic and histologic appearance of intestinal mucosa is identical to that seen in acute GVHD (see Fig. 34-4). Use of oral beclomethasone dipro­ pionate can be effective in treating patients with protracted acute GVHD involving the GI tract.242 Before the introduction of more effective immunosuppressive drugs, chronic GVHD resulted in extensive collagen deposition in submucosal and subserosal areas of the intestinal tract, resulting in refractory malabsorption243; this process has not been seen in recent years. There are sporadic cases of C. difficile, CMV,138 and rarely G. lamblia and cryptosporidiosis in longterm survivors. Chronic intestinal viral infection can be seen in patients who remain on immunosuppressive drugs, including rotavirus, norovirus, and adenovirus. Intestinal diseases in donors have been reported in their recipients,

such as idiopathic inflammatory bowel disease and celiac disease.244

Pancreatic Disease

Acute pancreatitis has been described in transplant survivors, usually related to passage of gallstones or sludge. Cyclosporine and tacrolimus also may cause pancreatitis.245-247 Diarrhea, steatorrhea, and weight loss secondary to pancreatic insufficiency have developed in some HCT survivors.248,249 The cause is unclear; previous pancreatic necrosis, prolonged glucocorticoid exposure, and tacrolimus toxicity are the leading possibilities.202 Transient pancreatic insufficiency has been noted in patients with gut and liver GVHD.250 Chronic GVHD and extreme iron overload also may contribute to pancreatic damage.

KEY REFERENCES

Abu-Elmagd K, Reyes J, Bond G, et al. Clinical intestinal transplantation: A decade of experience at a single center. Ann Surg 2001; 234:404-16. (Ref 61.) Appelbaum FR, Forman SJ, Negrin RS, Blume KG. Thomas’ Hemato­ poietic Cell Transplantation. 4th ed. Oxford, UK: Wiley-Blackwell Publishing; 2009. (Ref 94.) Assi MA, Pulido JS, Peters SG, et al. Graft-vs.-host disease in lung and other solid organ transplant recipients. Clin Transplant 2007; 21:1-6. (Ref 66.) Berenguer M. What determines the natural history of recurrent hepatitis C after liver transplantation? J Hepatol 2005; 42:448-56. (Ref 46.) Berkowitz N, Schulman LL, McGregor C, Markowitz D. Gastroparesis after lung transplantation. Potential role in postoperative respiratory complications. Chest 1995; 108:1602-7. (Ref 52.) Hockenbery DM, Cruickshank S, Rodell TC, et al. A randomized, placebo-controlled trial of oral beclomethasone dipropionate as a prednisone-sparing therapy for gastrointestinal graft-versus-host disease. Blood 2007; 109:4557-63. (Ref 135.) Hogan WJ, Maris M, Storer B, et al. Hepatic injury after nonmyeloa­ blative conditioning followed by allogeneic hematopoietic cell transplantation: A study of 193 patients. Blood 2004; 103:76-82. (Ref 118.) Lau GK, Suri D, Liang R, et al. Resolution of chronic hepatitis B and anti-HBs seroconversion in humans by adoptive transfer of immunity to hepatitis B core antigen. Gastroenterology 2002; 122:61424. (Ref 112.) McDonald GB, Slattery JT, Bouvier ME, et al. Cyclophosphamide metabolism, liver toxicity, and mortality following hematopoietic stem cell transplantation. Blood 2003; 101:2043-8. (Ref 121.) Ponec RJ, Hackman RC, McDonald GB. Endoscopic and histologic diagnosis of intestinal graft-vs.-host disease after marrow transplantation. Gastrointest Endosc 1999; 49:612-21. (Ref 134.) Ponticelli C, Passerini P. Gastrointestinal complications in renal transplant recipients. Transplant Int 2005; 18:643-50. (Ref 16.) Ruutu T, Eriksson B, Remes K, et al. Ursodeoxycholic acid for the prevention of hepatic complications in allogeneic stem cell transplantation. Blood 2002; 100:1977-83. (Ref 142.) Small LN, Lau J, Snydman DR. Preventing post-organ transplantation cytomegalovirus disease with ganciclovir: A meta-analysis comparing prophylactic and preemptive therapies. Clin Infect Dis 2006; 43:869-80. (Ref 11.) St Pierre TG, Clark PR, Chua-anusorn W, et al. Noninvasive measurement and imaging of liver iron concentrations using proton magnetic resonance. Blood 2005; 105:855-61. (Ref 129.) van Burik JA, Lawatsch EJ, DeFor TE, Weisdorf DJ. Cytomegalovirus enteritis among hematopoietic stem cell transplant recipients. Biol Blood Marrow Transplant 2001; 7:674-9. (Ref 138.) Full references for this chapter can be found on www.expertconsult.com.

555

CHAPTE R

35 Gastrointestinal and Hepatic Manifestations of Systemic Diseases Rajeev Jain and Dwain L. Thiele

CHAPTER OUTLINE Rheumatologic and Collagen Vascular Diseases  557 Rheumatoid Arthritis  558 Adult-Onset Still’s Disease  559 Progressive Systemic Sclerosis  560 Systemic Lupus Erythematosus  560 Polymyositis and Dermatomyositis  561 Mixed Connective Tissue Disease  561 Sjögren’s Syndrome  561 Polyarteritis Nodosa and Other Vasculitides  562 Behçet’s Disease  563 Seronegative Spondyloarthropathies (Reactive Arthritides)  563 Marfan’s and Ehlers-Danlos Syndromes  563 Familial Mediterranean Fever  563 Oncologic and Hematologic Diseases  564 Metastases  564 Paraneoplastic Syndromes  564 Hematologic Malignancies  564 Systemic Mastocytosis  567 Myeloproliferative and Myelophthisic Disorders  568 Dysproteinemias  568 Coagulation Disorders  569 Red Blood Cell Dyscrasias  570 Endocrine Diseases  572 Diabetes Mellitus  572 Thyroid Disease  575 Adrenal Disease  576

Numerous systemic and extraintestinal diseases have gastrointestinal and hepatic manifestations. Because it is impossible to discuss each entity in great detail in a single chapter, we endeavor here to emphasize frequently encountered diseases and those that may be of particular interest to the reader because of recent developments. For the sake of clarity, some diseases that result in similar manifestations are presented in tabular form. Some topics are taken up in detail in other chapters. The reader is referred to these chapters for a more complete discussion. Although not always a manifestation of systemic disease, nodular

Pituitary Disease  576 Parathyroid Disease  576 Disorders of Lipid Metabolism  577 Hyperlipoproteinemias and Dyslipidemias  577 Abetalipoproteinemia  577 Tangier Disease  577 Neutral Glycosphingolipidoses  577 Renal Diseases  578 Neurologic Diseases  578 Neurogenic Abdominal Pain  578 Gastrointestinal Complications of Acute Head Injury and Stroke  579 Gastrointestinal Problems after Spinal Cord Injury  579 Diseases of the Autonomic Nervous System  580 Extrapyramidal Disorders  580 Multiple Sclerosis  581 Neuromuscular Disorders  581 Pulmonary Diseases and Problems in Patients Who Require Critical Care  581 Intensive Care Unit Patients and Septic Patients  581 Cardiovascular Diseases  584 Infiltrative Diseases  584 Amyloidosis  584 Granulomatous Liver Disease  588 Sarcoidosis  588 Other Infiltrative Disorders  590 Nodular Disorders of the Liver  590

regenerative hyperplasia of the liver is also discussed in this chapter.

RHEUMATOLOGIC AND COLLAGEN VASCULAR DISEASES Rheumatologic diseases encompass a wide variety of clinical syndromes and are frequently associated with gastrointestinal abnormalities (Table 35-1). In addition,

557

558

Section IV  Topics Involving Multiple Organs Table 35-1  Gastrointestinal Manifestations of Rheumatologic Diseases DISEASE

ABNORMALITY/ASSOCIATION

CLINICAL MANIFESTATIONS

Rheumatoid arthritis

Temporomandibular arthritis Esophageal dysmotility Visceral vasculitis Amyloidosis

Impaired mastication Dysphagia, GERD Abdominal pain, cholecystitis, intestinal ulceration and infarction Pseudo-obstruction, malabsorption, protein-losing enteropathy, intestinal ulceration and infarction, gastric outlet obstruction Variceal hemorrhage Enteritis, diarrhea, fever, eosinophilia, megacolon Dysphagia, GERD, stricture, Barrett’s esophagus Gastric retention, GERD Constipation, pseudo-obstruction, malabsorption, intussusception, volvulus, pneumatosis intestinalis Hemorrhage, stasis, bacterial overgrowth Mesenteric thrombosis, infarction, pancreatic necrosis Calcific pancreatitis, pancreatic exocrine insufficiency Dysphagia, reflux GI ulceration, intestinal infarction, intussusception, pancreatitis, pneumatosis intestinalis Oral fissures, oropharyngeal dysphagia Dysphagia

Scleroderma

SLE Sjögren’s syndrome

Polymyositis-dermatomyositis MCTD

Portal hypertension (Felty’s syndrome) Gold enterocolitis Esophageal dysmotility Gastroparesis Intestinal fibrosis and dysmotility Pseudodiverticula Arteritis (rare) Pancreatitis Esophageal dysmotility Mesenteric vasculitis Desiccation of membranes Esophageal webs Gastric lymphoid infiltrates Pancreatitis Primary biliary cirrhosis Skeletal muscle dysfunction Dysmotility Mesenteric vasculitis (rare) Dysmotility

PAN

Mesenteric vasculitis (rare) Mesenteric vasculitis

CSS

Mesenteric vasculitis Eosinophilic gastritis Mesenteric vasculitis

Henoch-Schönlein purpura Kohlmeier-Degos disease Cogan’s syndrome Wegener’s granulomatosis Cryoglobulinemia Behçet’s disease Reactive arthritis Familial Mediterranean fever Marfan/Ehlers-Danlos syndromes

Mesenteric vasculitis Mesenteric vasculitis (infrequent) Crohn’s disease Mesenteric vasculitis Mesenteric vasculitis (rare) Mucosal ulcerations Ileocolonic inflammation Serositis/peritonitis, amyloidosis, PAN, Henoch-Schönlein purpura Defective collagen

Abdominal pain, pancreatic exocrine insufficiency Jaundice, hepatic failure, variceal hemorrhage Aspiration, impaired glutition Dysphagia, GERD, gastroparesis, constipation, diverticula GI ulceration, perforation, pneumatosis intestinalis Dysphagia, GERD, stricture, gastroparesis, bezoars, pseudoobstruction Ulceration, perforation, pancreatitis Cholecystitis, appendicitis, intestinal infarction, pancreatitis, perforation, strictures, mucosal hemorrhage, submucosal hematomas Hemorrhage, ulceration, intestinal infarction, perforation Gastric masses Intussusception, ulcers, cholecystitis, hemorrhage, intestinal infarction, appendicitis, perforation Hemorrhage, ulceration, intestinal infarction, malabsorption Hemorrhage, ulceration, intestinal infarction, intussusception Bloody diarrhea, abdominal pain, fissures, fistulas Cholecystitis, appendicitis, ileocolitis, intestinal infarction Intestinal infarction, ischemia Hemorrhage, perforation, pyloric stenosis Complications as in rheumatoid arthritis Usually asymptomatic Abdominal pain, fever, dysmotility Megaesophagus, hypomotility, diverticula, megacolon, malabsorption, perforation, arterial rupture

CSS, Churg-Strauss syndrome; GERD, gastroesophageal reflux disease; GI, gastrointestinal; MCTD, mixed connective tissue disease; PAN, polyarteritis nodosa; SLE, systemic lupus erythematosus.

the medications used to treat these diseases often produce gastrointestinal and hepatic toxicity. This section focuses on the more common abnormalities that may be encountered by the gastroenterologist.

RHEUMATOID ARTHRITIS

Approximately 0.8% of adults worldwide are affected with rheumatoid arthritis (RA), which is a chronic, inflammatory autoimmune disease primarily targeting the synovial tissues with systemic manifestations. Oropharyngeal symptoms may occur in patients with RA as a result of xerostomia, temporomandibular joint (TMJ) arthritis, cervical spine abnormalities, and laryngeal involvement.1 Esophageal dysmotility, characterized by low-amplitude peristaltic waves, has been described in the proximal, middle, and distal esophagus with reduced lower esophageal sphincter (LES) pressure.1,2 Rheumatoid vasculitis typically occurs

in the setting of severe RA with rare gastrointestinal manifestations such as ischemic cholecystitis or appendicitis, ulceration, pancolitis, infarction, or intra-abdominal hemorrhage due to a ruptured visceral aneurysm.3,4 Other gastrointes­tinal complications of RA include amyloidosis (discussed later) and malabsorption. Felty’s syndrome–RA, spleno­megaly, and leukopenia have been associated with severe infections, portal hypertension, and variceal hemorrhage.5

Hepatic Abnormalities

Abnormal liver function tests, especially elevations of serum alkaline phosphatase of hepatobiliary origin,6-8 are commonly observed in patients with RA. In one large series of patients with RA,6 18% had elevated levels of serum alkaline phosphatase and 11% were found to have hepatomegaly. Fluctuations in serum alkaline phosphatase levels

Chapter 35  Gastrointestinal and Hepatic Manifestations of Systemic Diseases have also been reported to correlate with disease activity.6,8 However, degrees of alkaline phosphatase elevations are usually modest, with the mean level being less than twofold abnormal.6 Furthermore, other clinical signs of liver disease are usually absent, and liver biopsy and autopsy studies have not revealed any consistent or specific findings, with the most common abnormalities being fatty change, Kupffer cell hyperplasia, and mild mononuclear cell infiltration of the portal tracts or rare parenchymal foci of hepatocyte necrosis.6,9-11 Periportal fibrosis is also present in a small minority of cases.11 Determination of etiology of hepatic dysfunction in patients with active rheumatoid arthritis is complicated by the fact that many of the agents commonly used as therapy for this disease have known potential for liver injury.7,10-12 In a small subset of patients with RA and/or Sjögren’s syndrome, antimitochondrial antibodies are present along with the biochemical and histologic features of primary biliary cirrhosis.13-15 The incidence of primary biliary cirrhosis or autoimmune hepatitis appears to be much higher in patients with Sjögren’s syndrome alone than in those with Sjögren’s RA.16,17 Because chronic hepatitis C and RA are relatively common diseases of adults, it is not surprising that these entities are found concurrently in some patients. However, in addition, it has been noted that 75% of individuals with chronic hepatitis C infection develop rheumatoid factors,18 and a subset of these rheumatoid factor–positive individuals develop essential mixed cryoglobulinemia that may be manifested in part by development of arthralgias.19,20 Liver disease in such individuals is often asymptomatic and biochemical abnormalities modest or even absent.19,20 Thus some individuals with essential mixed cryoglobulinemia associated with chronic hepatitis C infection may instead be labeled as having RA. However, anticyclic citrulinated peptide (CCP) antibodies are rarely found in subjects with chronic hepatitis C and nonspecific rheumatologic manifestations and thus anti-CCP antibodies appear to be reliable markers of RA.21 Most rheumatic disease patients with progressive liver disease have concomitant chronic viral or autoimmune hepatitis.22 In patients with concomitant hepatitis B infection, the intermittent use of tumor necrosis factor (TNF) inhibitors or other immunosuppressive therapy may be associated with severe flares of hepatitis B23-25 and prophylactic use of antiviral therapy should be considered.25 TNF inhibitor therapy in RA also has been associated with flares of severe liver disease, with characteristics of autoimmune hepatitis.26 Perhaps the most distinctive association between RA and hepatic abnormalities is seen in another subset of patients who develop splenomegaly and neutropenia (Felty’s syndrome). Felty’s syndrome is associated with an even higher incidence of hepatomegaly and liver function test abnormalities than seen in uncomplicated RA.27,28 However, there is little correlation between serum hepatic enzyme abnormalities and histopathologic findings.27,28 Nevertheless, more than half of patients with this syndrome have been found to have hepatic histologic abnormalities that range from sinusoidal lymphocytosis and portal fibrosis to the more distinctive picture of nodular regenerative hyper­ plasia, which has been reported on multiple occasions in patients with Felty’s syndrome27-30 and in one small prospective series was found to be present in 5 of 18 (28%) patients. Hepatic encephalopathy or other manifestations of liver failure have not been reported in patients with Felty’s syndrome, and nodular regenerative hyperplasia but portal hypertension and esophageal variceal hemorrhage may occur.28-30

Gastrointestinal Abnormalities

The most common gastrointestinal problems encountered in patients with RA are due to drug therapy with nonster­ oidal anti-inflammatory drugs (NSAIDs), glucocorticoids, and disease-modifying antirheumatic drugs (DMARDs). NSAIDs are most commonly associated with upper gastrointestinal complications such as perforation, ulcers, and bleeding (see Chapters 52 and 53). Less commonly recognized complications of NSAIDs include pill esophagitis (Chapter 45), small bowel ulceration (Chapter 115), strictures of the small and large intestine, and exacerbations of diverticular disease and inflammatory bowel disease.31 Significant risk factors for the development of serious upper gastrointestinal events in patients with RA include NSAIDs therapy, age older than 65 years, history of peptic ulcer disease, glucocorticoid therapy, and severe RA.32,33 In patients with RA, the use of certain selective cyclooxygenase-2 inhibitors results in a lower incidence of gastrointestinal complications than that seen with nonselective NSAIDs.34,35 Helicobacter pylori and NSAIDs are independent and possibly synergistic risk factors for peptic ulceration. As such, chronic NSAID users who develop ulcers should be assessed for H. pylori infection and undergo eradication therapy when infection is present.32,33 Although hypergastrinemia has been reported in patients with RA, the incidence of peptic ulcers is no greater than that seen in patients with osteoarthritis.36 As reviewed in Chapter 53, NSAID-associated gastric and duodenal ulcers can be prevented with misoprostol, high-dose histamine (H2) blockers, and proton pump inhibitors.37 Once identified, ulcers may be treated successfully using proton pump inhibitors despite continued NSAID therapy. In the subgroup of patients with a history of bleeding ulcers, therapy with cyclooxygenase-2 inhi­ bitors rather than NSAIDs may be cost effective and less expensive than combining an NSAID with a proton pump inhibitor.38,39 Synthetic DMARDs such as gold and penicillamine are rarely used because of toxicity and marginal efficacy.40 Gold, parenteral as well as oral forms, has been associated with diarrhea, enterocolitis, toxic megacolon, and death. The onset of gold colitis usually occurs within several weeks after the start of therapy and is manifested by nausea, vomiting, diarrhea, and fever. Although the colon is most commonly involved, gold-induced gastrointestinal toxicity may affect the esophagus, stomach, and small bowel, with 25% of patients developing a peripheral eosinophilia.41 Treatment includes dose reduction or discontinuation of gold, antidiarrheals, glucocorticoids, cromolyn sodium, or the chelating agent dimercaprol.41,42 Leflunomide, a synthetic DMARD that inhibits pyrimidine synthesis, can cause diarrhea in up to 32% of patients.43 It may also cause severe hepatic toxicity (see Chapter 86). Biologic DMARDs, which inhibit the action of TNF-α (infliximab, etanercept, and adalimumab) or interleukin-1 (IL-1; anakinra), have not shown significant gastrointestinal adverse effects but may cause hepatic toxicity on occasion (see later).

ADULT-ONSET STILL’S DISEASE

Adult-onset Still’s disease, the adult form of juvenile RA, often has gastrointestinal manifestations such as weight loss, sore throat, hepatosplenomegaly, elevated aminotransferases, and abdominal pain, in addition to fever.44 In contrast to the lack of significant hepatic dysfunction in classic rheumatoid arthritis, adults with Still’s disease present with features of mild hepatitis in the majority of cases and life-threatening acute liver failure in exceptional cases.45-49

559

560

Section IV  Topics Involving Multiple Organs Variable degrees of aminotransferase and alkaline phosphatase elevations are typically observed in such patients during symptomatic disease flares. Liver biopsies usually reveal moderate portal mononuclear cell infiltration with occasional evidence of focal hepatocyte necrosis.46 Biopsies obtained in patients with jaundice and biochemical evidence of severe hepatitis have been found to have interface and lobular hepatitis with lymphoplasmocytic inflammation reminiscent of autoimmune hepatitis.49 Most cases of severe hepatitis have been observed in patients previously treated with salicylates or other NSAIDs,46,49 but liver enzyme abnormalities are also commonly noted prior to therapy. Some patients with severe hepatitis have been reported to respond to immunosuppressive therapy,49 whereas others required liver transplantation or have died of liver failure.45,47-49 Although severe hepatitis is a rare complication of adult-onset Still’s disease, liver failure appears to be the most common cause of death related to this disease.45

PROGRESSIVE SYSTEMIC SCLEROSIS

Progressive systemic sclerosis (PSS, scleroderma) is a multisystem disorder characterized by obliterative small vessel vasculitis and proliferation of connective tissue with fibrosis of multiple organs. Patients with limited cutaneous involvement frequently display findings of the CREST syndrome (calcinosis, Raynaud’s phenomenon, esophageal dysmotility, sclerodactyly, and telangiectasias). Gastrointestinal manifestations occur in up to 90% of patients with PSS.50 Gastrointestinal tract involvement can occur from the mouth to the anus. Atrophy and fibrosis of the perioral skin may limit mandibular motion. The periodontal ligament may become hypertrophic, and the gingivae, tongue papillae, and buccal mucosa may become friable and atrophic, resulting in impaired sensation and taste. The esophagus is the most frequently involved gastro­ intestinal organ.50 On pathology, atrophy of smooth muscle layers and intimal proliferation of arterioles is seen in the distal esophagus.51 Dysphagia occurs as a result of impaired esophageal motility, and gastroesophageal reflux disease (GERD) is related to hypotensive LES pressures, impaired esophageal clearance of acid, and reduced acid-neutralizing capacity due to xerostomia with reduced saliva production.52 The incidence of esophagitis approaches 100% in patients with severe cutaneous involvement.53 The extent of hypomotility varies from occasional uncoordinated contractions to complete paralysis.54 The severity of esophageal dysmotility correlates with the development of interstitial lung disease.55,56 Stricture formation from GERD may contribute to dysphagia, affecting approximately 8% of patients.57 Upper gastrointestinal hemorrhage has been reported from esophageal ulcers, rare esophageoatrial fistulas, and esophageal telangiectasia.58,59 An increased risk of infectious esophagitis with Candida (see Chapter 45) has been attributed to esophageal dysmotility and concomitant immunosuppressive therapy.60 Severe esophagitis typically responds to proton pump inhibitors but may require higher doses for maximal effect.61 A neuropathic achalasia-like syndrome has also been reported.62 The prevalence of Barrett’s metaplasia of the esophagus in patients with PSS ranges between 7% and 38%.63,64 A recent cohort study reported an increased incidence of carcinoma of the tongue and esophagus in patients with PSS.65 Gastric involvement most commonly leads to gastroparesis but other manifestations may include dyspepsia, exacerbation of GERD, or gastric hemorrhage from gastric antral vascular ectasia (GAVE, watermelon stomach). Delayed

gastric emptying has been shown using radionucleotide scintigraphy or radiopaque pellets, with cutaneous electrogastrography demonstrating bradygastria and decreased amplitude of electrical activity.56,66,67 Prokinetic agents such as metoclopramide and erythromycin may increase LES pressures and improve gastric emptying in some patients with PSS.57 The pathologic changes in the small bowel of PSS patients consist of smooth muscle atrophy and deposition of collagen in submucosal, muscular, and serosal layers. Small bowel hypomotility is present in as many as 88% of cases.68 In the early stages of the disease, hypomotility is caused by neuropathic involvement, which may be more responsive to prokinetic agents. In advanced cases, hypomotility is more likely a result of “myopathic” and “fibrotic” changes.57 The interdigestive migrating motor complex (IMMC) is frequently absent or markedly diminished in amplitude in PSS patients with symptoms of intestinal dysmotility.69 Small bowel radiographic abnormalities are present in about 60% of PSS patients, but they may not correlate with symptoms. The duodenum is often dilated, especially in its second and third portions, often with prolonged retention of barium.70 Typically the jejunum is dilated and foreshortened because of mural fibrosis, but valvulae conniventes of normal thickness give rise to an accordion-like appearance. Pneumatosis cystoides intestinalis, pseudoobstruction, pseudodiverticula, sacculations, intussusception, acquired intestinal lymphangiectasia, and small bowel volvulus have been noted.71-73 Symptoms of small intestinal PSS include bloating, borborygmi, anorexia, nausea, and vomiting. Rarely, thrombosis of large mesenteric arteries with extensive bowel necrosis may occur.74 Malabsorption with steatorrhea is present in as many as a third of PSS patients68 and is caused by bacterial overgrowth (see Chapters 101 and 102). Although antibiotic therapy can be effective in these patients, d-xylose malabsorption is often incompletely reversed, suggesting that collagen deposition in PSS may also contribute to malabsorption.75 Although often disappointing, the use of prokinetic agents such as metoclopramide may be effective in some cases. Octreotide in low doses and erythromycin also may provide sustained relief from nausea, abdominal pain, and bloating in some patients with pseudo-obstruction.76 Delayed colonic transit and impaired anal sphincter function are frequently found in constipated patients with PSS.77,78 Cisapride (a drug no longer available in the United States) accelerates colonic transit,79 but refractory cases may require surgery.80 Colonic stricture, volvulus, and bleeding from mucosal telangiectasias have been reported.81,82 Widenecked diverticula can be seen, especially in the antimesenteric border of the transverse and descending colons. Rectal prolapse worsens anal sphincter function, aggravating fecal incontinence in patients with PSS.83 Rectal bleeding can occur from vascular ectasia.84 Pancreatic exocrine secretion is depressed in a third of patients with PSS, and idiopathic calcific pancreatitis has been reported.85 In addition, arteritis resulting in ischemic pancreatic necrosis has been described in patients with PSS.86,87 Gallbladder motility is not altered in PSS.88

SYSTEMIC LUPUS ERYTHEMATOSUS

Systemic lupus erythematosus (SLE) is a multisystem disease characterized by immune system abnormalities and the production of autoantibodies with tissue damage. Gastrointestinal symptoms are common in patients with active SLE. Oral ulcers (one of the criteria used to diagnose SLE) are most commonly seen in the buccal mucosa, hard palate,

Chapter 35  Gastrointestinal and Hepatic Manifestations of Systemic Diseases and vermilion border.89 In SLE, dysphagia (1% to 13% of patients) and GERD (11% to 50% of patients) poorly correlates with esophageal manometric abnormalities such as hypoperistalsis.90 Dysphagia is typically related to GERD or peptic stricture, with one report of esophageal epidermolysis bullosa acquisita.91 Malabsorption of d-xylose, steatorrhea, hyperplastic gastropathy, and protein-losing enteropathy have been described (see Chapter 28); the latter can be steroid responsive.92,93 Lupus peritonitis is a diagnosis that can be made only after other causes have been carefully excluded. Pneumatosis cystoides intestinalis may be an isolated benign condition or may accompany lupus vasculitis or necrotizing enterocolitis.94,95 One of the most devastating complications of lupus is gastrointestinal vasculitis. Affecting only 2% of patients, it has a fatality rate of more than 50%.96 Common sequelae include ulceration, hemorrhage, perforation, and infarction.97-99 Pancreatitis,100,101 gastritis, hemorrhagic ileocolitis resembling inflammatory bowel disease, and intussusception also have been reported. Although occasional case reports have documented polyarteritis-like changes on visceral arteriograms (described later), the typical pathologic changes are seen in the small vessels of the bowel wall rather than the medium-sized vessels of the bowel wall.94 Computed tomography (CT) scan may help establish the diagnosis of ischemic bowel disease in SLE if there are at least three of the following five CT findings: (1) bowel wall thickening, (2) target sign (a thickened bowel wall with peripheral rim enhancement or an enhancing inner and outer rim with hypoattenuation in the center), (3) dilatation of intestinal segments, (4) engorgement of mesenteric vessels, and (5) increased attenuation of mesenteric fat.102 Because visceral angiography is not routinely helpful, the diagnosis is difficult to establish. The role of endoscopy or upper gastrointestinal series in the diagnosis of lupus vasculitis is not well defined. The diagnosis currently rests on clinical judgment, findings on CT scans, and occasionally from surgical specimens when exploratory laparotomy is undertaken to rule out acute surgical emergencies.103 Treatment of abdominal lupus-induced vasculitis with glucocorticoids has been largely unsatisfactory. Although a controlled clinical trial comparing cyclophosphamide with glucocorticoids has not been performed, anecdotal reports of dramatic responses to intravenous cyclophosphamide are promising.94 Some investigators have suggested that cyclophosphamide be considered early in patients who have not shown significant improvement shortly after high-dose glucocorticoids are started. Patients with SLE have a 25% to 50% incidence of abnormal liver biochemical tests during the course of their disease, but clinically significant liver disease is rare.104 Abnormal liver tests are commonly associated either with medication use or with mild, predominantly lobular hepatitis associated with periods of SLE activity.104,105 Despite the shared association with antinuclear antibodies, the typical histologic and clinical features of autoimmune hepatitis are rarely observed in adult patients with SLE.104,106 Concurrent SLE and autoimmune hepatitis occur more frequently in pediatric patients.106 In addition, SLE patients with anticardiolipin antibodies or lupus anticoagulants may have thrombotic events in the liver leading to Budd-Chiari syndrome or nodular regenerative hyperplasia manifested by complications of portal hypertension.104,107

POLYMYOSITIS AND DERMATOMYOSITIS

Polymyositis is a syndrome characterized by weakness, high serum levels of striated muscle enzymes (creatine kinase,

aldolase), and electromyographic (EMG) or biopsy evidence of an inflammatory myopathy. When accompanied by a characteristic violaceous rash on the extensor surfaces of the hands and periorbital regions, the disease is termed dermatomyositis. The primary gastrointestinal symptoms are due to involvement of the cricopharyngeus, resulting in nasal regurgitation, tracheal aspiration, and impaired deglutition.108 Involvement is not limited to skeletal muscle fibers. Disordered esophageal motility, impaired gastric emptying, and poorly coordinated small intestinal peristalsis have been noted.109 Malabsorption, malnutrition, and pseudoobstruction rarely occur.110 Pathologically, edema of the bowel wall, muscle atrophy, fibrosis, and mucosal ulcerations or perforation due to vasculitis may be seen at any level of the gut. Symptoms include heartburn, bloating, constipation, and gastrointestinal hemorrhage. Pneumoperitoneum, pneumatosis intestinalis, colonic dilation, and pseudodiverticula also may be seen. Perforations of the esophagus and of duodenal diverticula have been described as rare complications.111,112 In middle-aged to older adult patients, dermatomyositis and possibly polmyositis are associated with an increased prevalence of malignancy.113 The possibility that gastrointestinal symptoms may be the resultl of an underlying malignancy should be considered when evaluating these patients (see Chapter 22).

MIXED CONNECTIVE TISSUE DISEASE

Mixed connective tissue disease (MCTD) is a syndrome with overlapping features of PSS, polymyositis, and SLE, often in the presence of high levels of antibody directed against ribonucleoprotein. Upper gastrointestinal symptoms are seen in most patients.114 Abnormalities include diminished esophageal peristalsis (48%), esophageal stricture (6%), abnormal gastric emptying (6%), and gastric bezoar (2%).114 Small intestinal and colonic involvement includes dilation of proximal bowel, slow transit, intestinal pseudoobstruction, diverticulosis, and, rarely, intestinal vasculitis. Pancreatitis also has been reported.114 Unlike PSS, the esophageal motility disturbances seen in MCTD appear to improve with the administration of glucocorticoids.

SJÖGREN’S SYNDROME

Sjögren’s syndrome (SS), occurring alone (primary SS) or in association with systemic autoimmune rheumatic diseases (secondary SS), is characterized by lymphocytic tissue infiltration of lacrimal and salivary glands with the clinical findings of keratoconjunctivitis sicca and xerostomia. As reviewed in Chapter 22, excessive dryness of the mouth and pharynx leads to oral symptoms of soreness, adherence of food to buccal surfaces, fissuring of the tongue, and periodontal disease.115 Dysphagia, reported by up to three quarters of patients with SS, can result from esophageal dysmotility and a lack of saliva; however, symptoms do not correlate with manometry or salivary secretion.116-118 Mild atrophic antral gastritis can be seen in 25% of patients with primary SS, but 31% were infected with H. pylori.119 Older studies that reported higher rates and greater severity of gastritis did not control for H. pylori infection. GAVE can occur in patients with SS and is responsive to fulguration therapy.120 A triad of sclerosing cholangitis, chronic pancreatitis, and SS has been reported in eight patients.121 Pancreatic exocrine function is frequently impaired.122 In primary SS, 7% of patients have positive antimitochondrial antibodies and among patients with primary biliary cirrhosis, clinical manifestations of SS are common (see Chapter 89).115

561

562

Section IV  Topics Involving Multiple Organs POLYARTERITIS NODOSA AND OTHER VASCULITIDES

Polyarteritis nodosa (PAN) is a necrotizing vasculitis of small and medium-sized muscular arteries, frequently with visceral involvement (Fig. 35-1). A characteristic feature of this condition is the finding of aneurysmal dilatations up to 1 cm in size seen on visceral angiography (Fig. 35-2). Abdominal complications occur in up to 50% of patients and carries a poor prognosis.123 Other clinical features of PAN include fever, myalgia, arthralgia of the large joints, mononeuritis multiplex, and livedo reticularis. Mesenteric visceral arteriograms are abnormal in up to 80% of patients

Figure 35-1.  Sigmoidoscopic examination in an intravenous drug user who presented with a foot drop and fecal occult blood. Although not shown, the mucosa in the rectum and proximal sigmoid colon was entirely normal. A sural nerve biopsy confirmed the diagnosis of polyarteritis nodosa.

Figure 35-2.  Celiac arteriogram in a patient with polyarteritis nodosa and hepatitis B surface antigenemia. Multiple saccular and fusiform aneurysms, as well as arterial tapering and beading, are seen throughout the celiac artery and its branches, especially the hepatic artery. (Courtesy Connie Wofsy, MD.)

with gastrointestinal involvement, with the superior mesenteric artery most commonly involved.123 Organ damage resulting from ischemia frequently underlies symptoms. The most common gastrointestinal manifestation is abdominal pain with other common symptoms including nausea, vomiting, and gastrointestinal bleeding.123 Bowel infarction and perforation, aneurysmal rupture, and acute cholecystitis are common causes of acute abdomen in PAN.123 Rarely, PAN can present as acalculous cholecystitis secondary to isolated vasculitis of the gallbladder.124 Pancreatitis,125 appendicitis,126 hemobilia,127 solitary biliary strictures,128 and hepatic infarcts129 also have been reported to complicate PAN. The frequency of hepatitis B infection in PAN has declined from more than 30% to less than 10% because of improved screening of the blood supply and vaccination against hepatitis B.130 Because of the frequent association with hepatitis B infection and potential association with hepatitis C infection (see Chapters 78 and 79), patients with clinical manifestations of PAN should be assessed for evidence of hepatitis B or C infection. Churg-Strauss syndrome (CSS, allergic granulomatous angiitis) is a small to medium-sized vessel vasculitis characteristically associated with eosinophilia, asthma, sinusitis, and rhinitis. Abdominal pain is the most common gastrointestinal symptom.131 Preceding the vasculitic phase of CSS, patients may present with an eosinophilic gastroenteritis associated with abdominal pain, nausea, vomiting, diarrhea, and bleeding with an absolute eosinophil count of greater than 1500 cells/mm3 (see Chapter 27).132 Additional gastrointestinal manifestations of CSS include pancreatitis, cholecystitis, ascites, small intestinal ulcerations, and perforation.131,133,134 Colonic involvement may present with multiple ulcers or obstruction.134,135 Henoch-Schönlein purpura (HSP) is a systemic vasculitis characterized by nonthrombocytopenic purpura, arthralgias, renal disease, and colicky abdominal pain. Although the disease is frequently seen in children and adolescents, adults of any age may be affected. Colicky abdominal pain and gastrointestinal bleeding are seen in two thirds of cases.136 Colonoscopic and endoscopic findings in bleeding patients include erosive duodenitis, small aphthous ulcerations, and petechial colonic lesions.137 In patients who undergo CT scan, common findings include bowel-wall thickening, dilated intestinal segments, mesenteric vascular engorgement, and regional lymphadenopathy.138 Other reported gastrointestinal complications of HSP include protein-losing enteropathy, esophageal and ileal structures, gastric and small bowel perforations, bowel infarction, pancreatitis, appendicitis, cholecystitis, intramural hematomas, and intussusception.139 Malignant atrophic papulosis (Kohlmeier-Degos disease) is a rare vasculitis that causes nausea, vomiting, bleeding, malabsorption, bowel ischemia, and perforation.140 Scattered on the skin are red papules that become hypopigmented atrophic scars (see Fig. 22-13). Cogan’s syndrome is characterized by nonsyphilitic interstitial keratitis, audiovestibular symptoms, and large-vessel vasculitis that may involve the gut. Gastrointestinal manifestations include abdominal pain, diarrhea, hepatomegaly, and splenomegaly.141 Crohn’s disease has been reported in association with this rare condition.142 Wegener’s granulomatosis, a systemic vasculitis characterized by pulmonary, sinus, and renal involvement, less commonly affects the gut.143 Inflammatory ileocolitis with hemorrhage, gangrenous cholecystitis, and bowel infarction have been reported.144 Wegener’s granulomatosis may mimic Crohn’s disease with granulomatous gastritis or ileitis.145,146

Chapter 35  Gastrointestinal and Hepatic Manifestations of Systemic Diseases Mixed immunoglobulin (IgG-IgM) cryoglobulinemia characterized by the triad of purpura, arthralgia, and asthenia may complicate chronic hepatitis C infection (see Chapter 79) and a variety of immune diseases, including inflammatory bowel disease, celiac disease, and postintestinal bypass syndrome. Cryoglobulinemia may cause severe visceral vasculitis with diarrhea, ischemia, and perforation of the small or large intestine.147

BEHÇET’S DISEASE

Behçet’s disease is an idiopathic inflammatory disorder characterized by oral aphthous ulcers, genital ulcers, uveitis, and skin lesions, with gastrointestinal involvement occurring in up to 50% of patients.148 As in Crohn’s disease, ulceration may occur throughout the alimentary tract, with the ileocecal region most commonly affected. Differentiating Behçet’s from Crohn’s disease can be difficult because of similarities in gastrointestinal symptoms, endoscopic findings, histology, and extraintestinal manifestations. Involvement of the esophagus includes ulcers (Fig. 35-3), varices, and perforation.149 In the stomach, which is infrequently involved, aphthous ulcers may be seen. The typical intestinal involvement in Behçet’s disease includes “punched-out” ileocecal ulcerations, the most common finding on colo­noscopy. Additional manifestations of Behçet’s disease include abdominal pain, diarrhea, bleeding, perforation, and fistulas (perianal, rectovaginal, and enteroenteric).150 Hepatic or portal vein thrombosis may occur in patients with Behçet’s, and this syndrome should be included in the differential diagnosis of patients presenting with Budd-Chiari syndrome.151,152 Medical therapy of the gastrointestinal lesions of Behçet’s disease includes mesalamine, gluco­corticoids, immunomodulators such as azathioprine and 6-mercaptopurine, infliximab, and thalidomide.153,154 Surgical intervention is associated with a high rate of recurrence, with nearly 50% requiring repeat surgery.155

Figure 35-3.  Aphthous ulcerations of the esophagus on an esophagogram in a patient with Behçet’s disease. (Courtesy of the Radiology Learning Center, University of California School of Medicine, San Francisco, Calif.)

SERONEGATIVE SPONDYLOARTHROPATHIES (REACTIVE ARTHRITIDES)

The term seronegative spondyloarthropathy is used to describe an interrelated group of inflammatory disorders that include ankylosing spondylitis, reactive arthritis (formerly called Reiter’s syndrome), and psoriatic arthritis. The term has also been used to describe the enteropathic spondylitis associated with Crohn’s disease and ulcerative colitis.156 These disorders are characterized by the absence of rheumatoid factor, an association with human leukocyte antigen-B27 (HLA-B27), and inflammation at the site of bony insertion of ligaments and tendons (enthesitis). There is a high prevalence of clinically silent inflammatory colon lesions in patients with these seronegative spondylo­ arthropathies.157 Capsule endoscopy may yield more small bowel abnormalities than ileocolonoscopy.158 Conversely, 22% of patients with inflammatory bowel disease have evidence of a seronegative spondyloarthropathy, with ankylosing spondylitis most commonly seen.159 Although infliximab has been shown to induce remissions in some patients with ankylosing spondylitis as well as in Crohn’s disease, the effect of infliximab on gastrointestinal inflammatory lesions in typical seronegative spondyloarthropathies has not yet been studied.

MARFAN’S AND EHLERS-DANLOS SYNDROMES

Owing to defective collagen synthesis, patients with Marfan’s or Ehlers-Danlos syndrome develop skin fragility, megaesophagus, small intestine hypomotility, giant jejunal diverticula, bacterial overgrowth, and megacolon.160 Mesenteric arterial rupture and intestinal perforation also can occur.161

FAMILIAL MEDITERRANEAN FEVER

Familial Mediterranean fever (FMF) is an autosomal recessive inherited disease characterized by recurrent selflimiting attacks of fever, joint pain, and abdominal pain. Acute attacks typically last three to five days. FMF is most commonly seen in people of Mediterranean origin including Sephardic Jews, Arabs, Turks, Greeks, Italians, and Armenians, although FMF has been described in Cubans, and Belgians. The gene responsible for FMF in Mediterranean patients, designated MEFV, has been mapped to chromosome 16, which encodes a 781-amino acid protein called pyrin or marenostrin.162 Gastrointestinal symptoms, typically manifest as episodic abdominal pain, are seen in 95% of patients, and this may be the presenting symptom in as many as 50% of cases.163 Abdominal pain may be diffuse or localized and may range from mild bloating to acute peritonitis with boardlike rigidity, rebound tenderness, and air-fluid levels on upright radiographs. The acute presentation may be confused with acute appendicitis, cholecystitis, or pelvic inflammatory disease, whereas relapsing and remitting attacks may be confused with other diseases such as porphyrias. Small bowel obstruction from adhesions may occur as a consequence of recurrent sterile peritonitis or as a result of previous exploratory surgery. In patients with obstruction due to adhesions, abdominal attacks without other typical symptoms (arthralgias, fever) should tip off the clinician to consider an obstruction.164 The diagnosis of FMF is based on validated clinical criteria including fever, serositis, location of pain, and response to colchicine.165 In FMF, the long-term prognosis is poor in patients who develop nephrotic syndrome and chronic kidney disease from amyloid A deposition163 (amyloidosis is discussed later in this chapter). Prophylactic colchicine has been shown to reduce the frequency of attacks, prevent amyloi-

563

564

Section IV  Topics Involving Multiple Organs dosis, and avoid renal failure.166 Vasculitis in the form of HSP, PAN, protracted febrile myalgia, or Behçet’s is encountered in 3% of FMF patients.163

ONCOLOGIC AND HEMATOLOGIC DISEASES METASTASES

Metastasis to the gut can occur by direct invasion from adjacent organs, by intraperitoneal seeding, or by hematogenous or lymphatic spread. About 20% of all patients with nongastrointestinal malignancies have metastases to the gastrointestinal tract, the most common of which are breast, lung, and ovarian cancers and melanoma (Fig. 35-4).167 Patterns of metastases are not random but reflect the location and histologic type of the primary tumor. The esophagus is most frequently affected by direct extension from tumors arising from adjacent structures (bronchus and stomach). The stomach is a particularly common site of breast cancer metastases, and the small intestine can be involved by tumor extension from the stomach, pancreas, biliary system, kidney, or retroperitoneum. The pancreas is usually an asymptomatic site of metastasis, with the most common primary tumors being lung, gastrointestinal, and renal.168 The ileum may be affected by cancers arising in the colon or pelvis. Metastases to the gut typically begin in the serosa or submucosa and produce intraluminal lesions that can lead to obstruction, submucosal polypoid masses that can result in intussusception or ulcerated mucosal lesions. The most common presenting clinical condition in patients with metastatic lesions to the gut is small bowel obstruction. Lobular breast cancer, malignant melanoma, and non–small cell lung cancer are the most common neoplasms to cause small bowel obstruction from isolated metastases.169 In addition, pain, fever, ascites, gastrointestinal bleeding, and perforation have been described. Metastases to the gastrointestinal tract may be difficult to diagnose. Barium contrast studies may reveal extramural masses, mucosal ulcerations, or a rigid stomach with the appearance of linitis plastica. CT may be helpful in determining the primary tumor, in tumor staging, and in detecting large serosal implants. Small bowel metastases,

however, are detectable radiographically in only 50% of cases.170 When feasible, surgical resection should be used to treat gastrointestinal metastases that result in obstruction, perforation, or significant hemorrhage. If a solitary bowel metastasis is the only evident site of disseminated malignancy, segmental bowel resection should be performed, offering a small chance for cure. In aggressive resections of melanoma metastases, the mesenteric nodes draining the involved segment of bowel should be resected because they frequently contain tumors.171

PARANEOPLASTIC SYNDROMES

Paraneoplastic syndromes affecting the gut include the hormonal effects of carcinoid tumors, vasoactive intestinal polypeptide-secreting tumors (VIPomas), gastrinomas, and somatostatinomas (see Chapters 31 and 32), as well as the gastrointestinal effects of hypercalcemia (constipation, nausea, and vomiting). A watery diarrhea syndrome with elevated serum immunoreactive VIP has been described accompanying nonpancreatic tumors such as bronchogenic carcinomas, ganglioneuromas, pheochromocytomas, and a rare mastocytoma.172 Elevated serum levels of somatostatin, calcitonin, gastrin, and corticotropin also have been reported in pheochromocytoma.173 Paraneoplastic gastrointestinal dysmotility may occur in some patients with occult or established malignancy and specific serum antibodies. Clinically the patient may present with pseudoachalasia, gastroparesis, intestinal pseudoobstruction, or constipation. Intestinal pseudo-obstruction (see Chapter 120) is most frequently associated with small cell carcinoma of the lung but has been described with other tumors such as squamous cell lung carcinoma, lymphoma, melanoma, and cancers of the kidney, breast, and prostate.174-176 Patients with paraneoplastic intestinal pseudo-obstruction characteristically suffer from consti­ pation and obstipation and from symptoms of intestinal obstruction. In addition, dysphagia, gastroparesis, early satiety, autonomic insufficiency, and peripheral neuropathy have been described.177 The onset of symptoms may precede the discovery of the primary tumor by several years. The gastrointestinal pathology in this syndrome is confined to the myenteric plexus, in which an inflammatory lymphocytic infiltrate is variably seen accompanying neuronal degeneration.178 Cross-reacting autoantibodies found in the sera of these patients bind to the primary tumor cells and to neural cells in the myenteric plexus, resulting in inflammation and destruction of the myenteric plexus.179 In the setting of pseudo-obstruction, detection in the serum of circulating antineuronal nuclear antibodies (ANNA-1 or anti-Hu), type 1 Purkinje cell antibodies (PCA-1), or N-type calcium channel binding antibodies should suggest a paraneoplastic process and prompt further evaluation for an underlying malignancy.177 ANNA-1 are postulated to induce neuronal apoptosis leading to gut dysmotility.180 Although the symptoms of paraneoplastic pseudo-obstruction may resolve with successful treatment of the primary tumor, persistence of gastrointestinal symptoms despite effective anticancer treatment is more common. Attempts to alleviate the symptoms of pseudo-obstruction with prokinetic agents have been disappointing.

HEMATOLOGIC MALIGNANCIES Figure 35-4.  Endoscopic view of an ulcerated metastatic melanoma lesion involving the second portion of the duodenum in a young man who presented with upper gastrointestinal bleeding.

Liver involvement during hematologic malignancies is only rarely life threatening or a source of great morbidity. Nevertheless, the liver is a major component of the reticuloendothelial system, and thus it is not surprising that malignant

Chapter 35  Gastrointestinal and Hepatic Manifestations of Systemic Diseases Table 35-2  Involvement of the Liver in Patients with Hematologic Malignancies Frequency of Liver Infiltration (%)

Hodgkin’s lymphoma Non-Hodgkin’s lymphoma Hepatosplenic γδ T cell lymphoma Multiple myeloma Leukemia   ALL   AML   CLL   HCL   LGLL

Clinical Evaluation

Postmortem

8-14

55

16-57 80

52 100

30-40

40-50

— — — 100 75-100

>95 75 98 100 —

OTHER NOTABLE HISTOLOGIC ABNORMALITIES (FREQUENCY) Portal lymphocytic infiltrates (32%), granulomas (9%-25%), steatosis (11%), hemosiderosis (9%), idiopathic cholestasis (110

5

S1, Proximal esophageal segment

Figure 42-12.  High-resolution esophageal pressure topography spanning from the pharynx (locations 0 to 2 cm) to the stomach (locations 29 to 35 cm) of a normal subject with normal peristalsis and normal esophagogastric junction (EGJ) relaxation. The transition zone, demarcating the end of the proximal esophageal segment S1 (striated muscle) and the beginning of the distal esophageal segment S2 (smooth muscle), is readily identified as a pressure minimum. Note that the distal esophageal segment, in fact, has three subsegments (S2, S3, S4) within it, each with an identifiable pressure peak. S4, the lower esophageal sphincter (LES), contracts at the termination of peristalsis and then descends back to the level of the crural diaphragm as the period of swallow-related esophageal shortening ends. The onset of the deglutitive relaxation window is at the onset of upper esophageal sphincter (UES) relaxation, whereas the offset is 10 seconds later. The spatial domain within which EGJ relaxation is assessed (the eSleeve range) is user defined, spanning at least 6 cm (in this example, labeled 0 and 6 cm), depending on the extent of esophageal shortening after the swallow. The contractile front velocity (CFV) is the slope of the line connecting points on the 30 mm Hg isobaric contour at the proximal margin of S2 and the distal margin of S3.

Distal esophageal segment

S2 15 Contractile front velocity (CFV) = 15 cm/5s = 3 cm/s

20

25

S3

70

50

S4 (LES)

30

0 cm EGJ

30

Deglutitive EGJ relaxation window

10

6 cm 0s

35

90

Center of transition zone

10 Location along lumen (cm)

696

0

5

Crural diaphragm contractions 10 Time (sec)

10 s 15

eSleeve range 20

–10

Chapter 42  Esophageal Neuromuscular Function and Motility Disorders ACHALASIA WITH COMPRESSION

Length along the esophagus (cm)

CLASSIC ACHALASIA 1

1 5 UES Relaxation

10

mm Hg 150 125

15

100

20

4s IRP = 42 mmHg

25

1.5s

2.5s

30 35

10

75

20

50

25

25 0 –10

30

Length along the esophagus (cm)

SPASTIC ACHALASIA (2D)

5

20

200

30

100 30 0

CFV = 25 10 cm/s

35 0

C

3

6

9

1

5 6

7 8 9 10

SPASTIC ACHALASIA (3D) mm Hg

5

>300 100

10 15

50 25

20

10

25

5 0

30 35

0

12 15 18 21

Time (sec)

50

Time (sec)

mm Hg

300

75 25 0 –10

B

15

100

35

400

10

125 30 mm Hg isobaric contour

0 1 2 3 4

Time (sec)

1

UES Relaxation

15

0 1 2 3 4 5 6 7 8 9 10 11

A

mm Hg 150

5

D

3

6

9

12

15

18

21

Time (sec)

Figure 42-13.  Achalasia subtypes are distinguished by three distinct manometric patterns of esophageal body contractility (panels A-C). In classic achalasia (panel A), there is no significant pressurization within the body of the esophagus and impaired esophagogastric junction (EGJ) relaxation. The integrated relaxation pressure (IRP) was 42 mm Hg in this example. Panel B represents a swallow from a patient with the “achalasia with compression” subtype exhibiting rapid panesophageal pressurization of the fluid column trapped between the sphincters as the esophagus shortens at seven to eight seconds. Panel C illustrates a pressure topography plot typical of spastic achalasia. Although this swallow is also associated with rapidly propagated pressurization, the pressurization is attributable to an abnormal lumen-obliterating contraction. A three-dimensional rendering of these same pressure data (panel D) illustrates the peaks and valleys of that spastic contraction (brown vs. red); this swallow would likely appear as a rosary-bead or corkscrew pattern on fluoroscopy. (Modified from Pandolfino JE, Kwiatek MA, Nealis T, et al. Achalasia: A new clinically relevant classification by high resolution manometry. Gastroenterology 2008; 135:1526-33.)

Apart from a rapid CFV, other common abnormalities of the distal esophageal contraction are weak or absent peristalsis. In such instances, the 30 mm Hg isobaric contour is either discontinuous or absent, reflective of either focal or diffuse hypotensive contraction within the distal segment. Each swallow is thus characterized as normal (intact 30 mm Hg isobaric contour and a CFV < 8 cm/second), hypotensive (3 cm or greater defect in the 30 mm Hg isobaric contour), or absent (complete failure of contraction with no pressure domain > 30 mm Hg). Weak or absent peristalsis is a risk factor for impaired bolus clearance, but, whether impaired bolus clearance occurs depends on the balance between the severity of weakness and the magnitude of outflow resistance at the EGJ.287 Once swallows are characterized by the integrity of deglutitive EGJ relaxation and normality of the CFV, the distal esophageal contraction is further analyzed for the vigor of contraction using a newly developed measure for high-resolution esophageal pressure topography, the distal contractile integral (DCI). The DCI integrates the

length, vigor, and persistence of the two subsegments of the distal esophageal segment contraction, expressed as mm Hg·s·cm. A DCI value greater than 5000 mm Hg·s·cm is considered elevated.288 Adopting the nomenclature “nutcracker esophagus” from conventional manometry, this is the high-resolution manometry criterion defining hyper­ tensive peristalsis and was seen in 9% of a 400-patient series.285 However, there was substantial heterogeneity as to the locus of the hypertensive contraction within this group, potentially involving either or both of the subsegments within the distal esophageal contraction. Similarly, the LES can exhibit a hypertensive postdeglutitive contraction, defined as exceeding 180 mm Hg. Furthermore, one particularly interesting subgroup, defined by having a higher threshold DCI (>8000 mm Hg·s·cm), exhibited repetitive high-amplitude contractions and was clinically distinguishable by the uniform association with dysphagia or chest pain. Similar to DES, this “spastic nutcracker” pattern is very rare, found in only 12 (3%) of this 400-patient series (Fig. 42-15).

697

Section V  Esophagus HIGH IBP, NORMAL CFV (at 50 mm Hg IBC)

Length along the esophagus (cm)

0 Pressure (mm Hg) 150

5 10

120 15

90

20

60 30 mm Hg isobaric contour

25

A

30 0

50 mm Hg isobaric contour

30 35

0

5

10 Time (sec)

15

20

RAPID CFV, SPASTIC CONTRACTION Figure 42-14.  Differentiating increased intrabolus pressure (IBP) from a rapidly pro­ pagated spastic contraction (bottom). A, Swallow with functional obstruction at the eso­phagogastric junction (EGJ). Note that the 30 mm Hg isobaric contour (IBC) line deviates quickly from the 50 mm Hg isobaric contour line (arrows). In this case the contraction front velocity (CFV) was normal, reflecting the propagation velocity of the 50 mm Hg isobaric contour rather than the 30 mm Hg isobaric contour. B, Swallow with rapid CFV attributable to spasm. EGJ relaxation is normal and the 30 and 50 mm Hg isobaric contours parallel each other, indicating that no compartmentalized esophageal pressurization has occurred. The entire distal esophagus is contracting simultaneously. (Modified from Pandolfino JE, Ghosh SK, Rice J, et al. Classifying esophageal motility by pressure topography characteristics: A study of 400 patients and 75 controls. Am J Gastroenterol 2008; 103:27-37.)

0 Length along the esophagus (cm)

698

Pressure (mm Hg) 150

5 10

120 15 20 25

0

B

Following analysis of individual swallows by the criteria outlined, the component results are synthesized into a global manometric diagnosis by the criteria detailed in Table 42-2. Patients with normal EGJ relaxation, normal CFV, and a DCI less than 5000 mm Hg·s·cm are normal. The abnormalities encountered are described in specific functional terms with the intent that these then be interpreted within the clinical context of the patient. The classification detailed in Table 42-2 represents an incremental update on the Chicago classification,289 the task of a newly convened international working group focused on the standardization of the performance and interpretation of high-resolution esophageal pressure topography studies.

Intraluminal Impedance Measurement

60 30

50 mm Hg isobaric contour

30 35

90

30 mm Hg isobaric contour

Intraluminal impedance monitoring was described more than a decade ago as a method to assess intraluminal bolus transit without using fluoroscopy. The technique uses an intraluminal catheter with multiple, closely spaced pairs of

0

5

10

15

20

Time (sec)

metal rings (see Fig. 42-10). An alternating current is applied across each pair of adjacent rings and the resultant current flow between the rings is dependent on the impedance of the tissue and luminal content between the rings. Impedance decreases when the electrodes are bridged by liquid and increases when they are surrounded by air thereby providing data on the direction, content, and completeness of bolus transit. Validation data suggest that liquid bolus entry at the level of an electrode pair is indicated by a 50% drop in impedance and return of the impedance tracing to 50% of baseline correlates with the passage of the tail of the bolus on fluoroscopy, also indicated by the contractile upstroke noted on manometry (see Fig. 42-10). Validation studies of intraluminal impedance measurement against videofluoroscopy have shown excellent concordance in ascertaining bolus transit, reporting agreement in 97% (83/86) of swallows analyzed.290 Intraluminal impedance measurement has also recently been combined with manometry to assess the efficacy of

Chapter 42  Esophageal Neuromuscular Function and Motility Disorders

Length along the esophagus (cm)

0 Pressure (mm Hg)

5

150

10

120 30 mm Hg

15

90

20

60

25 30 35

0

5

10

15

20

25

Table 42-2  The Chicago Classification of Distal Esophageal Motility Disorders With Normal EGJ Relaxation (Mean Integrated Relaxation Pressure 5000 and 180 mm Hg Normal CFV, mean DCI >8000 mm Hg⋅s⋅cm (Fig. 42-15) Normal EGJ relaxation and spasm (CFV >8 cm/s) with ≥20% of swallows (see Fig. 42-14B)

Achalasia with esophageal compression Spastic achalasia

Functional EGJ obstruction*

CRITERIA Impaired EGJ relaxation and aperistalsis (see Fig 42-13A) Impaired EGJ relaxation, aperistalsis, and panesophageal pressurization with ≥20% of swallows (see Fig. 42-13B) Impaired EGJ relaxation, aperistalsis, and spasm (CFV >8 cm/s) with ≥20% of swallows (see Fig. 42-13C and D) IBP >30 mm Hg compartmentalized between the peristaltic wavefront (normal or nutcracker) and EGJ (see Fig. 42-14A)

*May represent an achalasia variant. CFV, contractile front velocity; DCI, distal contractile integral; EGJ, esophagogastric junction; IBP, intrabolus pressure; LES, lower esophageal sphincter.

patients with esophageal chest pain tend to have lower threshold volumes for both first perception and first pain perception compared with controls.294,295 Combining impedance planimetry with balloon distention allowed other investigators to correlate biomechanical properties of the esophagus with the generation of chest pain.296,297 Results of those studies suggested that the tension-strain curve in chest pain patients was shifted to the left when compared with controls, a finding consistent with reduced compliance of the esophageal wall.269 The standard test of chemosensitivity is the Bernstein test wherein 0.1 normal hydrochloric acid is perfused in the esophagus to reproduce chest pain or heartburn. Typically, acid infusion is alternated with saline perfusion in a blinded fashion to increase the objectivity of the test, but no standardized protocol exists. Beyond the standard Bernstein perfusion test to assess esophageal sensitivity to acid, newer probes have been devised to test esophageal responsiveness to thermal challenges and transmucosal electrical nerve stimulation. However, although these tools have unquestionably been useful in improving our understanding of the

699

700

Section V  Esophagus interaction between peripheral receptors and central pain perception, their clinical utility remains limited owing to the lack of protocol standardization and the somewhat cumbersome nature of the studies. Currently, use of these devices is limited to subspecialty centers and further refinement will be required before mainstream clinical use can be advocated.

TREATMENT Oropharyngeal Dysphagia

Management of oropharyngeal dysphagia is focused on four specific issues: (1) identification of an underlying systemic disease, (2) characterization of a disorder amenable to surgery or dilation, (3) identification of specific patterns of dysphagia amenable to swallowing therapy, and (4) assessment of aspiration risk. Identifying Underlying Disease A potential outcome of the evaluation is the identification of an underlying neuromuscular, neoplastic, or metabolic disorder that dictates specific management. For example, dysphagia can be the presenting symptom in patients with myopathy, myasthenia, thyrotoxicosis, motor neuron disease, or Parkinson’s disease. In each instance, managing the underlying disease requires a specific treatment. Whether or not treatment of the underlying disorder improves swallowing function depends on the natural history of the specific disease and whether effective treatment exists. Disorders Amenable to Surgery The most common surgical treatment for oropharyngeal dysphagia is cricopharyngeal myotomy but the efficacy of myotomy in neurogenic or myogenic dysphagia is variable. Most series evaluating the efficacy of myotomy in these circumstances are uncontrolled and lack validated or even specific outcome measures. Thus, although an overall favorable response rate in excess of 60% is reported in this literature, there are no validated criteria for patient selection. Theoretically, the functional limitation faced by patients with neurogenic or myogenic dysphagia is of weak pharyngeal propulsion and the potential benefit of myotomy in that circumstance is less obvious than in the case of obstruction at the level of the cricopharyngeus.298 Patterns of Oropharyngeal Dysphagia Amenable to Swallow Therapy Identifying potential treatments for oropharyngeal dysphagia begins with definition of the aberrant physiology as categorized in Table 42-1. This is best accomplished with a videofluoroscopic swallowing study that first characterizes a patient’s swallow dysfunction and then proceeds to test the effectiveness of selected compensatory or therapeutic treatment strategies. Compensatory treatments include postural changes, modifying food delivery or consistency, or the use of prosthetics. For instance, head turning can eliminate aspiration or pharyngeal residue by favoring the more functional side in patients with hemiparesis.17 Similarly, diet modifications can reduce the “difficulty” of the swallow. Therapeutic strategies are designed to alter the physiology of the swallow, usually by improving the range of motion of oral or pharyngeal structures using voluntary control of oropharyngeal movement during a swallow. Depending on the severity of the impairment, level of motivation, and global neurologic integrity, defective elements of the swallow can be selectively rehabilitated. For a detailed description of the techniques and limitations of swallow therapy, the reader is referred to treatises on the topic.17,299

Evaluating Aspiration Risk Oropharyngeal dysphagia is responsible for an estimated 40,000 deaths a year due to aspiration pneumonia.300 Videoflouroscopy is considered the most sensitive test for detecting aspiration, reportedly detecting instances not evident by bedside evaluation in 42% to 60% of patients. However, despite the logical association between deglutitive aspiration and the subsequent development of pneumonia, this sequence is not inevitable. In fact, available data suggest that radiographic aspiration has a positive predictive value of only 19% to 68% and a negative predictive value of 55% to 97% for pneumonia.300 Nonetheless, the balance of evidence suggests that detection of aspiration is a predictor of pneumonia risk, and that its detection dictates that compensatory swallowing strategies, non-oral feeding or corrective surgery be instituted. Whether non-oral feeding eliminates the risk of aspiration is controversial. In one study of 22 patients with radiographic aspiration, pneumonia and death were more frequent among patients who received feeding tubes.185 This suggests that aspiration of oral secretions may be the essential element in pneumonia risk and has led some to consider procedures such as tracheostomy to protect the airway.

Hypopharyngeal (Zenker’s) Diverticulum and Cricopharyngeal Bar

The treatment of hypopharyngeal diverticulum is cricopharyngeal myotomy with or without a diverticulectomy (see Chapter 23). Cricopharyngeal myotomy reduces both the resting sphincter tone and resistance to flow across the UES. A study found that the compliance of the sphincter following diverticulectomy with myotomy was restored to normal following surgery, as indicated by normal hypopharyngeal intrabolus pressure during swallowing.301 Good or excellent results are reported in 80% to 100% of Zenker’s patients treated by transcervical myotomy combined with diverticulectomy or diverticulopexy.299 There are instances in which a limited procedure would be adequate, but a definitive approach to the problem of pulsion diverticula should generally involve myotomy and diverticulectomy. Diverticulectomy alone risks recurrence because the underlying stenosis at the level of the cricopharyngeus is not remedied. Similarly, myotomy alone may not solve the problem of food accumulation within the diverticulum, with attendant regurgitation and aspiration. Small diverticula may, however, disappear spontaneously following myotomy. A more recent trend is to treat Zenker’s diverticula via either rigid or flexible endoscopy. With both techniques, the principle is to divide the septum between the lumen of the diverticulum and the lumen of the esophagus. The division allows food and liquid to flow out of the diverticulum distal to the cricopharyngeus (which was within the septum) rather than to accumulate within the diverticulum. In the case of rigid endoscopy the procedure is performed under general anesthesia with a stapling device. In the case of flexible endoscopy the procedure is performed under light sedation with a needle knife, argon plasma coagulation, or hot biopsy forceps. Controlled trials have not been done comparing the two procedures, but a recent summary of 376 reported cases treated with flexible endoscopic methods found treatment to result in clinical resolution in 43% to 100% of cases among series.184 Whether a cricopharyngeal bar in the absence of a diverticulum requires treatment is less clear. Certainly, if dysphagia is present and combined fluoroscopic/manometric analysis demonstrates reduced sphincter opening in conjunction with elevated upstream intrabolus pressure,

Chapter 42  Esophageal Neuromuscular Function and Motility Disorders there is good rationale for treatment. One recent uncontrolled series suggests that in this scenario dilation with a large-caliber bougie may be efficacious in relieving dys­ phagia and this approach is certainly a reasonable treatment option prior to myotomy.302

Achalasia

Because the underlying neuropathology of achalasia cannot be corrected, treatment is directed at compensating for the poor esophageal emptying and preventing complications. In practical terms this amounts to reducing LES pressure so that gravity promotes esophageal emptying. Peristalsis is not restored with therapy. LES pressure can be reduced by pharmacologic therapy, forceful dilation, or surgical myotomy. Pharmacologic treatments, on the whole, are not very effective, making them more appropriate as temporizing maneuvers than definitive therapies. The definitive treatments of achalasia are disruption of the LES either surgically (Heller myotomy) or with a pneumatic dilator. Which of these is the optimal approach remains an issue of debate given the paucity of randomized controlled trials with accepted criteria for assessing efficacy. A further limitation of the previous studies is failing to stratify patients by disease severity or, as more recently defined, by disease subtype.224 High-resolution esophageal pressure topography allows the subtyping of achalasia into three distinct patterns: I, classic achalasia; II, achalasia with compression; and III, spastic achalasia (see Fig. 42-13). From a conceptual vantage point types I and II represent a continuum, with type II representing early disease before the progression of esophageal dilatation characteristic of type I. Type III, on the other hand, is a subtype characterized by spasm of the distal esophagus. The significance of these disease subtypes is in how differently they responded to therapy, be it botulinum toxin injection, pneumatic dilation, or Heller myotomy. In a series of 99 new cases of achalasia, the overall treatment response was 56% with type I, 96% with type II, and only 29% with type III. The literature pertinent to achalasia treatment is mainly composed of numerous uncontrolled case series using a variety of qualitative endpoints as indications of efficacy. As noted, there is also minimal standardization as to the criteria for defining achalasia, the disease severity included in one series versus another, or the technical details of how pneumatic dilation or Heller myotomy are performed. Furthermore, some series were collected prospectively, some retrospectively, and some a combination. Given all of these limitations, there is little merit to embarking on a detailed comparison of outcomes between techniques. The existing treatment data are summarized next. Pharmacologic Therapy Smooth muscle relaxants such as nitrates or calcium channel blockers, administered sublingually immediately prior to eating can relieve dysphagia in achalasia by reducing the LES pressure. Amyl nitrite,303 sublingual nitroglycerin, theophylline, and β2-adrenergic agonists304 have also been tried. The largest reported experience has been with isosorbide dinitrate (Isordil) and nifedipine.305 Isosorbide dinitrate, 5 to 10 mg sublingually before meals, reduces LES pressure by 66% for about 90 minutes, with the degree of dysphagia relief paralleling the magnitude of the LES response over the 19-month trial.306 Side effects, particularly headache, are common. Placebo-controlled trials have not been reported. Calcium channel blockers (diltiazem, nifedipine, verapamil) reduce LES pressure by 30% to 40% for more than

an hour.306,307 The largest clinical experience in achalasia has been with nifedipine (Procardia). Nifedipine, 10 mg sublingually (capsules are crushed in the mouth) administered before meals (30 to 40 mg per day) was studied in 29 patients with early achalasia (prior to esophageal dilatation) in a placebo-controlled trial. Nifedipine was significantly better that placebo (which had no benefit), with good results in 70% of achalasic patients followed for 6 to 18 months.305 However, subsequent placebo-controlled crossover trials have found only minimal benefit with nifedipine.308 Side effects of nifedipine include flushing, dizziness, headache, peripheral edema, and orthostasis. Sildenafil (Viagra) is another smooth muscle relaxant that can decrease LES pressure in patients with achalasia by blocking phosphodiesterase type 5, the enzyme that destroys cyclic guanosine monophosphate induced by NO. A double-blind placebo controlled trial found that 50 mg of sildenafil significantly reduced LES pressure and relaxation pressure when compared with placebo.309 The effect peaked at 15 to 20 minutes after administration and persisted for less than one hour. Although conceptually appealing, the practicality of using sildenafil clinically is limited by its cost that is rarely, if ever, covered by health care insurance. Botulinum Toxin Injection The initial landmark study of botulinum toxin in achalasia reported that intrasphincteric injection of 80 units of botulinum toxin decreased LES pressure by 33% and improved dysphagia in 66% of patients for a six-month period.310 Botulinum toxin irreversibly inhibits the release of acetylcholine from presynaptic cholinergic terminals, effectively eliminating the neurogenic component of LES pressure. However, because this inhibitory effect is eventually reversed by the growth of new axons, botulinum toxin is not a long-lasting therapy. The technique involves injecting divided doses of botulinum toxin into four quadrants of the LES with a sclerotherapy catheter. Side effects are rare, but include chest discomfort for several days and rash. Although many patients initially experience a good response, there is minimal continued efficacy at one year.311-313 Repeat injection can be effective for a reasonable subset of patients, but the injection leads to a local inflammatory reaction and fibrosis, ultimately limiting this strategy. Doses greater than 100 units do not have increased efficacy.314 Studies comparing botulinum toxin injection to pneumatic dilation suggest that the expense of repeated injection outweighs the potential economic benefits of added safety, unless the patient’s life expectancy is minimal.315 Thus, this option is mainly reserved for older adults or frail individuals who are poor risks for definitive treatments. Pneumatic Dilation Therapeutic dilation for achalasia requires distention of the LES to a diameter of at least 3 cm to produce a lasting reduction of LES pressure, presumably by partially disrupting the circular muscle of the sphincter. Dilation with an endoscope, standard bougies (up to 60 French), or with throughthe-scope balloon dilators (up to 2 cm) provides very temporary benefit at best. Only dilators specifically designed to treat achalasia achieve adequate diameter for lasting effectiveness. The basic element of an achalasia dilator is a long, noncompliant, cylindrical balloon that can be positioned across the LES fluoroscopically (Rigiflex dilator) or endoscopically (Witzel dilator) and then inflated to a characteristic diameter in a controlled fashion using a handheld manometer. There is general agreement that pneumatic dilation can be done on an outpatient basis with the patient

701

702

Section V  Esophagus under conscious sedation. The technique of pneumatic dilation is variable among practitioners in terms of patient preparation, parameters of balloon inflation, and postdilation monitoring. In patients with substantial esophageal retention, it is useful to impose a liquid diet for one or more days prior to the procedure. Reported balloon inflation periods range from several seconds to five minutes.316 Although there is minimal methodologic consistency among authors, a cautious approach of beginning with a smalldiameter dilator (3 cm) and progressing to larger diameters (3.5 and 4 cm) only when the smaller dilator proved ineffective is fairly universal. As for inflation pressures, these are of minimal relevance with modern noncompliant balloon dilators because they do not distend beyond their specified diameter regardless of inflation pressure. Hence, it is simply necessary to observe under fluoroscopy that the balloon is properly positioned to capture the LES, observed as the “waist” of the hourglass-shaped balloon silhouette and that the waist fully effaces as the inflation proceeds. As for technical details of the procedure other than balloon diameter, there is minimal evidence that they influence outcome. The major complication of pneumatic dilation is esophageal perforation (see Chapter 40); mortality is fortunately rare.317 The reported incidence of esophageal perforation consequent from pneumatic dilation ranges between 1% and 5%261,316 with a global average of 3%. Because most perforations are readily evident or at least suspected within an hour of the procedure, patients should be observed closely for signs of an esophageal leak for at least two hours after pneumatic dilation. Alternatively, some practitioners routinely obtain a fluoroscopic examination of the esophagus following pneumatic dilation to ensure that perforation has not occurred. Usually, water-soluble contrast is given first, followed by barium. If a perforation appears small and contained or intramural, conservative management in the hospital consisting of close observation while maintaining the patient on nothing per mouth status and administering intravenous antibiotics is appropriate.261 If a perforation is substantial, or if worsening pain and fever occur during observation of what was thought to be a small perforation, surgical repair should be pursued expediently. Patients with a perforation from pneumatic dilation that is recognized and promptly treated surgically (within six to eight hours) have outcomes comparable with those of patients undergoing elective Heller myotomy.318 The best predictor of efficacy following a pneumatic dilation is the postdilation LES pressure; neither sphincter relaxation nor peristaltic function is significantly changed. A postdilation LES pressure less than 10 mm Hg is associated with prolonged remission, whereas a postdilation LES pressure greater than 20 mm Hg predicts that little benefit will occur from the procedure.319 In instances of an unsatisfactory result, it is reasonable to perform a subsequent dilation within a matter of weeks using an incrementally larger dilator. If the benefit of dilation persisted for a year or more, it is neither unusual nor dangerous to repeat pneumatic dilation as necessary. The clinical efficacy of dilation has been reported to range from 32% to 98%.311 Patients having a poor initial result or rapid recurrence of symptoms have diminished likelihood of responding to additional dilations.311 Subsequent response to surgical myotomy is not influenced by the history of previous dilations.261 Heller Myotomy Current surgical procedures for treating achalasia are variations on the esophagomyotomy described by Heller in 1913 consisting of an anterior and posterior myotomy performed

through either a laparotomy or a thoracotomy.311 Subsequently, this was modified to an anterior myotomy via thoracotomy. The appeal of myotomy is that it offers a more predictable method of reducing LES pressure than does pneumatic dilation.320 Although clearly efficacious, open Heller myotomy is associated with considerable morbidity related to thoracotomy, which led most patients to pursue pneumatic dilation as the initial intervention. However, adoption of the laparoscopic approach for achalasia surgery has led many practitioners to reconsider this. Published series of the efficacy of Heller myotomy in treating achalasia report good to excellent results in 62% to 100% of patients, with persistent dysphagia troubling less than 10% of patients.311 Recent studies suggest that a laparoscopic approach is associated with similar efficacy, reduced morbidity, and shorter hospital stay when compared with myotomy via thoracotomy, laparotomy, or thoracoscopy.311,321-325 The overall mortality from Heller myotomy is less than 2%. Historically, postmyotomy reflux in achalasic patients could be particularly severe, making this a hotly disputed detail of the surgical technique.326 However, with the broad use of proton pump inhibitors, reflux is usually easily controlled, making these complications very unlikely. Thus, laparoscopic Heller myotomy combined with a partial fundoplication (Toupet or Dor) has become the preferred surgical procedure for achalasia. An unsatisfactory result following Heller myotomy can result from incomplete myotomy, scarring of the myotomy, functional esophageal obstruction from the antireflux component of the operation, paraesophageal hernia, or severe esophageal dilatation. Heller Myotomy versus Medical Treatment Although pharmacologic therapy is simple and safe, it is increasingly clear that this should be reserved for use as a temporizing measure while more definitive therapy is being considered. Thus, practically speaking, the therapeutic choice is between pneumatic dilation and laparoscopic Heller myotomy as the primary therapy for achalasia. However, there are as yet no prospective controlled trials comparing these treatments. One controlled trial compares pneumatic dilation to myotomy via thoracotomy. That study reported 95% symptom resolution with myotomy and 51% symptom resolution in the dilation group, but the study was criticized for the methodology of pneumatic dilation used.327 Most case series report symptom resolution in approximately 70% of patients with pneumatic dilation, substantially higher than the 51% reported in the controlled trial, but still substantially lower than that reported in uncontrolled series of laparoscopic Heller myotomy (85% to 91%). Furthermore, although laparoscopic Heller myotomy is invasive, its morbidity and mortality are low. On the other hand, pneumatic dilation has a reported perforation rate that averages 3%, an incidence that probably exceeds that sustained by clinicians with substantial experience. Even though these patients do well if the perforation is recognized and addressed promptly, they may require a thoracotomy. Thus, it appears that cogent arguments can be made for each of these therapies and likely one should assess the local skills available, as well a patient preference, in selecting the most appropriate initial therapy. Treatment Failures Persistent dysphagia after treatment suggests treatment failure and should be evaluated with some combination of endoscopy, esophageal manometry, and fluoroscopic imaging. Endoscopy may detect esophagitis, stricture, para-

Chapter 42  Esophageal Neuromuscular Function and Motility Disorders esophageal hernia, or anatomic deformity. Manometry may be useful to quantify residual LES pressure, with values exceeding 10 mm Hg arguing for further therapy targeting the LES. Fluoroscopy is useful to identify anatomic problems as well as to evaluate esophageal emptying by using a timed barium swallow, a standardized method of measuring the height of the esophageal barium column one and five minutes after ingestion.328 In some instances these evaluations will lead to further intervention. In the case of a patient not previously operated on this could potentially be either repeat dilation or Heller myotomy. In patients who have already undergone myotomy, detection of an excessively short myotomy or functional esophageal obstruction from the antireflux component of the surgery usually requires reoperation, but pneumatic dilation can be pursued as an alternative. Reoperation, in general, is less effective than an initial operation for any indication in achalasia.329 Occasionally patients fail to respond to optimally performed dilation or myotomy and require alternative approaches. In extremely advanced or refractory cases of achalasia, esophageal resection with gastric pull-up or interposition of a segment of transverse colon or small bowel may be the only surgical option.330 Indications for this intervention include unresolvable obstructive symptoms, starvation, chronic aspiration, cancer, and perforation during dilation. Although excellent long-term functional results can be achieved, the reported mortality rate of this surgery is about 4%, consistent with the mortality rate of esophagectomy done for other indications. Risk of Squamous Cell Cancer Numerous series report cases of squamous cell carcinoma developing in the achalasic esophagus (see Chapter 46).331 The relative risk of developing squamous cell cancer has been estimated to be 33-fold relative to the non-achalasic population.332 The pathogenesis of the carcinoma is obscure, but stasis esophagitis is the likely precipitating factor. The tumors develop many years after the diagnosis of achalasia and usually arise in a greatly dilated esophagus, often in the middle third of the esophagus. Symptoms attributable to the cancer can be delayed, and the neoplasms are often large and advanced at the time of detection. These considerations raise the issue of surveillance endoscopy in achalasic individuals to detect early squamous cell cancer. However, an elegant analysis of a database encompassing the entire Swedish population of 1062 achalasic patients suggests that after discounting incident carcinomas, the overall odds ratio of squamous cell cancer for these people compared with age-matched controls was 17, corresponding to a 0.15% incidence of squamous cell cancer among the achalasic subjects.333 The authors calculated that if surveillance endoscopy was done annually, 406 examinations would need to be done in men and 2220 in women before one potentially treatable tumor was found. However, even that calculation is optimistic given that detection of a small cancer in a massively dilated esophagus with retained food and stasis esophagitis is far from ensured. Given these considerations, a surveillance program is currently not the standard of practice.

Distal Esophageal Spasm

Despite the dogma of treatment with smooth muscle relaxants, minimal controlled data exist regarding pharmacologic therapy of DES. Long-term studies are not available, and the entire basis for this therapy is anecdotal. Furthermore, most instances of esophageal chest pain are attributable to reflux rather than DES, and reflux symptoms will likely be made

worse by treating with smooth muscle relaxants. Uncontrolled trials of small numbers of DES patients report clinical response to nitrates,334 calcium channel blockers,335 hydralazine,336 botulinum toxin,337 and anxiolytics. The only controlled trial showing efficacy was with the anxiolytic trazodone, suggesting that reassurance and control of anxiety are important therapeutic goals.338 Also consistent with that conclusion, success has been reported using behavioral modification and biofeedback.339 Although the rationale for dilation is unclear, use of bougie dilators has been suggested as a therapy for dys­ phagia or chest pain in patients with spastic disorders. However, in the only controlled trial of this therapy, dilation with an 8-mm “placebo” dilator was as effective as an 18-mm “therapeutic” dilator in producing transient symptom relief.340 Alternatively, pneumatic dilation has been used in DES patients with severe dysphagia. In one practitioner’s experience, 45% of DES patients noted relief from pneumatic dilation, compared with 80% of achalasic patients.261 In another series of nine patients with DES and LES dysfunction treated with pneumatic dilation, dysphagia but not chest pain was improved during 37 months of observation.341 However, it is not clear that the patients who benefited by pneumatic dilation in these series would not be more properly categorized as spastic achalasia, emphasizing the need for accurate manometric classification.224 If dysphagia becomes so severe in DES that weight loss is observed or if pain becomes unbearable, surgical therapy consisting of a Heller myotomy across the LES with proximal extension of the incision up the distal esophagus to include the involved area of spasm or even esoph­ agectomy should be considered.186,342 However, there are no controlled studies of these procedures in welldefined DES patients and the indication is, fortunately, extremely rare.

Esophageal Hypersensitivity

Therapies for esophageal motor disorders have traditionally centered on improving esophageal contractility and emptying. However, the efficacy of these therapies is very limited except in the instance of achalasia. More recently, there has been a paradigm shift with the realization that minor manometric findings formerly interpreted as indicative of symptomatic hypercontractile conditions were often an epiphenomenon indicative of hypersensitivity syndromes. Hence, there is substantial interest in developing treatments directed at reducing esophageal hypersensitivity, and a number of pharmacologic and behavioral therapies have been identified with the potential to modulate pain perception and improve esophageal symptoms associated with swallowing. Pharmacologic Treatments Numerous neuropeptides and pharmacologic agents can reduce chemical and mechanical visceral sensitivity suggesting a possible role in the treatment of esophageal pain syndromes. Although data on these agents specific to eso­ phageal motor disorders are sparse, there is substantial literature focused on the treatment of noncardiac chest pain with or without motor abnormalities, and it is reasonable to generalize these findings to the treatment of esophageal hypersensitivity (see Chapter 12). Antidepressants are the most common medications prescribed for visceral pain modulation or chest pain of esophageal origin. Among antidepressants, the tricyclic antidepressants (TCAs) are the best studied. The mechanism of action for this therapeutic benefit is unknown

703

704

Section V  Esophagus because these agents act centrally as well as peripherally and have multiple receptor targets (acetylcholine, histamine, α-adrenergic). In a randomized placebo-controlled study, imipramine at a dose of 50 mg nightly was shown to be effective in reducing chest pain in patients with normal coronary angiograms.343 Similar results have been reported with other TCAs, and treatment with these agents at doses lower than those used for mood altering effects is common. Typical starting doses for TCAs (amitriptyline, nortriptyline) are 10 to 25 mg at bedtime with escalation of 10- to 25-mg increments to a target of 50 to 75 mg.344 Low-dose trazodone also has been used to treat noncardiac chest pain associated with esophageal dysmotility.338 In a double-blind placebo-controlled study in patients with noncardiac chest pain, the group taking 100 to 150 mg of trazodone had significant symptomatic improvement and less residual distress related to their esophageal symptoms. Esophageal motor function was not altered. Recent data also support the effectiveness of selective serotonin reuptake inhibitors (SSRIs) in the treatment of esophageal hypersensitivity. Intravenous citalopram at a dose of 20 mg was studied in a randomized, double-blinded, crossover study and found to significantly reduce both chemical (acid perfusion) and mechanical (balloon distention) esophageal sensitivity.345 Although clinical trials are not yet available, mechanistic studies assessing other SSRIs have also yielded encouraging results. Along similar lines, there has been a substantial interest in developing serotonin (5-HT) medications.293,346 5-HT3 antagonists and 5-HT4 agonists have been the most extensively studied given their effects on gut motility and as treatments for nausea. Unfortunately, several of these medications have proven to have unacceptable risks related to cardiac dysrhythmias or gut ischemia that led to their withdrawal. Theophylline has shown promising effects in the treatment of noncardiac chest pain, presumably by adenosine receptor blockade. In a recent placebo-controlled doubleblind study, sensory and biomechanical properties of the esophagus were assessed using impedance planimetry in 16 patients with esophageal hypersensitivity.347 Chest pain thresholds increased after intravenous theophylline and the esophageal wall was shown to relax and become more distensible. In a parallel study using oral theophylline and placebo in 24 chest pain patients there was a significant reduction in chest pain episodes, chest pain duration, and chest pain severity in the theophylline group.347 Although limited, these are very promising results for patients

with symptoms thought to be attributable to mechanical hypersensitivity. Nonpharmacologic Treatments Although the link between esophageal hypersensitivity, psychological factors, and psychiatric abnormalities is unclear, therapy focused on reassurance, behavioral modification, and relaxation techniques may be helpful. These therapies will most likely benefit patients with comorbidities such as panic disorder, generalized anxiety, and depression. However, it is also possible that therapies using controlled breathing, relaxation techniques, or hypnotherapy may benefit patients with hypersensitivity by diverting mental attention and reducing hypervigilance for visceral stimuli. Well-performed prospective trials are necessary to define the clinical role of these therapies.

KEY REFERENCES

Behar J, Biancani P. Pathogenesis of simultaneous esophageal contractions in patients with motility disorders. Gastroenterology 1993; 105:111-18. (Ref 251.) Cook IJ, Kahrilas PJ. AGA technical review on management of oropharyngeal dysphagia. Gastroenterology 1999; 116:455-78. (Ref 299.) Kahrilas PJ, Clouse RE, Hogan WJ. American Gastroenterological Association technical review on the clinical use of esophageal manometry [comment]. Gastroenterology. 1994; 107:1865-84. (Ref 197.) Logemann J. Evaluation and treatment of swallowing disorders. Austin, Tex: Pro-Ed Inc; 1998. (Ref 17.) Massey BT, Dodds WJ, Hogan WJ, et al. Abnormal esophageal motility. An analysis of concurrent radiographic and manometric findings. Gastroenterology 1991; 101:344-54. (Ref 258.) Ott DJ, Richter JE, Chen YM, et al. Esophageal radiography and manometry: Correlation in 172 patients with dysphagia. AJR Am J Roentgenol 1987; 149:307-11. (Ref 282.) Pandolfino JE, Ghosh SK, Rice J, et al. Classifying esophageal motility by pressure topography characteristics: A study of 400 patients and 75 controls. Am J Gastroenterol 2008; 103:27-37. (Ref 285.) Pandolfino JE, Kwiatek MA, Nealis T, et al. Achalasia: A new clinically relevant classification by high-resolution manometry. Gastroenterology 2008; 135:1526-33. (Ref 224.) Sifrim D, Janssens J, Vantrappen G. A wave of inhibition precedes primary peristaltic contractions in the human esophagus. Gastroenterology 1992; 103:876-82. (Ref 46.) Vaezi MF, Richter JE, Wilcox CM, et al. Botulinum toxin versus pneumatic dilatation in the treatment of achalasia: A randomised trial. Gut 1999; 44:231-9. (Ref 312.) Vela MF, Richter JE, Wachsberger D, et al. Complexities of managing achalasia at a tertiary referral center: Use of pneumatic dilatation, Heller myotomy, and botulinum toxin injection. Am J Gastroenterol 2004; 99:1029-36. (Ref 329.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

43 Gastroesophageal Reflux Disease Joel E. Richter and Frank K. Friedenberg

CHAPTER OUTLINE Epidemiology  705 Health Care Impact  707 Pathogenesis  707 Antireflux Barriers  707 Mechanisms of Reflux  708 Hiatal Hernia  710 Esophageal Acid Clearance  710 Gastric Factors  713 Clinical Features  713 Classic Reflux Symptoms  713 Extraesophageal Manifestations  714 Differential Diagnosis  715 Associated Conditions  715 Diagnosis  715 Empirical Trial of Acid Suppression  715 Endoscopy  716 Esophageal Biopsy  716 Esophageal pH Monitoring  717

Gastroesophageal reflux disease (GERD) is a consequence of the failure of the normal antireflux barrier to protect against frequent and abnormal amounts of gastroesophageal reflux (GER; i.e., gastric contents moving retrograde effortlessly from the stomach to the esophagus). GER itself is not a disease but rather a normal physiologic process. It occurs multiple times each day, especially after large meals, without producing symptoms or mucosal damage. In contrast, GERD is a spectrum of disease usually producing symptoms of heartburn and acid regurgitation. Most patients have no visible mucosal damage at the time of endoscopy (nonerosive GERD), whereas others have esophagitis, peptic strictures, or Barrett’s esophagus. Symptoms may include chest pain or evidence of extraesophageal manifestations such as pulmonary, ear, nose, or throat symptoms. GERD is a multifactorial process and one of the most common diseases of mankind. It greatly affects health care, contributing to the expenditure in the United States of nearly 12 billion dollars per year for antacid medications.

EPIDEMIOLOGY Although GERD is widely reported to be one of the most prevalent diseases of the gastrointestinal tract, prevalence data are based primarily on estimates rather than actual data. Furthermore, estimates differ depending on whether the analysis is based on symptoms (usually heartburn) or signs of disease (i.e., esophagitis).

Barium Esophagogram  719 Esophageal Manometry  719 Clinical Course  719 Nonerosive Reflux Disease  719 Erosive Reflux Disease  720 Complications  720 Hemorrhage, Ulcers, and Perforation  720 Peptic Esophageal Strictures  720 Barrett’s Esophagus  720 Treatment of Uncomplicated Disease  720 Nonprescription Therapies  721 Prescription Medications  721 Maintenance Therapies  723 Surgical Therapy  724 Endoscopic Therapy  725 Treatment of Complications  725 Chest Pain and Extraesophageal Manifestations  725 Peptic Esophageal Strictures  726

On the basis of symptoms, GERD is common in Western countries. In a nationwide population-based study by the Gallup Organization in the United States, 44% of the respondents reported heartburn at least once a month.1 More convincing data were obtained from a mailing of 2200 validated self-report questionnaires to a predominantly white population living in Olmsted County, Minnesota.2 The prevalence of heartburn and acid regurgitation in the past year was 42% and 45%, respectively. Symptoms that occurred at least weekly were reported by 20% of respondents, with an equal gender distribution across all ages. Most subjects reported their heartburn as being moderately severe, with a duration of 5 years or more, and only 5.4% had seen a physician for their reflux symptoms within the past year. More varying prevalence rates for symptomatic GERD have been reported from Europe, ranging from 5% in Switzerland to 27% in Finland.3 In contrast, the true prevalence of esophagitis is very difficult to define because healthy subjects rarely undergo upper endoscopy. Studies suggest that 7% of Americans have erosive esophagitis, whereas European studies identify prevalence rates ranging from 2% to 10%.4 GERD affects nearly equal proportions of men and women, but a male predominance occurs in esophagitis and Barrett’s esophagus.4 Increasing age is an important factor in the prevalence of GERD complications, probably the result of cumulative acid injury over time to the esophagus (Fig. 43-1).5,6 The prevalence of GERD only recently has been studied in multiracial populations. In a cross-sectional survey among employees at a Houston Veterans Affairs hospital,

705

Section V  Esophagus 50

400

400

350

350

300

300

250

250

200

200

Symptom severity

25 Esophagitis

15 10 5 0

100

>70

Age (years)

Figure 43-1.  Relationship of gastroesophageal reflux disease (GERD) symptom severity or presence of esophagitis with age in a large population study. Although the severity of GERD symptoms decreased with older age, the prevalence of esophagitis increased with age. (From Johnson DA, Fennerty MB. Heartburn severity underestimates erosive esophagitis severity in elderly patients with gastroesophageal reflux disease. Gastroenterology 2004; 126:660-4.)

the prevalence of heartburn was similar (23% to 27%) across ethnic groups including African Americans, Hispanics, Asians, and whites. However, African Americans had significantly less esophagitis than whites (24% versus 50%) for the same severity of symptoms (weekly or more).7 A study from Boston reviewed endoscopic reports from nearly 2500 consecutive patients, finding complicated GERD in 12% of white patients, 3% of African American patients, and 2% of Asian patients.8 The prevalence of GERD is relatively low among residents of Africa and Asia. For example, a cross-sectional study in Singapore reported prevalence rates for reflux symptoms of 7.5% in Indians, 0.8% in Chinese, and 3% in Malays.9 There have been exceptions such as the remarkable increase in the frequency of reflux symptoms seen in Japan and Singapore.10 An endoscopic, population-based study from South Korea encompassing more than 25,000 individuals, found the prevalence of erosive esophagitis to be 8%, whereas nonerosive reflux disease occurred in 4% of examined individuals.11 More than 90 % of subjects with erosions had mild disease, consistent with previous endoscopic studies from Asia. A recent systematic review that summarized trend data from longitudinal population-based studies performed in Asia failed to demonstrate an increase in prevalence over the past decade.12 Possible reasons for the lower GERD prevalence include low dietary fat; low body mass index (BMI); and lower gastric acid output, possibly related to Helicobacter pylori infection.11,13 The prevalence of GERD has been increasing in Western countries over the past 30 years.14 El Serag and Sonnenberg observed opposing time trends in the prevalence of peptic ulcer disease and GERD in the United States. Rates of duodenal ulcer fell between 1970 and 1995, while the prevalence of GERD and esophageal adenocarcinoma rose significantly (Fig. 43-2).15 The authors speculated that the decreasing prevalence of H. pylori may be playing a con-

50

Non-white men

50

Non-white men

0

0 1975–79

61–70

100

1970–74

41–50 51–60

150

1990–95

31–40

150

1985–89

21–30

White men

1980–84

10 seconds) than swallow-induced LESRs, and are accompanied by inhibition of the crural diaphragm.50 tLESRs account for nearly all reflux episodes in healthy subjects and 50% to 80% of episodes in GERD patients, depending on the severity of associated esoph­ agitis (Fig. 43-5).51 However, one study suggests that low basal LES pressure, rather than tLESRs, may be the primary mechanism of GER in patients with nonreducible hiatal hernias.52 tLESRs are not always associated with GER. In normal subjects 40% to 60% of tLESRs are accompanied by reflux episodes, compared with 60% to 70% in GERD patients.45,51,53 Possible factors determining whether reflux occurs include abdominal straining, presence of a hiatal hernia, degree of esophageal shortening, and duration of tLESRs. The dominant stimulus for tLESR is distention of the proximal stomach by either food or gas,54,55 which is not surprising given that a tLESR is the mechanism of belching. More varying stimuli are fat, stress, and subthreshold (for swallowing) stimulation of the pharynx.49 Various drugs may impair tLESRs including cholecystokinin A (CCK-1) receptor antagonists, anticholinergic drugs, morphine, somato­ statin, nitric oxide inhibitors, 5-hydroxytryptamine (5-HT)3 antagonists, and γ-aminobutyric acid (GABAB) agonists.56 Evidence indicates that tLESRs are mediated through vagal pathways.54 Gastric distention activates mechanoreceptors (intraganglionic lamellar endings) adjacent to the gastric cardia, sending signals to the brainstem center via vagal afferent pathways.57 The structured sequence of motor events including LESR, crural diaphragm inhibition, and secondary esophageal peristalsis suggests that this process occurs in a programmed manner, probably controlled by a pattern generator within the vagal nuclei. The motor arm is the vagus nerve sharing common elements with swallowinduced LESR.56

Chapter 43  Gastroesophageal Reflux Disease 7 Distal Esophageal 4 pH 1 + Submandibular 0 EMG –

Sw

mm Hg

Esophageal body

12 cm above LES 9 cm above LES 6 cm above LES

3 cm above LES

50

0 50

0 50

0 50

0 70

Sw tLESR

LES 0 40 Gastric 0

1 minute Figure 43-4.  Example of a transient lower esophageal sphincter (LES) relaxation (tLESR) on an esophageal manometry study. LES pressure is referenced to gastric pressure, which is indicated by the horizontal dashed line. Note that the tLESR persisted for almost 30 seconds, whereas the swallow-induced LES relaxation (LESR) to the right (Sw) persisted for only 5 seconds. Also note the absence of a submandibular electromyographic (EMG) signal during the tLESR, which indicates absence of a pharyngeal swallow. Finally, the associated esophageal motor activity is different in the two types of LESR: the swallow-induced relaxation is associated with primary peristalsis, whereas the tLESR is associated with a vigorous, repetitive “off contraction” throughout the esophageal body. (From Kahrilas PJ, Gupta RR. Mechanisms of reflux of acid associated with cigarette smoking. Gut 1990; 31:4.)

Swallow-Induced Lower Esophageal Sphincter Relaxations

About 5% to 10% of reflux episodes occur during swallowinduced LESRs. Most episodes are associated with defective or incomplete peristalsis.53 During a normal swallowinduced LESR, reflux is uncommon because (1) the crural diaphragm does not relax, (2) the duration of LESR is relatively short (5 to 10 seconds), and (3) reflux is prevented by the oncoming peristaltic wave (see Fig. 43-4). Reflux during swallow-induced LESRs is more common with a hiatal hernia. This may be due to the lower compliance of the esophagogastric junction in hernia patients, permitting it to open at pressures equal to or lower than intragastric pressure, thereby allowing reflux of gastric juices accumulating in the hiatal hernia.58,59

Hypotensive Lower Esophageal Sphincter Pressure

GER can occur in the context of a hypotensive LES by either strain-induced or free reflux.44,51 Strain-induced reflux

occurs when a relatively hypotensive LES is overcome and “blown open” by an abrupt increase in intra-abdominal pressure from coughing, straining, or bending over. This type of reflux is unlikely when the LES pressure is greater than 10  mm  Hg. Free reflux is characterized by a fall in intraesophageal pH without an identifiable change in intragastric pressure, usually occurring when LES pressure is less than 5 mm Hg. Reflux due to a low or absent LES pressure is uncommon. Mostly it occurs in patients with severe esophagitis and may account for up to 25% of reflux episodes (see Fig. 43-5); it rarely occurs in patients without esophagitis.45,51,60 The mechanisms responsible for idiopathic low LES pressure (i.e., not part of a systemic disease such as scleroderma) are poorly understood. The presence of a hiatal hernia reduces the pressure measured in the LES due to losing the intrinsic support of the crural diaphragm.44 Some LES weakness may be secondary to esophagitis impairing the excitatory cholinergic pathways to the LES. Induction of experimental esophagitis in cats attenuates the

709

Section V  Esophagus

100 90 Percentage of reflux episodes

710

Weakened and shortened LES

80 70

Loss of diaphragmatic support for the LES

60 50 40

Retention of gastric fluid in hernial sac

30 Control GERD Mild esophagitis Severe esophagitis

20 10 0

Swallow- Absent basal Straining induced LES pressure LESR Figure 43-5.  Proportion of reflux episodes in control subjects and in patients with gastroesophageal reflux disease (GERD) occurring by the following mechanisms: transient lower esophageal sphincter relaxation (tLESR), swallow-induced lower esophageal sphincter relaxation (LESR), absent basal LES pressure, and straining in the presence of low LES pressure. (From Holloway RH. The anti-reflux barrier and mechanisms of gastro-oesophageal reflux Ballieres Clin Gastroenterol 2000; 14:681.) tLESR

release of acetylcholine and lowers LES pressures—changes that are reversible on healing of the esophagitis.60 However, healing of esophagitis in humans is rarely accompanied by an increase in LES pressure.61

HIATAL HERNIA

The contribution of the hiatal hernia to GERD is controversial. Opinion has shifted widely from one that virtually equated hiatal hernia with reflux disease to one that denied it a causal role. Epidemiologic and physiologic data confirm the importance of the hiatal hernia in patients with more severe esophagitis, peptic stricture, or Barrett’s esophagus.62 Hiatal hernia occurs in 54% to 94% of patients with reflux esophagitis, a rate strikingly higher than that in the healthy population.63 Two studies have also found that in individuals with reflux symptoms, the presence of hiatal hernia confers a significantly increased risk of erosive esophageal injury.64 The hiatal hernia impairs LES function through several mechanisms, as well as impairing esophageal acid clearance (Fig. 43-6).65 Reflux is worse in patients having a “nonreducible” as opposed to a “reducible” hiatal hernia. Nonreducing hernias are those in which the gastric rugal folds remain above the diaphragm between swallows.62 Statistical modeling has revealed a significant interaction between hiatal hernia and LES pressure, such that the likelihood of GER is increased as basal LES pressure decreases, an effect substantially amplified by the presence of a hernia and as the hernia size increases.43 Displacement of the LES from the crural diaphragm into the chest reduces basal LES pressure and shortens the length of the high-pressure zone primarily due to the loss of the intra-abdominal LES segment.62 Hiatal hernia eliminates the increase of LES pressure that occurs during straining and increases tLESRs during gastric distention with gas.65,66 Large, nonreducible hernias also impair esophageal acid clearance because of an increased tendency for reflux

Loss of intraabdominal LES segment

Stretching and rupture of the phrenoesophageal ligament Widened diaphragmatic hiatus

Figure 43-6.  Schematic diagram showing the effect of a hiatal hernia on the antireflux barrier. LES, lower esophageal sphincter.

to occur from the hernia sac during swallow-induced LESRs.53 Finally, an alteration of esophagogastric junction compliance, especially in GERD patients with hiatal hernia, has been identified.59 For the same degree of intragastric pressure, the esophageal junction opens at a lower pressure and the cross-sectional area is greater and more symmetrical as intragastric pressure increases. These changes in compliance simulated a 10-fold increase in air and 6-fold increase in liquid reflux across the esophageal junction. The etiology of a hiatal hernia remains unclear. Familial clustering of GERD suggests the possibilities of an inherited smooth muscle disorder. Animal studies propose that reflux itself causes esophageal shortening promoting the development of a hiatal hernia.67 Other studies find an association with obesity68 and heavy lifting,69 raising the possibilities that over time chronic intra-abdominal stressors may weaken the esophageal hiatus, causing the development of a hiatal hernia. This theory is attractive as it helps to reconcile the increased prevalence of hiatal hernias as the population grows older.63

ESOPHAGEAL ACID CLEARANCE

The second tier against reflux damage is esophageal acid clearance. This phenomenon involves two related but separate processes: volume clearance, which is the actual removal of the reflux material from the esophagus, and acid clearance, which is the restoration of normal esophageal pH following acid exposure through titration with base from saliva and esophageal gland secretions. Although the competency of the antireflux barrier determines the frequency and volume of GER, esophageal acid clearance determines the duration of acid exposure to the mucosa and probably the severity of mucosal damage.

Volume Clearance

Esophageal peristalsis clears acid volume in the upright and supine positions but is inoperative during deep rapid-

Chapter 43  Gastroesophageal Reflux Disease clearance time and is most beneficial in patients with aperistalsis (i.e., scleroderma).72

Bolus volume (mL)

15

Salivary and Esophageal Gland Secretions

10

5

Pressure (mm Hg)

Esophageal pH

0 8 4 0

140

0 Acid bolus 0

DS

DS 1

DS

DS 2

DS

DS 3

Time (minutes) Figure 43-7.  Relationship between esophageal peristalsis, distal esophageal pH, esophageal emptying, and esophageal acid clearance during an acid clearance test performed with radiolabeled 0.1  N hydrochloric acid. Bolus volume within the esophagus is derived from scintiscanning over the chest. Note that, although all but 1 mL of the infused fluid is cleared from the esophagus by the first peristaltic contraction, the distal esophageal pH remains low for at least 3 minutes. Stepwise increases in distal esophageal pH occur with subsequent swallows. DS, dry swallow. (From Helm JF, Dodds WJ, Pek LR, et al. Effect of esophageal emptying and saliva on clearance of acid from the esophagus. N Engl J Med 1984; 310:284.)

Saliva is the second essential factor required for normal esophageal acid clearance. Compared with gastric acid, saliva is a weak base with a pH of 6.4 to 7.8.73 Although saliva is ineffective in neutralizing large acid volumes (5 to 10  mL), it easily neutralizes the small amount of acid remaining in the esophagus after several peristaltic contractions (see Fig. 43-7).70 The importance of saliva is supported by observation that increased salivation induced by oral lozenges or bethanechol significantly decreases acid clearance time. In contrast, suction aspiration of saliva markedly prolongs acid clearance, despite the presence of normal peristaltic contractions.73 Modulation of salivation may contribute to GERD. Decreased salivation during sleep is the reason that nocturnal reflux episodes are associated with markedly prolonged acid clearance times.74 Xerostomia (see Chapter 22) is associated with prolonged esophageal acid exposure and esophagitis.75 Cigarette smoking promotes GER. Originally attributed to nicotine’s effect on lowering LES pressure, cigarette smokers also have prolonged esophageal acid clearance times due to hyposalivation.76 Finally, the esophagosalivary reflex is impaired in patients with reflux esophagitis and individuals with strictures.77 This is a vasovagal reflex demonstrated by perfusing acid into the esophagus, which stimulates salivation. This reflex explains the symptoms of water brash (copious salivation) observed in some reflux patients. In addition to saliva, the aqueous bicarbonate-rich secretions of the esophageal submucosal glands dilute and neutralize residual esophageal acid.78 Acid refluxing into the esophageal lumen stimulates these glands and helps neutralize the acid, even if swallowing does not occur.79

Tissue Resistance

eye-movement (REM) sleep. Helm and colleagues70 showed that one or two primary peristaltic contractions completely clear a 15-mL fluid bolus from the esophagus (Fig. 43-7). Primary peristalsis is elicited by swallowing. Secondary peristalsis, initiated by esophageal distention from acid reflux, is much less effective in clearing the refluxate, thus offering only an ancillary protective role. Peristaltic dysfunction (i.e., failed peristaltic contractions and hypotensive [5 mm confined to folds but not continuous between tops of mucosal folds Grade C Mucosal breaks continuous between tops of two or more mucosal folds but not circumferential Grade D Circumferential mucosal break

thoroughly evaluated esophagitis classification is the Los Angeles (LA) system, which is gaining acceptance in the United States and Europe (Table 43-3) (Fig. 43-10A to D).169 Esophageal capsule endoscopy for the evaluation of reflux symptoms has thus far been disappointing. The capsule is 11 by 26  mm and acquires video images at 14 frames per second. After swallowing images are transmitted to a portable receiver via digital radiofrequency. In one study, compared with standard upper endoscopy, the capsule has a sensitivity of 50% for erosive esophagitis, 54% for the presence of a hiatal hernia, and 79% for the presence of Barrett’s esophagus.170 As mentioned, most patients with GERD are treated initially with PPIs and without endoscopy. The important exception is the patient experiencing “alarm” symptoms: dysphagia, odynophagia, weight loss, and gastrointestinal bleeding. Here endoscopy should be performed early to diagnose complications of GERD (e.g., strictures) and to rule out other entities such as infections, ulcers, cancer, or varices. Current guidelines suggest the major role of endoscopy is to diagnose and treat GERD complications, especially peptic strictures, and to define Barrett’s esophagus.171 Using this rationale, the majority of patients with chronic GERD need only one endoscopy while on therapy.

ESOPHAGEAL BIOPSY

Like endoscopy, the role of esophageal biopsies in evaluating GERD has evolved over the years. Microscopic changes of reflux may occur even when the mucosa endoscopically appears normal.172 These classic changes of basal cell hyperplasia and increased height of the rete peg, both representing increased epithelial turnover of the squamous mucosa, are sensitive but not specific histologic findings for GERD.173 Acute inflammation characterized by the presence of neutrophils and often eosinophils (Fig. 43-11) is very specific for esophagitis; however, the sensitivity is low, in the range of 15% to 40%.174 Thus, there is little value for histologic examination of normal-appearing squamous mucosa to either confirm or exclude pathologic acid reflux.175 In patients with classic esophagitis, biopsies are usually not taken unless necessary to exclude neoplasm, infection, pill injury, or bullous skin disease. Therefore, the current primary indication for esophageal biopsies is to determine the presence of Barrett’s epithelium.171 When this diagnosis is suspected, biopsies are mandatory and best done when esophagitis is healed (see Chapter 44).

Chapter 43  Gastroesophageal Reflux Disease

A

B

C

D

Figure 43-10.  Endoscopic photographs of the four grades of esophagitis (A to D) using the Los Angeles classification system as outlined in Table 42-3.

ESOPHAGEAL pH MONITORING

Figure 43-11.  Histopathology of gastroesophageal reflux disease. Inflammatory cells (eosinophils and neutrophils) are interspersed between squamous epithelial cells. (Courtesy Edward Lee, MD, Washington, D.C.)

Ambulatory intraesophageal pH monitoring is the standard test for establishing pathologic reflux.176,177 For catheterbased pH testing, the probe is passed nasally, positioned 5 cm above the manometrically determined LES, and connected to a battery-powered data logger capable of collecting pH values every four to six seconds. An event marker is activated by the patient when symptoms, meals, and body position changes occur. Patients are encouraged to eat normally and engage in regular daily activities, with monitoring carried out for 18 to 24 hours. Reflux episodes are defined by a pH drop of less than 4. Conventionally measured parameters include percent of total time when pH is less than 4; percent of time, upright and supine when pH is less than 4; total number of reflux episodes; duration of longest reflux episode; and number of episodes greater than five minutes.176 The percent of total time pH is less than 4 is the most reproducible measurement for GERD, with reported upper limits of normal ranging from 4% to 5.5%.176 Ambulatory pH testing discerns positional variations in

717

Section V  Esophagus Physiologic reflux pattern M

M S

M

S M

8

pH

6 4 2

04:00 PM

10:00 AM

10:00 PM

04:00 AM

10:00 AM

A Upright reflux pattern M

S

M

SM

8

pH

6 4 2

11:00 AM

05:00 PM

11:00 PM

05:00 AM

11:00 AM

B Combined reflux pattern M

S

M S

8 6 pH

718

4 2

11:00 AM

05:00 PM

11:00 PM

05:00 AM

11:00 AM

C Figure 43-12.  Common patterns of 24-hour esophageal pH monitoring. A, The physiologic pattern of gastroesophageal reflux (GER) seen in healthy subjects. Reflux is noted after meals (M) but not during sleep (S). A reflux episode is defined as a pH drop to less than 4. B, Upright reflux pattern with extensive GER during the day but not at night. These patients have frequent symptoms, but esophagitis is uncommon. C, Combined pattern with GER during the day and at night. Most of these patients have esophagitis.

GER, meals, and sleep-related episodes and helps relate symptoms to reflux events (Fig. 43-12). A critical limitation of esophageal pH monitoring is that there exists no absolute threshold value that reliably identifies GERD patients. Studies comparing patients with endoscopic esophagitis who underwent pH tests report sensitivities from 77% to 100% with specificities from 85% to 100%.178 However, these patients with esophagitis rarely need pH testing; rather, patients with normal endoscopy and suspected GERD might benefit most from this test. Unfortunately, data on these patients are less conclusive, with considerable overlap between controls and nonerosive refluxers.177 Other drawbacks of pH testing include possible equipment failure, pH probe missing reflux events because the probe is buried in a mucosal fold, and false-negative studies due to dietary or activity limitations from poor tolerability of the nasal probe.178 A major advance in esophageal pH testing has been the development of a catheter-free system.179 This system uses

a wireless pH capsule that is affixed to the esophageal mucosa with a delivery system that drives a small needle into the epithelium. The capsule then transmits pH data to a portable receiver using radiofrequency signals. Catheterfree testing is rapidly becoming the preferred method of pH testing because monitoring can be extended beyond 24 hours and limitations on normal daily activities and meals are negligible.179 Ambulatory esophageal pH monitoring is the only test that records and correlates symptoms with reflux episodes over extended periods of time. However, because only 10% to 20% of reflux episodes are associated with symptoms, different statistical analyses have evolved, attempting to define a significant association between symptoms and reflux episodes, including the symptom index, symptom sensitivity index, and symptom association probability.146 Unfortunately, no studies have defined the accuracy of these symptom scores in predicting response to therapy. Therefore, pH testing can define an association between

Chapter 43  Gastroesophageal Reflux Disease complaints and GER, but only treatment trials address the critical clinical issue of causality. Clinical indications for ambulatory pH monitoring are established.171 Before fundoplication, pH testing should be done in patients with normal endoscopy to ensure the presence of pathologic acid reflux. After antireflux surgery, persistent or recurrent symptoms warrant repeat pH testing. In these situations, pH monitoring is performed with the patient off antireflux medications. Esophageal pH testing is particularly helpful in evaluating patients with reflux symptoms who are resistant to treatment and who have normal or equivocal endoscopic findings. For this indication, pH testing is usually done on PPI therapy to define two populations: those with and those without continued abnormal acid exposure times. The group with persistent GER needs intensified medical therapy, whereas patients with symptoms and good acid control have another etiology for their complaints. Finally, ambulatory pH testing may help in defining patients with extraesophageal manifestations of GERD. In this situation, pH testing is often done with additional pH probes in the proximal esophagus or pharynx.147 Initially most of these studies were done off antireflux medications to confirm the coexistence of GERD; however, this does not guarantee symptom causality. Therefore, one approach is to first treat aggressively with PPIs, reserving pH testing for those patients not responding after 4 to 12 weeks of therapy.164 A relatively new method of evaluating GERD has been combined impedance and acid testing, which allows the measurement of acid and nonacid (volume) reflux. Nonacid reflux is measured by the detection of a retrograde bolus of ion-rich fluid in the esophagus. Refluxates that are a mixture of liquid and air are also readily detected. In a large group of normal subjects, roughly 40% of reflux episodes were either weakly acidic (pH 4 to 6.5) or alkaline (nadir esophageal pH during episode >6.5).180 In a multicenter study using impedance, 37% of patients experienced continued reflux symptoms despite twice-daily PPI therapy due to nonacid reflux.181 These patients would have been interpreted as negative for reflux had they been studied using conventional pH only. Another study using 24-hour ambulatory pH impedance found a temporal relationship between symptoms and nonacid reflux in 4.1% and 16.7% of subjects off and on PPI therapy, respectively.182 Regurgitation and cough were the most prevalent symptoms associated with nonacid reflux.

BARIUM ESOPHAGOGRAM

The barium esophagogram is an inexpensive, readily available, and noninvasive esophageal test. It is most useful in demonstrating anatomic narrowing of the esophagus and assessing the presence and reducibility of a hiatal hernia. Schatzki’s rings, webs, or minimally narrowed peptic strictures may only be seen with an esophagogram, being missed by endoscopy, which may not adequately distend the esophagus. Giving a 13-mm radiopaque pill or marshmallow along with the barium liquid can help to identify these subtle narrowings.183 The barium esophagogram allows good assessment of peristalsis and is helpful preoperatively in identifying a weak esophageal pump.184 The ability of barium esophagogram to detect esophagitis varies, with sensitivities of 79% to 100% for moderate to severe esophagitis, whereas mild esophagitis is usually missed.185,186 Barium testing also falls short when addressing the presence of Barrett’s esophagus. The spontaneous reflux of barium into the proximal esophagus is very specific for reflux, but it is not sensitive. Provocative maneuvers (e.g., leg lifting, coughing, Valsalva, or water siphon) can elicit stress reflux

and improve the sensitivity of the barium esophagogram, but some argue that these maneuvers also decrease its specificity.185,186

ESOPHAGEAL MANOMETRY

Esophageal manometry allows assessment of LES pressure and relaxation, as well as peristaltic activity, including contraction amplitude, duration, and velocity. However, esophageal manometry is generally not indicated in the evaluation of the uncomplicated GERD patient because most have a normal resting LES pressure.51 (It is an integral component of pH testing to accurately define LES location; see earlier.) Esophageal manometry to document adequate esophageal peristalsis is traditionally recommended before antireflux surgery.187 If the study identifies ineffective peristalsis (low amplitude or frequent failed peristalsis),188 then a complete fundoplication may be contraindicated. However, this assumption has recently been challenged by several studies finding that reflux control was better and dysphagia no more common in patients with weak peristalsis after a complete, as opposed to a partial, fundoplication.189 An improvement of traditional manometry, combining it with impedance testing, is helping to clarify this controversy. Using this technique, a study found that less than 50% of patients with ineffective peristalsis had a significant delay in esophageal bolus transit measured by impedance.190 Therefore, potentially only these patients with a significant physiologic defect in motility will require a modified fundoplication.

CLINICAL COURSE The clinical course of GERD depends to a great extent on whether the patient has erosive or nonerosive disease. There is controversy as to whether GERD exists as a spectrum of disease severity or as a categorical disease in three distinct groups, including Barrett’s esophagus. Patients tend not to cross over from one group to another; in follow-ups ranging from six months to longer than 22 years, less than 25% of patients with nonerosive disease evolved over time to having esophagitis, nearly all to LA grade A/B disease, or to having complications of GERD.191-194

NONEROSIVE REFLUX DISEASE

Early studies from tertiary referral centers suggested that the majority of GERD patients had esophagitis.195 However, studies carried out in community practices reveal that up to 70% of GERD patients had a normal endoscopic examination.196-198 Endoscopy-negative GERD patients are more likely to be female, younger, thin, and without hiatal hernia, and they have a higher prevalence of functional GI disorders.199 Despite their mild mucosal damage, these patients demonstrate a chronic pattern of symptoms with periods of exacerbation and remission.200 Nonerosive GERD is suspected in the patient with typical reflux symptoms and a normal endoscopy and confirmed by the patient’s response to antisecretory therapy. Esophageal pH testing identifies three distinct subsets of nonerosive GERD patients. First are the patients with excessive acid reflux who usually respond to PPI therapy. Second are the patients with normal reflux parameters but a good correlation between symptoms and acid reflux episodes. This group represents 30% to 50% of nonerosive GERD patients and has “functional heartburn.”200 These patients probably have heightened esophageal sen­ sitivity to acid and are less likely to respond to antireflux therapy.201 The third group is characterized by normal acid exposure times and poor symptom correlation. Whether

719

720

Section V  Esophagus they truly represent a subset of nonerosive GERD is questionable.

EROSIVE REFLUX DISEASE

Patients with erosive esophagitis tend to be male, older, and overweight and are more likely to have hiatal hernias.199 The clinical course of these patients with erosive esophagitis is more predictable and associated with complications of GERD. Longitudinal studies have shown that up to 85% of patients with erosive GERD, on no maintenance reflux therapy, will relapse within six months of stopping PPI therapy, and the relapse rate is highest in patients with more severe grades of esophagitis (see Table 43-3).202,203 Several studies confirm that erosive esophagitis patients are prone to reflux complications, including ulcers, strictures, and Barrett’s esophagus. In a Finnish study, 20 patients with erosive GERD treated with lifestyle changes, antacids, and prokinetic drugs were followed for a median of 19 years. Fourteen patients continued to have erosions, and six new cases of Barrett’s esophagus were detected.192 In another more recent European study193 over two years, patients with LA grade C/D esophagitis developed Barrett’s esophagus at a rate of 5.8% compared with only 1.4% for LA grade A/B and 0.5% for nonerosive GERD. However, these data must be contrasted with a two-year U.S. trial in which no patient with erosive esophagitis developed Barrett’s esophagus204 and in another study in which stricture was reported in only 0.9% of 957 patients over 7.6 years of symptom driven antireflux treatment.194

COMPLICATIONS HEMORRHAGE, ULCERS, AND PERFORATION

GERD-related non-cancer deaths are rare (0.46 per 100,000 persons). The most common fatal causes are hemorrhagic esophagitis, aspiration pneumonia, ulcer perforation, and rupture with severe esophagitis.205 Major hemorrhage and esophageal perforation are usually associated with deep esophageal ulcers or severe esophagitis.206 Esophageal per-

Figure 43-13.  Classic peptic stricture demonstrated by barium esophagogram (A) and endoscopy (B). The film shows a large hiatal hernia (HH) common to all GERD strictures. Dark arrow points to short thick fibrous stricture with mul­tiple pseudodiverticula (white arrows). Although not seen on barium examination, the endoscopic view also demonstrates circumferential esophagitis (Los Angeles grade D). GERD, gastroesophageal reflux disease.

forations are very rare in the PPI era but can result in mediastinitis and death. Clinically important hemorrhage has been reported in 7% to 18% of GERD patients207 and may result in iron deficiency anemia.

PEPTIC ESOPHAGEAL STRICTURES

Strictures occur in 7% to 23% of patients with untreated reflux esophagitis, and are especially seen in older men.208 They may be linked to chronic nonsteroidal anti-inflammatory drug (NSAID) use.209 Stricture formation is complex, starting as reversible inflammation with edema, cellular infiltration, and vascular congestion, progressing to collagen deposition and ending in irreversible fibrosis. As dysphagia progresses, heartburn often decreases, reflecting the stricture acting as a barrier to further reflux. Dysphagia is usually limited to solids. Unlike malignant strictures, patients with peptic strictures have a good appetite, alter their diet, and lose little weight. Peptic strictures are smooth-walled, tapered, circumferential narrowings in the lower esophagus, usually less than 1 cm long but occasionally extend to 8 cm (Fig. 43-13). In these unusual cases, the clinician should suspect a predisposing condition, such as Zollinger-Ellison syndrome, or another condition such as pill esophagitis or a stricture from prolonged nasogastric intubation.208 A mid- to upper esophageal stricture should raise concern for Barrett’s esophagus or malignancy. Although once controversial, today a Schatzki’s ring is considered a forme fruste of an early peptic stricture.210 All stricture patients should undergo endoscopy, at least initially, to confirm the benign nature of the lesion and, if necessary, take biopsies to exclude cancer and Barrett’s esophagus.

BARRETT’S ESOPHAGUS (see Chapter 44) TREATMENT OF UNCOMPLICATED DISEASE The rationale for GERD therapy depends on a careful definition of specific aims. In patients without esophagitis, the therapeutic goals are to relieve reflux symptoms and prevent

HH

B

A

Chapter 43  Gastroesophageal Reflux Disease frequent symptomatic relapses. In patients with esophagitis, the goals are to relieve symptoms and heal esophagitis while preventing further relapses and complications.

NONPRESCRIPTION THERAPIES

Although GERD is common, many sufferers do not seek medical care, instead choosing to change their lifestyles and self-medicate with OTC antacids, histamine-2 receptor antagonists (H2RAs), and even PPIs. These observations have led to the “iceberg” model of the GERD population. The vast majority of heartburn suffers are invisible because they self-medicate and do not seek professional help; only those at the tip of the iceberg, typically patients with severe symptoms or reflux complications, are seen by physicians.211

Lifestyle Modifications

Selective lifestyle changes, carefully explained to the patient, should be part of the initial management plan and are especially helpful in those with mild, intermittent complaints. These include elevating the head of the bed, avoiding tight-fitting clothes, losing weight if overweight, restricting alcohol and smoking, making dietary changes, refraining from lying down after meals, and avoiding bedtime snacks. Physiologic studies show that these maneuvers enhance esophageal acid clearance, decrease acid reflux–related events, or ease heartburn symptoms.212 Headof-the-bed elevation can be done by using 6- to 8-inch blocks or a foam wedge under the mattress to elevate the upper torso. Eating several hours before retiring and avoiding bedtime snacks keeps the stomach empty at night, thereby decreasing nocturnal reflux episodes. Avoiding tight-fitting clothes and losing weight are interventions aimed at reducing the incidence of reflux by the “abdominal stress” mechanism. Targeted weight loss may be helpful, whereas discrete periods of weight gain can be associated with exacerbation of reflux symptoms.30 Cessation of smoking and alcohol reduction is valuable because both agents lower LES pressure, reduce acid clearance, and impair intrinsic squamous epithelial protective functions.76,211 Reducing meal size and avoiding fats, carminatives, and chocolate reduces reflux frequency by decreasing episodes of tLESRs, as well lowering LES pressure.211 Additionally, some patients complain of heartburn after citrus drinks, spicy foods, tomato-based products, coffee, tea, or cola drinks. Stimulation of gastric acid secretion or esophageal sensitivity to low pH (or perhaps hyperosmolar solutions) may account for these symptoms.213 However, indiscriminate food prohibition should be avoided but rather tailored to individual sensitivity to better promote compliance. Finally, patients should avoid, if possible, drugs that lower LES pressure (see Table 43-1) or promote localized esophagitis, such as certain bisphosphonates (see Chapter 45). How good are the clinical studies assessing the efficacy of these commonly prescribed lifestyle changes? In an evidence-based review,211 studies of smoking, alcohol, chocolate, fatty foods, and citrus products had sound phy­ siologic data that their intake can adversely effect symptoms or promote reflux on esophageal pH tests. However, there was little convincing evidence that cessation of these pro­ ducts predictably improved reflux symptoms. Only elevation of the head of the bed, left lateral decubitus positioning, and weight loss were associated with GERD improvement in case-controlled studies.211

Over-the-Counter Medications

These drugs are used in treating mild, infrequent heartburn symptoms triggered by lifestyle indiscretions. Antacids

increase LES pressure but work primarily by buffering gastric acid, albeit for short periods. Heartburn symptoms are rapidly relieved, but patients need to take antacids frequently, usually 1 to 3 hours after meals. Gaviscon, containing alginic acid and antacids, mixes with saliva to form a highly viscous solution that floats on the gastric pool, acting as a mechanical barrier. Both antacids214 and Gaviscon215 are more effective than placebo in relieving symptoms induced by a heartburn-promoting meal. However, they do not heal esophagitis, and long-term trials suggest symptom relief in only 20% of patients.216,217 OTC H2RAs are available at doses usually one half the standard prescription dose. Although onset of relief is not as rapid as antacids, the OTC H2RAs relieve symptoms for 6 to 10 hours.218 Therefore, they are particularly useful when taken before potentially refluxogenic activities. Like antacids, OTC H2RAs are ineffective in healing esophagitis.218 OTC combinations of antacids and H2RAs are available. The long-term safety and efficacy of PPIs led the U.S. Food and Drug Administration (FDA) to approve omeprazole at full dose (20 mg) for OTC use in 2003. Drug labeling suggested daily use for only two weeks and recommended physician follow-up for persistent symptoms. Despite initial “real world” concerns of abusing this drug, early actual-use data support that consumers accurately self-select if OTC omeprazole is appropriate for use, comply with a two-week regimen, and seek physician care for longer-term management of frequent heartburn.219

PRESCRIPTION MEDICATIONS

Patients with frequent heartburn, esophagitis, or complications usually see a physician and receive prescription medications. Prokinetic drugs attempt to correct the GERD-related motility disorders associated with GERD. However, the most clinically effective drugs for short- and long-term reflux treatment are acid suppressive drugs.

Prokinetic Drugs

Until recently three prokinetic drugs were available for treating GERD: bethanechol, a cholinergic agonist; metoclopramide, a dopamine antagonist; and cisapride, a serotonin (5-HT4) receptor agonist that increases acetylcholine release in the myenteric plexus. These drugs improve reflux symptoms by increasing LES pressure, acid clearance, or gastric emptying. However, none alters tLESRs, and their effectiveness decreases with disease severity.220 Current prokinetics provide modest benefit in controlling heartburn but have unreliable efficacy in healing esophagitis unless combined with acid-inhibiting drugs.220 Prokinetic drugs are limited by their side effect profiles. Bethanechol commonly causes flushing, blurred vision, headaches, abdominal cramps, and urinary frequency. Metoclopramide, which crosses the blood-brain barrier, has a 20% to 50% incidence of fatigue, lethargy, anxiety, and restlessness and rarely causes tremor, parkinsonism, dystonia, or tardive dyskinesia, especially in older patients. Side effects may be decreased by reducing the dosing regimen to twice a day, taking a larger single dose before dinner or at bedtime, or using a sustained release tablet. Domperidone, another dopamine antagonist not crossing the blood-brain barrier, has fewer side effects but is not available in the United States. Cisapride was the best prokinetic drug for treating GERD but was withdrawn from the U.S. market because of reports of serious cardiac dysrhythmias (ventricular tachycardia, ventricular fibrillation, torsades de pointes, and QT prolongation) with associated cardiac arrest and deaths related to possible drug interactions.221

721

Section V  Esophagus Regulating the frequency of tLESRs is an attractive target for GERD treatment because of its pivotal role in most reflux episodes. Potential agents to modify this vagovagal reflex include atropine, morphine, CCKA (CCK-1) receptor antagonists, NO synthase inhibitors, and γ-aminobutyric acid (GABAB) agonists.222 Only the latter category of drug, baclofen, has been extensively studied in humans and found to be a potent inhibitory neurotransmitter in the central nervous system, antagonizing the release of neurotransmitters from vagal nerve afferents. The frequency of tLESRs, especially after meals, is decreased, reducing exposure time for acid and duodenal reflux. This correlates with improvement of acute and chronic symptoms in GERD patients.223,224 Baclofen needs to be titrated upward slowly (5  mg three or four times daily initially and increased as needed over 10 days to 40 to 60  mg per day). Side effects including drowsiness, nausea, and lowering of the threshold for seizures require discontinuation in up to 20% of patients. Potential applications could be in patients with nonerosive GERD or as adjunct therapy in patients with persistent symptoms on PPIs possibly related to nonacid duodenal reflux.224

that the overall esophagitis healing rates with H2RAs rarely exceeded 60% after up to 12 weeks of treatment, even when higher doses were used.226,227 Healing rates differ in individual trials depending primarily on the severity of esophagitis being treated: grades I and II esophagitis heal in 60% to 90% of patients, whereas grades III and IV heal in only 30% to 50% despite high-dose regimens.227 Although PPIs are more effective than H2RAs (Fig. 43-15; discussed following), nocturnal gastric acid breakthrough while on PPI therapy may cause reflux symptoms in some patients. H2RAs given at bedtime successfully eliminated this problem in one study, suggesting a new indication for H2RAs in the PPI era.228 However, this study used only a single evening dose and did not account for the tolerance that frequently develops to H2RAs over weeks to months.229 This tolerance impairs the effectiveness of chronic nocturnal dosing of H2RAs to eliminate nocturnal acid breakthrough,230 but suggests a useful role in as-needed medications in situations in which lifestyle indiscretions may promote nocturnal complaints. The H2RAs are very safe with a side effect rate (most of which are minor and reversible) of about 4%.225 Serum concentrations of phenytoin, procainamide, theophylline, and warfarin are higher after the administration of cimetidine and, to a lesser degree, ranitidine, whereas these interactions are not reported with the other two H2RAs.

Histamine-2 Receptor Antagonists (H2RAs)

(see also Chapter 53) These drugs (cimetidine, ranitidine, famotidine, and nizatidine) are more effective in controlling nocturnal, as compared with meal-related, acid secretion because the parietal cell is stimulated postprandially by gastrin acting via histamine and by acetylcholine (see Chapter 49).225 The four H2RAs are equally effective when used in proper doses, usually twice a day before meals. GERD trials find that heartburn can be significantly decreased by H2RAs, when compared with placebo, although symptoms are rarely abolished (Fig. 43-14). A comprehensive meta-analysis found

Proton Pump Inhibitors (see also Chapter 53) PPIs inhibit meal-stimulated and nocturnal acid secretion to a significantly greater degree than H2RAs231 but rarely make patients achlorhydric. After oral ingestion, acid inhibition is delayed because PPIs need to accumulate in the parietal cell secretory canaliculus to bind irreversibly to actively secreting proton pumps.232 Therefore, the slower a PPI is cleared from plasma, the more it is available for delivery to the proton pumps. PPIs should be taken before

100

100

(3) 80

(17) (13)

60

(12)

PPI (10) (10)

40 (3)

H2RA

(27) (4)

(22)

PPI

60

(23)

40 (2)

(25) (25) (9)

H2RA (5)

20

20

0

(5)

(8) Placebo

0 0

A

(2) (26)

80 Percentage of patients healed

Percentage of patients free of heartburn

722

1–2

3–4 Weeks

0

6–8

B

2

4

6

8

12

Weeks

Figure 43-14.  A, Symptom relief time curve over 8 weeks for a proton pump inhibitor (PPI) or histamine-2 receptor antagonist (H2RA) corrected for patients free of heartburn at baseline. By week 2, more patients treated with a PPI were asymptomatic compared with those treated with a H2RA, even after a much longer duration of treatment with the H2RA. B, Esophagitis healing time curve for PPI, H2RA, and placebo over 12 weeks. By 4 weeks treatment with a PPI healed esophagitis in more patients than the other two drug classes over 12 weeks, implying a substantial therapeutic gain. The numbers of studies included for each time point and treatment are shown in parentheses. (Data based on meta-analysis from Chiba N, Gara CJ, Wilkinson JM, Hunt RH. Speed of healing and symptom relief in grade II to IV gastroesophageal reflux disease: A meta-analysis. Gastroenterology 1997; 112:1798.)

Chapter 43  Gastroesophageal Reflux Disease

Wrap Esophagus

A

Wrap

140˚± 20 Esophagus

Esophagus

B

180˚

C

Wrap

Figure 43-15.  The most common surgical fundoplications used during antireflux surgery. A, The most popular worldwide is the 360-degree Nissen fundoplication. B, An anterior wrap (e.g., Thal, Dor) is commonly used to prevent gastroesophageal reflux after a Heller myotomy for achalasia. The experience with this repair is limited in patients with classical gastroesophageal reflux disease. C, The posterior wrap (Toupet) is popular in patients with poor esophageal motility because postoperative dysphagia is less frequent than after other operations. This is a 220- to 250-degree wrap. (From Oelschlager BK, Eubanks TR, Pellegrini CA. Hiatal hernia and gastroesophageal reflux disease. In: Townsend CM, Beauchamp RD, Foshee JC, et al, editors. Sabiston Textbook of Surgery: The Biological Basis of Modern Surgical Practice. 18th ed. Philadelphia: Saunders; 2007.)

the first meal of the day, when most proton pumps become active. Because not all pumps are active at any given time, a single PPI dose will not inhibit all pumps. A second dose, if necessary, can be taken before the evening meal. PPIs (omeprazole, lansoprazole, rabeprazole, pantoprazole, and esomeprazole) have superior efficacy compared with H2RAs on the basis of their ability to maintain an intragastric pH greater than 4 from 10 to 14 hours daily compared with approximately 6 to 8 hours daily with the H2RAs.233,234 PPIs are superior to H2RAs in completely relieving heartburn symptoms in patients with severe GERD, usually within one to two weeks (see Fig. 43-15A).227 PPI therapy has been shown in a Cochrane meta-analysis to be superior to placebo and H2RAs in nonerosive GERD and for undiagnosed reflux symptoms in primary care, although the effect is 20% to 30% lower than in patients with esophagitis.235 Controlled studies and a large metaanalysis report complete healing of even severe ulcerative esophagitis after eight weeks in more than 80% of patients taking PPIs compared with 51% on H2RAs and 28% receiving placebo (see Fig. 43-15B).227,236-239 In another recent Cochrane review involving 4064 patients in 26 trials,240 PPIs were superior to H2RAs in healing esophagitis at four to eight weeks (risk ratio, 0.47) with a number-to-treat of 3. In patients not healing initially, prolonged therapy with the same dose or an increased PPI dose usually resulted in 100% healing.241 Until recently, therapeutic efficacy among PPIs was similar. However, large studies have found the newest PPI esomeprazole 40  mg superior to omeprazole 20  mg and to lansoprazole 30  mg in healing esophagitis.242,243 A meta-analysis of 10 randomized clinical trials244 comparing esomeprazole to all other PPIs found the therapeutic advantage is minimal with LA grade A/B esophagitis (number-to-treat 50 and 33, respectively), and greater with severe LA grade C/D esophagitis (number-to-treat 14 and 8, res­pectively). This superiority is related to higher systemic bioavailability and less interpatient variability with esomeprazole. Several PPIs are available in the United States for intravenous use.245

PPIs are well tolerated, with headaches and diarrhea described as the most common side effects. Increased fasting serum gastrin levels are reported with all the PPIs, but the elevations generally do not exceed the normal range for gastrin and return to normal values within one to four weeks of drug discontinuation. Omeprazole decreases the clearance of diazepam and warfarin due to competition for the cytochrome P-450 isoenzyme P2C19.246 The four newer PPIs have minimal or no important drug-drug interactions.

MAINTENANCE THERAPIES

GERD may be a chronic relapsing disease, especially in patients with low LES pressure, severe grades of esophagitis, and difficult-to-manage symptoms.217 After esophagitis is healed, recurrence within six months of stopping medication occurs in more than 80% of patients with severe esophagitis and in 15% to 30% of those with milder esophagitis.202,247 Cochrane reviews have identified the superiority of PPIs over H2RAs in maintaining the remission of esophagitis over 6 to 12 months.248 Among 10 randomized trials, the relapse rate for esophagitis was 22% on PPIs versus 58% with H2RAs, with a number-to-treat of 2.5. The FDA has approved all the PPIs, sometimes at one half the acute dose, for maintenance therapy, but only ranitidine 150  mg twice a day among the H2RAs has maintenance indications for mild esophagitis. Many clinicians now place their patients with severe disease (daily symptoms, severe esophagitis, or complications) on chronic PPI therapy indefinitely. The efficacy of this approach is supported by open, compassionate use data primarily from the Netherlands and Australia.249 In a study of 230 patients with severe esophagitis healed with 40  mg omeprazole, all subjects remained in remission for up to 11 years. More than 60% were maintained on omeprazole 20  mg a day, whereas higher doses of 60  mg or more were necessary in only 12% of patients, confirming a lack of tolerance to PPIs. Relapses were rare (one per 9.4 years of follow-up), strictures did not occur, and Barrett’s esophagus did not progress.245

723

724

Section V  Esophagus Although PPIs offer the best symptom relief and esophagitis healing, many patients do well long term on lesser treatments after having their complaints initially alleviated with PPIs. Using this “step-down approach,” a recent Veterans Affairs study reported that 58% of 71 patients on chronic PPIs could be switched to H2RAs and/or prokinetics or taken off medication completely.250 Younger age and severe heartburn symptoms predicted PPI requirement. Overall, this approach saved money for the health care system. A similar study by the same investigators found that 80% of patients using multiple-dose PPIs could be stepped down to single-dose PPI, remaining symptom free for six months with considerable cost savings.251 Hence the adage “once on a PPI, always on a PPI” is not true. The initial safety concerns with the PPIs was the potential to develop stomach or other cancers with long-term use. This was initiated by reports of omeprazole producing hypergastrinemia and gastric carcinoid tumors in rats, changes also subsequently demonstrated with chronic ranitidine therapy and subtotal resection of the gastric fundus.252 However, rats have a high density of enterochromaffin-like (ECL) cells and an exaggerated gastrin response to achlorhydria; chronic omeprazole therapy in other species with lower densities of ECL cells (mice, dog, man) has not caused carcinoid tumors. Another review found no evidence of an increased risk of colon cancer with chronic PPI use.253 Fundic gland polyps are the most common gastric polyp found at endoscopy. Their association with chronic PPI use has been a topic of debate since these drugs were first described. A recent study evaluated 599 patients of whom 322 used PPIs and 107 had fundic gland polyps.254 Longterm PPI use was associated with up to a four-fold increase in the risk of fundic gland polyps. Low-grade dysplasia was found in one fundic gland polyp. Etiologically, these polyps seem to arise because of parietal cell hyperplasia and parietal cell protrusions resulting from acid suppression. Recent studies confirm that chronic acid suppression may be associated with an increased risk of community-acquired pneumonias and enteric infections. In a large Scandinavian population-based study,255 the adjusted relative risk for pneumonia among current PPI users, compared with those who stopped using PPIs, was 1.89. Current users of H2RAs had a 1.63-fold increased risk of pneumonia compared with those who stopped. A significant positive dose-response relationship was observed in the PPI users. Likewise, a recent systematic review found an increased risk of enteric infections with acid suppression.256 The correlation was stronger with Salmonella, Campylobacter, and other enteric infections, compared with Clostridium difficile, and greater with PPI compared with H2RA therapy. Chronic high-dose use of PPIs may also affect the absorption of calcium and vitamin B12. A nested case-controlled study from the United Kingdom among 13,556 patients found that the risk of hip fractures increased with chronic PPI use over one year (adjusted odds ratio, 1.44), especially in those patients receiving high-dose PPIs (adjusted odds ratio, 2.65). A smaller but still significant risk was observed in chronic H2RA users.257 A large Canadian study258 reached similar conclusions, but found the risk for hip fractures became apparent after five years of treatment (adjusted odds ratio, 1.62) and after seven years for all osteoporotic fractures (adjusted odds ratio, 1.92). The mechanism for this association is unknown and cause-and-effect is not proven as yet. It has been suggested that if PPIs cause osteoporosis, they may interfere with insoluble calcium absorption or possibly inhibit the osteoclastic proton transport system, potentially reducing bone resorption. PPIs could retard the

absorption of vitamin B12 by decreasing gastric acidity; reducing the release of cobalamin from dietary protein; or by promoting small bowel bacterial overgrowth, thereby increasing luminal cobalamin consumption. However, cohort and case-control studies have not shown a convincing link between PPI use and vitamin B12 deficiency.259 Finally, a study suggested that patients on long-term omeprazole who are infected with H. pylori develop atrophic gastritis, a precursor to gastric adenocarcinoma, at a more rapid rate than noninfected patients.260 Nevertheless, a subsequent FDA panel determined that the available data were insufficient for recommending screening and treatment of H. pylori infection in patients on long-term PPI therapy.261

SURGICAL THERAPY

Only surgical fundoplication can correct the physiologic factors contributing to GERD and potentially eliminate the need for long-term medications. Antireflux surgery reduces GER by increasing basal LES pressure, decreasing episodes of tLESRs, and inhibiting complete LESR.262 This is done by reducing the hiatal hernia into the abdomen, reconstructing the diaphragmatic hiatus, and reinforcing the LES.263 Before laparoscopic surgery, the three most common operations were the Nissen fundoplication, Belsey Mark IV repair, and Hill posterior gastropexy. Since the introduction of minimally invasive surgery, the two most popular procedures, performed laparoscopically through the abdomen are the Nissen 360-degree fundoplication and the Toupet partial fundoplication (see Fig. 43-15).264 The former is a superior operation with better long-term durability, but it causes more postoperative dysphagia and gas bloat symptoms.265,266 The typical hospital stay is 1 to 2 days, and many patients return to normal activity in 7 to 10 days. Patients with more severe disease and a short esophagus suggested by a large nonreducible hernia, tight stricture, or long-segment Barrett’s esophagus will require a Collis lengthening procedure to create a 3- to 5-cm neoesophagus that allows the fundoplication to be placed in the abdomen under minimal tension.267 The popularity of antireflux surgery has undergone an explosion since the advent of the laparoscopic operation. The numbers of antireflux operations performed in the United States nearly tripled from 11,000 per year in 1985 (open surgery) to a peak of nearly 32,000 in 1999 but have leveled off at around 24,000 cases per year (11 cases per 100,000 adult population).268,269 A systematic review identified six randomized controlled trials involving 449 patients that compared open and laparoscopic fundo­ plication.270 There was no significant difference in recurrence rates between the procedures, and laparoscopic fundoplication was associated with lower operative morbidity (number-to-treat to prevent complication, 8) and shorter hospital stay. In the PPI era, symptom resolution on treatment helps predicate the success of antireflux surgery for classic as well as atypical symptoms.271 Antireflux surgery is a reasonable option in (1) the healthy patient with typical or atypical GERD symptoms well controlled on PPIs desiring alternative therapy because of drug expense, poor medication compliance, or fear of unknown long-term side effects; (2) patients with volume regurgitation and aspiration symptoms not controlled on PPIs; and (3) recurrent peptic strictures in younger patients.265 Patients recalcitrant to PPI therapy may well have another etiology for their complaints (e.g., pill esophagitis, gastroparesis, functional heartburn) and should be approached cautiously with surgery.

Chapter 43  Gastroesophageal Reflux Disease Testing must be done before antireflux surgery. Endoscopy is necessary to exclude stricture, Barrett’s esophagus, and dysplasia or carcinoma. A barium esophagogram can help define a nonreducible hiatal hernia, a shortened esophagus, and poor esophageal motility. Esophageal manometry, possibly combined with impedance if available, will identify ineffective esophageal peristalsis and previously misdiagnosed achalasia or scleroderma. Twenty-four-hour pH testing is necessary in patients with nonerosive GERD or those with esophagitis not responding to PPI therapy. Gastric analysis and gastric emptying studies may be indicated in select patients. Careful testing will result in modification of the original operation or an alternative diagnosis in approximately 25% of patients.187 Antireflux surgery relieves reflux symptoms and reduces the need for stricture dilation in more than 90% of patients,272 but Barrett’s esophagus rarely regresses and the risk of developing esophageal cancer is unchanged.273 Older studies found antireflux surgery superior to antacids, H2RA, and prokinetic therapy,204 but not PPI therapy, especially when dose titration is permitted.274 Mortality is rare ( CHO Tryptophan Undigestible fibers Small Intestinal Factors Fatty acids in duodenum Fatty acids in ileum Colonic Factors Constipation, IBS Other Factors Hyperglycemia Hypoglycemia Illusory self-motion (vection)

Proximal

Effect on Rate of Gastric Emptying (Symptoms) Delay (nausea) Acceleration (early satiety) Delay (prolonged fullness) Delay (prolonged fullness) Delay (right upper quadrant pain) Delay (prolonged fullness) Proportional to meal volume Delay Delay Delay

Middle

Distal 0.0

2.5

A

5.0

7.5

10.0

7.5

10.0

Time (min)

Proximal

Middle

Delay Delay

Distal

Delay (“duodenal tasting,” “duodenal brake”) Delay (“ileal brake”)

0.0

2.5

5.0

Delay

B

Delay Acceleration Delay

Figure 48-16.  Electrical recordings from electrodes secured to the mucosa of the proximal, middle, and distal antrum in a healthy subject. A, 3 cycle per minute (cpm) electrical slow waves in the proximal, middle, and distal electrode leads. The slow waves are propagated in an aborad direction as indicated by the dotted lines. B, Disruption of propagation and the onset of a 5- to 6-cpm tachygastria in the distal lead during hyperglycemia (glucose clamping), with a blood glucose level of 240 mg/dL. (Modified from Coleski R, Hasler WL. Coupling and propagation of normal dysrhythmic gastric slow waves during acute hyperglycemia in healthy humans. Neurogastroenterol Motil 2009; 21:492-99.

CHO, carbohydrate; IBS, irritable bowel syndrome.

CCK relaxes fundic tone, decreases antral contraction, and increases pyloric tone, all of which result in delay in gastric emptying. In contrast, short- and medium-chain fatty acids (shorter than C12) do not have these neuromuscular effects on gastric emptying rates.54,64 CCK released from the duodenum also activates CCKA receptors on the vagal affer­ ent neurons with synapses in the nucleus tractus solitar­ ius.65 Neurons from the nucleus tractus solitarius ascend to the periventricular nucleus that participate in mechanisms of satiation, and descending vagal efferent neurons from the dorsal motor nucleus of the vagus inhibit gastric emptying and maintain fundic relaxation. The sensitivity of the duo­ denal mucosa to fat and other nutrients led to the concept of duodenal tasting and duodenal brake, sensorimotor events that modulate gastric emptying of nutrients.66-68 Monosaccharides in the duodenum stimulate the release of incretins such as glucagon-like polypeptide-1 (GLP-1), which promotes insulin secretion to match increasing post­ prandial blood glucose levels and decreases antral contrac­ tions.69-71 In order to harmonize the relationships between glucose absorption, glycemia, and insulin secretion, the gastric emptying of carbohydrates is highly regulated.72 Hasler and colleagues showed that hyperglycemia decreases antral contractions and increases gastric dysrhythmias, a “physiologic” gastric dysrhythmia that decreases the rate of gastric emptying (Fig. 48-16).6,73 Glucagon infusions also induce bradygastrias.74 Hyperglycemia increases fundic compliance and decreases sensations related to fundic distention.75 Blood glucose levels greater than 220 mg/dL result in decreased antral contractions, decreased gastric emptying, and induced gastric dysrhythmias,6,73 all of which are gastric neuromuscular activities that reduce gastric emp­ tying and reduce further exposure of the duodenum to nutri­ ents. Hypoglycemia on the other hand increases gastric contractility and emptying.76

Time (min)

The interaction between nutrients in the lumen and the regulation of the rate of gastric emptying continues in the later postprandial period as digestion and absorption of nutrients occur throughout the small intestine. For example, if diet-derived fatty acids or carbohydrates reach the lumen of the ileum, the so-called ileal break is activated and gastric emptying is delayed. Infusion of nutrients into the lumen of the ileum delays gastric emptying,77 an enterogastric reflex mediated in part by peptide YY, CCK, and GLP-1 (see Chapter 1).61,68,77 Regulation of stomach emptying also is achieved by vagus nerve and splanchnic nerve activity that modulates the neu­ romuscular activities of the stomach described earlier. Vagal afferent nerves “monitor” neuromuscular function in the stomach moment by moment, and interactions between afferent vagal nerve activity and the nucleus tractus soli­ tarius and synapses with the efferent vagal nerve output from the dorsal motor nucleus produce an ongoing inter­action of CNS excitatory and inhibitory effects on the stomach. Gastric emptying is delayed during stress. Corticotropin-releasing factor (CRF) plays a role in the mediation of stress and inhibits gastric emptying through central dopamine1 and 2 and vasopressin (AVP) pathways in the periventricular nucleus.78 Other factors that affect the rate of gastric emptying not already mentioned include rec­ tocolonic distention, nausea and vomiting of pregnancy, and vection-induced motion sickness.79 Stimulation of various areas in the CNS affects gastric neuromuscular func­ tion. Illusory self-motion (vection) induces antral hypomo­

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders tility, tachygastria, and decreased gastric emptying.79,80 A series of studies using the experience of illusory self-motion, a unique CNS sensory stimulation, showed that the onset of nausea was associated with tachygastria and increased levels of plasma vasopressin.80,81 Gender affects the gastric emptying rate of a standard meal. Gastric emptying is significantly slower in healthy women compared with men.43,82,83 Gender differences in gastric emptying rates may be related to fluctuations in sex hormones, but phases of the menstrual cycle (variations in estradiol and progesterone concentrations) have not shown consistent relationships with emptying measurements.84 The rate of gastric emptying increases as body mass index rises, a relationship that may be relevant to the onset and maintenance of obesity.

GASTRIC SENSORY ACTIVITIES Free nerve endings in the stomach act as polymodal sensory receptors that respond to light touch or pressure, acid, and other chemical stimuli. Afferent neurons within the stomach

CNS perceptions Discomfort, Nausea Pain

are termed intrinsic primary afferent neurons, or IPANs.85 Cell bodies of IPANs reside in the submucosal or the myen­ teric plexus areas of the stomach wall. IPANs may be acti­ vated by serotonin release from local enterochromaffin cells.32,86 The afferent information in the IPANs is used in local reflexes and provides input to vagal and splanchnic afferent neurons for vagovagal and spinal reflexes, respec­ tively, to subserve transmission of visceral sensory informa­ tion to CNS centers. Vagal afferent neurons whose cell bodies reside in the nodose ganglia connect with the nucleus of the tractus solitarius and second-order neurons connect with higher center of the hypothalamus, and some inputs reach the cortex, where they are consciously perceived as visceral sensations (stomach emptiness or fullness) or symptoms such as nausea or abdominal pain (Fig. 48-17). From the SNS, splanchnic or spinal primary afferent neurons in the gastric wall mediate pain sensations. Cell bodies of these neurons lie in the dorsal horn of the spinal cord with second-order neurons that ascend via the spino­ thalamic and spinoreticular tracks in the dorsal columns. Sensory neurons are thin, myelinated A-delta or unmyelin­ ated C fibers. Spinal afferents include a population of unmy­ elinated C fibers. Capsaicin-sensitive unmyelinated fibers

Visceral perceptions Discomfort, Nausea Pain

Dorsal column

Dorsal root ganglion A-delta fiber

Spinothalamic tract

C fiber

To CNS

IML

Motor n.

Somatic nerves

Vagus nerve

Splanchnic nerves SPINAL CORD

Somatic perceptions Discomfort Pain

SKIN

Celiac ganglia Vertebral ganglia T5-T9

Visceral and motor reflexes

C fiber A-delta fiber Afferent n.

A-delta fiber C fiber Pacemaker region

Efferent n.

Afferent n. Efferent n. Gastric dysrhythmias Bradygastria Tachygastria

STOMACH

60 s

Figure 48-17.  Afferent and efferent neural connections between the stomach and central nervous system (CNS). The vagus nerve contains afferent nerves with A-delta and C pain fibers with cell bodies in the nodose ganglia with connections to the nucleus tractus solitarius (not shown). Low threshold mechano- and chemoreceptors stimulate visceral sensations such as stomach emptiness or fullness and symptoms such as nausea and discomfort. These stimuli are mediated through vagal pathways and become conscious perceptions of visceral sensations if sensory inputs reach the cortex. The splanchnic nerves also contain afferent nerves with A-delta and C fibers that synapse in the celiac ganglia with some cells bodies in the vertebral ganglia (T5-T9). Interneurons in the white rami in the dorsal horn of the spinal cord cross to the dorsal columns and spinothalamic tracts and ascend to sensory areas of the medulla oblongata. These splanchnic afferent fibers are thought to mediate high-threshold stimuli for visceral pain. In contrast to visceral sensations, somatic nerves such as from the skin carry sensory information via A-delta and C fibers through the dorsal root ganglia and into the dorsal horn and then through dorsal columns and spinothalamic tracts to cortical areas of somatic representation. Changes in gastric electrical rhythm, excess amplitude contractions, or stretch on the gastric wall are peripheral mechanisms that elicit changes in afferent neural activity (via vagal and/or splanchnic nerves) that may reach consciousness to be perceived as visceral perceptions (symptoms) emanating from the stomach. IML, intermediolateral nucleus; n., nerve.

801

802

Section VI  Stomach and Duodenum contain neuropeptides such as CGRP, VIP, somatostatin, substance P, and neurokinin A. These fibers are considered to be the primary route of transmission for various pain stimuli from the gut to the CNS. These nerve fibers may respond to inflammatory mediators that also awaken “silent” nociceptive fibers.86 In addition to interacting with IPANs, vagal afferent axons have multiple connections with the enteric neurons and innervate the circular muscle fiber bundles via connections with ICCs.87 Vagal afferent neurons are also sensitive to chemostimuli via mucosal neurons and mechanosensitive neurons and ICCs in the muscle layers. CCK receptors on vagal afferent neurons are primarily activated by physio­ logic mechanical and chemical stimuli from the stomach during fasting and fed conditions. These vagal afferents mediate the sensory response to intraluminal acid and fat. Acid may have a direct action on the nerve endings themselves.88 Nausea is a common sensation that is often attributable to stomach dysfunction. During the illusion of self-motion, gastric dysrhythmias develop as healthy individuals report nausea.89 Plasma vasopressin levels increase in the subjects who develop nausea, but do not increase in those who experience no nausea.90 This brain-gut, gutbrain interaction during illusory self-motion illustrates the temporal relationships between the onset of gastric dysrhythmias in the periphery and acute, severe nausea experience of the subject. On the other hand, mechanical or physical distention of the antrum, but not the fundus, using a balloon induces nausea sensations and gastric dysrhythmias in healthy individuals.91 These studies show that gastric dysrhythmias originate in the antrum in humans and that stretch of the antral wall is another mechanism that elicits gastric dysrhythmias and nausea sensations from the stomach. Distention of the gastric antrum and corpus by the water-load test (rather than a balloon) also elicits the gastric dysrhythmias and nausea in susceptible individuals.61

THE STOMACH AND THE REGULATION OF FOOD INTAKE, HUNGER, AND SATIETY Hunger is a basic human drive, a stressful condition that is eliminated or reduced by the ingestion of food. Hunger is also described as an uncomfortable “emptiness” of the stomach. The ingestion of food elicits relaxation of the stomach musculature (receptive relaxation) and accommo­ dation of the physical volume of the meal; as these gastric neuromuscular events occur, hunger disappears and the comfortable, postprandial sensations of stomach fullness are experienced. The volume of food ingested suppresses hunger and stim­ ulates the sense of fullness more than the calorie content of the meal.92-94 Infusion of nutrients into the stomach induces a greater intensity of fullness or satiety compared with infu­ sion of the same nutrients into the duodenum.67 The sup­ pression of hunger is greater when nutrients are taken by mouth, indicating that CNS, oropharyngeal, and gastric neu­ romuscular factors are integrated to produce the comforts of normal postprandial stomach fullness.95 Healthy individuals usually eat until they are reasonably full. The physiologic attributes of fullness are not com­ pletely known, but the physical stretch on the stomach walls induced by the volume of food ingested and the gastric juice secreted are responsible, in part, for the sense

of postprandial fullness.93,96 Subjects experience a dramatic change from the sensation of stomach emptiness at baseline to the sensation of stomach fullness after ingesting water over a five-minute period. The average volume of water ingested to achieve fullness is 600 mL of water; in contrast, patients with functional dyspepsia ingest, on the average, 350 mL of water on average and feel full, indicating a dis­ turbance in stomach wall relaxation and/or wall tension.61 Similarly, fullness and satiety can be achieved by ingesting a nutrient drink until achieving maximum tolerated satiety.97 The presence of acid or nutrients in the duodenum or an elevated blood glucose level decreases the stomach wall tension.98,99 The ingestion of a solid meal initially elicits fundic relax­ ation, and little emptying of the food occurs during the lag phase. Sensations of fullness continue during the lag phase when the food is being triturated. Once the linear phase of gastric emptying begins, there is a progressive perception of decreasing stomach fullness and increasing stomach empti­ ness over time. Four or five hours after a solid meal, the stomach is indeed empty and the healthy individual feels hungry once again. The physiologic mechanisms of hunger and satiety (and stomach emptiness and fullness) are under intense investi­ gation. In the fasting state plasma motilin levels increase during the phase 3 of the MMC, but correlations between the sensation of hunger and increases of motilin or onset of phase 3 have not been described. As discussed in other chapters, ghrelin is a 28 amino acid peptide secreted from endocrine cells of the oxyntic glands in the gastric fundus.100 Ghrelin levels increase in the plasma during fasting (hunger) and stimulate food intake, probably acting via vagal afferent nerves.101 Orexins or appetite-stimulating peptides are syn­ thesized by neurons in the lateral hypothalamus, promote food intake, and stimulate gastric contractility (in the rat) by actions on the dorsal motor nucleus of the vagus with projections to the gastric fundus and corpus.102 After inges­ tion of food, ghrelin levels decrease103 and are profoundly suppressed after gastric bypass surgery.104 Ghrelin also has promotility effects on the stomach and is being evaluated for the treatment of gastroparesis.105,106 Other hormones are candidates for important roles in the sensation of fullness or satiety, and these hormones are released after the ingestion of meals. CCK is released from the duodenal mucosa exposed to fatty acids as described previously. CCK receptors participate in fullness and nausea sensations elicited by intraduodenal lipid and gastric dis­ tention.107,108 Leptin is synthesized in the stomach and released after food ingestion; circulating leptin reduces food intake via CNS regulation of the arcuate nucleus.71,109 GLP-1 enhances fullness after a standard meal, reduces antral motility, and increases gastric volume.71,110 Apolipoprotein A-IV is released from the small intestine during absorption of triglycerides and decreases food intake and gastric motil­ ity, in part, via CCK and vagal afferent pathways.71,111 Polypeptide-YY (PYY) is released from the ileocolonic area after meals and is an important mediator of the “ileal brake” effect112 and appetite suppression.113,114 The brain and these gut hormones are clearly linked in the regulation of food intake and the regulation of gastric neuromuscular activity that produces stomach empty­ ing.71,114 The cephalic phase of gastric physiology is well known but has not been re-explored for many years. The sight, smell, and taste of food stimulate central vagal effer­ ent activity that increases gastric acid secretion, gastric contractility, and increases 3 cpm gastric myoelectrical activity.115-117 Sham feeding, during which the subject chews

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders

Meal

Meal

Sham feeding

Time in 64 second intervals

1

A

3

5

7

9

11

13

Sham feeding

15

Frequency (cpm)

Time in 64 second intervals

1

B

3

5

7

9

11

13

15

Frequency (cpm)

Figure 48-18.  Gastric myoelectrical response to sham feeding with tasty food and “disgusting” food. A, Running spectral analysis (RSA) of the electrogastrogram (EGG) signal recorded while a healthy subject chewed a warm hot dog and spit it out into a paper bag (sham feeding). The increase in amplitude of the peaks in the normal 3 cycle per minute (cpm) range during sham feeding is the normal response. “Meal” indicates the actual ingestion of a hot dog. B, RSA of the EGG recorded while a healthy subject chewed a cold tofu dog and spit it out (sham feeding). The subject felt “disgusted” during the sham feeding effort. Note the lack in increase of 3 cpm peaks during sham feeding the tofu dog compared with (A). The subject then ingested a warm hot dog on a bun at “Meal”; note the subsequent increase in peaks at 3 cpm. (Modified from Stern RM, Crawford HE, Stewart WR, et al. Sham feeding. Cephalic-vagal influences on gastric myoelectric activity. Dig Dis Sci 1989; 34:521-7.)

and spits out the test meal rather than swallowing it, elicits the cephalic-vagal reflex. Sham feeding a warm hot dog on a bun elicits enhanced 3 cpm EGG activity, whereas sham feeding a cold tofu dog, a food that the subjects considered disgusting, resulted in blunted or no increase in the 3 cpm myoelectrical activity (Fig. 48-18).117 Thus, sensory and emotional attributes of food during the cephalic phase of ingestive behavior also affect the neuromuscular activity of the stomach.

DEVELOPMENTAL ASPECTS OF GASTRIC NEUROMUSCULAR FUNCTION Gastric peristalsis appears between 14 and 23 weeks of gestation. Grouped or clustered peristaltic waves are evident by 24 weeks.118 The neuroregulatory mechanisms respon­ sible for the coordination of antropyloroduodenal motility in gastric emptying are well developed by 30 weeks of gesta­ tion.119 EGG recordings show normal 3 cpm activity in preterm infants delivered at 35 weeks that are similar to EGG signals recorded in full-term infants.120,121 On the other hand EGG recordings from premature infants (less than 35 weeks’ gestation) showed considerable tachygastria.120 Gastric myoelectrical activity matures further over the first 6 to 24 months of life and achieves full adult values by the end of the first decade.121,122 The development of ICCs has been studied intensely because of the interest in gastric electrical rhythmicity, smooth muscle contractions, and gastric dysrhythmias. Labels for the tyrosine kinase receptor (c-Kit) and the avail­ ability of knock-out mice lacking c-Kit have led to increased understanding of the development of ICCs.123 The ICCs dem­ onstrate differential development, with c-Kit expression on ICCs in the MY-ICCs developing before birth, whereas ICCs in the deep muscular plexus (IM-ICCs) develop after birth.124 ENS and ICC networks are not fully developed and are poorly coupled at birth, but progressive maturity of gastric

rhythmicity and contractility occur during perinatal devel­ opment.120,121 The ENS and ICCs in the deep muscular plexus are closely related, whereas ICCs in the myenteric plexus can develop normally in the absence of the enteric nervous system.124 Loss of ICCs in the pylorus is associated with loss of the inhibitory neural activity that may contrib­ ute to the development of pyloric stenosis in infants (see Chapter 47).125

ASSESSMENT OF GASTRIC NEUROMUSCULAR FUNCTION GASTRIC EMPTYING RATES

Clinical tests are available to assess the neuromuscular functions of the stomach, including emptying, contractions, and electrical rhythm, and they are discussed following.

Scintigraphy

Test meals labeled with radioisotope are available to assess the rate of gastric emptying. The seminal solid-phase gastric emptying protocol was a multinational study that used a 255-kcal technetium-99m (99mTc)–labeled egg substitute with bread and jam as the standard meal.43 Scans were obtained for 1 minute immediately after ingestion of the meal and at 30 minutes, 60 minutes, 120 minutes, 180 minutes, and 240 minutes in 123 healthy individuals. Delayed gastric emptying was defined as greater than 60% retention of the meal at 120 minutes and greater than 10% retention of the meal in the stomach 240 minutes after ingestion (see Figs. 48-10 and 48-11). The four-hour empty­ ing test was superior to the two-hour test because almost 20% of patients with suspected gastroparesis had normal emptying at two hours, but abnormal emptying at four hours.126 Pitfalls in the scintigraphic method for solid-phase gastric emptying studies include improper binding of the isotope with the test meal, which results in rapid or normal empty­

803

804

Section VI  Stomach and Duodenum ing and continuance of medications that may stimulate (e.g., metoclopramide) or inhibit (e.g., narcotics, anticholin­ ergic agents) gastric smooth muscle contractions. These medications should be stopped five to seven days before all gastric neuromuscular tests, if possible. Radiation exposure for the subject occurs with the scintigraphic tests, and mul­ tiple tests in the same subject are not advisable. Liquidphase gastric emptying tests can be performed with indium 111-diethylenetriaminepentaacetic acid (111In-DTPA) 99mTclabeled water or other liquids. Patients with unexplained nausea symptoms may have altered emptying of liquid meals, even if solid phase emptying is normal.127,128

Capsule Technology

Gastric emptying time of test meals is obtained from a small capsule that measures intraluminal pH and contractions (see Fig. 48-12). The capsule is swallowed with a standard test meal. During the postprandial period, measurements of luminal pH and contractions are transmitted to a receiver worn by the subject. In healthy subjects the capsule is emptied from the stomach into the duodenum approxi­ mately five hours after ingestion of the egg substitute meal. Emptying of the capsule correlated with 90% emptying of the technetium-labeled egg substitute solid meal. The test had very good sensitivity and specificity in detecting gastroparesis.83

Breath Tests

Breath tests indirectly reflect gastric emptying of solid and liquid test meals. The solid meals are labeled with C13 and include C13 octanoic acid, C13 acetate, or C13 Spirulina pla­ tensis. The C13 octanoic acid breath test has been performed in many experimental protocols and is used widely in Europe for research and clinical studies.129,130 The C13labeled food is emptied from the stomach and absorbed in the small intestine. The labeled nutrients are metabolized in the liver to C13O2, excreted in the lungs, and detected in breath samples. Breath samples are collected at 45, 90, 120, 150, and 180 minutes after the meal in the C13 Spirulina test. C13 is a stable isotope with no radiation risks. The C13 breath tests are generally comparable to scintigraphy.131 Pitfalls include spurious results in patients with malabsorp­ tive conditions, liver diseases or lung diseases that may preclude normal oxidation and excretion of the C13-labeled foods.

Ultrasonography

Transabdominal ultrasonographic techniques are used to measure antral diameter and antropyloroduodenal func­ tion.132,133 Three-dimensional ultrasound methods show the intragastric distribution of the test meal and regional varia­ tions in gastric volume (see Fig. 48-13).58 The technique is ideal with a liquid meal, but solids can also be identified and gastric emptying rates can be determined. The clinical application is limited by the high level of expertise required by the ultrasound operators.

Computed Tomography and Magnetic Resonance Imaging

These two techniques have been used to measure gastric emptying and demonstrate intragastric distribution of test meals. Computed tomography (CT) and magnetic resonance imaging (MRI) technologies offer unique anatomic and func­ tional views of the stomach in the fasting and postprandial periods.134 Sequential antral contractions can be visualized. Because of expense and availability, these techniques are not used in clinical practice.

GASTRIC CONTRACTIONS Antroduodenal Manometry

Antroduodenal manometry is an invasive technique wherein a water-perfused multilumen catheter is placed either through the nose or the mouth and advanced to a position where the proximal catheter ports are in the distal antrum and the distal ports are in the duodenum.40 Placement of the catheter requires endoscopic or fluoroscopic aid. The recordings typically last for several hours in order to record phases 1, 2, and 3 of the MMC and several more hours to record postprandial contractions after the subject ingests a test meal (see Fig. 48-7A and B). Antroduodenal manometry testing is not only invasive but also time intensive and requires extensive assistant or physician time for perfor­ mance of the test and interpretation of the data. Intraluminal manometry catheters detect only lumen-occluding contrac­ tions.136 Intraluminal pressure transducer devices fail to record almost 50% of contractions in the corpus and antrum because the majority of postprandial peristaltic waves are not lumen-occluding contractions. Manometry catheters positioned in the duodenum can detect patterns of neuro­ pathic or myopathic dysfunction.

Capsule Technology

The ingestible capsule described previously measures contractions of the stomach wall. After a standard test meal, irregular contractions occur at one to three per minute and are consistent with manometric recordings from the antrum.137 Several minutes of sustained high-amplitude, antral contractions occur prior to the emptying of the capsule into the duodenum (see Fig. 48-12) and some pat­ terns are consistent with phase 3–like contractions recorded by antroduodenal manometry. In patients with gastropare­ sis, capsule studies showed a decreased motility index during the 20 to 30 minutes before emptying of the capsule, but a normal motility index in the 10 minutes before the capsule was emptied into the duodenum. These studies suggest that normal terminal antral contractions may be maintained even in patients with gastroparesis, whereas antral contractility required for trituration of digestible foodstuffs is abnormal.138 Determination of the unique gastric contraction patterns after ingestion of a variety of common foods is possible using the capsule motility device.

GASTRIC MYOELECTRICAL ACTIVITY

Electrogastrography (EGG) refers to the recording methods used to noninvasively measure gastric myoelectrical activity using electrodes positioned on the abdominal surface.139,140 The EGG signal summates the ongoing gastric myoelectrical activity. The EGG reflects the slow wave activity during fasting and the summation of slow wave activity linked to plateau and action potential activity during the postprandial period (see Figs. 48-1 and 48-2).140 In response to a water load or a nutrient load, the amplitude of the EGG signal increases in the normal 2.5 to 3.7 cpm range, as determined by visual and computer analysis (e.g., an increase in the percentage of EGG power or in the post­ prandial power ratio in the normal frequency range) (see Fig. 48-15).60,140,141 Pitfalls for recording and analyzing EGGs include failure to identify artifact in the signal and harmonics in the computer analyses.142 Changes in the EGG frequency and amplitude are key measures. After ingestion of most solid or liquid meals, a so-called frequency dip occurs in the first 10 to 15 minutes after the meal. The frequency dip reflects changes resulting from marked gastric relaxation and accommodation of the test meal related to the volume or the temperature of the

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders TACHYGASTRIA 500 µV

A

60 s BRADYGASTRIA (LOW-AMPLITUDE) 500 µV

60 s

B

BRADYGASTRIA (HIGH-AMPLITUDE)

500 µV

sures after meals are not useful. The barostat balloon was designed to measure changes in tone (or gastric relaxation) and volume in the more spherical areas of the proximal stomach.153 Barostat methods involve a large, collapsed thin-walled balloon that is mounted on a catheter and passed through the mouth into the stomach. Intraballoon pressure is maintained with infused air during the baseline or fasting period with the balloon slightly distended. A test meal is then ingested. As the fundus and proximal stomach relax in response to the meal, more air is concomitantly infused into the balloon to maintain the established baseline intraballoon pressure (see Fig. 48-9).154 The volume of air that is infused to maintain baseline pressure is measured and is an estimate of the increased gastric volume that occurs as the proximal stomach relaxes. Barostat studies demonstrate abnormalities in fundic relaxation in almost 30% of patients with functional dys­ pepsia.155 The failure of fundic relaxation correlates with early satiety. Failure of fundic relaxation has also been recorded in patients with gastroparesis of diverse causes.156,157 Because the barostat method is invasive and uncomfortable for patients, these studies have been limited to the research laboratory.

Scintigraphy and Other Tests

60 s

C

NORMAL 1000 µV

D

60 s

Figure 48-19.  Gastric dysrhythmias recorded with electrogastrogram (EGG) methods are shown. A, Tachygastria, an abnormally rapid signal at 6 cycles per minute (cpm), shown by dots. B and C, Bradygastria at low- or high-amplitude 1 cpm wave, respectively. The one-per-minute waves are indicated by the solid curved lines and the smaller waves in the EGG signal represent respiratory activity. D, Normal 3 cpm EGG signal is identified by dots.

meal.143,144 Several minutes after ingestion of the meal, the frequency of the EGG signal returns to the middle of the normal 2.5 to 3.7 cpm range (see Fig. 48-15). Gastric dysrhythmias are associated with symptoms of nausea in subjects with motion sickness,89 nausea and vom­ iting of pregnancy,145,146 functional dyspepsia,60,147 and gas­ troparesis.148-150 Gastric dysrhythmias include 0.5 to 2.5 cpm signals termed bradygastrias and 3.7 to 10 cpm signals termed tachygastrias (Fig. 48-19). Recordings that have com­ binations of tachygastria and bradygastrias are termed mixed or nonspecific gastric dysrhythmias.151 Gastric myo­ electrical activity can also be recorded from serosal elec­ trodes placed during surgery or with mucosal electrodes placed during endoscopy.6,74,152

GASTRIC RELAXATION, ACCOMMODATION, AND VOLUME Barostat Tests

Due to the spherical shape of the fundus and the proximal stomach, manometric catheters to record intraluminal pres­

Excessive or poor fundic relaxation in response to liquids and solids can be demonstrated with scintigraphy, ultra­ sound, and MRI. Single photon emission computed tomog­ raphy (SPECT) is a method that outlines the gastric wall before and after ingestion of a meal to determine changes in volume of the stomach. This method requires intravenous injection of 99mTc-pertechnetate to outline the gastric wall. The accommodation response can be identified with SPECT.158,159

Non-Nutrient Liquid and Nutrient Drink Tests

Non-nutrient liquids (the water-load test) and nutrient drink tests are used to assess overall gastric volume or gastric capacity and visceral sensations such as nausea, stomach fullness, or satiety in response to ingestion of these liquids,60,97,160 often in conjunction with measures of gastric myoelectrical activity, accommodation, or emptying. In the water-load test, water is consumed over a 5-minute period until the subject feels full. In the typical caloric drink test, subjects drink 150 mL of the liquid (e.g., Ensure) every 5 minutes until maximum tolerated satiety is achieved, an endpoint that requires almost 30 minutes and the con­ sumption of 800 to 1000 mL of the nutrient drink.161 In many healthy subjects, nausea and gastric dysrhythmia are evoked by the satiety drink test. Subjects with functional dyspepsia or gastroparesis ingest much smaller volumes of water or nutrient drink and report fullness, indicating impaired relaxation and accommodation of the stomach.60,97

HISTOPATHOLOGIC STUDIES IN GASTRIC NEUROMUSCULAR DISORDERS

Efforts to define the histopathologic basis of gastric neuro­ muscular disorders have provided basic knowledge for the evolving field of neurogastroenterology. Most of the fullthickness specimens from the gastric wall have been har­ vested during the placement of gastric electrical stimulation devices or the placement of jejunostomy feeding tubes in patients with severe gastroparesis. Mucosal biopsies do not typically contain smooth muscle, ICCs, or ENS neurons, but an endoscopic biopsy technique has been described that provides a full-thickness specimen that contains elements of circular muscle, ENS neurons, and ICCs.162 Histologic

805

806

Section VI  Stomach and Duodenum abnormalities in smooth muscle, neurons of the ENS, number or location of ICCs, and distribution of key neural or muscular receptors are possible underlying mechanisms of disordered gastric smooth muscle relaxation, peristaltic contraction, and gastric slow wave activity. Loss of ICCs (MY-ICCs) was reported in patients with type 1 and type 2 diabetic gastroparesis.25,163,164 Interestingly, similar loss of ICCs was shown in diabetic mice, and the ICCs were restored with intense insulin therapy, suggesting that the ICCs are not entirely destroyed in diabetes but dedifferentiate into immature myoblasts during prolonged hyperglycemia.164 Damage to ICCs, ENS, and smooth muscle can also be seen in inflammatory and neoplastic condi­ tions.165-168 Other studies in humans have shown fibrosis of smooth muscle layers but intact myenteric plexus and vagus nerve in patients with diabetic gastroparesis, loss of ICCs but minimal smooth muscle fibrosis, and inflammatory T lymphocyte infiltration in myenteric neurons.169-171 Pitfalls with neurohistologic studies include tissue sampling of variably affected regions of the stomach, with patchy distri­ bution abnormalities of ICCs, ENS neurons, or smooth muscle. Nevertheless, results from histochemical studies are needed to provide new directions for understanding the neuromuscular dysfunction of the stomach and to stimulate ideas for novel therapeutic approaches.

NEUROMUSCULAR DISORDERS OF THE STOMACH Gastric neuromuscular disorders encompass a continuum of electrical and contractile dysfunction.172 At one end of the spectrum are gastric dysrhythmias, which are subtle electri­ cal disturbances associated with mild to severe nausea symptoms (Fig. 48-20). Abnormalities in relaxation of the

Vagal afferent nerve Impaired fundic relaxation Abnormal fundic emptying

Pacemaker region Weak 3 cpm rhythm

Pylorospasm

Gastric dysrhythmias Tachygastria Bradygastria 60 s

Dilated gastric antrum Antral hypomotility Gastroparesis Figure 48-20.  Spectrum of gastric neuromuscular disorders. Gastric neuromuscular disorders range from abnormal fundic relaxation and emptying to gastric dysrhythmias and antral hypomotility and gastroparesis. Pyloric sphincter dysfunction, as well as duodenal dysfunction, antroduodenal dyscoordination, and vagal neurohypersensitivity, may all be present in some patients with gastric neuromuscular disorders. cpm, cycles per minute. See text for details. (Modified from Koch KL, Stern RM. Functional disorders of the stomach. Semin Gastrointest Dis 1996; 7:185-95.)

fundus are associated with early satiety. At the severe end of the spectrum, antral hypomotility and profound gastro­ paresis are associated with prolonged postprandial fullness, vomiting, bloating, weight loss, and malnutrition that may require enteral or parenteral nutritional support. Patients with gastroparesis may also have gastric dysrhythmias, a dilated antrum, poor fundic relaxation, and gastric hyper- or hyposensitivity due to vagal or sphlanchnic nerve dysfunc­ tion.172,173 Clinical tests currently approved by the U.S. Food and Drug Administration (FDA) to assess gastric neuromus­ cular function are the scintigraphy tests to measure the rate of gastric emptying, the capsule motility device to measure gastric emptying, gastric pH, and contraction patterns of the stomach, and electrogastrography devices to measure gastric myoelectrical activity before and after provocative test meals. These tests provide objective assess­ ments of different aspects of the neuromuscular activity of the stomach in health and disease. Results of gastric empty­ ing and gastric myoelectrical activity tests provide objective diagnoses of gastric dysrhythmias and gastroparesis.

GASTROPARESIS

Gastroparesis means “paralysis” of the stomach as defined by the delayed rate of emptying of a standard test meal. Approximately 90% of the patients with gastroparesis have either diabetic, postsurgical, or idiopathic gastroparesis, but the less common forms of obstructive and ischemic gastro­ paresis are reversible. Major categories of gastroparesis are described below.

Diabetic Gastroparesis

Fifty percent of patients with long-standing type 1 diabetes mellitus develop gastroparesis. These patients often have diabetes for more than 10 years, erratic and elevated glucose levels, peripheral neuropathy, nephropathy, and cardiovascular disease.174-176 In a minority of patients the gastroparesis is the initial complication of their diabetes. Upper GI symptoms, however, may be minimal and non­ specific. This is not surprising because patients with diabe­ tes often have ischemic heart disease with minimal chest discomfort or cholecystitis with minimal or no right upper quadrant pain. One important manifestation of gastric emptying dysfunc­ tion in patients with diabetes is erratic glucose control, especially with unexpected hypoglycemic episodes in the postprandial period. Because the patient may have no per­ ception of delayed emptying of food, the usual insulin doses are administered before meals. Insulin levels increase in the blood, but if gastric emptying is delayed, then nutrient delivery into the duodenum is delayed. Thus, plasma glucose levels decrease in response to the insulin treatment and symptomatic hypoglycemia develops unex­ pectedly in the postprandial period because of delayed emptying. Acute hyperglycemia (>220 mg/dL) is associated with tachygastrias and delayed gastric emptying in healthy sub­ jects and in patients with diabetes.73,175,177-179 Hyperglycemia is also associated with loss of ICCs,25,163 antral hypomoti­l­ ity,180 isolated pyloric contractions,181 gastric dysrhyth­ mias,6,177,178 and impaired prokinetic action of prokinetic drugs like erythromycin.182 Relatively minor increases in the blood sugar level, even elevations within the physiologic range, delay gastric emptying in normal volunteers and diabetic patients.179 Acute hyperglycemia produced by glucose clamp methodology elicits fullness, decreased antral contractility, blunts the contractile pyloric response to intraduodenal lipid infusion, and modifies upper GI sensations.183,184

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders RUNNING SPECTRAL ANALYSIS Power

EGG rhythm strips

60 s 1 mV

60 s

3

6

9

Frequency (cpm)

Power

Time (1 min between lines)

Time (1 min between lines)

1 mV

0

A

RUNNING SPECTRAL ANALYSIS

12

15

EGG rhythm strips B

B1

60 s

500 µV

0

B

3

6

Waterload (250 mL)

A

A1

9

60 s

12

15

Frequency (cpm)

Figure 48-21.  Obstructive and idiopathic gastroparesis. A, Running spectral analysis (RSA) of the electrogastrogram (EGG) recording from a patient with gastroparesis due to mechanical obstruction at the pylorus secondary to chronic peptic ulcer disease. Note the persistent high-amplitude 3 cycles per minute (cpm) waves in the EGG rhythm strips and the uniform and unvarying peaks at 3 cpm in the RSA, findings that would not be expected in a patient with gastroparesis due to electrical and contractile dysfunction. B, RSA and EGG rhythm strips in a patient with idiopathic gastroparesis. Note that the EGG rhythm strips show a 7 to 8 cpm tachygastria before (B) and after (A) the water-load test. The RSA shows multiple peaks in the 7 to 8 tachygastria range and few peaks in the normal 3 cpm range. This patient had electrical and contractile abnormalities of the stomach as documented by the tachygastria and gastroparesis. (Modified from Brzana RJ, Koch KL, Bingaman S. Gastric myoelectrical activity in patients with gastric outlet obstruction and idiopathic gastroparesis. Am J Gastroenterol 1998; 93:1803-9.)

Gastric neuromuscular abnormalities documented in patients with chronic type 1 diabetes include abnormal intragastric distribution of food,185 reduced receptive relaxation and accommodation,186 reduced incidence of the antral component of the MMC, antral dilation, postprandial antral hypomotility,176 and electrical dysrhythmias.148,177,187 Loss of the antral phase 3 contractions results in poor emp­ tying of fibrous debris in the stomach and is the neuromus­ cular basis for the formation of bezoars (see Chapter 25). Thus, the progressive neuromuscular dysfunction of the diabetic stomach reflects the effects of chronic hyperglyce­ mia and superimposed, intermittent hyperglycemia epi­ sodes that occur repeatedly over many years. Gastric smooth muscle dysfunction is another mechanism of delayed gastric emptying in some patients with diabetes. In rats with diabetes gastric smooth muscle contractility is reduced in response to electrical stimulation.188 The inhi­ bitory effect of hyperglycemia on stem cell factor has a role in the reduction in ICCs and smooth muscle contractility.189 Pylorospasm as a cause of gastroparesis was documented in one study.190 Some patients with pylorospasm have a form of obstructive gastroparesis; that is, the neuromuscular function of the corpus and antrum is intact, but persistent functional obstruction at the pylorus (e.g., pylorospasm) causes delayed emptying. These patients have normal or increased 3 cpm myoelectrical activity, a discordant finding in patients with gastroparesis that also suggests the possibil­ ity of gastric outlet obstruction (Fig. 48-21A).149 In type 2 diabetes mellitus the incidence of gastroparesis ranges from 30% to almost 50%.191 Patients with type 2 diabetes differ significantly from patients with type 1 dia­ betes in that they have insulin resistance, the diabetes appears later in life, and the hyperglycemia has been present longer before diagnosis compared with patients with type 1 diabetes.193 Almost 20 million people in the United States have type 2 diabetes mellitus, and many have unsuspected gastroparesis.192 Gastroparesis in patients with type 2 diabe­ tes presents with subtle dyspepsia-like symptoms, but a minority of patients may present acutely with nausea, vom­

iting, and severe gastroparesis. Gastric dysrhythmias have been recorded in up to 75% of patients with type 2 diabetes with an average hemoglobin A1C of 8.2.193 In the early stages of type 2 diabetes mellitus gastric emptying may be accelerated.194 Gastroparesis has been also described in experimental diabetes, specifically in nonobese diabetic mice and in the type 2–like diabetes db/db mouse. Studies by Watkins and associates showed that neuronal nitric oxide synthase (nNOS) was reduced in the diabetic mouse stomach and was associated with antropylorospasm and gastroparesis.195 Treatment with insulin or sildenafil restored NOS and was associated with reversal of the delay in gastric emptying. Hypomotility of the fundus and hypercontractility of the pylorus were found in db/db mice.196 In other studies of diabetic mice the loss of ICCs was associated with electrical dysrhythmias, delayed gastric emptying, and reduced neu­ rotransmission in gastric smooth muscle.197 Chronic hyper­ glycemia also leads to glycosolation metabolic end products that interfere with neural function and smooth muscle contraction.

Postsurgical Gastroparesis

Gastroparesis occurs in a subset of patients undergoing subtle or radical stomach operations that range from vagot­ omy to fundoplication to antrectomy. Truncal vagotomy produces complex effects on the neuromuscular function of the stomach. After vagotomy, the fundus fails to relax nor­ mally after meals, resulting in rapid filling of the antrum.198 Vagotomy is also associated with gastric dysrhythmias,3 decreased antral contractions, and poor antropyloroduode­ nal coordination.199 Truncal vagotomy performed during ulcer operations required a pyloroplasty to reduce sphinc­ teric resistance to outflow because antral contractions were weak and peristalsis was disrupted.200 Most patients recover from the effects of vagotomy, but in patients undergoing extensive resection of the antrum and corpus, prolonged symptoms and chronic gastric neuromuscular dysfunction are likely.

807

808

Section VI  Stomach and Duodenum During fundectomy and lower esophageal resection for esophageal cancer, the vagus nerves are transected and the fundic reservoir is lost. Although pyloroplasty is performed to facilitate gastric emptying, the loss of the fundus and variable amounts of the corpus (pacemaker region) often leads to chronic nausea, gastric dysrhythmias, and gastro­ paresis. Antral resections with Billroth I, Billroth II, or Roux-en-Y gastrojejunostomy are performed to treat gastric tumors or peptic ulcer disease, but these operations may lead to profound neuromuscular dysfunction of the stomach.201,202 Critical amounts of the corpus-antrum (including unknown amounts of the gastric wall containing the pacemaker region) required for normal gastric neuro­ muscular activity may be resected and the efficacy of tritura­ tion by the corpus and antrum may be markedly reduced. Ingested food is retained in the remnant fundus and fails to empty into the corpus203; the corpus fails to mix and empty gastric contents even though the anastomosis is widely patent. Liquids empty poorly into the duodenal or jejunal segments if no peristaltic contractions are present. The Roux-en-Y gastroenterostomy operation may result in the Roux syndrome in which postprandial pain, bloating, and nausea develop. Delayed gastric emptying is due to “func­ tional obstruction” by the Roux limb as the neuromuscular dyssynchrony within the Roux limb prevents emptying of the stomach.204,205 After vagotomy and antrectomy, a minor­ ity of patients develop the dumping syndrome described below. Fundoplication is commonly performed to treat gastro­ esophageal reflux disease that fails to respond to medical therapy (see Chapter 43). Postfundoplication gastroparesis and early satiety, bloating, prolonged fullness, and nausea occur in a minority of patients.206 These patients have altered fundic relaxation, delayed gastric emptying, and gastric dysrhythmias, possibly on the basis of vagal nerve injury during or after the fundoplication procedure.207-209 Because gastric emptying studies are infrequently per­ formed before this operation, it is not known how many of the patients already had gastroparesis before the fundoplica­ tion. In some patients the fundoplication results in rapid filing of the antrum (due to poor relaxation of the fundus) and the rate of gastric emptying is increased.210

Ischemic Gastroparesis

Chronic mesenteric ischemia may cause ischemic gastropa­ resis.211 These patients present with chronic symptoms of gastroparesis. Ischemic gastroparesis is distinct from acute mesenteric ischemia, which presents as an abdominal catas­ trophe with an acute abdomen and gangrenous small intes­ tine (see Chapter 114). Chronic mesenteric ischemia is usually due to progressive atherosclerosis or hyperplasia of the intima of the arteries of the celiac, superior mesenteric, or inferior mesenteric artery. Collaterals of these obstructed arteries form over time so that neuromuscular function of the stomach is preserved, at least for some time. Bypass graft surgery or dilatation of the stenotic arteries results in resolu­ tion of symptoms, eradication of gastric dysrhythmias, and reversal of gastroparesis.211 Thus, ischemic gastroparesis is a reversible form of gastroparesis and should be suspected in patients with gastroparesis, weight loss, and a history of peripheral vascular disease, cerebral vascular disease, or myocardial infarction. An abdominal bruit is present in approximately 50% of patients. There are other uncommon forms of mesenteric vascular compromise (e.g., the median arcuate ligament syndrome) that may result in decreased blood flow to the stomach.212 Release of the arcuate ligament and restoration of blood flow has been associated with improvement in gastric emptying.

On the other hand, superior mesenteric artery syndrome is not accepted as a cause of mechanical obstruction that leads to gastroparesis, nausea, and vomiting.

Obstructive Gastroparesis

Obstructive gastroparesis refers to mechanical obstruction at the pylorus, duodenum, or postduodenal area by tumor, chronic peptic ulcer or inflammation, rings, or webs.149 These patients have documented gastroparesis, but the EGG signal is normal and the amplitude of the 3 cpm EGG wave after the water-load test is increased.149 Figure 48-21A shows an example of high-amplitude 3 cpm EGG waves in a patient with pyloric outlet obstruction due to chronic peptic ulcer disease and fixed narrowing of the pylorus. In these patients the smooth muscle, ENS, and ICCs of the antrum are intact, but the recurrent gastric peristaltic waves meet sustained resistance at the point of obstruction (the pylorus) and the emptying of solid food is delayed. In contrast, patients with idiopathic gastroparesis have bradygastria, tachygastria, or mixed gastric dysrhythmias indicating electrical and con­ tractile dysfunction (see Fig. 48-21B). Surgical correction of the obstruction with Billroth I or Billroth II gastrojejunos­ tomy is necessary to correct obstructive gastroparesis, although some patients may respond to balloon dilation of strictures (see Chapter 53). In patients who have had prolonged obstruction, the stomach may dilate, smooth muscle contractions are weak, and gastric dysrhythmias are present. A more subtle type of gastric outlet obstruction occurs in pylorospasm. The sustained pyloric “spasm” or tone may cause right upper quadrant abdominal pain and prevents normal gastric peristaltic waves from empting chyme into the duodenum. Thus, the rate of emptying is delayed and gastroparesis is present. These patients also have normal 3 cpm EGG rhythm because the neuromuscular apparatus of the corpus-antrum is intact. Dilatation of the pylorus with a 20-mm balloon for two minutes decreased postprandial symptoms and improved the rate of gastric emptying.213 Finally, some patients have normal 3 cpm EGG signals and poor gastric emptying on the basis of “electromechanical dissociation.” In these situations it is possible that MY-ICCs that generate slow waves may be normal, but IM-ICCs and/ or smooth muscle dysfunction is present.

Idiopathic Gastroparesis

Almost one third of patients with gastroparesis have idio­ pathic gastroparesis.214 In many of these patients an acute febrile illness preceded the diagnosis of gastroparesis by many months.215,216 These “herald” illnesses are frequently described as flu-like with fever and nausea and vomiting. Patients often date the onset of their nausea from that point. Norwalk virus, herpes simplex virus, and EpsteinBarr virus infections have been documented in patients with sudden onset gastroparesis and normal immune systems, whereas other patients are immunocompro­ mised.217 Postviral gastroparesis may resolve completely over one to two years. It is unclear if the offending virus affects ICCs, smooth muscle or the ENS of the stomach, or vagal or splanchnic nerves, but these cases are analogous to postviral cardiomypathy or postviral neuropathy. Degen­ eration and loss of ICCs has been reported in patients with severe idiopathic gastroparesis.165,218 Loss of ICCs in knock­ out mice is associated with gastric dysrhythmias.219 Gastric dysrhythmias are common in patients with idiopathic gastroparesis (see Fig. 48-21B).149 Other historical clues to the genesis of idiopathic gastroparesis are exposures to (1) multiple courses of antibiotics (for example, treatments for chronic ear or sinus infections), (2) anesthetic agents for a

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders variety of common operations, and (3) food poisoning–like illnesses of unknown cause. Idiopathic gastroparesis is diagnosed in patients who have delayed gastric emptying and gastric dysrhythmias, but no primary causes of gastroparesis such as diabetes, ischemia, or gastric surgery. Many other diseases and dis­ orders, including intestinal pseudo-obstruction (discussed later), autonomic nervous system abnormalities, thyroid disease, or CNS diseases, may cause or are associated with gastroparesis. If these disorders are identified, then the gas­ troparesis is considered secondary to these diseases.

GASTRIC NEUROMUSCULAR DYSFUNCTION ASSOCIATED WITH OTHER GASTROINTESTINAL DISORDERS Functional Dyspepsia

Functional dyspepsia (FD) symptoms include epigastric dis­ comfort, early satiety, fullness, nausea, and vomiting in the setting of normal upper endoscopy and routine laboratory tests. FD is divided further into an epigastric pain syndrome and postprandial distress syndrome, the latter comprising 80% of the FD patients.220 One of the dominant symptoms that patients with FD have is recurrent and unexplained nausea. This is a debilitating symptom with a large differ­ ential diagnosis (Table 48-2) and these disease categories should be considered in the evaluation of the patient. If the patient has one of these disorders, then FD is not the diagnosis. The pathophysiology of the FD symptoms remains an area of intensive investigation (see Chapter 13). The FD symp­ toms may be caused by one of several neuromuscular dis­ orders of the stomach as shown in Figure 48-20. Gastroparesis is found in 17% to 40% and gastric dysrhythmias in 40% to 55% of patients with FD symptoms.60,147 If secondary causes for gastroparesis have been excluded, then these patients have idiopathic gastroparesis and/or gastric dys­ rhythmias, not FD.

Table 48-2  Differential Diagnosis of Chronic Nausea and Vomiting Mechanical gastrointestinal tract obstruction Pylorus, small intestine, colon Mucosal inflammation Peritoneal irritation Carcinomas Gastric, ovarian, renal, bronchogenic Metabolic/endocrine disorders Diabetic mellitus, hypothyroidism, hyperthyroidism, adrenal insufficiency, uremia Medications Anticholinergics, narcotics, l-dopa, progesterone, calcium channel blockers, digitalis, nonsteroidal anti-inflammatory agents, antidysrhythmia agents, lubiprostone Gastroparesis Obstructive Diabetic Ischemic Postsurgical Miscellaneous, including pseudo-obstruction Idiopathic Gastric dysrhythmias Tachygastria, bradygastria, mixed Central nervous system disorders Tumors, migraine, seizures, stroke Psychogenic disorders Anorexia nervosa, bulimia nervosa

The rate of gastric emptying is normal in the majority of patients with FD, but gastric dysrhythmias may be present in these patients. In one series of patients with FD, 35% had gastric dysrhythmias and normal gastric emptying.63 Improvement in gastric dysrhythmia and FD symptoms was reported in patients treated with cisapride,221,222 and similar results were reported in children and adults with dys­ pepsia.223,224 Thus, gastric dysrhythmias themselves are the cause of some of the FD symptoms and are a therapeutic target. Gastric neuromuscular disorders such as abnormalities in gastric accommodation or gastric hypersensitivity to disten­ tion may account for FD symptoms in other patients.225-230 Thus, a variety of neuromuscular dysfunctions in different regions of the stomach may be present in patients with FD symptoms.231 Exposure of the duodenal mucosa to acid is also a proposed mechanism of FD symptoms.232 Increasing data that link certain dyspepsia symptoms such as nausea or bloating or discomfort to distinct neuromuscular abnor­ malities of the stomach will aid in improved diagnosis and treatment of these patients.

Gastroesophageal Reflux Disease

Gastroesophageal reflux disease (GERD) typically presents with heartburn and regurgitation, but many patients have additional symptoms of early satiety, postprandial fullness, and nausea (see Chapter 43). Approximately 30% of patients with GERD also have FD and represent an “overlap syn­ drome.”233 Furthermore, 20% to 30% of patients with GERD have delayed gastric emptying.234,235 Twenty-five percent of patients with GERD and dyspepsia had delayed gastric emp­ tying and almost 70% had gastric dysrhythmias.236 The FD-like symptoms do not resolve when the GERD symptoms are treated with proton pump inhibitors (PPIs) because the gastric neuromuscular disorder is not treated by PPIs. Lower esophageal sphincter abnormalities, including transient lower esophageal sphincter relaxations, may con­ tribute to inefficient gastric emptying and gastric dysrhyth­ mias reported in these patients.237 In response to a meal, the fundus relaxes more in GERD patients compared with controls.238 GERD patients also tended to retain solids and liquids in the proximal stomach compared with control subjects,239 an abnormality that may stimulate additional transient lower esophageal sphincter relaxations. Radiofrequency ablation (RFA) treatment at the lower esophageal sphincter region in patients with heartburn and dyspepsia improved heartburn, gastric emptying rates, and gastric dysrhythmias.240,241 These data suggest in some patients with GERD and gastroparesis, treatments of GERD by RFA or fundoplication improves the gastric electrical rhythms and the rate of gastric emptying.

Constipation, Irritable Bowel Syndrome, and Pseudo-Obstruction

Gastroparesis and gastric dysrhythmias after the water-load test have been reported in 20% to 30% and 60% of patients with constipation-predominant irritable bowel syndrome, respectively.242,243 These patients represent another overlap syndrome of gastrointestinal neuromuscular disorders (see Chapter 118). GERD, gastroparesis/gastric dysrhythmias, and irritable bowel syndrome are all present in some patients and indicate a disorder of diffuse gastrointestinal neuromuscular dysfunction.233 Patients with intestinal pseudo-obstruction syndromes often have GERD, gastroparesis, small bowel dysmotility, and colonic inertia and reflect the severest form of general­ ized GI neuromuscular disorders.244,245 Pseudo-obstruction

809

810

Section VI  Stomach and Duodenum secondary to scleroderma commonly involves the esopha­ gus and stomach but may also cause small bowel dysmo­ tility and subsequent bacterial overgrowth.246 Intestinal pseudo-obstruction may be due to idiopathic degenerative or inflammatory processes involving the smooth muscle or enteric nervous system, as discussed in Chapter 120. A variety of neurologic diseases may also involve the stomach or other regions of the digestive tract and cause pseudo-obstruction–like symptoms. These neurologic disorders include spinal cord injuries, head injuries, amyo­ trophic lateral sclerosis, myasthenia gravis, a variety of muscular dystrophies, and Parkinson’s disease. Chapter 35 discusses stomach neuromuscular dysfunction and auto­ nomic neuropathies.

bowel transit and poor absorption of the ingested nutrients lead to an osmotic form of diarrhea.260 Patients with no history of gastric operations may also have rapid gastric emptying (dumping syndrome),259 includ­ ing patients with FD.260 Rapid gastric emptying is present when more than 30% of the meal (a low-fat egg substitute meal) is emptied in 30 minutes, or more than 70% is emptied at 60 minutes.128 Idiopathic abnormally rapid emptying is diagnosed in patients with no history of gastric operations or other causes. Rapid gastric emptying is also associated with early stages of type 2 diabetes mellitus (see earlier), Zollinger-Ellison syndrome, and the variety of gastric sur­ geries described earlier. Details of these entities is reviewed under the postsurgical syndromes in Chapter 53.

Miscellaneous Conditions

Patients with FD symptoms and cirrhosis with portal hyper­ tension,247 chronic kidney disease,248 chronic pancreatitis,249 or advanced human immunodeficiency virus (HIV) infec­ tions250 may have underlying gastroparesis. Patients with the Rett syndrome, which includes lack of development, autistic behavior, ataxia, and dementia in young girls, fre­ quently have failure to thrive and significant gastroparesis and esophageal contraction abnormalities.251 Patients with a variety of cancers may have local and systemic effects on the stomach neuromuscular apparatus that results in gastro­ paresis.252,253 The neoplastic neuropathic syndromes and the effects of cytotoxic chemotherapy and radiation may result in gastroparesis and affect the patient’s nutrition and intra­ vascular volume status. H. pylori infection does not affect the rate of gastric emp­ tying. On the other hand, abnormal gastric myoelectrical activity has been reported in patients with H. pylori infec­ tion, and the dysrhythmias disappeared after eradication of H. pylori.254 Gastric emptying is delayed in patients with anorexia nervosa,255 but is normal in patients with bulimia.256 Cyclic vomiting syndrome (CVS) is an unusual entity in that days of profound and unremitting nausea and vomiting (that requires hospitalization) are followed by many days or months with virtually no GI symptoms.257 CVS occurs in adults as well as in infants.257,258 When patients are well, EGG abnormalities are present and some of these patients have gastroparesis.

DUMPING SYNDROME AND RAPID GASTRIC EMPTYING

Dumping syndrome occurs in some patients who have had vagotomy and pyloroplasty or Billroth I or Billroth II gas­ trojejunostomy.259 In these patients the ingested foods are not accommodated and retained in the fundus (because of poor fundic relaxation) and the foods are not normally tritu­ rated because the corpus and antrum have been resected. Thus, liquid and solid nutrients are rapidly emptied or “dumped” into the duodenum or jejunum. Symptoms of dumping syndrome include nonspecific abdominal discom­ fort, bloating, and nausea that may precipitate vomiting. These symptoms are usually experienced in the first hour after ingestion of foods and can mimic symptoms of gastro­ paresis. Sweating and lightheadedness, however, may occur and be followed by abdominal cramps and diarrhea that occur two to four hours after the meal and are clues to the dumping syndrome. Early symptoms are due to the disten­ tion of the small bowel, whereas symptoms that occur later are due to rapid absorption of carbohydrates and hypergly­ cemia that is poorly matched in time with secretion of insulin. This mismatch of plasma glucose and insulin results in symptomatic hypoglycemia. The rapid small

DIAGNOSIS SYMPTOMS

Most affected patients have few upper GI symptoms when they are fasting. However, the ingestion of meals stimulates the disordered gastric neuromuscular apparatus, and early satiety, prolonged epigastric fullness, nonspecific epigas­ tric discomfort, mild to severe nausea, and vomiting are experienced.233 These are also the symptoms associated with functional dyspepsia (Chapter 13). Vomitus that con­ tains undigested, chewed food is strong evidence for gas­ troparesis. Prolonged postprandial fullness, weight loss, and female gender are predictive factors for gastropare­ sis.261 By adjusting their diet, patients learn to reduce these postprandial symptoms and carry on for months or years before they seek medical attention or their physicians rec­ ognize the possibility of gastroparesis. The Gastroparesis Cardinal Symptom Index is a useful validated question­ naire to quantify gastroparesis symptoms.262 Persistent nausea is one of the most noxious symptoms of the gastric neuromuscular disorders. A thorough review of the causes of nausea and vomiting is required (Chapter 14) and an appropriate differential diagnosis should be considered (see Table 48-2).263 In the evaluation of unexplained nausea and vomiting, gastric neuromuscular dysfunction must be distinguished from esophageal diseases and rumination syndrome. Occult GERD may present as unexplained nausea because these patients report little or no heartburn.264 Regurgitation is the gentle delivery of gastric content into the esophagus and pharynx (and the content is sometimes reswallowed), whereas vomiting is the forceful ejection of gastric contents from the mouth. Rumination refers to the effortless return of ingested liquids and solid foods into the mouth without burning, bitter taste, or nausea. Patients with rumination have impaired gastric accommodation and a more sensitive relaxation of the lower esophageal sphincter pressure in response to gastric distention.265-267 Rumination occurs in healthy adolescents and young adults, but was previously recognized among children with neural and developmental disorders. Abdominal pain, in contrast to the abdominal discomfort of bloating and nausea, occurs infrequently in patients with gastric neuromuscular disorders. Any recurrent abdominal pain syndrome should be worked up extensively because the nausea and vomiting may be secondary to the specific cause of the pain (e.g., burning epigastric pain associated with nausea and vomiting may be due to active peptic ulcer disease). Once the peptic ulcer disease is diagnosed and treated, the pain and nausea disappear. A specific pain syndrome may suggest diagnoses such as cholecystitis,

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders pancreatitis, or sphincter of Oddi dysfunction. On the other hand, the epigastric discomfort or pain in some patients may originate from excessive muscle tone of the fundus, high-amplitude antral contractions, pylorospasm, or hyper­ sensitivity of the stomach.268 Recurrent retching and vomit­ ing may result in abdominal pain, which is a result of abdominal muscle and rib tenderness or the abdominal wall syndrome.269,270

PHYSICAL EXAMINATION

The general examination may be normal or reveal signs of volume depletion, weight loss, and poor nutrition. Inspec­ tion of dentition may show erosion of enamel associated with chronic GERD or bulimia. Abdominal examination may detect masses, organomegaly, and areas of tenderness. Abdominal distention and succession splash may be present. Auscultation over the epigastrium may detect bruits that indicate stenoses of the celiac or superior mesenteric arter­ ies. Tenderness that is well localized at healed incisions and persists when the anterior abdominal muscles are con­ tracted (Carnett’s sign) suggests an abdominal wall syn­ drome.269,270 Neurologic examination may reveal nystagmus, facial weakness, ataxia, or other abnormalities.

STANDARD TESTS

Common causes of nausea and vomiting and dyspepsia symptoms are excluded by normal upper gastrointestinal (GI) series, CT of the abdomen and head, routine laboratory studies, and upper endoscopy.271 In patients with dyspep­ sia, gastric cancers are detected in less than 1% of the patients undergoing upper endoscopy.272 Most patients seen by gastroenterologists for dyspepsia or chronic nausea have already received acid-suppression therapy with PPIs, but symptoms persist. In such patients, gastric neuromuscular disorders should be considered.

SPECIALIZED NONINVASIVE TESTS

An objective diagnosis of neuromuscular disorders of the stomach can be established by the results of gastric empty­ ing tests and EGG. Gastric emptying tests and EGG are

complementary in defining different aspects of gastric neuromuscular disorders (Table 48-3).147,271 By combining the results of gastric emptying tests and EGG, four patho­ physiologic categories of gastric neuromuscular function are defined and rational treatment approaches can be designed. The categories are (1) gastroparesis with gastric dysrhyth­ mia; (2) normal gastric emptying with gastric dysrhythmia; (3) normal gastric emptying with normal gastric electrical rhythm; and (4) gastroparesis with normal gastric electrical rhythm. The EGG testing device is cleared by the FDA, but the combination of EGG and gastric emptying testing is infrequently performed. Nevertheless, the four categories provide a conceptual framework for understanding the spectrum of gastric neuromuscular disorders and providing an approach to therapy (see Table 48-3). Category 1 patients have severe neuromuscular dysfunc­ tion with gastroparesis and gastric dysrhythmias such as tachygastria. Among patients with “functional dyspepsia,” 17% to 32% have gastric dysrhythmias and gastropare­ sis.60,147,271 Patients in category 1 often have more severe symptoms, require many drugs, and may require venting gastrostomies, enteral feeding, and gastric pacing, as dis­ cussed following. Among patients with functional dyspepsia, 40% to 60% have gastric dysrhythmias despite normal gastric emptying. Such category 2 patients had a significantly better response to the prokinetic agent cisapride than patients with normal EGG recordings.221 In patients with normal gastric myoelectrical activity and normal gastric emptying (category 3), the nausea symptoms are likely due to visceral hypersensitivity or nongastric causes. Nongastric diagnoses should be considered in this patient group. Atypical GERD may cause nausea, and a 24-hour pH study will confirm the relationship.264 A CCKstimulated gallbladder emptying study may document gall­ bladder dysfunction in the absence of cholelithiasis. If postprandial abdominal discomfort and disturbed bowel function components are present, then irritable bowel syn­ drome should be considered. Central nervous system causes of nausea also should be assessed.

Table 48-3  Categories of Gastric Neuromuscular Disorders and Treatment Approaches Based on Gastric Electrical and Emptying Test Results CATEGORY 1

CATEGORY 2

CATEGORY 3

CATEGORY 4

Test Results Gastric dysrhythmia and gastroparesis Diagnosis Severe gastric myoelectrical contractile disorder Treatment Nausea/vomiting diet† Prokinetic therapy Anti-nauseant therapy G tube/J tube Total parenteral nutrition Acustimulation Endoscopic therapies* Gastric electrical stimulation Gastric pacemaker

Test Results Gastric dysrhythmia and normal emptying Diagnosis Gastric myoelectrical disorder

Test Results Normal gastric rhythm and emptying Diagnosis Visceral hypersensitivity Nongastric causes

Test Results Normal gastric rhythm and gastroparesis

Treatment Nausea/vomiting diet† Prokinetic therapy Anti-nauseant therapy

Treatment Nausea/vomiting diet† Antidepressant therapy Drugs for fundic/antrum relaxation (Further workup for nongastric causes)

Treatment Surgery for obstruction Nausea/vomiting diet† Prokinetic therapy Anti-nauseant therapy

Diagnosis Mechanical obstruction Electro-contractile dissociation

*See Table 48-4. † See Table 48-5. G tube, gastrostomy tube; J tube, jejunostomy tube. Modified from Koch KL. Nausea and Vomiting. In: Wolfe MM, editor. Therapy of Digestive Disorders. 2nd ed. Philadelphia, Pa: Elsevier; 2006. pp 1003-17.

811

812

Section VI  Stomach and Duodenum Patients who have gastroparesis and normal or highamplitude 3 cpm EGG signals (category 4) may have mechanical obstructions of the stomach and duodenum that are reversible with operation.149 Alternatively, patients with a normal-amplitude 3 cpm pattern and gastroparesis may have a form of “electromechanical dissociation,” requiring medical therapy.

TREATMENT By grouping patients on the basis of the gastric emptying and myoelectrical results, the pathophysiologic findings can help in the understanding of symptoms and assist in devel­ oping an approach to treatment (see Table 48-3). If delayed gastric emptying is confirmed, then the major causes of gastroparesis should be reviewed to specifically define the cause of the gastroparesis (see Table 48-2). The reversible causes of gastroparesis, gastric outlet obstruction, and chronic mesenteric ischemia, should be excluded. If gastric emptying is normal, then gastric dysrhythmias and gastric accommodation disorders may be the neuromuscular disor­ ders that are relevant to the symptoms. Treatment of gastric neuromuscular disorders is problematic. Scant specific ther­ apies are available to address specific pathophysiologic dis­ orders. Treatments listed in Table 48-4 reflect the broad but limited armamentarium that ranges from prokinetic agents to diet counseling and electrical therapies.

DRUG THERAPY Prokinetic Agents for Corpus-Antrum

Drugs with prokinetic effects on gastric contractility and gastric dysrhythmias are usually prescribed for patients in categories 1 and 2 (see Table 48-4). Patients who have gastroparesis and tachygastria have severe electrical and contractile abnormalities of the stomach. The treatment includes prokinetic, antinauseant therapies, and dietary counseling.271 Erythromycin is a macrolide antibiotic and motilin-like molecule that increases gastric emptying by stimulating strong phase 3–like antral contractions.273,274 Erythromycin, however, often increases nausea and vomiting symptoms. Metoclopramide, a substituted benzamide related to pro­ cainamide, is a useful prokinetic antiemetic. However, metoclopramide also causes depression, extrapyramidal side effects, and irreversible tardive dyskinesia.275,276 Dom­ peridone is a dopamine antagonist that decreases nausea, corrects gastric dysrhythmias, and increases gastric empty­ ing rates.277,278 Domperidone may be obtained through a new drug application process with the FDA. Cisapride and tega­ serod are 5-HT4 agonists and were not approved for gastric emptying disorders, but these drugs increase gastric empty­ ing rates and decrease dyspepsia symptoms in some patients.279-281 Cisapride and tegaserod were withdrawn from the market but are available through special FDA programs.

Prorelaxant Agents for Fundus and Pylorus

Drugs that relax the fundus are few. Sumitriptan, a 5-HT1 antagonist, decreases fundic tone but was not better than placebo in reducing symptoms in patients with FD.282 Trials of dicyclomine or calcium channel blockers to decrease fundic tone have not been reported. Botulinum toxin relaxes the pyloric sphincter pressure and is described below.

Antinauseant Therapy

Nonspecific approaches to treating nausea and vomiting from gastric neuromuscular disorders include 5-HT3 sero­ tonin antagonists such as ondansetron and granisetron (see Table 48-4). These agents, as well as the phenothiazines and antihistamines such as promethazine, dimenhydrinate, and cyclizine, are often used for these symptoms, but there are no controlled trials in patients with gastric neuromuscular disorders. Lorazepam or alprazolam or other antianxiety medications reduce nausea in some patients.283 An uncon­ trolled trial of tricyclic antidepressants such as amitripty­ line alleviated nausea in approximately 70% of patients with unexplained nausea.284

ELECTRICAL THERAPY Acustimulation

Acustimulation (mild electrical stimulation of acupuncture points) reduces nausea of pregnancy, nausea due to chemo­ therapy agents, postoperative nausea, and the nausea of motion sickness.285,286 These treatment methods have not been systematically studied in patients with gastric neuro­ muscular disorders.

Gastric Electrical Therapies

Three different methods are being investigated to treat gastroparesis: (1) gastric electrical stimulation (GES) with high-frequency (e.g., 12 cpm) and short duration (300 microsecond) stimulation; (2) gastric pacing with lowfrequency (e.g., 3 cpm) and long duration (300 millisecond) stimulation; and (3) sequential neural electrical stimulation with multiple pairs of electrodes positioned on the corpus-antrum. Gastric Electrical Stimulation GES at 12 cpm during a 12-month period of continuous treatment significantly decreased nausea and vomiting in patients with refractory, idiopathic, diabetic gastroparesis or postsurgical gastroparesis.287-289 Stimulating the stomach at 12 cpm (four times the normal slow wave frequency) resulted in improvement in nausea and vomiting in patients with gastroparesis.290 No placebo-controlled studies are available to document these benefits of GES. Gastric empty­ ing rates and gastric dysrhythmias did not improve in those patients reporting improvement in symptoms. Vagal affer­ ent nerve stimulation and CNS changes in response to the GES have been proposed as the mechanisms of benefit. Uncontrolled studies from several centers indicate that 50% to 70% of patients treated GES report a decrease in their chronic nausea and vomiting symptoms. Complications include a 10% incidence of infections in the subcutaneous pocket where the device is positioned. There is also a small incidence of fracture of the electrode wires. Gastric Pacing Lin and coleagues291 studied low-frequency gastric stimula­ tion using a 3 cpm stimulus to pace or entrain the normal slow wave rhythm in patients with gastroparesis. Gastric pacing seeks to stimulate three gastric peristaltic contrac­ tions per minute and improve gastric emptying. Electrodes were positioned on the serosal surface of the stomach during surgery in 13 patients. Stimulation at a frequency up to 10% higher than the normal 3 cpm, at an amplitude of 4mA and a pulse width of 300 milliseconds was used to entrain the slow waves. Gastric dysrhythmias were eradicated in some patients. In a similar study of 9 additional patients, electri­ cal stimulation was used to entrain the slow wave and

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders Table 48-4  Drug and Nondrug Therapies Used to Treat Symptoms in Patients with Gastric Neuromuscular Disorders THERAPY

MECHANISMS/SITES OF ACTION

DOSAGE

ADVERSE EFFECTS

Prokinetic Therapy Macrolides Erythyromycin

Motilin receptor agonist

125-250 mg four times daily

Nausea, diarrhea abdominal cramps, rash

Dopamine (D2) receptor antagonist; 5-HT3-receptor antagonist; 5-HT4 receptor agonist D2-receptor antagonist (peripheral)

5-20 mg before meals and at bedtime

Extrapyramidal symptoms, dystonic reactions, anxiety, depression, hyperprolactinemia, tardive dyskinesia Hyperprolactinemia, breast tenderness, galactorrhea

5-HT4 receptor agonist

5-20 mg before meals

Partial 5-HT4 receptor agonist

2-6 mg three times daily

Cardiac dysrhythmias, diarrhea, abdominal discomfort Diarrhea, abdominal pain

Muscarinic antagonist see Endoscopic Therapies

10-20 mg before meals

Drowsiness, dry mouth

5-HT3 receptor antagonist

4-8 mg twice daily, either orally or intravenously 2 mg once daily

Headache, increased liver enzymes

Substituted Benzamides Metoclopramide

Domperidone* Serotonin Agonists Cisapride* Tegaserod* Prorelaxant Therapies Dicyclomine Botulinum toxin (Botox) Anti-nauseant Therapy Serotonin Antagonists Ondansetron Granisetron Phenothiazines Prochlorperazine Antihistamines Promethazine Dimenhydrinate Cyclizine Butyrophenones Droperidol Antidepressants Amitriptyline Nortriptyline Benzodiazepines Lorazepam Alprazolam Electrical Therapies Acustimulation Acupressure/Acupuncture Gastric electrical stimulation Gastric pacing Endoscopic Therapies Botox injection into pylorus Balloon dilation of pylorus Radiofrequency ablation at LES Diet Therapies Gastroparesis diet High-protein drinks Gastrostomy Jejunostomy Total parenteral nutrition

5-HT3 receptor antagonist

10-20 mg before meals and at bedtime

Headache, increased liver enzymes

Central nervous system (CNS) sites

5-10 mg three times daily

Hypotension, extrapyramidal symptoms

CNS, H1 receptor antagonist H1 receptor antagonist H1 receptor antagonist

25 mg twice daily 50 mg four times daily 50 mg four times daily

Drowsiness Drowsiness Drowsiness

Central dopamine receptor antagonist

2.5-5 mg intravenously every 2 hours

Sedation, hypotension

CNS sites CNS sites

25-100 mg at bedtime 10-75 mg at bedtime

Constipation Constipation

CNS sites CNS sites

0.5-1 mg four times daily 0.25-0.5 mg three times daily

Drowsiness, lightheadedness Drowsiness, lightheadedness

Spinal/vagal afferents? Endorphins ?Vagal afferents Control dysrhythmias, improve gastric emptying

Variable NA 12 cpm, 330 milliseconds, 5 mÅ 3 cpm, 300 milliseconds, 4 mÅ

Local tenderness Pocket infection Pocket infection

Relax pyloric muscle Stretch pyloric muscle Improve GEPG, improve gastric myoelectrical activity

25-50 units per quadrant 20 mm balloon, 2 min NA

None Perforation Transient dysphagia

Diet based on gastric emptying physiology Decreases gastric dysrhythmias Venting paretic stomach Enteral nutritional support Bypass paretic stomach

see Table 46-5

None

Unknown As needed As needed As needed

None See Chapter 5 See Chapter 5 Sepsis, thrombosis of central veins

*Compassionate clearance use only. CNS, central nervous system; D2, dopamine2; 5-HT, 5-hydroxytryptamine; GEPG, gastroesophageal pressure gradient; H1, histamine1; LES, lower esophageal sphincter; NA, not applicable.

tachygastria was converted to normal 3 cpm rhythms in 2 patients.292 After one month of gastric pacing treatment, gastric emptying was significantly improved, symptoms of gastroparesis were significantly reduced, and 8 of 9 patients no longer required jejunostomy tube feedings. These pacing

devices require more electrical energy compared with GES; thus, battery life is a major limitation at this time. Adverse events due to gastric pacing devices are discomfort at the site of electrical stimulation and fracture or dislodgement of the electrodes.

813

814

Section VI  Stomach and Duodenum Table 48-5  Diet for Nausea and Vomiting in Patients with Gastric Neuromuscular Disorders DIET Step 1: Sports Drinks and Bouillon For severe nausea and vomiting: Small volumes of salty liquids, with some caloric content to avoid volume depletion Multiple vitamin Step 2: Soups If Step 1 is tolerated: Soup with noodles or rice and crackers Peanut butter, cheese, and crackers in small amounts Caramels or other chewy confection Ingest above foods in at least six small-volume meals/day Multiple vitamin Step 3: Starches, Chicken, Fish If Step 2 is tolerated: Noodles, pastas, potatoes (mashed or baked), rice, baked chicken breast, fish (all easily mixed and emptied by the stomach) Ingest solids in at least six small-volume meals/day Multiple vitamin (liquid or dissolvable)

GOAL

AVOID

1000-1500 mL/day in multiple servings (e.g., 12, 120-mL servings over 12-14 hr) Patient can sip 30-60 mL at a time to reach approximately 120 mL/hr

Citrus drinks of all kinds; highly sweetened drinks

Approximately 1500 calories/day to avoid volume depletion and maintain weight (often more realistic than weight gain)

Creamy, milk-based liquids

Common foods that patient finds interesting and satisfying and that provoke minimal nausea/vomiting symptoms

Fatty foods that delay gastric emptying; red meats and fresh vegetables that require considerable trituration; pulpy fibrous foods that promote formation of bezoars

Modified from Koch KL. Nausea and vomiting. In Wolfe MM, editor. Therapy of Digestive Disorders. 2nd ed. Philadelphia: Elsevier; 2006. pp 1003-17.

Sequential Neural Electrical Stimulation Sequential neural electrical gastric stimulation is gastric pacing that uses a microprocessor to sequentially activate a series of electrodes that encircle the distal two thirds of the stomach. The stimulation sequence induces propagated contractions that cause a forceful emptying of the gastric content.293

soups containing noodles or rice. Milk-based creamy soups are avoided. A dissolvable multiple vitamin is taken daily. Step 3 emphasizes starches and chicken and turkey breast. These solid foods require less gastric work to mix and empty compared with fresh vegetables or red meats. Fried and greasy foods are avoided because fats delay gastric emptying.

ENDOSCOPIC THERAPY

Nutraceuticals

Endoscopic therapies refer to drug or device therapies deliv­ ered to the relevant regions of the stomach via endoscopes. The injection of botulinum toxin into the pylorus to decrease sphincter pressure and to improve gastric emptying and nausea and vomiting in patients with gastroparesis pro­ duced results similar to placebo injections.293-296 Balloon dilation of the pylorus improves gastric emptying in patients with gastroparesis and normal 3 cpm EGG patterns.213 RFA treatment applied to the lower esophageal sphincter (LES) area in patients with GERD and dyspepsia improve gastric dysrhythmias and emptying.240,241

DIET THERAPY Dietary Counseling

Many patients with acute or chronic nausea and vomiting from gastric neuromuscular disorders do not know what to eat and may benefit from dietary counseling.297 In the Nausea/Vomiting (Gastroparesis) Diet, liquid and solid foods that are easy for the stomach to mix and empty are prescribed.271 The Nausea/Vomiting Diet is based on gastric emptying principles and is a three-step diet that requires minimal neuromuscular work of the stomach as the diet is advanced (Table 48-5). Step 1 is primarily electrolyte solutions that are con­ sumed in small amounts over a 24-hour period in order to avoid volume depletion. Liquids require less gastric neuro­ muscular work to empty than solid foods. If patients tolerate step 1, then step 2 may be tried next. Step 2 diet includes

Liquid protein meals with or without ginger decrease nausea associated with motion sickness and gastric dysrhythmias, nausea of pregnancy, and delayed nausea after chemother­ apy.298-301 These protein-based meal therapies have not been formally evaluated in patients with gastroparesis or gastric dysrhythmias. A rationale for the nutraceutical approach to treatment of nausea in patients with gastric neuromuscular disorders has evolved and deserves further study.

Other Approaches to Nutritional Support

For patients with chronic nausea and vomiting, percutane­ ous endoscopic gastrostomy (PEG) tubes may be placed for venting gastric contents in order to avoid frequent vomiting episodes and thereby improve quality of life.302 The venting PEG does not treat the underlying gastric neuromuscular disorder, but it allows the patients to empty the stomach rather than suffer repeated episodes of emesis and discom­ fort. Medications and some nutritional liquids may be toler­ ated when given through the gastrostomy tube. Jejunal feeding tubes for enteral nutrition may be needed to provide basic caloric support for patients with severe nausea and vomiting from gastric neuromuscular disorders. A PEG with J-tube extension usually fails because a single vomiting episode may propel the extension tube into the stomach. A J-tube placed endoscopically or surgically is required (see Chapters 4 and 5).303 Total parenteral nutrition (TPN) via central intravenous catheters should be avoided if at all possible because of the frequent development of line sepsis and occasional venous thrombosis.

Chapter 48  Gastric Neuromuscular Function and Neuromuscular Disorders KEY REFERENCES

Abell T, McCallum R, Hocking M, et al. Gastric electrical stimulation for medically refractory gastroparesis. Gastroenterology 2003; 125:421-8. (Ref 287.) Brzana RJ, Koch KL, Bingaman S. Gastric myoelectrical activity in patients with gastric outlet obstruction and idiopathic gastroparesis. Am J Gastroenterol 1998; 93:1803-9. (Ref 149.) Camilleri M. Integrated upper gastrointestinal response to food intake. Gastroenterology 2006; 131:640-58. (Ref 54.) De Giorgio R, Sarnelli G, Corinaldesi R, Stranghellini V. Advances in our understanding of the pathology of chronic intestinal pseudoobstruction. Gut 2004; 53:1549-52. (Ref 245.) Hinder RA, Kelley KA. Human gastric pacesetter potential: Site of origin, spread, and response to gastric transection and proximal vagotomy. Am J Surg 1997; 133:29-33. (Ref 3.) Huizinga JD. Physiology and pathophysiology of the interstitial cell of Cajal: From bench to bedside II. Gastric motility: Lessons from mutant mice on slow waves and innervation. Am J Physiol 2001; 281:G112934. (Ref 17.) Ladabaum U, Koshy SS, Woods ML, et al. Differential symptomatic and electrogastrographic effects of distal and proximal human gastric distension. Am J Physiol 1998; 275:G418-24. (Ref 91.) Murray CD, Martin NM, Patterson M, et al. Ghrelin enhances gastric emptying in diabetic gastroparesis: a double-blind, placebocontrolled, crossover study. Gut 2005; 54:1693-8. (Ref 105.)

Ordog T. Interstitial cells of Cajal in diabetic gastroenteropathy. Neuro­ gastroenterol Motil 2008; 20:8-18. (Ref 164.) Owyang C, Hasler WL. Physiology and pathophysiology of the intersti­ tial cells of Cajal: From bench to bedside. VI. Pathogenesis and thera­ peutic approaches to human gastric dysrhythmias. Am J Physiol 2002; 283:G8-15. (Ref 29.) Sanders KM, Koh SD, Ward SM. Interstitial cells of Cajal as pacemakers in the gastrointestinal tract. Annu Rev Physiol 2006; 68:307-43. (Ref 2.) Schulze K. Imaging and modeling of digestion in the stomach and the duodenum. Neurogastroenterol Motil 2006; 18:172-83. (Ref 53.) Stern RM, Jokerst MD, Levine ME, Koch KL. The stomach’s response to unappetizing food: Cephalic-vagal effects on gastric myoelectrical activity. Neurogastroenterol Motil 2001; 13:151-4. (Ref 117.) Tougas G, Eaker EY, Abell TL, et al. Assessment of gastric emptying using a low fat meal: Establishment of international control values. Am J Gastroenterol 2000; 95:1456-62. (Ref 43.) Woods SC. Gastrointestinal satiety signals. I. An overview of gastroin­ testinal signals that influence food intake. Am J Physiol Gastrointest Liver Physiol 2004; 286:G7-13. (Ref 114.) Full references for this chapter can be found on www.expertconsult.com.

815

CHAPTE R

49 Gastric Secretion Mitchell L. Schubert and Jonathan D. Kaunitz

CHAPTER OUTLINE Functional Anatomy  817 Acid Secretion: Paracrine, Hormonal, Neural, and Intracellular Regulation  821 Histamine  821 Gastrin  821 Acetylcholine  822 Somatostatin  823 Miscellaneous Peptides  823 Parietal Cell Intracellular Pathways  823 Proton Pump Inhibitors  824 Integrated Response to a Meal: Interplay of Neural, Paracrine, and Hormonal Pathways  825 Helicobacter pylori–Induced Perturbations in Acid Secretion  826

The stomach is an active reservoir that stores, grinds, and slowly dispenses partially digested food to the intestine for further digestion and absorption. Its main secretory function is the production of hydrochloric acid. Gastric acid secre­ tion is present on the first day of life and increases as infants become more mature.1 By two years of age, acid secretion is similar to that of adults, when corrected for body weight.2 Most studies indicate that the rate of acid secretion changes little after the second decade of life unless there is coexi­ sting disease of the acid-secreting glandular mucosa such as infection with Helicobacter pylori (HP) or atrophic gastritis.3-5 Acid facilitates the digestion of protein by converting the proenzyme pepsinogen to the active proteolytic enzyme pepsin. It also facilitates the absorption of iron, calcium, vitamin B12, and certain medications (e.g., thyroxin) as well as prevents bacterial overgrowth, enteric infection, and possibly community-acquired pneumonia.6-18 The stomach also secretes lipase, intrinsic factor, elec­ trolytes (e.g., HCO3−, K+, and Cl−), and mucins in addition to a variety of neurocrine, paracrine, and hormonal agents (Fig. 49-1). Neurocrine agents are released from nerve ter­ minals and reach their targets via synaptic diffusion (e.g., acetylcholine [ACh], gastrin-releasing peptide [GRP], and vasoactive intestinal peptide [VIP]). Paracrine agents are released in proximity to their targets and reach them via diffusion (e.g., histamine and somatostatin). Hormones are released into the circulation and reach their targets via the bloodstream (e.g., gastrin). Gastric mucosal integrity depends on a delicate balance between secretion of aggressive (e.g., acid and pepsin) and defensive (e.g. bicarbonate and mucin) substances (Fig. 49-2).19 When mucosal defense mechanisms are over­ whelmed, ulceration may occur. In order to reap the benefits of acid without untoward effects, gastric exocrine and endocrine secretion must be precisely regulated. This is

Measurement of Gastric Acid Secretion  827 Indications for Secretory Testing  827 Methods for Measuring Acid Secretion  827 Basal Acid Output  827 Maximal Acid output and Peak Acid Output  827 Sham Feeding–Stimulated Acid Output  828 Meal-Stimulated Acid Output  828 Diseases Associated with Increased Gastric Acid Secretion  828 Pepsinogen Secretion  829 Gastric Lipase Secretion  829 Intrinsic Factor Secretion  829 Bicarbonate Secretion  830 Mucus Secretion  831

accomplished by a highly coordinated interaction among a multitude of neural, paracrine, and hormonal pathways.

FUNCTIONAL ANATOMY The stomach consists of three anatomic (fundus, corpus or body, and antrum) and two functional (oxyntic and pyloric gland) areas (Fig. 49-3). The oxyntic gland area, the hall­ mark of which is the oxyntic (oxys, Greek for acid) or parietal cell, comprises 80% of the organ (fundus and corpus). The pyloric gland area, the hallmark of which is the G or gastrin cell, comprises 20% of the organ (antrum). The human stomach contains approximately 1 × 109 parietal cells and 9 × 106 gastrin cells.20 There is debate as to whether the cardia, a transition zone of 0 to 9 mm between the squamous mucosa of the esophagus and the oxyntic mucosa of the stomach, exists as a normal anatomic structure or develops as a result of abnormal reflux. Autopsy and endoscopic studies suggest that cardiac mucosa is absent in more than 50% of the general population.21 Gastric anatomy is discussed in greater detail in Chapter 47. The glandular area is organized in vertical tubular units that consist of an apical pit region, an isthmus, and the actual gland region that forms the lower part of the unit; the latter consists of a neck and a base (Fig. 49-4). The progeni­ tor cell of the gastric unit, located in the isthmus, gives rise to all gastric epithelial cells. In the oxyntic gland area, the mucus-producing pit cells migrate upward from the pro­ genitor cell toward the gastric lumen. Acid-secreting pari­ etal cells migrate downward to the middle and lower regions of the gland; those at the bases are less active acid secretors. The turnover time for parietal cells is 54 days in mice and 164 days in rats.20 Chief (zymogenic) cells predominate at the base and secrete pepsinogen and leptin22; the latter is

817

Section VI  Stomach and Duodenum H+

ECL cell (Histamine)

Gastrin cell

M

+

H+

H+

Parietal cell oo

d

ve

ss

el H2

Bl

818

3

+

K-2

CC

ACh

Neurocrine pathway

+

SST − R2

D cell (SST)

Hormonal pathway

Paracrine pathway

Figure 49-1.  Neural, hormonal, and paracrine pathways directly regulating parietal cell acid (H+) secretion. Left: Acetylcholine (ACh), released from postganglionic intramural neurons within the oxyntic mucosa, stimulates the parietal cell directly via M3 receptors coupled to release of intracellular calcium. Center: Gastrin, released from antral G cells, travels in the circulation to reach the parietal cell and directly activates cholecystokinin-2 (CCK-2) receptors coupled to release of intracellular calcium. Right: Histamine, released from oxyntic enterochromaffin-like (ECL) cells, diffuses to the parietal cell and directly activates histamine (H2) receptors coupled to generation of cyclic adenosine monophosphate (cAMP). Somatostatin (SST), released from oxyntic D cells, diffuses to the parietal cell and directly activates SSTR2 receptors coupled to inhibition of acid secretion. Ach, gastrin, and SST also have important indirect actions affecting acid secretion not shown here. +, stimulatory; −, inhibitory.

OFFENSE Toxins

Steroids NSAIDs Microbes

ROS

H+ Pepsin

Mucus HCO3−

Phospholipids Figure 49-2.  Gastroduodenal offense and defense. Mucosal integrity depends on a delicate balance between aggressive and defensive factors. When mucosal defense mechanisms are overwhelmed, ulceration may occur. CA, carbonic anhydrase; CGRP, calcitonin gene-related peptide; H2S, hydrogen sulfide; HCO3−, bicar­ bonate; NO, nitric oxide; NSAIDs, nonsteroidal anti-inflammatory drugs; ROS, reactive oxygen species.

Mucosal blood flow

Neutrophils Cytokines

H+ Pepsin

H+ Pepsin

H+ Pepsin

H+ Pepsin

Mucus HCO3−

Mucus HCO3−

Mucus HCO3−

Mucus HCO3−

Epithelial integrity

Regulatory peptides Traditional growth factors Cytokines

Nerves (NO; CGRP) Myofibrolasts Immune cells DEFENSE

H2S CA

Prostaglandins

Chapter 49  Gastric Secretion also present in parietal cells.23 Several distinct neuroendo­ crine cell types are contained within the gland but only some of their products have been assigned physiologic func­ tions (see Chapter 1). The cells include enterochromaffin (EC) cells (atrial natriuretic peptide [ANP], serotonin, and adrenomedullin), enterochromaffin-like (ECL) cells (hista­ mine), D cells (somatostatin and amylin), and A-like or Gr cells (ghrelin and obestatin).24-29 ECL cells constitute 66% of the neuroendocrine cell population in rats and 30% in

Esophagus Fundus

ia

rd

Ca

Corpus (body) Angularis incisura Lesser curve

Antrum Greater curve Duodenum G cell (gastrin)

Parietal cell (H+)

Figure 49-3.  Functional gastric anatomy. The stomach consists of three anatomic (fundus, corpus or body, and antrum) and two functional (oxyntic and pyloric gland) areas. The hallmark of the oxyntic gland area is the parietal cell. The hallmark of the pyloric gland area is the G or gastrin cell.

humans. Somatostatin-containing D cells possess neurallike cytoplasmic processes that terminate in the vicinity of parietal and ECL cells (see Fig. 49-4). The functional correlate of this anatomic coupling is a tonic paracrine restraint exerted by somatostatin on acid secretion directly as well as indirectly by inhibiting histamine secretion (Fig. 49-5).30-32 Somatostatin-containing D cells are also present in the pyloric gland area (see Figs. 49-4 and 49-5); in this region they exert a tonic paracrine restraint on gastrin secretion from G cells.33,34 The pyloric gland also contains EC cells (ANP and serotonin), A-like or Gr cells (ghrelin and obestatin), and endocrine cells containing orexin.26,35,36 The stomach is innervated by a neural network, the enteric nervous system (ENS), that contains intrinsic neurons, that is, neurons whose cell bodies are contained within the gastric wall (Figs. 49-6 to 49-8). The ENS, the third division of the autonomic nervous system (the other two being the sympathetic and parasympathetic), is often referred to as the “little brain” because it contains as many neurons as the spinal cord, approximately 106, and can function autonomous of central input (see Fig. 49-6).37 Nevertheless, the ENS receives input from and sends input to the central nervous system via sympathetic and parasym­ pathetic neurons. In rats and guinea pigs, most of the intrin­ sic neural innervation of the stomach originates in the myenteric plexus, located between the circular and longitu­ dinal muscle layers; the submucosal plexus in these species, adjacent to the mucosal layer, contains only a small number of neurons. Humans, in contrast, have a clearly defined submucosal plexus that regulates gastric secretion and con­ tains a variety of neurotransmitters (see Figs. 49-7 and 49-8). It should be noted that the vagus nerve is predominantly afferent, containing 80% to 90% afferent fibers and 10% to 20% efferent fibers. The efferent fibers arise from the dorsal motor nucleus of the brainstem. They are preganglionic and do not directly innervate parietal or neuroendocrine cells but rather synapse with postganglionic neurons of the ENS. The postganglionic neurons contain a variety of transmitters including ACh, GRP, nitric oxide (NO), VIP, and pituitary

OXYNTIC GLAND

Surface mucous cell Parietal cell Mucous neck cell Enterochromaffin-like cell (histamine)

PYLORIC GLAND

GASTRIC PIT (FOVEOLUS) ISTHMUS (PROGENITOR ZONE)

Mucous neck cell

NECK G cell (gastrin)

D cell (somatostatin) Chief cell (pepsinogen) Enterochromaffin cell (ANP)

Surface mucous cell

BASE

D cell (somatostatin)

Enterochromaffin cell (ANP) Figure 49-4.  Gastric gland anatomy. Somatostatin-containing D cells contain cytoplasmic processes that terminate in the vicinity of acid-secreting parietal and histamine-secreting enterochromaffin-like cells in the oxyntic gland area (fundus and corpus) and gastrin-secreting G cells in the pyloric gland area (antrum). The functional correlate of this anatomic coupling is a tonic paracrine restraint on acid secretion by somatostatin that is exerted directly on the parietal cell as well as indirectly by inhibiting histamine and gastrin secretion. ANP, atrial natriuretic peptide. (From Schubert ML, Peura DA. Control of gastric acid secretion in health and disease. Gastroenterology 2008; 134:1842-60.)

819

Section VI  Stomach and Duodenum H+ D cell (SST)

−

Gastrin SSTR2 cell

D cell (SST)

Parietal SSTR2 − cell

H2

− R2 SST

Antrum

+ Fundus

820

ECL cell (Histamine)

Figure 49-5.  Model illustrating the inhibitory actions of somatostatin (SST) on gastric acid secretion in the oxyntic gland area (fundus and body) and the pyloric gland area (antrum). SST-containing D cells are structurally and functionally coupled to their target cells: parietal, enterochromaffin-like (ECL), and gastrin cells. SST, acting via SSTR2 receptors, tonically restrains acid secretion. This restraint is exerted directly on the parietal cell as well as indirectly by inhibiting histamine secretion from ECL cells and gastrin secretion from G cells.

AUTONOMIC NERVOUS SYSTEM

Sympathetic division

Vagus nerve (preganglionic) Parasympathetic division

Enteric Division Myenteric plexus Submucosal plexus

Figure 49-6.  The autonomic nervous system consists of three divisions: sympathetic, parasympathetic, and enteric. The enteric division consists of the myenteric plexus, which regulates motility and the submucosal plexus, which regulates secretion. Although the enteric division can function auto­nomously, it receives input from and sends projections to the other divisions.

ACh

ACh

ACh

ACh

ACh

Intramural gastric nerves (postganglionic)

ENTERIC NERVOUS SYSTEM Myenteric plexus

Circular muscle ACh

GRP

NO

VIP

PACAP

Target cell G/D/ECL/parietal

Longitudinal muscle

Mucosa

Submucosal plexus

Figure 49-7.  The enteric nervous system contains intrinsic neurons, the cell bodies of which are contained within the gastric wall. The myenteric plexus, which innervates the circular and longitudinal muscle layers, regulates motility. The submucosal plexus, which innervates the mucosa, regulates secretion.

Figure 49-8.  Functional neural anatomy. The vagus nerve contains preganglionic neurons that synapse with postganglionic neurons within the wall of the stomach and that are part of the enteric nervous system. The postganglionic neurons contain a variety of transmitters including acetylcholine (ACh), gastrin-releasing peptide (GRP, or mammalian bombesin), nitric oxide (NO), vasoactive intestinal peptide (VIP), and pituitary adenylate cyclase–activating by the parietal cell polypeptide (PACAP). The postganglionic neurons regulate acid secretion by the parietal cell directly and/or indirectly by modulating the secretion of gastrin from G cells, somatostatin from D cells, and possibly histamine from enterochromaffin-like (ECL) cells.

TR

Somatostatin −

M

3

HISTAMINE

Histamine, produced in ECL cells by decarboxylation of L-histidine by histidine decarboxylase (HDC), stimulates the parietal cell directly by binding to H2 receptors coupled to activation of adenylate cyclase and generation of cAMP (see Fig. 49-9).43 Histamine also stimulates acid secretion indirectly by binding to H3 receptors coupled to inhibition of somatostatin release from oxyntic D cells, thus resulting in stimulation of histamine release and acid secretion (see Fig. 49-10).44,45 Gastrin, PACAP, VIP, and ghrelin stimulate, whereas somatostatin, CGRP, prostaglandins, peptide YY (PYY), and galanin inhibit histamine secretion.46,47 As dis­ cussed following, gastrin also exerts a direct proliferative effect on ECL cells. ACh has no direct effect on histamine secretion.48-50

Gas trin

AC

Ca++ Parietal cell

ATP

+

K-2

Parietal cells secrete hydrochloric acid at a concentration of approximately 160 mM or pH 0.8. Acid is thought to gain access to the lumen via channels in the mucus layer created by the relatively high intraglandular hydrostatic pressures generated during secretion, about 17 mm Hg.40 Acid facilitates the digestion of protein and absorption of iron, calcium, and vitamin B12 as well as prevents bacterial overgrowth, enteric infection, and possibly community acquired pneumonia.6-14,18,41 However, when levels of acid (and pepsin) overwhelm mucosal defense mechanisms, ulcers occur. To prevent such damage, gastric acid must be precisely regulated and produced according to need. This is accomplished by a highly coordinated interaction among a number of neural, hormonal, and paracrine pathways. These pathways can be activated directly by stimuli origi­ nating in the brain or reflexively by stimuli originating in the stomach after ingestion of a meal such as mechanical stimulation (e.g., distention) or chemical stimulation (e.g., protein and acid). The principal stimulants of acid secretion are (1) ACh, released from postganglionic enteric neurons (neurocrine), (2) gastrin, released from antral G cells (hormonal), and (3) histamine, released from oxyntic ECL cells (paracrine) (see Fig. 49-1; Fig. 49-9). These agents interact with specific G protein–binding receptors (M3, CCK-2, and H2, respectively) that are coupled to two major signal transduction pathways: intracellular calcium in the case of gastrin and ACh, and adenylate cyclase or adenosine 3′,5′-cyclic monophosphate (cAMP) in the case of histamine (see Fig. 49-9). There is evidence for potentiation (or synergism) between histamine and either ACh or gastrin, probably as a result of postrecep­ tor interaction between the two signaling pathways.42 The main inhibitor of acid secretion is somatostatin, released from oxyntic and antral D cells (paracrine) (see Figs. 49-1, 49-5, and 49-9). Each of these agents acts directly on the parietal cell as well as indirectly by modulating the secre­ tion of neuroendocrine cells (Fig. 49-10).

n

T SS

CC

ACID SECRETION: PARACRINE, HORMONAL, NEURAL, AND INTRACELLULAR REGULATION

Ga str i

H2

R2

+

+

Histamine +

− AC h

ECL cell

SS

D cell

CCK-2

adenylate cyclase–activating polypeptide (PACAP) (see Fig. 49-8).38 In the stomach, afferent nerve fibers containing calcitonin gene–related peptide (CGRP) are of extrinsic origin, that is, the cell bodies are located outside the stomach wall.39 Postganglionic neurons of the ENS regulate acid secretion directly, as is the case for ACh, and/or indirectly by modulating the secretion of gastrin from G cells, soma­ tostatin from D cells, and possibly histamine from ECL cells (see Fig. 49-8).

2

Chapter 49  Gastric Secretion

cAMP

H+,K+ATPase

Ca++

K+

Acid Figure 49-9.  Model illustrating parietal cell receptors and transduction pathways. The principal stimulants of acid secretion at the level of the parietal cell are histamine (paracrine), gastrin (hormonal), and acetylcholine (ACh; neurocrine). Histamine, released from enterochromaffin-like (ECL) cells, binds to H2 receptors that activate adenylate cyclase (AC) and generate adenosine 3´,5´-cyclic monophosphate (cAMP). Gastrin, released from G cells, binds to cholecystokinin-2 (CCK-2) receptors that activate phospholipase C (not shown) to induce release of cytosolic calcium (Ca++). Gastrin stimulates the parietal cell directly and, more importantly, indirectly by releasing histamine from ECL cells. ACh, released from intramural neurons, binds to M3 receptors that are coupled to an increase in intracellular calcium. The intracellular cAMP- and calcium-dependent signaling systems activate downstream protein kinases, ultimately leading to fusion and activation of H+,K+-ATPase, the proton pump. Somatostatin, released from oxyntic D cells, is the principal inhibitor of acid secretion. Somato­ statin, acting via the SSTR2 receptor, inhibits the parietal cell directly as well as indirectly by inhibiting histamine release from ECL cells. +, stimulatory; −, inhibitory.

GASTRIN

Gastrin, the main stimulant of acid secretion during meal ingestion, is produced in G cells of the gastric antrum and, in much lower and variable amounts, in the proximal small intestine, colon, and pancreas. Gastrin is synthesized as a large precursor molecule of 101 amino acids, which is con­ verted to progastrin (80 amino acids) by cleavage of the N-terminal signal peptide. Progastrin is further processed to yield peptides with C-terminal glycine, that is, G34gly and G17gly. The final processing step involves amidation to yield G34amide and G17amide. The plasma half-life of G34amide is 30 minutes and that of G17amide is three to seven minutes; they are metabolized primarily by the kidney and, in addition, by the intestine and liver.51,52 Conse­ quently, most gastrin in the circulation during fasting is G34, whereas after a meal it is G17. In patients with renal insuffiency as well as massive small bowel resection, fasting blood levels of G17 and G34 are elevated.53,54 It should be noted that the commercially available test substance penta­ gastrin is not a naturally occurring peptide but rather is a manufactured analog that contains the biologically active C-terminus sequence Trp-Met-Asp-Phe-NH2. Gastrin and cholecystokinin (CCK) belong to the same family of peptides and possess an identical carboxylterminal pentapeptide sequence (-Gly-Trp-Met-Asp-PheNH2). Two main classes of gastrin/CCK receptors have been characterized: CCK1 (formerly CCK-A) and CCK-2 (formerly CCKB or CCKB/gastrin). CCK1 receptors are specific for CCK, whereas CCK-2 receptors recognize both CCK and gastrin with high affinity. Gastrin, acting via CCK-2 receptors that

821

822

Section VI  Stomach and Duodenum VAGUS

CGRP

VIP ACh +

ACh GRP

−

+

D cell (SST) +

−

+

ACh

+

G cell (GASTRIN)

+ +

PARIETAL cell H2

−

+

+

−

−

D cell (SST)

+

H3 −

ECL cell (Histamine)

HP (Chronic) Acid HP (Acute)

ACh

ANTRUM

−

FUNDUS

Figure 49-10.  Model illustrating the neural, paracrine, and hormonal regulation of gastric acid secretion. Efferent vagal fibers synapse with intramural gastric cholinergic (ACh) and peptidergic (gastrin-releasing peptide [GRP] and vasoactive intestinal peptide [VIP]) neurons. In the fundus (oxyntic mucosa), ACh neurons stimulate acid secretion directly as well as indirectly by inhibiting somatostatin (SST) secretion, thus eliminating its restraint on parietal cells and histamine-containing enterochromaffin-like (ECL) cells. In the antrum (pyloric mucosa), ACh neurons stimulate gastrin secretion directly as well as indirectly by inhibiting SST secretion, thus eliminating its restraint on gastrin-containing G cells. GRP neurons, activated by luminal protein, also stimulate gastrin secretion. VIP neurons, activated by low-grade gastric distention, stimulate SST and thus inhibit gastrin secretion. Dual paracrine pathways link SST-containing D cells to parietal cells and to ECL cells in the fundus. Histamine released from ECL cells acts via H3 receptors to inhibit SST secretion. This serves to accentuate the decrease in SST secretion induced by cholinergic stimuli and thus augments acid secretion. In the antrum, dual paracrine pathways link SST-containing D cells to gastrin cells. Release of acid into the lumen of the stomach restores SST secretion in both the fundus and antrum; the latter is mediated via release of calcitonin gene-related peptide (CGRP) from extrinsic sensory neurons. Acute infection with Helicobacter pylori (HP) also activates CGRP neurons to stimulate SST and thus inhibit gastrin secretion. In duodenal ulcer patients who are chronically infected with HP, the organism or cytokines released from the inflammatory infiltrate inhibit SST and thus stimulate gastrin (and acid) secretion.

activate phospholipase C to induce release of intracellular calcium, stimulates the parietal cell directly and, more importantly, indirectly by releasing histamine from ECL cells (see Figs. 49-9 and 49-10).55,56 Gastrin regulates the secretion and synthesis of histamine in a biphasic manner. The first phase involves release of stored histamine. The second phase relates to the replenishment of histamine stores and involves an increase in HDC activity followed by an increase in HDC gene transcription.57 H2 receptor, HDC, and CCK-2 receptor knockout mice manifest decreased acid secretion, especially in response to gastrin.58-60 Although amidated gastrins had been thought to be the only forms with biological activity, glycine-extended gas­ trins may regulate the capacity of the parietal cell to respond to secretagogues, release histamine from ECL cells, and stimulate proliferation of colonic mucosa and colorectal cancers.61,62 ACh, GRP, secretin, β2/β3-adrenergic agonists, calcium, protein, and alcoholic beverages produced by fer­ mentation stimulate, whereas somatostatin, galanin, and adenosine inhibit gastrin secretion. In addition, at least two negative-feedback pathways, mediated via release of soma­ tostatin, regulate gastrin secretion. The first is activated by luminal acidity and involves sensory CGRP neurons (see Fig. 49-10). Low intragastric pH (high intragastric acidity) activates CGRP neurons that, via an axon reflex, stimulate somatostatin and thus inhibit gastrin secretion.63-65 Con­ versely, when intragastric pH rises (low intragastric acidity), for example, by administering antisecretory medications such as proton pump inhibitors (PPIs) or by developing gastric atrophy, somatostatin secretion is inhibited and patients develop hypergastrinemia. There is some evidence that bacterial overgrowth induced by hypochlorhydria may also contribute to hypergastrinemia.66 The second negativefeedback pathway involves a paracrine action whereby

gastrin directly stimulates somatostatin and thus attenuates its own secretion (see Fig. 49-10).67 Gastrin also functions as a trophic hormone to stimulate mucosal proliferation. CCK-2 receptors have been localized to the progenitor zone in oxyntic glands, and chronic hyper­ gastrinemia induces proliferation of ECL and parietal cells directly as well as indirectly via the autocrine or paracrine action of growth factors such as heparin-binding epidermal growth factor, amphiregulin, transforming growth factor-α, metalloproteinases, and regenerating islet-derived 1.68,69 Rats rendered hypergastrinemic by a PPI demonstrate a five-fold increase in the number of ECL cells and a 1.5-fold increase in the number of parietal cells.70 Gastrin acts directly on ECL cells to induce hyperplasia, dysplasia, and eventually neoplasia (carcinoids).71 In con­ trast to rodents, humans rarely develop carcinoid tumors in response to hypergastrinemia unless other factors are present such as chronic active gastritis or gastrinoma asso­ ciated with multiple endocrine neoplasia type 1 (see Chapter 31).72 Because ECL cells contain somatostatin subtype 2 receptors (SSTR2), somatostatin scintigraphy with 111indiumdiethylenetriamine pentaacetic acid [111In-DTPA]octreotide is the preferred imaging method to detect carcinoid tumors (see Chapter 31).73,74

ACETYLCHOLINE

ACh, released from postganglionic neurons whose cell bodies are located primarily in the submucosal (Meissner’s) plexus, stimulates the parietal cell directly as well as indirectly by inhibiting somatostatin secretion (see Fig. 49-10). The parietal cell muscarinic receptor is of the M3 subtype.75,76 Like CCK-2 receptors, M3 receptors are coupled to activation of phospholipase C with generation of inositol trisphosphate and release of intracellular calcium

Chapter 49  Gastric Secretion

SOMATOSTATIN

The main inhibitor of acid secretion is somatostatin. Soma­ tostatin is synthesized from a 92-amino acid preproso­ matostatin precursor molecule that is processed to yield somatostatin-14 and somatostatin-28. Somatostatin-14 is predominantly found in stomach, pancreatic islets, and enteric neurons, whereas somatostatin-28 is the major form in small intestine. The half-life of somatostatin-14 is 1 to 3 minutes, whereas the half-life of somatostatin-28 is about 15 minutes. In the stomach, somatostatin cells are closely coupled to their target cells (gastrin cells in the antrum, or parietal and ECL cells in the fundus/body) either directly via cytoplas­ mic processes or indirectly via the local circulation.28,79 The functional correlate of this anatomic coupling is a tonic restraint exerted by somatostatin on acid secretion from the parietal cell, histamine secretion from the ECL cell, and gastrin secretion from the G cell (see Figs. 49-5 and 49-10).30,31,34,80,81 Removing this restraint (i.e., disinhibition or elimination of the influence of an inhibitor), by activation of cholinergic neurons, is an important physiologic mecha­ nism for stimulating acid secretion (see Fig. 49-10). In the stomach, the actions of somatostatin are mediated primarily via the somatostatin subtype 2 receptor (SSTR2).82-84 Gastrin, GRP, VIP, PACAP, β2/β3-adrenergic agonists, secretin, ANP, adrenomedullin, amylin, adenosine, and CGRP stimu­ late, whereas ACh and interferon-γ inhibit somatostatin secretion. As mentioned, an increase in luminal acidity acts to attenuate acid secretion via a pathway involving release of somatostatin in the antrum and the fundus. The change in gastric somatostatin secretion can be demonstrated over a range of pH 3 to pH 5, which is within the range observed after ingestion of a meal (Fig. 49-11).85

MISCELLANEOUS PEPTIDES

Ghrelin, the natural ligand for the growth hormone secreta­ gogue receptor, is present in greatest concentrations in gastric oxyntic mucosa and is localized to A-like (or Gr) cells.86,87 Lesser amounts are present in the antrum, small intestine, and colon (see Chapter 1). Plasma ghrelin concen­ trations increase before meals and decrease postprandi­ ally.88 It is postulated that ghrelin triggers premeal hunger and promotes feeding. Its suppression after Roux-en-Y gastric bypass may, in part, contribute to weight loss.89 Most studies report that exogenously administered ghrelin stimu­ lates acid secretion.90,91 The stimulatory effect appears to involve the vagus nerve and histamine release because it is abolished by vagotomy and is associated with an increase in HDC messenger ribonucleic acid (mRNA).92,93 Orexin-A, derived from propro-orexin by posttranslational processing, is co-localized with gastrin in human pyloric mucosa.91,94 Intracerebroventricular and peripherally administered orexin-A stimulate gastric acid secretion.95 In rats equipped with gastric fistulas, an orexin receptor 1 antagonist inhibits basal and pentagas­ trin-stimulated acid secretion, implying that endogenous orexin-A stimulates acid secretion.94,95 ANP, CCK, secretin, neurotensin, glucagon-like peptide 1 (GLP-1), glicentin, oxyntomodulin, peptide YY, adreno­

180 160 140 Somatostatin secretion % of basal

(see Fig. 49-9).77 Alcoholic beverages produced by fermen­ tation stimulate gastric acid secretion and the effect may be mediated via activation of M3 receptors.78 ACh also stimulates acid secretion indirectly by activating M2 and M4 receptors on D cells coupled to inhibition of somato­ statin secretion, thus removing the tonic restraint exerted by this peptide on gastrin, ECL, and parietal cells (see Fig. 49-10).

120 100

_

ADD H+

ADD HCO3

80 60 40 20 0 4.8

4.4

4.0

3.6

3.2

pH luminal perfusate Figure 49-11.  Relationship between luminal pH and gastric somatostatin secretion. In isolated mouse stomach, addition of bicarbonate (HCO3−) to neutralize basal acid secretion or HCl to augment luminal acidity causes a corresponding change in somatostatin secretion. (From Schubert ML, Edwards NF, Makhlouf GM. Regulation of gastric somatostatin secretion in the mouse by luminal acid: A local feedback mechanism. Gastroen­ terology 1988; 94:317-22).

medullin, amylin, glucose-dependent insulinotropic poly­ peptide (GIP), leptin, epidermal growth factor, and interleukin-1β (IL-1β) inhibit acid secretion. The effect of each, except perhaps for IL-1β, is mediated via release of somatostatin.24-26,96 The term enterogastrone has been used to describe the intestinal factor or factors responsible for inhibiting acid secretion in response to nutrients in the intestine. Prime candidates include CCK, secretin, neuro­ tensin, GLP-1, glicentin, and oxyntomodulin because they are present in intestinal mucosa, released into the circula­ tion in response to luminal nutrients, and capable of inhibit­ ing acid secretion at “physiologic” concentrations.97-101 Although it is likely that enterogastrone activity represents the combined influence of several of these peptides, the strongest evidence favors CCK. CCK, produced in I cells in the proximal small intestine, is released by luminal protein and fat. The acid-inhibitory response to intraduodenal fat is abolished by pretreatment with a CCK-1 receptor antagonist in dogs, and the response is blocked in CCK-1 receptor knockout mice as well.102-105

PARIETAL CELL INTRACELLULAR PATHWAYS

In parietal cells, acid secretion is increased by activation of intracellular cAMP- and calcium-dependent signaling pathways that activate downstream protein kinases, ulti­ mately leading to fusion and activation of H+,K+-ATPase (the proton pump) with concomitant activation of luminal mem­ brane conductances for K+ and Cl− (see Fig. 49-9; Fig. 49-12). The H+,K+-ATPase actively pumps out H+ against a tremen­ dous concentration gradient (cell interior pH 7.4 or 40 nM; acid secreted at pH 0.8 or 160 million nM) in exchange for luminal K+. The energy required comes from adenosine triphosphate (ATP) produced by the parietal cell’s extensive

823

Section VI  Stomach and Duodenum HCO3− CI− 7

AE2

H+ + HCO3− 3Na+

H2CO3 CA CO2 + H2O

2K+

H+,K+-ATPase α-Subunit β-Subunit

CI− SL C2 6A

824

2CI− CI−

Tubulovesicle

Na+ NKCCI

K+

K+ Resting parietal cell

H+,K+ATPase

H+

K+

CI−

HCI Figure 49-12.  Model illustrating ion transport in the parietal cell. Acid secretion requires a functional H+,K+-ATPase as well as apical K+ and Cl− channels and basolateral transporters and/or channels for K+, Cl−, and HCO3−. Acid is produced from the hydration of CO2 to form H+ and HCO3−, a reaction catalyzed by cytoplasmic carbonic anhydrase (CA). In the presence of luminal K+, H+,K+-ATPase pumps H+ into the lumen in exchange for K+. Luminal K+ channels (KCNE2/KCNQ1 and ROMK [not shown]) recycle K+ across the luminal membrane. The source of intracellular K+ is the basolateral Na+,K+-ATPase and the sodium-2 chloride potassiumcotransporter-1 (NKCC1). For each H+ secreted, a HCO3− exits the cell across the basolateral membrane via the anion exchanger (AE2, or Slc4a2). Concurrently with H+, Cl− is extruded across the luminal membrane via an apical chloride channel. The sources of intracellular Cl− are AE2, NKCCl, and the SLC26A7 channel.

mitochondrial network. H+,K+-ATPase consists of an αsubunit that carries out the catalytic and transport function of the enzyme and contains sequences responsible for apical membrane localization106 as well as a β-subunit, which is heavily glycosylated, and protects the enzyme from degra­ dation and is necessary for trafficking to and from the luminal membrane.107 In the resting unstimulated state, H+,K+-ATPase activity is sequestered within cytoplasmic tubulovesicles. On stimula­ tion, there is a dramatic morphologic transformation as these vesicles fuse with the apical plasma membrane, result­ ing in a 6- to 10-fold increase in the membrane and the formation of the canalicular system (Fig. 49-13). Transloca­ tion of the H+,K+-ATPase into the canalicular membrane together with the presence of luminal K+ activates the enzyme.108 On cessation of secretion, the H+,K+-ATPase is retrieved from the apical membrane and the tubulovesicular compartment is reestablished. The precise mechanisms regulating trafficking are not known, but data suggest that it involves actin-based microfilaments, small GTPases, docking/fusion proteins, ezrin, and clathrin.109-111 Acid secretion requires not only a functional H+,K+ATPase but also apical K+ and Cl− channels and basolateral HCO3− and Cl− exchangers. Acid is produced from the hydra­ tion of CO2 to form H+ and HCO3−, a reaction catalyzed by cytoplasmic carbonic anhydrase (see Fig. 49-12). Because the H+,K+-ATPase is unable to pump H+ into the lumen without a parallel uptake of K+, sufficient quantities of K+ must be delivered to the lumen. This K+ recycling is accom­ plished by luminal KCNE2/KCNQ1 and ROMK (KCNJ1)

Stimulated parietal cell

Figure 49-13.  Model illustrating translocation and activation of H+,K+ATPase. In the resting state, H+,K+-ATPase is sequestered within cyto­ plasmic tubulovesicles and is inactive. On stimulation, the tubulovesicles move to and fuse with the apical membrane, forming an extensive canalicular system. Translocation of H+,K+-ATPase into the canalicular membrane together with the presence of luminal K+ activates the enzyme.

Benzimidazole Imidazole Benzene H N

Pyridine O

CH2 S Methylsulphinyl bridge N

N

Figure 49-14.  Structure of proton pump inhibitors (PPIs). PPIs consist of two heterocyclic moieties, a benzimidazole ring, and a pyridine, connected by a methylsulphinyl bridge. PPIs are weak bases (pKa 4-5) that accumulate and activate in acidic spaces within the body that have a pH less than 4. Once activated within the parietal cell canaliculus, the PPI binds covalently with certain cysteine residues within the α-subunit of the inserted H+,K+-ATPase.

potassium channels. KCNQ1 is a voltage-activated K+ channel, which, when modified by the small regulatory subunit, KCNE2, becomes voltage insensitive, constitutively open, and acid activated.108,112,113 ROMK may be regulated by the cystic fibrosis transmembrane conductance regulator (CFTR).108,112-114 The concentration of K+ in gastric juice (8 to 12 mM) exceeds plasma K+ by two- to four-fold. For each H+ secreted, an HCO3− ion exits the cell across the basolateral membrane via the anion exchanger 2 (AE2), Slc4a2 (see Fig. 49-12).115 As a result of this HCO3−/Cl− exchange, the pH within the parietal cell remains only slightly alkaline during acid secretion.116 Rapid entry of HCO3− from parietal cells into blood has been referred to as the alkaline tide. Some of this HCO3− may be taken up and secreted by surface epithelial cells. Concurrently with H+, Cl− is extruded across the luminal membrane via an apical chloride channel, the precise identity of which is not known. The sources of intra­ cellular Cl− are the basolateral anion exchanger 2 (AE2), sodium-2 chloride potassium-cotransporter-1 (NKCC1), and the SLC26A7 channel (see Fig. 49-12).117,118

PROTON PUMP INHIBITORS

Current PPIs (e.g., omeprazole, lansoprazole, rabeprazole, pantoprazole, and esomeprazole) consist of two heterocy­ clic moieties, a pyridine and a benzimidazole ring, con­ nected by a methylsulphinyl group (Fig. 49-14). They are weak bases (pKa 4 or 5) that concentrate in acidic spaces within the body that have a pH less than 4. The pKa of a

Chapter 49  Gastric Secretion disulfide bonds by reducing agents such as glutathione (15 hours in rat) could also play a role.123

PPI

PPI

PPI+ +

+

H ,K - Sulfenamide+ ATPase S-S-Sulfenamide+

Blood (pH 7.4)

Cytosol (pH 7.3)

Secretory canaliculus (pH 8 weeks) use of PPIs is discontinued; this may unleash or exacerbate GERD, particularly in patients with large hiatal hernias.170,171 Acid hypersecretion persists at least eight weeks and appears to be due to hypergastrinemia-induced increases in parietal and ECL cell masses.172 The reason the phenomenon does not occur in HP-positive individuals who stop taking PPIs may be due to the fact that HP as well as the cytokines produced by the inflammatory infiltrate inhibit acid secre­ tion and thus mask the rebound. With time, atrophy of oxyntic glands with loss of parietal cells may occur in patients chronically infected with HP, resulting in irrevers­ ible achlorhydria. Autoimmune gastritis is an inflammatory disorder of the oxyntic mucosa often associated with antiparietal cell autoantibodies directed against H+,K+-ATPase with sub­ sequent loss of parietal cells.173 H+,K+-ATPase is a major autoantigen in a subset of patients infected with HP and these antibodies may play a role in the subsequent develop­ ment of atrophic gastritis. It is postulated that antibodies are acquired due to molecular mimicry between HP lipopoly­ saccharide and H+,K+-ATPase, both of which contain Lewis epitopes.174 About 10% to 15% of patients chronically infected with HP have antral-predominant inflammation. These patients, who are predisposed to duodenal ulcer (see Chapters 50, 51, and 52), produce increased amounts of acid as a result of reduced antral somatostatin content and elevated basal and stimulated gastrin secretion (see Fig. 49-16).175-177 Gastrin stimulates histamine secretion from fundic/body ECL cells and induces ECL hyperplasia.178 The mechanism by which somatostatin secretion is decreased is not known but may

Chapter 49  Gastric Secretion involve cytokines induced by the inflammation and/or the production of Na-methyl histamine, a selective H3-receptor agonist, by HP.179,180 One may speculate that the H3-receptor agonist could diffuse across the antral mucosa to interact with H3 receptors on antral somatostatin cells, causing inhibition of somatostatin secretion, and, thus, stimulation of gastrin secretion.45 In addition, IL-8 and platelet activat­ ing factor are upregulated in HP-infected mucosa and are capable of stimulating gastrin release from isolated G cells.181,182

MEASUREMENT OF GASTRIC ACID SECRETION INDICATIONS FOR SECRETORY TESTING

Gastric secretory testing assesses the basal and maximal capacity of the stomach to produce acid. Clinically, its utility has diminished but it may assist in the diagnosis and management of patients with hypergastrinemia (e.g., gastri­ noma) and in the diagnosis of incomplete vagotomy in patients with postoperative recurrent ulcer. Demonstrating fasting acid secretion or an acidic fasting gastric pH excludes achlorhydria as a cause of elevated fasting serum gastrin concentration. Patients with gastrinoma (Zollinger-Ellison syndrome; ZES) demonstrate hypergastrinemia with ele­ vated basal acid output (see Chapter 32).

METHODS FOR MEASURING ACID SECRETION

Aspiration of gastric juice is the most widely used method for measuring acid secretion in humans. Traditionally, this is performed by positioning a nasogastric tube into the most dependent portion of the stomach of a fasted individual. Proper positioning may be verified fluoroscopically or by recovery of more than 90 mL after injection of 100 mL water. Gastric juice is collected by suction. When the tube is pro­perly positioned, only 5% to 10% of gastric juice escapes collection and enters the duodenum. Neutralization by bicarbonate and diffusion of tiny amounts of acid back into the mucosa result in a small underestimation of the true rate of secretion. More recently, an endoscopic technique has been described to measure acid secretion in patients with gastrinoma. In this technique, all gastric contents are aspirated and discarded and then a single 15-minute sample of gastric juice is collected under direct endoscopic visualization.183 The H+ concentration in a sample of gastric juice can be determined by one of two methods. First, the specimen

can be titrated in vitro with a base (e.g., NaOH). The millimoles (mmol) of base needed to titrate a volume of gastric juice to an arbitrary pH endpoint (e.g., 7) represents the “titratable” acidity in mmol per liter of the sample. The other method is to measure the pH of the sample with an electrode. Because pH electrodes measure H+ activity and not concentration, it is necessary to convert activity to concentration using a table of activity coefficients for H+ in gastric juice.184 Once the H+ concentration of the sample in mmol per liter is determined by either of these methods, it is multiplied by the volume of the sample in liters to determine the acid output during the collection period (e.g., mmol per hour or mmol per kilogram of body weight per hour).

BASAL ACID OUTPUT

Basal acid output (BAO) estimates resting acid secretion in the absence of intentional and avoidable stimulation. It is expressed as the sum of the measured acid output, expressed as mmol H+ per hour, for four consecutive 15-minute periods. The upper limit of normal for BAO is about 10 mmol H+ per hour in men and 5 mmol H+ per hour in women (Table 49-1).185 BAO fluctuates from hour to hour in the same person. The lowest BAO occurs between 6 and 11 am and the highest occurs between 2 and 11 pm. Variation is also related to cyclic gastric motor activity with increased BAO in late gastric phase III (migrating motor complex).186

MAXIMAL ACID OUTPUT AND PEAK ACID OUTPUT

Maximal acid output (MAO) and peak acid output (PAO) estimate the acid secretory response to an exogenous secre­ tagogue, usually pentagastrin (6 µg/kg subcutaneous or intramuscular or 6 µg/kg/hr continuous intravenous infu­ sion). Pentagastrin is a manufactured analog of gastrin that contains its biologically active C-terminus sequence. Pos­ sible side effects include flushing, nausea, abdominal pain, dizziness, and palpitations. MAO is the sum of acid output of four consecutive 15-minute collection periods, and PAO is calculated by multiplying by two the sum of the two highest outputs recorded in the four test periods. The expected range for MAO is 5 to 50 mmol H+ per hour and for PAO is 10 to 60 mmol H+ per hour. MAO and PAO are higher in men and in smokers; they correlate with pari­ etal cell mass (i.e., the total number of parietal cells). Typical results for MAO in healthy subjects and in disease are shown in Table 49-1.

Table 49-1  Typical Results of Gastric Secretory Testing in Health and Disease Basal Acid Output (mmol H+/hr)

Normal subjects   Men   Women Duodenal ulcer   Men   Women Gastric ulcer   Men   Women Gastrinoma   Both sexes

Maximal Acid Output (mmol H+/hr)

Average

Range

Average

Range

2.5 1.5

0-10 0-5

25 15

7-50 5-30

5.0 3.0

0-15 0-15

40 30

15-60 10-45

1.5 1.0

0-8 0-5

20 12

5-40 3-25

10-90

65

30-120

40

827

828

Section VI  Stomach and Duodenum SHAM FEEDING–STIMULATED ACID OUTPUT

The cephalic phase of acid secretion whereby the smell, sight, and thought of appetizing food, transmitted via the vagus nerve, stimulates acid secretion can be studied by sham feeding. Sham feeding, in which foods are chewed then spit out, increases acid secretion to about 50% of PAO. Thought and taste appear to play more important roles than sight and smell. Cholinergic and GRP neurons are involved because the response can be abolished by atropine or a selective GRP antagonist.187

MEAL-STIMULATED ACID OUTPUT

Continuous intragastric titration is primarily a research tool used to measure acid secretion in response to food in the stomach.188,189 It measures the cephalic and gastric phases of acid secretion. A double-lumen tube is placed in the most dependent part of the stomach and a homogenized meal buffered to pH 5.5 or 5 is infused into the stomach. Small volumes of gastric contents are sampled from one lumen, the pH is measured, and the contents are returned to the stomach. The second lumen is used to infuse sodium bicar­ bonate to maintain gastric pH at the meal pH. The amount of bicarbonate required to keep the pH of gastric contents constant is a measure of the postprandial acid secretory response. Rates of gastric acid secretion after eating increase rapidly and approach the PAO.

DISEASES ASSOCIATED WITH INCREASED GASTRIC ACID SECRETION Duodenal ulcer patients, as a group, manifest increased basal and stimulated gastrin and acid production (see Table 49-1).190 It is recognized that most cases of duodenal ulcer are due to infection with HP (Chapter 52) and that this infec­ tion is responsible for the perturbations in acid secretion observed in these patients. Pentagastrin-stimulated PAO, an indicator of functional parietal cell mass, is increased in HP-infected duodenal ulcer patients as is GRP-stimulated peak acid output, an indicator of the stomach’s functional response to endogenous gastrin.177,191,192 Suppression of somatostatin secretion by the infection may be the root cause for these changes (see Fig. 49-10; Fig. 49-16). Eradica­ tion of HP restores somatostatin as well as basal and stimu­ lated gastrin and acid secretion, over time, to normal in most individuals, thus providing a permanent cure for duo­ denal ulcer disease.176,177,191,193-195 In contrast to duodenal ulcer, gastric ulcer patients, as a group, exhibit normal or decreased basal and stimulated acid production (see Table 49-1), even though they too are often infected with HP. This suggests that altered gastric mucosal defense may be of primary pathophysiologic importance. Gastric ulcers have been classified according to their location and concomitant association with duodenal ulcer.196 Type I ulcers occur in the gastric body and are generally characterized by low acid secretion. These find­ ings may reflect a greater degree and more generalized mucosal inflammation of the oxyntic mucosa with reduced functional parietal cell mass. Type II ulcers occur in the antrum and are characterized by low, normal, or high acid secretion. Type III ulcers occur within 3 cm of the pylorus, commonly accompany duodenal ulcer, and are characterized by high acid output. Type IV ulcers occur in the gastric cardia and are characterized by low acid secretion.197 Accordingly, the more distant a gastric ulcer is from the pylorus the more likely acid secretion will be low.

A number of uncommon conditions are marked by gastric acid hypersecretion and subsequent ulceration (see Chapter 52). In patients with systemic mastocytosis, high histamine levels, as a consequence of increased numbers of mast cells, continuously stimulate parietal cells to secrete acid.198 When a portion of gastric antrum is retained in the afferent remnant after antrectomy with Billroth II anastomosis, it is bathed in alkaline secretions leading to decreased soma­ tostatin secretion, hypergastrinemia, increased acid produc­ tion, and anastomotic ulceration.65,85,199 Acid hypersecretion also can result from chronic hypercalcemia of any cause because calcium directly stimulates gastrin secretion from G cells and acid secretion from parietal cells.200,201 The best characterized acid hypersecretory condition is ZES,202-204 as discussed in Chapter 32. The BAO is almost always higher than 15 mmol/hr and the BAO/PAO ratio is usually 0.6 or greater (see Table 49-1). Gastrin, synthesized by the tumor, is secreted into the bloodstream, where it binds to CCK-2 receptors on acid-producing parietal and histamine-containing ECL cells to induce secretion as well as proliferation. The clinical correlate of the proliferation is rugal hypertrophy with prominent gastric folds. Diagnosis and treatment of gastrinoma are discussed in detail in Chapter 32. The basis of the secretin test to diag­ nose gastrinoma is that normally somatostatin cells in the antrum tonically restrain gastrin secretion from G cells. Secretin stimulates the G cell directly and, at the same time, inhibits the G cell indirectly by stimulating somatostatin secretion; the effect of the latter usually dominates and gastrin is not stimulated to a significant degree (Fig. 49-17).

+

Gastrin +

CCK-2 ECL cell (histamine) +

Secretin + + SST −

CCK-2 Parietal H2 cell

H+ Fundus and body (oxyntic mucosa)

G

Antrum (pyloric mucosa) Figure 49-17.  Zollinger-Ellison syndrome (ZES). Model illustrating the action of gastrin in oxyntic mucosa and secretin in pyloric mucosa of stomach in patients with ZES. Gastrin, synthesized and secreted by the gastrinoma into the bloodstream, acts via cholecystokinin-2 (CCK-2) receptors on acid-secreting parietal and histamine-secreting enterochromaffinlike (ECL) cells to increase acid secretion and induce cell proliferation. In the antrum, exogenous secretin (i.e., secretin stimulation test [see text and Chapter 32]) stimulates gastrin secretion directly and concomitantly inhibits gastrin secretion by stimulating somatostatin (SST) secretion resulting in little or no gastrin release. Because the gastrinoma does not contain functionally coupled SST cells, the effect of secretin in ZES patients is solely to stimulate gastrin secretion from the tumor. (From Hung PD, Schubert ML, Mihas AA. Zollinger-Ellison syndrome. Curr Treat Options Gastroenterol 2003; 6:163-70.)

Chapter 49  Gastric Secretion Because the gastrinoma does not contain functionally coupled somatostatin cells, the effect of secretin is solely stimulation of gastrin secretion from the tumor.205-207 Almost all gastrinomas contain somatostatin receptors and soma­ tostatin receptor scintigraphy (SRS) using [111In-DPTADphe1]-octreotide is considered the initial localization study of choice, with a 71% sensitivity and 86% specificity for primary tumors and 92% detection for metastatic disease.208,209 PPIs are the antisecretory therapies of choice and are able to control acid secretion and prevent complica­ tions in most patients with ZES.210

PEPSINOGEN SECRETION Pepsinogens, which belong to a family of enzymes called gastric aspartic proteases, are inactive polypeptide proen­ zymes known as zymogens. They are synthesized primarily in chief cells but also in mucous neck cells. Pepsinogens are converted in the gastric lumen by gastric acid to pepsins, which contain two active-site aspartate residues. Once this reaction begins, pepsins can autocatalyze the conversion of pepsinogens to pepsins.211 Pepsins are optimally active at pH 1.8 to 3.5, reversibly inactivated at pH 5, and irreversibly denatured at pH 7. Gastric acid not only provides an optimum pH for peptic activity but itself denatures dietary protein, making it more susceptible to peptic hydrolysis. Thus, acid and pepsin work in concert to promote digestion of dietary protein. As discussed, partially digested protein stimulates gastrin and thus acid secretion.142 More recent data suggest that pepsins may also be important for killing ingested bacteria.17,212 Pepsinogens have been electrophoretically separated into seven isozymogens. The five fractions (pepsinogen 1 to pep­ sinogen 5) that migrate toward the anode most rapidly at pH 5 are similar immunologically and are referred to as group I pepsinogens (PGI; old term, pepsinogen A).213 PGI is expressed in chief and mucous cells of the oxyntic mucosa. Migrating slightly behind the PGIs are two immu­ nologically similar isozymogens, pepsinogen 6 and pep­ sinogen 7, that are referred to as group II pepsinogens (PGII; old term, pepsinogen C). PGII, which represents approxi­ mately 20% of total pepsin content, is expressed in oxyntic and pyloric mucosa as well as in duodenal Brunner’s glands. The most important physiologic stimulant for pepsinogen secretion is ACh released from intramural cholinergic neurons. ACh, acting via both M1 and M3 muscarinic recep­ tors on chief cells, induces an increase in cytosolic calcium.214 Calcium, in turn, activates cytosolic kinases, phosphatases, and nitric oxide synthase that induce pep­ sinogen secretion.215 Other agents capable of stimulating pepsinogen secretion from chief cells via the calcium signal­ ing pathway include CCK, gastrin, and GRP.216-219 Agents that increase cAMP within chief cells, such as isoprotere­ nol, secretin, VIP, and histamine also augment pepsinogen secretion as do agents that activate tyrosine kinase such as epidermal growth factor and transforming growth factor-α.218 Inhibitors of pepsinogen secretion include somatostatin, neuropeptide Y, PYY, and IL-1β. Optimal pepsinogen secre­ tion requires a functional sodium-2 chloride potassium cotransporter-1 (NKCC1), a basolateral cotransporter respon­ sible for chloride uptake into secretory epithelia,118 includ­ ing the chief cell (and the parietal cell; see Fig. 49-12). It has been postulated that loss of the flushing action of neutral gastric secretion may impair pepsinogen secretion. Serum levels of PGI correlate with maximal acid output. A linear correlation exists between the loss of chief cells in

patients with oxyntic atrophy and serum PGI220,221; a serum PGI/PGII ratio of 2.5 or less has been used as a noninvasive test to detect gastric mucosal atrophy.5,222 Serum PGI is also increased in humans treated with PPIs, but the precise mechanism is not known.223 Both PGI and PGII are filtered and metabolized by the kidney, but serum PGI concentra­ tion is increased more than PGII concentration in patients with renal insufficiency.224,225

GASTRIC LIPASE SECRETION Gastric lipase, secreted by chief cells of the oxyntic mucosa, helps initiate the digestion of dietary triglycerides by hydro­ lyzing them to free fatty acids, diglycerides, and 2monoglycerides. The properties of gastric lipase are quite distinct from those of pancreatic lipase. Gastric lipase has a pH optimum of 4.5 to 5.5 (compared with 6.5 to 7.4 for pan­ creatic lipase) and does not require colipase. Furthermore, protection from peptic proteolysis by an N-glycosylated asparagine at residue 308 permits gastric lipase to retains its full activity in acidic gastric juice (pH 2) despite a high gastric juice peptic activity.226,227 Stimulants and inhibitors of gastric lipase are similar to those for pepsinogen secretion. Aging has been reported to decrease gastric lipase secretion, but data are controver­ sial.228 In humans, increasing the amount of lipid in the diet causes a corresponding increase in gastric lipase secretion into gastric juice.229 The amount of gastric lipase secreted after a meal is small relative to the amount of pancreatic lipase. However, the specific activity of gastric lipase is equal to or greater than that of pancreatic lipase. Thus, gastric lipase is capable of digesting 10% to 25% of dietary triglyceride.230 In patients with chronic pancreatitis, gastric lipase secretion is increased three- to four-fold and can partly, but incompletely, compensate for loss of pancreatic lipase (Fig. 49-18).231 A feedback mechanism exists whereby fat in the small intestine inhibits gastric lipase secretion by a humoral mechanism, with GLP-1 a prime candidate to be the mediator.232 Orlistat (tetrahydrolipstatin) is a lipophilic compound derived from lipstatin produced by Streptomyces toxitricini and is a potent inhibitor of all lipases.233 Orlistat reacts covalently with the catalytic serine residue of lipases.234 Orlistat decreases lipid digestion by about 35% and has been used clinically to induce modest weight loss.235 Adverse gastrointestinal events include loss of lipid-soluble vitamins, accelerated gastric emptying, fecal urgency, abdominal discomfort, flatus with discharge, and fecal urgency (see Chapter 6).236,237

INTRINSIC FACTOR SECRETION Intrinsic factor (IF), a 50-kd glycoprotein secreted by human parietal cells and, to a lesser degree, chief cells, is necessary for the absorption of cobalamin (vitamin B12).238 Although all stimulants (e.g., gastrin, histamine, and ace­ tylcholine) and inhibitors (e.g., somatostatin) of gastric acid secretion discussed previously have similar effects on IF secretion, the secretion of IF is not coupled to acid secre­ tion. PPIs, for example, inhibit acid but have no significant effect on basal or stimulated IF secretion.239 Several reports indicate, however, that chronic use of PPIs may result in low serum levels of cobalamin, probably as a result of impaired acid-facilitated release of cobalamin from

829

Section VI  Stomach and Duodenum 365% of mean control values

2% of mean control values

food.240,241 The recommended daily allowance of cobalamin is 2 µg/day and total body stores are 2.5 mg. Cobalamin deficiency due to antisecretory therapy is therefore rare because acid secretion is not completely suppressed and because the body contains relatively large stores.10,242,243 The delivery of cobalamin from food to tissues begins with release of cobalamin from dietary protein by the pHdependent activity of pepsin, followed by binding of cobala­ min to two binding proteins that are secreted into gastric juice: IF and haptocorrin (R binder).244,245 Haptocorrin is also secreted in saliva and bile. Haptocorrin binds cobalamin more avidly than IF in the acidic stomach and, therefore, most cobalamin initially becomes attached to haptocorrin. In the duodenum, cobalamin is released from haptocorrin by pancreatic trypsin and the free cobalamin then binds to IF. IF-cobalamin complexes are resistant to pancreatic pro­ teolysis and eventually bind to a specific receptor on the distal ileal mucosa. This receptor, cubilin, is expressed in clefts between microvilli and mediates endocytosis of the IF-cobalamin complex.246 Once within the ileal enterocyte, IF is degraded by lysosomal enzymes and cobalamin binds to transcobalamin. The cobalamin-transcobalamin complex is released into the circulation and from there enters cells by receptor-mediated endocytosis. Once within cells, cobalamin is dissociated from its transport protein and converted to its active forms, methylcobalamin and 5-deoxyadenosyl cobalamin. The active forms serve as coenzymes for methionine synthase and methylmalonylcoenzyme A mutase, enzymes involved in methylation of homocysteine to methionine and the catabolism of branchedchain amino acids and odd-chain fatty acids in mitochon­ dria, respectively.247 When radiolabeled cobalamin is administered orally after a large dose of nonradioactive cobalamin is given parenter­ ally, patients with IF deficiency excrete much lower amounts of radioactive cobalamin in a 24-hour urine collection than do normal controls (Schilling test, part I). If IF is administered orally together with radioactive cobalamin to IF-deficient patients, urinary radioactive cobalamin excre­ tion normalizes (Schilling test, part II). In addition to IF

50 40 30 20 10 0

80 Lipolytic activity (%) in duodenal contents

60

Human pancreatic lipase outputs (mg)

Figure 49-18.  Relative outputs of human gastric lipase (left graph) human pancreatic lipase (middle graph), and lipolytic activity in duodenal contents (right graph) in healthy volunteers (violet columns) and patients with chronic pancreatitis with exocrine insufficiency (CP; beige columns). Note that despite the increase of human gastric lipase output in patients with CP, the overall magnitude of the rise is insufficient to fully correct the reduction in duodenal lipolytic activity. (From Carrière F, Grandval P, Renou C, et al. Quantitative study of digestive enzyme secretion and gastrointestinal lipolysis in chronic pancreatitis. Clin Gastroenterol Hepatol 2005; 3:28-38.)

70

31% of mean control values

600

80 Human gastric lipase outputs (mg)

830

500 400 300 200 100

70 60 50 40 30 20 10 0

0 Healthy volunteers Patients with CP

deficiency, cobalamin malabsorption may result from achlorhydria or hypochlorhydria (reduced peptic hydroly­ sis of cobalamin from food protein), bacterial overgrowth (cobalamin competed for by bacteria), pancreatic insuffi­ ciency (impaired tryptic cleavage of haptocorrin-cobalamin complex), ileal receptor defect (cubilin mutation), and ileitis or ileal resection (absent IF-cobalamin absorptive site).248-250 Secretion of IF far exceeds the amount necessary for cobalamin absorption. Thus, in most patients with hypo­ chlorhydria, continued IF secretion in low amounts is suf­ ficient to prevent cobalamin deficiency. Patients with pernicious anemia, however, which affects 2% of indivi­ duals older than the age of 60, do develop cobalamin defi­ ciency.251 The pathology involves a chronic inflammatory, mainly lymphocytic, infiltrate of the oxyntic mucosa accom­ panied by loss of parietal and chief cells. The pathogenesis involves proinflammatory TH1 CD4 T cells directed toward the α- and β-subunits of the parietal cell H+,K+-ATPase, as well as circulating antibodies directed against H+,K+-ATPase or IF.252 Chronic infection with HP may play a primary role in the immunologic response. It is proposed that HP-induced inflammation results in breakdown of tolerance for self antigens such as H+,K+-ATPase in genetically susceptible individuals or that antibodies are acquired due to molecular mimicry between HP lipopolysaccharide and H+,K+-ATPase, both of which contain Lewis epitopes.173,174 Antibodies directed against H+,K+-ATPase are found in 90% of patients with pernicious anemia but the incidence of these anti­ bodies may decrease to about 55% to 80% with progression of autoimmune gastritis presumably because of disappear­ ance of HP and the loss of antigenic drive.252-254

BICARBONATE SECRETION Regulation of gastric bicarbonate (HCO3−) secretion has been studied much less intensively than has duodenal HCO3− secretion. The precise function of gastric HCO3− secretion is

Chapter 49  Gastric Secretion uncertain given the overwhelming simultaneous secretion of H+. Nevertheless, HCO3− secretion has been implicated in the formation of a protective pre-epithelial alkaline layer (see Fig. 49-2). Measurement of gastric HCO3− secretion has been impeded by the presence of considerable H+ secretion, necessitating the use of potent antisecretory compounds or measurement methods unaffected by the presence of acid. One means of measuring bulk HCO3− secretion is with the use of inline pH and CO2 electrodes, in which HCO3− concentration is calcu­ lated using the Henderson-Hasselbach equation.255,256 An electrophysiologic method has been used to measure alka­ line secretion in the lumen of individual gastric glands isolated from the frog Rana esculenta.257 Gastric HCO3− secretion is an energy-dependent process. The finding that there is virtually no change in gastric electrical potential difference during HCO3− secretion sug­ gests that HCO3− transport takes place via an electroneutral ion exchange mechanism, probably an exchange of HCO3− for Cl− at the luminal surface.258 Although several candidate anion exchangers, solute carriers, and anion transporters have been localized to gastric surface cells including AE2, AE4, Slc4a2, Slc4a4, PAT1, and Slc26a6, there is little evidence that gastric surface cells actually secrete HCO3−.259-263 The cell responsible for HCO3− secretion is not known. The source of some of the HCO3− secreted during H+ secretion may actually be the parietal cell. As discussed previously, for each H+ secreted, a HCO3− ion exits the parietal cell across the basolateral membrane via the anion exchanger 2 (AE2), Slc4a2 (see Fig. 49-12).115 This HCO3− may alkalinize the blood that perfuses surface epithelial cells, be taken up by sodium bicarbonate cotransporters (NBC1 and NBC2), and then be secreted by epithelial cells in an effort to protect them from luminal acid.264 The pari­ etal cell, however, may not be the only source of HCO3− for surface cells, as marked inhibition of gastric H+ secretion by PPIs does not significantly diminish gastric HCO3− secre­ tion in patients with duodenal ulcer.265 Prostaglandin E2 analogs stimulate gastric HCO3− and mucus secretion266,267 and blockade of endogenous prosta­ glandin synthesis reduces gastric HCO3− secretion.268 In rats and mice, the prostaglandin E receptor subtype involved in stimulation of gastric HCO3− secretion is EP1.269,270 Gastric mucosal prostaglandin synthesis and HCO3− secretion decline in older adults.271,272

MUCUS SECRETION A firmly adherent viscous mucus gel overlies the gastric surface. It is composed of 95% water and 5% extensively cross-linked mucin glycoproteins that are products of MUC genes.273,274 Ultrastructural studies reveal alternating layers of two distinct mucin classes, MUC5AC (secreted by surface and pit area epithelial cells) and MUC6 (secreted from neck and gland cells).275-278 Observations that the gastroprotective compound, geranylgeranylacetone (GGA), increases MUC6 expression in rat gastric mucosa279 and that lafutidine, a gastroprotective compound used clinically in Japan, increases mucus thickness and mucin content in humans280 suggest that the mucus gel may contribute to mucosal defense.273,281,282 Furthermore, electron microscopy studies demonstrate that HP accumulates within and disrupts the MUC5AC-enriched gel layer.277,283,284 MUC5AC secretion is increased in first-degree relatives of subjects with gastric cancer infected with HP, suggesting that mucus secretion may be a marker for a more severe inflammatory response to the organism.285

Mucus gel thickness can be measured continuously and noninvasively in living rodents by alternately focusing between fluorescently labeled surface epithelial cells and the mucus gel surface, as delineated by carbon particles or by fluorescent microspheres.286-288 A pH gradient at the gastric mucosal surface has been observed in a variety of species including humans.289-292 In most cases, the gradient is relatively alkaline at the tissue surface and gradually more acidic at distances further from the surface. The gradient is due to active bicarbonate secre­ tion and is considered a defense against luminal acid.290,293 Although most measurements using microelectrodes have reported values near neutrality at the gastric mucosal surface, more recent measurements using pH-sensitive fluo­ rescent dyes with ex vivo confocal microscopy report a surface pH near pH 4 in guinea pigs and frogs.294,295 The observation that the steady-state surface pH gradient extends beyond the thickness of the mucus gel layer suggests that the unstirred layer formed at the interface between the mucus and the aqueous lumen or the interface between the epithelial surface and the luminal contents may also play a role in mucosal defense by restricting mixing of molecules within the unstirred layer.283,293,296 Although it has been suggested that gastric mucus may physically retard diffusion of protons, the evidence for this is equivocal. In vitro measurements of proton and bicar­ bonate diffusion have yielded diffusion coefficients that are up to 10 times slower in isolated mucus compared with saline solution (≈0.5 − 2 × 10−5 cm2/sec).287,297 However, these diffusion coefficients in mucus are comparable to those found in saline for ions such as Na+, K+, and Cl−. Even accepting the slowest reported diffusion of protons, one-third of protons will theoretically diffuse 80  µm or greater within 10 seconds, a distance greater than the average mucus thickness in rats and mice.281,290,296 There are also concerns regarding diffusion measurements per­ formed in vitro as nondestructive removal of the tightly adherent mucus is extremely difficult, mucus structure and function is sensitive to environmental conditions, and measurement instruments may produce artifactual unstirred layers.298,299 Trefoil factor (TFF) peptides are cosecreted with mucins.300 These are 7- to 12-kd peptides sharing a common structure of three internal disulfide bonds that yield the signature “trefoil” structure of three internal loops and are designated TFF1, TFF2, and TFF3. TFFs are markedly pepsin resistant and able to survive intact in the gastric lumen301; the con­ centration of TFF2 in rat gastric mucus has been estimated at 10 µm.302,303 The distribution of each trefoil peptide in the normal GI tract is distinct. TFF1 is stored in gastric pit and surface mucous cells, TFF2 is present in gastric gland mucous cells, and TFF3 is stored in intestinal goblet cells.303-307 The relative abundance of TFF1 and TFF2 expression is reciprocally regulated by gastrin, H+,K+ATPase, and inflammation.308 The localization and coordinated secretion of trefoil peptides with mucins suggest that they too may be involved in mucosal defense.305,309,310 In support of this notion (1) increased TFF1 expression is observed after administration of gastroprotective agents311; (2) addition of TFF2 to mucin solutions significantly increases viscosity and elasticity in vitro and in vivo312-313; (3) chronic treatment with PPIs increases TFF2 concentration in gastric secretions coincident with promoting repair and preventing injury in response to luminal noxious agents314; and (4) TFF2 -/- knockout mice exhibit shortened gastric glands with decreased epithelial migration, increased net acid secretion, and a four-fold increase in number

831

832

Section VI  Stomach and Duodenum of lesions after a 12 hour-exposure to the nonselective cyclooxygenase (COX) inhi­bitor indomethacin.315

Acknowledgment

The authors thank Mary Beatty-Brooks for the artwork and Mark Feldman, John Walsh, Andrew Soll, and Gabriel Makhlouf, pioneers in the field of gastric exocrine and endocrine secretion, for their encouragement, support, and friendship.

KEY REFERENCES

Carrière F, Grandval P, Renou C, et al. Quantitative study of digestive enzyme secretion and gastrointestinal lipolysis in chronic pancrea­ titis. Clin Gastroenterol Hepatol 2005; 3:28-38. (Ref 231.) Flemström G, Isenberg JI. Gastroduodenal mucosal alkaline secretion and mucosal protection. News Physiol Sci 2001; 16:23-8. (Ref 258.) Fossmark R, Johnsen G, Johanessen E, Waldum HL. Rebound hyperse­ cretion after long-term inhibition of gastric acid secretion. Aliment Pharmacol Therapeut 2005; 21:149-54. (Ref 172.) Gillen D, Wirz AA, Neithercut WD, et al. Helicobacter pylori infection potentiates the inhibition of gastric acid secretion by omeprazole. Gut 1999; 44:468-75. (Ref 169.) Hatlebakk JG, Katz PO, Camacho-Lobato L, Castell DO. Proton pump inhibitors: Better acid suppression when taken before a meal than without a meal. Aliment Pharmacol Therapeut 2000; 14:1267-72. (Ref 122.)

Heitzmann D, Warth R. No potassium, no acid: K+ channels and gastric acid secretion. Physiology 2007; 22:335-41. (Ref 108.) Jain RN, Samuelson LC. Differentiation of the gastric mucosa: Role of gastrin in gastric epithelial cell proliferation and maturation. Am J Physiol Gastrointest Liver Physiol 2006; 291:G762-5. (Ref 68.) Prinz C, Zanner R, Gratzl M. Physiology of gastric enterochromaffin-like cells. Ann Rev Physiol 2003; 65:371-82. (Ref 46.) Manela FD, Ren J, Gao J, et al. Calcitonin gene-related peptide modu­ lates acid-mediated regulation of somatostatin and gastrin release from rat antrum. Gastroenterology 1995; 109:701-6. (Ref 65.) Moss SF, Legon S, Bishop AE, et al. Effect of Helicobacter pylori on gastric somatostatin in duodenal ulcer disease. Lancet 1992; 340:9302. (Ref 176.) Schubert ML, Edwards NF, Makhlouf GM. Regulation of gastric soma­ tostatin secretion in the mouse by luminal acid: A local feedback mechanism. Gastroenterology 1988; 94:317-22. (Ref 85.) Wang TC, Dockray GJ. Lessons from genetically engineered animal models I. Physiological studies with gastrin in transgenic mice. Am J Physiol Gastrointest Liver Physiol 1999; 277:G6-11. (Ref 60.) Whited KL, Thao D, Lloyd KCK, et al. Targeted disruption of the murine CCK1 receptor gene reduces intestinal lipid-induced feedback inhibi­ tion of gastric function. Am J Physiol Gastrointest Liver Physiol 2006; 291:G156-62. (Ref 104.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

50 Helicobacter pylori David A. Peura and Sheila E. Crowe

CHAPTER OUTLINE Epidemiology  833 Pathogenesis  834 Colonization and Virulence Factors  834 Host Response to Infection  835

Helicobacter pylori are unique bacteria ideally suited to live in the acidic environment of the human stomach. Their spiral shape and multiple unipolar flagella allow them to move freely through the gastric mucous layer, where they remain protected from low gastric pH.1 Organisms produce large amounts of urease, an enzyme that hydrolyzes urea to alkaline ammonia and CO2. This permits the bacteria to further control the pH of their microenvironment. Urease is also the basis of clinical diagnostic tests (urea breath test and rapid urea biopsy tests) for infection. H. pylori remain difficult and tedious to culture because they grow slowly and require specialized culture media and a controlled microaerophilic environment. When they gain access to a human host, H. pylori recognize and attach to various gastric epithelial surface receptors, thereby chronically colonizing the mucosa, disrupting cell function, inciting an intense local inflammatory and systemic immune response, and altering acid secretory physiology.2,3 The ultimate clinical manifestations of H. pylori infection include gastric and duodenal ulcer, gastric mucosa–associated lymphoid tissue (MALT) lymphoma, and adenocarcinoma; yet most infected individuals remain asymptomatic for life despite developing chronic histologic gastritis.1,2,4,5 What factors determine why some develop disease in response to infection and others do not remains a mystery, but host genetics, bacterial characteristics, and environmental features undoubtedly can influence clinical outcome.5,6 Research focusing on epidemiology, pathogenesis, management, and prevention of H. pylori infection and associated clinical conditions continues to be fueled by the tremendous worldwide prevalence of infection (especially in less-developed countries), the huge health and economic burden imposed by ulcer disease and gastric cancer, and heightened prominence and awareness of the bacteria accompanying Drs. Robin Warren and Barry Marshall’s receipt of the 2005 Nobel Prize in Physiology or Medicine for seminal contributions to the field.7

EPIDEMIOLOGY Helicobacter pylori infection remains one of the most common chronic bacterial infections in humans. Estimates suggest that more than 50% of the world’s population is infected with the bacterium and genetic sequence analysis proposes that humans have been infected for

Conditions Arising from Infection  838 Diagnosis  839 Treatment  841

more than 58,000 years at a time when they first migrated from Africa.8 While H. pylori have been demonstrated worldwide in individuals of all ages, infection is more common and acquired at an earlier age in developing countries compared with industrialized nations.9,10 In developing nations, the majority of children become infected before the age of 10, and during early childhood spontaneous elimination of bacteria and subsequent reinfection is quite common. Infection persists in older children and adults so that in the developing areas of the world H. pylori prevalence can reach more than 80% by age 50. In developed countries, such as the United States, young children can also acquire H. pylori, but usually before age 5.11 Spontaneous clearance often occurs and there is less chance of reinfection; thus, persistent childhood infection is much less frequently seen than in less-developed countries.9 In fact, serologic evidence of H. pylori is uncommon in children before age 10, but rises to 10% in adults between 18 and 30 years of age and further increases to 50% in those 60 or older.9 This increased prevalence of infection with age was initially thought to represent continuing acquisition throughout adult life. However, new adult infection and reinfection are quite uncommon, especially in developed countries. Epidemiologic evidence supports childhood-acquired infection even in developed nations, so the frequency of H. pylori infection for any age group in any locality reflects that particular birth cohort’s rate of bacterial acquisition early in life.9 In the United States, within any age group, infection appears to be more common in blacks than whites.12 Also Hispanic immigrants and their first-generation children are more likely to harbor H. pylori than their second-generation relatives.13 These differences probably relate to factors early in life that are linked to acquiring infection. The risk of acquiring H. pylori is associated with living conditions and the family’s socioeconomic status during one’s childhood.9 Housing density, crowded conditions in the home, number of siblings, sharing a bed, and lack of hot or running water have been linked to higher rates of infection. In Japan, the rapidly declining prevalence of H. pylori appears to parallel the nation’s postwar economic progress and improvement in hygiene and sanitation. Of the Japanese born before 1950, more than 70% are infected compared with 45% born between 1950 and 1960 and 25% born between 1960 and 1970.14 Presently, childhood infection in Japan is rare. Predictions based on a similar declining prevalence in the United States suggest that the organism could

833

834

Section VI  Stomach and Duodenum eventually become extinct here and in other areas of the world, which would affect infection-related illness.15 Twin studies support genetic susceptibility to H. pylori infection because monozygotic twins who were raised in different households have a greater concordance of infection than dizygotic twins also raised separately.16 However, twins growing up together have a higher concordance of H. pylori status than twins growing up separately, suggesting childhood environmental factors are also important for acquisition. Humans appear to be the major reservoir of H. pylori. However, domestic cats and captive primates and sheep can harbor these organisms,10 although it is possible that these animals actually acquired H. pylori from a human source. In the case of cats, isolation of viable bacteria from saliva and gastric juice suggests at least the possibility of transmission to humans.10 Especially in developing countries, contaminated water might serve as an environmental source of bacteria because the organism can remain viable for several days in water.17 Bacterial deoxyribonucleic acid (DNA) can be found in samples of municipal water from endemic areas of infection but whether viable H. pylori are present remains to be proven.5 In countries where infection is common, children who drink untreated stream water, eat uncooked vegetables, or swim in rivers and streams are more likely to harbor the bacteria, providing further indirect evidence of an environmental source of organisms. Person-to-person transmission of bacteria from fecal-oral, oral-oral, or gastro-oral exposure seems the most probable explanation for infection.5,10 Within-family clustering of infection (often with genetically identical strains of H. pylori) supports person-to-person transmission.9 Also infected individuals more often have infected spouses or children than uninfected individuals. Support for siblingto-sibling transmission comes from studies reporting that likelihood of infection correlated with number of children in the household and that younger children were more apt to be infected if older siblings were also infected.9 Motherto-child transmission is also quite likely.11,18 Fecal-oral transmission of bacteria is a possible mechanism by which H. pylori gain access to the human host. The bacterium can be cultured from diarrheal stools and vomitus, suggesting the potential for transmission.19 Exposure to an infected family member during an acute gastrointestinal illness, especially with vomiting, appears to be a risk factor for subsequent infection.20 How frequently bacteria are transmitted through oral-oral contact is not known. Although organisms can be identified in dental plaque and saliva, the prevalence is low and it is questionable if the mouth serves as a source or reservoir for H. pylori.21 Also dentists and oral hygienists who continually have occupational exposure to dental plaque and oral secretions do not have a higher prevalence of H. pylori.22 In developed countries, spousal transmission of infection also appears to be uncommon. Infected gastric secretions can serve as a source of bacterial transmission. Iatrogenic infection has occurred during the use of a variety of inadequately disinfected gastric devices, endoscopes, and endoscopic accessories.10 Also gastroenterologists and nurses appear to be at greater risk for acquiring H. pylori, presumably due to occupational contact with infected gastric secretions.23 Mandated universal precautions, standardized equipment disinfection, and use of video-endoscopes that reposition the instrument channel away from the mouth should reduce such iatrogenic and occupational transmission. Natural transmission could occur through contact with infected vomitus during

an acute illness20 or with regurgitated material from an infected child. Such contact could explain the higher concordance of maternal/child H. pylori infection and the presumed child-to-child transmission that occurs in an infant daycare setting.24

PATHOGENESIS Specific genetic or phenotypic factors in infectious agents have been implicated as single causal factors in a variety of infectious diseases and associated outbreaks. However, H. pylori infection alone appears insufficient to fully explain the spectrum of diseases that is associated with chronic infection. Research over the past quarter century suggests that the pathogenicity of H. pylori depends on bacterial and host factors in addition to less well-defined environmental conditions. Virulence of this infectious pathogen is based on bacterial properties that allow colonization and adaptation to the gastric environment and a host response that contributes to the host physiologic and histologic changes.

COLONIZATION AND VIRULENCE FACTORS One of the interesting aspects of this pathogen is how it confers disease when the organism resides, for the most part, in the lumen. Studies describing the genome of two distinct strains of H. pylori have helped to advance our understanding of the ecology of the organism and the potential gene expression patterns that can affect the pathogenesis of disease.25,26 Importantly, bacterial genes expressed in gastric mucosa differ from the pattern observed in vitro,27 whereas exposure of the bacterium to low pH increases its expression of genes encoding proteins involved in the motility apparatus and of genes encoding urease.28 The urease helps the organism adapt to the gastric milieu as it retains optimal function at two different pH values: usually pH 7.2 and pH 3.29 H. pylori is further adapted to the gastric pH by producing the molecular machinery required to migrate rapidly to a more favorable environment below the mucus layer. Motility is one of the few H. pylori characteristics shown to be necessary for successful colonization of the host. H. pylori show a strict tropism for the gastric mucosa or intestinal sites in which there is gastric metaplasia. The corollary is also true, as H. pylori do not colonize epithelium in the stomach that has undergone intestinal metaplastic change, possibly due to the production of antimicrobial factors that select against colonization. This possibility is supported by the fact that H. pylori rarely colonize the deeper portions of the gastric glandular mucosa, where O-glycans that impair H. pylori growth are found.30 This concept is also supported by another study that shows H. pylori decreases the expression of the antibacterial molecule, secretory leukocyte protease inhibitor,31 thereby removing an element of the host response that would be detrimental to the persistent infection. Another important factor that controls colonization is the expression of receptors on host cells that allow H. pylori to bind. Lewis (Le) antigens expressed by host cells serve as a receptor for bacterial binding.32 Specific bacterial gene pro­ ducts, most notably BabA, act as the bacterial ligand for the Leb receptor.33 The Leb receptor may be located in the gastric mucus because strains bearing BabA also bind to the mucin MUC5AC.34 Some studies suggest that the babA2

Chapter 50  Helicobacter pylori genotype is more frequently associated with inflammation, duodenal ulcer and gastric cancer.2 However, observations that binding of H. pylori to epithelial cells freshly isolated from human gastric biopsy specimens is unaffected by the expression of Le antigen, that infection is not increased in subjects with the Leb phenotype, and that individuals who do not express Leb can clearly be infected with H. pylori cast doubt on the true role of Leb and BabA.2 Additionally, in one report, the majority of strains infecting the individuals studied did not induce ulcers or cancer, despite expressing the babA2 gene.35 H. pylori also bind to the molecular complex of invariant chain and class II human leukocyte antigens (HLAs) expressed on the surface of gastric epithelial cells.36 Class II major histocompatibility complex (MHC) molecules, with their expression increased by infection, were the first epithelial cell receptor for H. pylori demonstrated to directly affect signaling in host cells. Binding of urease to epithelial cells via class II MHC was sufficient to induce apoptosis.37 More recently, the gastric trefoil protein TFF1 was shown to serve as a receptor for H. pylori.38 This molecule is predominantly expressed in the gastric mucosa and found in association with gastric mucus. In genetically engineered mice with TFF1 deficiency, a spontaneous, antral adenoma develops, suggesting that this molecule provides an element of control over gastric epithelial cell growth.39,40 The pathogen-associated molecular receptors (PAMPS) have also been examined for their role in binding of H. pylori to the host epithelial cells. The Toll-like receptors (TLRs) are a family of PAMPS, with an apparent different specificity for various bacterial molecules.41 For example, TLR4 is able to recognize the LPS of many bacteria, with cytokines and H. pylori particles increasing the expression of TLR4.42 H. pylori LPS stimulates monocytes and gastric epithelial cell responses via TLR4.2 Other PAMPs, including TLR2, are also activated by highly purified H. pylori LPS and have even been described as being more important than TLR4.43,44 These receptors may bind bacterial products, and thereby enhance bacterial binding and cell signaling.42 TLR5 binds bacterial flagellins and, similar to TLR2, induces a signaling response that can trigger acute inflammation. H. pylori produce flagellin that binds TLR5 and activates a response in vitro.43 There is disagreement on which TLR is stimulated by H. pylori LPS or whether H. pylori flagellin is ever able to bind TLR5,45 and this may reflect the different conditions used for bacterial culture and LPS or flagellin purification. Further studies are needed to establish the significance of H. pylori binding to TLRs in the pathogenesis of infection. After H. pylori migrate to the gastric epithelium, the organism attaches to host cells and may damage them in order to obtain nutrients from the subsequent inflammatory exudate or transudate. A key interaction between the bacteria and gastric epithelium involves a segment of bacterial DNA referred to as the cag pathogenicity island (cag PAI). Genes within the cag PAI encode proteins that provide a type IV secretion apparatus (i.e., cagE) that allows bacterial macromolecules to translocate into the host cell (i.e., cagA).26,46 cag PAI plays an important role in the pathogenesis of gastritis in humans26,47 because H. pylori bearing the cag PAI are associated with increased interleukin-8 (IL-8) expression and inflammation in gastric mucosal biopsy specimens and increased IL-8 expression and apoptosis in vitro.48 Infecting gerbils with mutated strains lacking cagE reduces the severity of gastritis and the development of gastric ulcers, intestinal metaplasia, and gastric cancer compared with gerbils infected with the wild-type strain.48,49 Human studies in which duodenal ulceration occurred

more frequently in children carrying strains expressing cagE associated with higher levels of gastric IL-850 corroborate animal and in vitro studies. All strains of H. pylori possess the vacA gene, with more than half expressing the vacuolating cytotoxin (VacA), which attaches to epithelial cells via an interaction with protein-tyrosine phosphatases.51 Although the majority of the VacA is secreted, some may remain on the surface of the bacteria and serve as a ligand for bacterial attachment via this protein-tyrosine phosphatase receptor. Several studies have examined the structure and function of VacA and its association with disease.26,52,53 For example, mice deficient in protein-tyrosine phosphatase beta do not develop ulceration when exposed to VacA.54 Different vacA alleles have been classified in the 5′ signal region (s-region) and the middle region (m-region) of the vacA gene.52 The s-region is present as s1 (which can be further distinguished as s1a, s1b, s1c) or s2, whereas the m-region is present as m1 or m2. Production of VacA is designated by the allelic combination s1/m1 and s1/m2. Specific vacA alleles (s1 and m1) are associated with disease55 and the induction of epithelial cell apoptosis.56 The interaction between VacA and its receptor(s) appears to be important in the pathogenesis of gastroduodenal disease, whether it serves as a ligand for bacterial attachment or as a secreted virulence factor.

HOST RESPONSE TO INFECTION Increasing evidence suggests that the host response to H. pylori infection is an intrinsic component of the patho­ genesis of gastrointestinal disease (Chapters 29, 52, and 54). The possibility that the host response may play a direct role in gastric cancer is supported by the observation that heterogeneity in the regions of the genome that control the magnitude of inflammation is associated with cancer linked to H. pylori infection.57 Polymorphisms in the regions controlling IL-1β58 were shown to be associated with an increased incidence of hypochlorhydria and gastric cancer. This seminal observation has been replicated in other studies in which IL-1β polymorphisms were associated not only with gastric cancer but also a decrease in recurrence of duodenal ulcer.59 An increase in IL-1 may not only drive inflammation but also lead to a physiologic state known to precede gastric cancer development because IL-1 potently inhibits gastric acid secretion. Increased gastric IL-1, more severe gastritis, gastric atrophy, and greater colonization with H. pylori strains have been associated with gastric cancer.2 Other genes that regulate the magnitude of the inflammatory response, including IL-10, tumor necrosis factor-α (TNF-α), and IL-8, have also been associated with the sequence of events leading to cancer.60,61 It is apparent that epithelial cells play an integral part in the host response to H. pylori infection as well as being the target of infection. The epithelial response to infection is complex, as it is driven by several variables: bacterial virulence factors; the signaling linked to specific receptors that recognize the bacterial components; and the local milieu of hormones, neurotransmitters, immune or inflammatory cytokines and mediators, as well as stromal factors. The epithelial cell responses include changes in epithelial cell morphology (the hummingbird phenotype),62 disruption of the tight junctional complexes,63 the production of cytokines,46 increased epithelial cell proliferation, increased rates of epithelial cell death via apoptosis, and the induction of numerous genes associated with the stress encountered in response to infection.2 The best overview of the

835

836

Section VI  Stomach and Duodenum epithelial cell response can be ascertained from the broad changes in the expression of hundreds of genes demonstrated with high-throughput gene expression systems.2 Detailed analyses of the effects of infection, different cell lines and various inflammatory mediators have yet to be carried out over the wide range of time points that would be relevant to the conditions epithelial cells face in vivo. The expression of genes in epithelial cells stimulated with H. pylori is regulated by transcription factors that are controlled by a series of signaling mechanisms. While many transcription factors are likely activated by infection, the most studied are nuclear factor-kappa B (NF-κB) and activator protein-1 (AP-1) which regulate the expression of a wide variety of proinflammatory cytokines and cellular adhesion molecules in response to infection or the local cytokine milieu. H pylori activates NF-κB in gastric epithelial cells, both in vitro and in vivo in patients with H. pylori gastritis; gastric epithelial cell NF-κB activity is markedly enhanced, correlating with the intensity of neutrophil infiltration and IL-8 protein levels.46 This pathway is of particular interest given the recent report that polymorphisms in the IL-8 gene lead to increased mucosal IL-8 expression, inflammation, and other premalignant changes associated with gastric cancer. H. pylori infection appears to activate NF-κB and AP-1 in gastric epithelial cell lines through various signaling mechanisms including mitogen-activated protein (MAP) kinases.47,64 The MAP kinase cascades regulate a wide range of cell functions, including proliferation, inflammatory responses, and cell survival. For example, cag PAI-positive H. pylori activate the ERK, JNK, and p38 MAP kinase pathways, and ERK and p38 regulate IL-8 production in gastric epithelial cells.65,66 An inhibitor of p38α MAP kinase, FR167653, reduced both neutrophil infiltration and gastric mucosal injury in H. pylori–infected Mongolian gerbils.67 One of the novel concepts emerging is the role of oxidative stress in regulating gene expression during H. pylori infection. Apurinic-apyrimidinic endonuclease-1 (also referred to as redox factor-1) plays a key role in the regulation of redox-sensitive signaling and is expressed in gastric epithelial cells during infection with H. pylori,68 contributing to activation of AP-1 and NF-κB required for the host response to infection, including IL-8 production.69 However, the role it and other redox-regulatory molecules play in the pathogenesis of diseases associated with H. pylori has yet to be clearly defined. Oxidation of DNA by reactive oxygen species such as hydroxyl radicals are thought to play a causal role in malignant transformation through the induction of DNA damage. Oxidative DNA damage is increased in gastric epithelial cells by H. pylori infection. There is growing interest in the role of antioxidants in disease prevention or treatment because infection is associated with decreased levels of a tissue antioxidant scavenger, vitamin C. Moreover, there is evidence that diets high in antioxidants70 or “nutraceuticals” of the isothiocyanate group, such as sulforaphane,71 can antagonize oxidative stress and protect the host from gastric cancer, perhaps by decreasing inflammation and attenuating bacterial load. CagA protein translocates into the host cell cytoplasm where it is tyrosine phosphorylated by host Src kinases, and through other interactions, regulates epithelial cell morphology.72-76 CagA in strains from distinct geographic populations appears to be phosphorylated in a different manner resulting in different effects on intracellular signaling.77,78 It is intriguing to speculate that heterogeneity in the CagA protein may lead to distinct effects on the host response that account for some of the geographic differences in disease. Although phosphorylation of the CagA protein may be important, it is not the only mechanism whereby this

molecule regulates the host response. A phosphorylationindependent effect on gene transcription, which is also attributed to the CagA protein, also has been reported.79 Moreover, greater than 30% of host gene expression is altered independently of the phosphorylation of CagA, at least in intestinal epithelial cell lines.80 Outer inflammatory protein A (OipA) is another bacterial product that induces epithelial cell IL-8 production81 and strains that express OipA are associated with increased bacterial density, mucosal IL-8 levels, and neutrophil infiltration, as well as more severe clinical consequences.82 Peptidoglycan has been reported to translocate into gastric epithelial cells via the type IV secretion system encoded by the cag PAI. Once inside the cell these bacterial compounds are recognized by nucleotide-binding oligomerization domain-1 (NOD1), thereby providing a novel mechanism of bacterial sensing.83 Binding to NOD1 can lead to activation of NF-κB and the subsequent expression of various host genes encoding proinflammatory molecules. As discussed in Chapter 49, gastric acid secretion is a major function of the gastric mucosa that is regulated by a variety of neural, endocrine, and immune factors.3 Elevated fasting and meal- or hormone-stimulated levels of gastrin are well documented in H. pylori infection, and there is evidence that gastrin expression is regulated by bacterial factors and cytokines. Expression of somatostatin, an acidinhibitory peptide, is diminished in infected individuals as is duodenal bicarbonate secretion. The net effect of H. pylori infection on acid secretion is complex and varies depending on the duration and distribution of infection and presence of mucosal atrophy. Secretion of mucus is also affected by H. pylori infection with decreased amounts of mucus and gastric mucosal hydrophobicity; these abnormalities reverse after eradication of infection. Epithelial barrier function is altered during H. pylori infection as a consequence of both direct effects of H. pylori and the accompanying inflammatory response that collectively increase epithelial cell proliferation and programmed cell death.2 Infection with H. pylori results in a unique inflammatory response in which infection persists despite the recruitment and activation of T and B lymphocytes, phagocytic cells, and other immune cell populations. Whereas several epithelial cell responses to H. pylori have been described earlier, they do not appear to fully account for the magnitude of the inflammatory response to an organism that resides predominantly in the lumen. Some bacteria may infect epithelial cells, and significant amounts of bacterial material may “leak” around epithelial cells and reach the lamina propria, where it can activate underlying phagocytes, including neutrophils and macrophages. One bacterial factor is the H. pylori neutrophil-activating protein (HP-NAP). This 150-kd decamer protein promotes neutrophil adhesion to endothelial cells and stimulates chemotaxis of monocytes and neutrophils, NADPH oxidase complex assembly at the plasma membrane, and the subsequent production of reactive oxygen intermediates (ROIs).84 In the inflammatory environment present during H. pylori gastritis, TNF-α and interferon-γ (IFN-γ) can prime neutrophils and potentate the effects of HP-NAP. When cells undergo apoptosis and die, they are removed by phagocytes. This response occurs in the digestive tract as epithelial cells migrate toward the lumen and undergo apoptosis, providing another means whereby the host can sample the antigenic challenges facing epithelial cells. Viral antigens are presented to T cells when infected apoptotic epithelial cells overlie the Peyer’s patch. Engulfment of H. pylori infected epithelial cells by phagocytes may also be an important mechanism by which H. pylori can activate

Chapter 50  Helicobacter pylori the host response, and it has been shown that macrophages bind and then engulf gastric epithelial cells that undergo apoptosis due to infection.2 Recruitment and activation of macrophages and neutrophils cause the release of other inflammatory mediators. Increased expression of inducible nitric oxide synthase (iNOS) is observed in the gastric mucosa during infection with H. pylori.2 Nitric oxide (NO) and superoxide (O2−), which may be produced by infiltrating neutrophils, react to form peroxynitrite (ONOO−), a potent oxidant and reducing agent. Although these products have potent antimicrobial effects, uncontrolled or inappropriate production could play a role in the gastric mucosal damage observed during H. pylori infection. The catabolism of urea by urease provides CO2, which rapidly neutralizes the bactericidal activity of the peroxynitrate by reacting with it to form ONO-OCO2. Urease may favor bacterial colonization by neutralizing some host responses but this also enhances the nitration potential of ONOO− and may favor mutagenesis of host cell DNA. Cytokines secreted by epithelial cells complement those released in the lamina propria. For example, neutrophils are not only activated by IL-8 but also by chemokines such as ENA-7885 and Gro-α,86 which can derive from the epithelium, the adjacent myofibroblasts or the macrophages within the lamina propria. Cytokine induction in macrophages includes induction of TNF-α and IL-6 by urease,87 whereas heat shock protein 60 induces IL-6.88 Intact bacteria can induce the production of chemokines that recruit T cells,89 as well as IL-1290,91 and IL-18,92 two cytokines that favor the selection of Th1 cells. Thus, intact bacteria or bacterial factors trigger a broad cytokine response within the lamina propria. As adaptive responses develop, different T lymphocyte helper (Th) cell subsets emerge, with characteristic patterns of cytokine secretion. Th1 cells promote cell-mediated immune responses through the production of IFN-γ and TNF-α, whereas Th2 cells produce IL-4, IL-5, IL-10, and transforming growth factor-β (TGF-β). Th2 cells can promote mucosal IgA or IgE responses, as well as diminish the inflammation caused by Th1 cytokines. Previous studies suggest that the infected gastric mucosa is preconditioned to favor Th1 cell development.90 One possible hypothesis to account for this tendency is that infection selectively blocks Th2 development. H. pylori can interfere with STAT6 activation by IL-4,93 which could impair Th2 development, and IL-12 and IL-18 induced in response to infection may positively select for the predominant Th1 response. Other cytokines that may enhance Th1 responses, such as IL-23 and IL-27, have yet to be studied in human tissue. T cell activation by H. pylori infection may contribute to more severe inflammation and gastroduodenal diseases. Increased levels of biologically active IL-17, a cytokine produced by activated CD4+ T lymphocytes, are found in the mucosa of H. pylori–infected patients.94,95 IL-17, in turn, induces IL-8 expression by gastric epithelial cells, thereby enhancing neutrophil recruitment. Activation of transcription factors by IL-17 may also contribute to the increased levels of numerous other proinflammatory cytokines and enzymes observed during H. pylori infection, such as IL-1β, TNF-α, and cyclooxygenase-2 (COX-2). IFN-γ and TNF-α produced by Th1 cells can increase the expression of many genes in the epithelium, including IL-8. These cytokines also enhance bacterial binding36 and may contribute to enhanced bacterial load.96 In animal models Th1 cells increase epithelial cell apoptosis36 as well as inflammation, atrophy, and dysplasia.97 TNF-α, IFN-γ, and IL-1β upregulate gastric mucosal Fas antigen expression.98 Since Th1

cells express higher levels of Fas ligand (FasL) than Th2 cells, the relative increase in Th1 cells during H. pylori infection may induce epithelial cell death through Fas-FasL interactions.98,99 This notion is substantiated by the observation that proton pump H+,K+-ATPase–specific Th1 cells in the gastric mucosa kill target cells via Fas-FasL interactions and may act as effector cells in autoimmune gastritis.100 Because Th1 cells cannot clear H. pylori, some other T cell subset may have to be stimulated in order to confer immunity. Studies in animal models indicate that protective immunity was induced by vaccines for Helicobacter spp. via Th cells other than Th1 cells, possibly including Th2 cells. The anti-inflammatory cytokines associated with Th2 cells or other regulatory subsets of Th cells can attenuate the pathogenic effects of Th1 cells.101 More direct evidence suggests that IL-4 can decrease gastritis, an effect that may be mediated by the release of somatostatin.97 As gastric responses can be modified by Th2 cells, the role of other T cell subsets, such as regulatory T cells (Treg), in the pathogenesis of disease associated with H. pylori infection is being addressed. Depletion of Treg in neonatal mice leads to autoimmune gastritis,102 and infection with H. pylori alleviates autoimmune gastritis induced in neonatal mice.103 This suggests that infection may stimulate a subset of antiinflammatory T cells that impair excessive inflammation, which could otherwise lead to the spontaneous clearance of the organism, an effect that appears to occur in the human mucosa in response to H. pylori infection.104 Antibodies in the gastrointestinal tract are normally of the immunoglobulin A (IgA) isotype, which are highly adapted for mucosal protection, conferring protective immunity without activating complement and stimulating deleterious amounts of inflammation. During infection with H. pylori, the number of IgA producing cells increases. IgG and IgM are also detected, along with activated complement. It has been suggested that the level of autoantibodies in humans correlates with the severity of gastritis.2 Local immune complexes contribute to gastroduodenal inflammation and tissue damage during infection and may contribute to autoimmune gastritis. Monoclonal antibodies that recognize H. pylori cross-react with human and murine gastric epithelial cells.105,106 Adoptive transfer of these antibodies to recipient mice induces gastritis,105 as does the transfer of B cells that recognize heat shock proteins from individuals with maltoma.107 Anti-Le antibodies have been described in humans and occur independently of the Le phenotype of the host but they do not appear to be autoreactive. Autoantibodies induced in mice may recognize different targets within the gastric mucosa and even though they may cross-react with human gastric tissue, autoantibodies induced in humans may have a completely different specificity. With few exceptions, infection with H. pylori persists for the life of the host unless there is some intervention with antibiotics. This observation has led to investigations as to whether immunity is impaired by immunologic avoidance or tolerance. Several bacterial factors, including catalase and urease, antagonize innate host responses. Production of the enzyme arginase by H. pylori inhibits NO production and may favor bacterial survival,108 whereas virulent strains of H. pylori have also been shown to alter mucus production109 and phagocytosis.110 The VacA toxin can impair antigen presentation by macrophages by inhibiting the Iidependent pathway, which is mediated by newly synthesized class II MHC molecules.111 Moreover, H. pylori express antigenic molecules that mimic host molecules, such as Lewis antigens that theoretically could stimulate T cells to release cytokines that inhibit autoimmune reactions.

837

838

Section VI  Stomach and Duodenum However, the cytokine profile associated with H. pylori infection is not one that would be expected to occur in a tolerant environment. For example, IL-4, IL-10, and TGF-β (which could mediate an anti-inflammatory effect) are not expressed to the same levels as proinflammatory cytokines such as IFN-γ and TNF-α.2 Because the infected gastric mucosa is characterized by chronic active inflammation, tolerance, if it has occurred, may favor persistent infection even though it cannot prevent the chronic inflammatory response. An alternative possibility is that rather than the T cell responses being inappropriate, they lack a level of coordination necessary to achieve immunity. Under certain circumstances, bacteria can inhibit the growth of T cells and actually induce their apoptosis.112-114 This process has been demonstrated using cell lines, and attributed to very specific patterns of gene expression, including the induction of FasL and Fas on T cells, which enables apoptosis.98,114 The studies by Wang and associates suggest this can occur directly,114 but peptides from H. pylori may activate monocytes to produce oxygen radicals which impair the expression of CD3 zeta chain of the T cell receptor complex and the induction of T cell apoptosis.113 The loss of T cells by apoptosis as an explanation is further supported by the presence of apoptotic T cells in the gastric mucosa,112 as well as the expression of Fas and FasL on biopsy specimens.98 More recently, VacA has been shown to impair the growth of a T cell line, independently of any effect on apoptosis, by inhibiting nuclear translocation of a transcription factor, NFAT, that is required for IL-2 production.115,116 However, these events are complex, with VacA toxin impairing IL2–dependent expansion in T cells freshly isolated from human peripheral blood, but not their survival or ability to produce IL-2.117 H. pylori can also impair T cell function directly due to arginase118 as well as indirectly, as a consequence of inducing hydrogen peroxide in adjacent macrophages that leads to T cell apoptosis.113 These observations support a model in which H. pylori interferes with normal T cell activation in several ways. Despite the evidence that H. pylori can impair T cell responses, some antigen-specific T cell responses are found in the gastric mucosa.100 Nonetheless, it is still possible that a combination of the effects of H. pylori and the cytokine milieu disrupt the coordination required for the development of an effective, antigenspecific T cell response.

CONDITIONS ARISING FROM INFECTION Infection of the human stomach by H. pylori leads to gastritis, which initially affects the superficial layers of the mucosa (see Chapter 51). In some instances the infection is short lived, but typically the infection results in a unique pattern of gastritis, so-called chronic active gastritis, which is essentially a lifelong condition. As discussed, chemokines induced by infection lead to a persistent acute inflammatory infiltrate with neutrophils and other cells (active inflammation) coexisting with cells characteristic of chronic inflammation (lymphocytes, macrophages). In spite of this robust immune response bacteria persist in most infected humans, and the factors that result in clearance of some cases of acute infection remain largely unknown.2 Such insights might have augmented the efforts of the past two decades to develop vaccines to prevent chronic H. pylori infection and enhance eradication that have been unfor­ tunately, largely unsuccessful. Most chronically infected

individuals are asymptomatic with somewhere between 10% and 15% going on to develop peptic ulcer disease, gastric adenocarcinoma, and lymphoma of the gastric mucosa–associated lymphoid tissue. As discussed in the previous section, factors that contribute to whether an individual develops these complications of chronic infection include bacterial factors, although these are not yet well defined; host factors including genes that regulate the host immune response (cytokines and their receptors); and environmental factors. As mentioned, H. pylori organisms colonize only gastrictype epithelium within the human host and may colonize tissues outside the stomach when there is gastric metaplasia of the esophagus or duodenum, or in a Meckel’s diver­ ticulum. The pattern of colonization within the stomach appears to be an important determinant of H. pylori disease manifestations. It is unclear exactly what leads to duodenal ulcers associated with H. pylori infection, but it is thought that hyperacidity associated with antral colonization leads to gastric metaplasia of the duodenum, which can then become colonized, leading to duodenal ulcer in some instances. Distal gastric infection may also present with erosions and small ulcers in the gastric antrum, similar in appearance to lesions associated with anti-inflammatory drug use. Gastric ulcers and gastric adenocarcinoma occur more often when there is proximal colonization of the stomach (pan-gastritis), which results in injury to the gastric glands, leading to atrophic gastritis and associated hypoor achlorhydria (see Chapter 54). Precursor lesions of gastric cancer, including atrophic gastritis, intestinal metaplasia, and dysplasia,119 result from infection although most individuals with intestinal metaplasia in North American populations do not have evidence of infection.120 As such, there are no general recommendations to screen individuals with intestinal metaplasia in the United States. The patterns of infection-associated gastritis and resultant effects on acid secretion are thought to account for the reported differences of the impact of infection and eradication on pre-existing and new symptoms of gastroesophageal acid reflux.121 Those with antral-predominant infection have hyperacidity, which may promote acid reflux that improves after eradication therapy. In contrast, those with pangastritis do not apparently suffer from heartburn but after clearance of infection there is a slow return of secretory function, which may be associated with the development of heartburn. Perhaps even more contentious is the inverse association of H. pylori infection with Barrett’s esophagus and esophageal adenocarcinoma (see Chapters 43, 44, and 46). This may be a related to acid exposure but studies indicate that Cag A–positive strains are more strongly associated with a reduced frequency of Barrett’s esophagus and esophageal adenocarcinoma than Cag A–negative strains. This negative association seems to be greatest in Eastern countries.122 Of note, gastric cancer remains the second major cause of cancer death in the world. The burden of risk of gastric cancer is considered largely attributable to H. pylori infection, with cag PAI–bearing strains having a higher association with gastric cancer than cag PAI–negative strains.123 Given the burden of gastric cancer worldwide, the risk of infection seemingly outweighs the benefits in terms of the development of proximal gastrointestinal tract cancer. Over the past two decades a large number of associations with nongastric diseases and H. pylori infection have been reported including Raynaud’s, scleroderma, idiopathic urticaria, acne rosacea, migraines, thyroiditis, and GuillainBarré syndrome, but the data supporting an association

Chapter 50  Helicobacter pylori for this group of conditions are weak or nonexistent.124 Associations that have somewhat better levels of evidence for an association include coronary artery disease, immune thrombocytopenic purpura,125,126 and iron deficiency anemia127 and for the latter two conditions, eradication of infection may be considered when other treatments have failed. The proposed mechanisms leading to these various conditions range from systemic immune reactions, crossreactivity of bacterial and host proteins, and events secondary to gastric mucosal injury.

DIAGNOSIS The National Institutes of Health (NIH) proposed consensus recommendations for H. pylori testing applicable to the United States in 1994.128 H. pylori management in other parts of the world is generally concordant with the United States, but there are some exceptions related to regional differences in clinical practice and specific infection-related disease prevalence (gastric cancer and premalignant gastric histology).129,130 The American College of Gastroenterology published updated U.S. guidelines in 2007 that recommend testing for H. pylori only if a clinician is prepared to treat a patient with a positive test result.131 Specific indications for testing include patients with active or documented history of uncomplicated or complicated peptic ulcer, early gastric cancer, or gastric MALT lymphoma. Testing for H. pylori is often recommended in younger patients with uninvestigated dyspepsia and no “alarm features” (i.e., early satiety, unexplained weight loss, dysphagia, recurrent vomiting, family history of gastric cancer)132 and in patients with functional dyspepsia (symptoms and negative endoscopy).133 However, the clinical and cost benefits of H. pylori in the setting of dyspepsia remain controversial (see Chapter 13), especially in regions where prevalence of infection is relatively low and gastroesophageal reflux disease (GERD) as a cause of symptoms is high.134 Testing for infection prior to starting nonsteroidal anti-inflammatory drugs (NSAIDs) may reduce subsequent ulcers, but this not generally recommended or often done in the United States, where prevalence of H. pylori is low.131 Also there is no general recommendation to test asymptomatic persons, with the possible exception of those with a family history of gastric cancer,135 particularly individuals of Asian, Eastern European, or Mesoamerican descent, for whom the risk of gastric malignancy is highest. Occasionally immune thrombocytopenic purpura126,136 and refractory iron deficiency anemia127 respond to eradication of infection, so decisions to test for H. pylori in these conditions are made on a case-by-case and regional basis. Finally a potpourri of conditions is linked to H. pylori infection, but causal evidence and biologic plausibility are sufficiently weak or lacking in these clinical situations to recommend testing.124 Indications for testing and treatment for H. pylori are summarized in Table 50-1. There are endoscopic and nonendoscopic means to diagnose infection, and techniques can directly (histologic demonstration of organisms, presence of bacterial antigen in the stool, culture) or indirectly (using urease or an antibody response as a marker of bacteria) detect H. pylori.137,138 The appropriate method to choose depends on the clinical situation, population prevalence, and pretest probability of infection as well as test availability and cost. In addition, recent use of antibiotics or proton pump inhibitors can influence results of certain tests.131 The advantages and dis-

Table 50-1  Indications for Testing and Treatment of Helicobacter pylori Infection Supported by evidence Active peptic ulcer disease (gastric or duodenal ulcer) Confirmed history of peptic ulcer (not previously treated for H. pylori infection) Gastric MALT-lymphoma (low grade) Following endoscopic resection of early gastric cancer Uninvestigated dyspepsia (if H. pylori population prevalence high) Controversial Functional dyspepsia GERD Persons using NSAIDs, especially when first initiating NSAID treatment Unexplained iron deficiency anemia or immune thrombocytopenic purpura Populations at higher risk of gastric cancer (e.g. Asians, Eastern Europeans, Mesoamericans) GERD, gastroesophageal reflux disease; MALT, mucosa-associated lymphoid tissue; NSAIDs, nonsteroidal anti-inflammatory drugs. Adapted from references 129 and 131.

advantages of commonly used diagnostic tests for H. pylori are summarized in Table 50-2. Performing endoscopy solely to diagnosis H. pylori infection is not appropriate; there are three methods—biopsy urease test, histology, and (less often) culture—to identify the organism during an otherwise indicated endoscopic procedure. The choice of method depends on the clinical situation, cost, and test accuracy.131 Guidelines propose initially using a biopsy urease test because the method is quick, easy to perform, relatively inexpensive, and generally accurate. Gastric biopsy material is tested for urease activity by placing several pieces of tissue in a medium containing urea and a pH reagent. Bacterial urease hydrolyzes urea-liberating ammonia, producing an alkaline pH and a resultant color change of the test medium.137 Test results are often positive within minutes to hours. Several urease test kits are commercially available based on the methodology described here, differing only with regard to medium (agar gel or membrane pad) and testing reagents.137 These test kits are generally inexpensive but there are added costs associated with obtaining gastric tissue samples, for example, up-coding diagnostic esophagogastroduodenoscopy (EGD), to EGD with biopsy. Nevertheless, biopsy urease testing is less expensive than histology so one proposed cost-saving measure is to obtain specimens for histology but delay sending them to the laboratory pending urease test results. Specificity of the urease tests is 95% to 100% with falsepositive tests uncommon.137,139 Although reported sensi­ tivity of urease tests is 90% to 95%, accuracy can be negatively affected by blood in the stomach,140 and current or recent use of medications such as antibiotics, bismuthcontaining compounds, or acid inhibitors, especially PPIs.141 Therefore, a negative urease test does not necessarily exclude H. pylori infection in an individual taking anti­ secretory medication, a common scenario in patients referred for endoscopy. Testing samples from multiple regions of the stomach or stopping offending medication and delaying endoscopy for several weeks may improve test sensitivity in such patients. Evaluation of gastric mucosal histology is generally not necessary to diagnose H. pylori, but it can provide infor­ mation regarding the activity and severity of mucosal inflammation (see Fig. 51-3C). Histology can also detect metaplasia, dysplasia, and neoplasia.137 In addition to biopsying “clinically suspicious” areas, taking multiple biopsies

839

840

Section VI  Stomach and Duodenum Table 50-2  Diagnostic Tests for Helicobacter pylori NONENDOSCOPIC TESTs

ADVANTAGES

DISADVANTAGES

Serology (qualitative or quantitative immunoglobulin G [IgG]) Urea breath test (13C or 14C)

Widely available, inexpensive, good NPV

Poor PPV if HP prevalence is low, not useful after treatment Availability and reimbursement inconsistent, accuracy affected by PPI and antibiotic use, small radiation dose with 14C test Fewer data available for polyclonal test, accuracy affected by PPI and antibiotic use

Stool antigen test

Identifies active infection, accuracy (PPV, NPV) not affected by H. pylori prevalence, useful both before and after treatment Identifies active infection; accuracy (PPV, NPV) not affected by H. pylori prevalence; useful both before and after treatment (monoclonal test)

ENDOSCOPIC TESTs

ADVANTAGES

DISADVANTAGES

Histology

Excellent sensitivity and specificity, especially with special and immune stains; provides additional information about gastric mucosa Rapid results, accurate in patients not using PPIs or antibiotics, no added histopathology cost Specificity 100%, allows antibiotic sensitivity testing Excellent sensitivity and specificity, permits detection of antibiotic resistance

Expensive (endoscopy and histopathology costs), interobserver variability, accuracy affected by PPI and antibiotic use Requires endoscopy, less accurate after treatment or in patients using PPIs Difficult and tedious to perform; not widely available; expensive Not widely available; technique not standardized; expensive

Rapid urease test Culture Polymerase chain reaction (PCR) assay

NPV, negative predictive value; PPI, proton pump inhibitor (see Table 50-3); PPV, positive predictive value. Adapted from Chey WD, Wong BC. American College of Gastroenterology guideline on the management of Helicobacter pylori infection. Am J Gastroenterol 2007; 102:1808-25.

Figure 50-1.  Photomicrograph of a gastric mucosal biopsy specimen, stained with the Genta stain, from a patient with Helicobacter pylori gastritis. The bacteria are well seen. (Courtesy of Hala El-Zimaity, Houston, Tex.)

and sampling lesser and greater curvatures of gastric antrum and body are important, especially when looking for evidence of multifocal atrophic gastritis and/or intestinal metaplasia (see Chapter 51). Histologic examination had been considered the gold standard for identifying infection, with reported sensitivity and specificity as high as 95% and 98%, respectively.142 However, the distribution and density of organisms can vary within the stomach resulting in sampling error, particularly in patients taking antisecretory medications.131,137 Detecting organisms can be difficult when standard hematoxylin and eosin staining is used alone, but is less of an issue when processing tissue with special stains such as Giemsa, silver, or Genta (Fig. 50-1) or specific immune stains.138,142,143 H. pylori are difficult to culture because the organism is fastidious, slow growing, and requires specialized media

and growth environment.137,138 In fact the initial isolation of H. pylori occurred by happenstance when plated cultures incubated over a long holiday weekend. When culturing for H. pylori, tissue should be obtained before biopsy forceps become contaminated with formalin and placed in a container with only a few drops of saline to preserve the specimen during transport to a local or offsite microbiology facility.138 Although culture is not generally recommended, in those with refractory disease culture with antibiotic sensitivity testing can guide subsequent treatment, although in vitro sensitivity testing does not always predict clinical treatment outcome.138,144 Most often nonendoscopic tests are used to diagnose H. pylori infection, and serology remains the most popular method used, although use of other noninvasive methods that can detect active infection has increased. Infection incites a systemic immune response, and enzyme-linked immunosorbent assay (ELISA) technology can detect IgG antibodies to a variety of bacterial antigens in serum samples.137,138 Tests for IgA and IgM class antibodies are less reliable and not recommended.137 Office-based kits that test whole blood can provide results within 30 minutes and permit “point of service” testing. Although serology is inexpensive, noninvasive and ideally suited to a primary care setting, the prevalence of H. pylori in the population being tested influences its accuracy.131 The sensitivity of serology is generally quite high (90% to 100%) but specificity is variable (76% to 96%), especially if prevalence of H. pylori is low. Therefore, in places where infection is less common (most areas of the United States), the negative predictive value of serology is high. On the other hand, the corresponding positive predictive value is poor, suggesting most often positive results are actually falsely positive.131 So it is best to confirm positive serology results with another method such as a stool antigen or urea breath test before starting treatment or to use a test that detects active infection in the first place. Conversion of a positive serology to negative after treatment suggests bacterial cure, but in most instances serology remains positive for months to years even after successful treatment of infection.145 This “serologic scar” effectively precludes use of serology to confirm

Chapter 50  Helicobacter pylori 13C-urea 13CO 2 (µmol)

Positive urea breath test

NH2 H2O 2NH3 + 13CO 2

+ 13C

O

NH2 Urease Breath 13CO 2

Negative urea breath test 2 1 Time (hour)

Blood Figure 50-2.  The urea breath test. (From Walsh JH, Peterson WL. Drug therapy: The treatment of Helicobacter pylori infection in the management of peptic ulcer disease. N Engl J Med 1995; 333:984.)

bacterial eradication after treatment, a practice that is unfortunately still quite common in the primary care setting even though better tests to confirm eradication are more widely available. The urea breath test (UBT) detects active H. pylori infection and so it is useful for making the primary diagnosis, confirming the accuracy of serology, and documenting successful treatment.131 UBT relies on bacterial hydrolysis of orally administered urea tagged with a carbon isotope, either 13C or 14C (Fig. 50-2). Hydrolysis generates ammonia and tagged CO2, which can be detected in breath samples.137,138 The 13C test is best for children and pregnant women because it uses a nonradioactive isotope, whereas the radiation dose with the 14C test is 1 microCi138 equivalent to one day of background radiation exposure. The specificity of UBT is more than 95%131,137; therefore, false-positive results are uncommon. The sensitivity of the test is 88% to 95% with false-negative results reported in patients taking antisecretory therapy such as PPIs,131,141 bismuth, or antibiotics. To improve accuracy, antibiotics should be stopped at least four weeks and PPIs at least one week before breath testing.131 UBT is not accurate in patients who have had gastric resective surgery. An immunoassay that detects the presence of bacterial antigens in stool of infected patients is an alternative nonendoscopic method to diagnose active H. pylori infection as well as confirm eradication following treatment. Overall sensitivity and specificity of the stool test are comparable to the UBT (94% and 97%, respectively).131,138,140 A rapid H. pylori stool antigen test is available that permits testing during a clinic visit but it is slightly less accurate than a traditional laboratory based stool test.146 The sensitivity of stool testing is negatively affected by PPIs, bismuth, and antibiotics, which can decrease bacterial load, so similar precautions as described earlier for UBT are appropriate when using stool tests.131,138,141 Polymerase chain reaction is a sensitive method to detect H. pylori in gastric mucosal biopsies, but it is not practical for routine clinical diagnosis. It is, however, used for research purposes to identify bacteria when ordinary culture is difficult, as when testing stool or drinking water in a community setting, to type organisms during epidemiologic or transmission studies or for “real time” antibiotic resistance testing of tissue.138,147 Current recommendations for testing are as follows. A stool antigen assay or UBT is the preferred noninvasive

method for initial diagnosis of H. pylori because it can detect active infection. Serology is only useful to exclude H. pylori infection, and positive serology results should be confirmed by a test for active infection before starting treatment. Endoscopic biopsy is suitable for patients undergoing a diagnostic endoscopy who are found to have an abnormality such as an ulcer or for those requiring endoscopy to follow-up a gastric ulcer or suspected MALT lymphoma. Biopsy urease testing can be used in patients not taking a PPI or antibiotics when histopathology is not clinically necessary. When clinically indicated it is appropriate to confirm successful eradication of infection with either a UBT or stool antigen test. These tests should not be performed sooner than four to six weeks after completion of treatment because earlier testing might yield false-negative results. Also medications that could affect test results such as PPIs should be discontinued at least one week prior to testing to improve accuracy. Post-treatment endoscopy with biopsy is only necessary if a repeat procedure is clinically indicated to follow up complicated ulcer disease or other mucosal abnormality, but this should be delayed for at least four to six weeks after therapy. Sampling multiple areas of the stomach is important to avoid missing persistent infection because of density and distribution of bacteria by prior antibiotics and concomitant antisecretory medications. Serology is not useful for follow-up because the test remains positive in most patients for months or even years after infection is gone.

TREATMENT Because there is currently no “pylori specific” or single antibiotic available to cure infection, treatment requires combining several medications. Recommended regimens usually include two antibiotics dosed several times daily for 7 to 14 days along with acid-suppressive medication.131,148 Attempts to simplify regimens or shorten treatment duration generally reduce effectiveness. Compliance can be a problem because taking multiple medications is difficult, and minor medication-related side effects are frequent. Treatment success can vary among countries and even within regions of countries, possibly related to antibioticresistant organisms that are more common than previously appreciated.148,149 Despite these concerns, treatment regimens are available that cure H. pylori infection in more 75% of individuals.131,148,150,151 After cure, annual adult reinfection especially in developed countries is uncommon, probably less than 1%. Higher rates of reinfection are reported, but these often include cases that actually represent recrudescence of the original infection that failed to clear during antibiotic treatment.152 Reinfection tends to be higher in children especially after spontaneous clearance of a primary infection, and it is reported to be higher in adults living in areas of the world with high H. pylori prevalence.9,11,153 Commonly used treatment regimens are summarized in Table 50-3.129,131,148,150,151 Triple therapy, composed of two antibiotics, clarithromycin (500 mg twice daily) and amoxicillin (1 g twice daily) along with a PPI (e.g., omeprazole 20 mg twice daily, lansoprazole 30 mg twice daily, pantoprazole 40 mg twice daily, rabeprazole 20 mg twice daily, or esomeprazole 40 mg every day) for 7 to 14 days, is currently the most popular initial treatment for H. pylori. PPI triple therapy consistently cures more than 80% of infections, especially if organisms are sensitive to clarithromycin

841

842

Section VI  Stomach and Duodenum Table 50-3  First-Line Treatment of Helicobacter pylori Infection* TREATMENT REGIMEN* †

DURATION

ERADICATION RATE

COMMENTS

PPI , clarithromycin 500 mg, amoxicillin 1000 mg (each twice daily)

10-14 days

70%-85%

Macrolide resistance affects eradication success; not appropriate for penicillin allergic individuals or those who have received a clarithromycin regimen in the past

PPI†, clarithromycin 500 mg, metronidazole 500 mg (each twice daily)

10-14 days

70%-85%

Appropriate for penicillin-allergic individuals who have not received a clarithromycin-containing regimen in the past

PPI†, amoxicillin 1000 mg (each twice daily)   followed by PPI†, clarithromycin 500 mg, tinidazole 500 mg (each twice daily)

5 days 5 days

90%

Appears highly effective despite clarithromycin resistance; limited experience to date in the United States

Bismuth subsalicylate 525 mg, metronidazole 500 mg, tetracycline 500 mg (each four times daily)   plus PPI† or H2RA (twice daily)

10-14 days

75%-90%

Inexpensive but complicated regimen; consider in penicillin allergic individual or if clarithromycin resistance is suspected; can be used for retreatment (see Table 50-4)

*Note that not all of these regimens are currently approved by the U.S. Food and Drug Administration (FDA). † Lansoprazole 30 mg, pantoprazole 40 mg, rabeprazole 20 mg, omeprazole 20 mg, or esomeprazole 40 mg (esomeprazole can be taken once daily). H2RA, histamine H2-receptor antagonist; PPI, proton pump inhibitor. Adapted from references 131, 150, and 151.

and longer treatment duration (14 or 10 days vs. 7 days) is used. Metronidazole (500 mg twice daily) can be substituted for either amoxicillin or clarithromycin, but this is appropriate only for penicillin-allergic or macrolide-intolerant individuals because metronidazole resistance is common and can reduce treatment success.129,131,148,150 A 10-day sequential regimen (a PPI and amoxicillin 1 g, each given twice daily for the first 5 days, followed by the PPI, clarithromycin 500 mg, and tinidazole 500 mg, each given twice daily for the remaining 5 days) improved overall eradication rates compared with standard PPI triple therapy (89% vs. 77 %), but was particularly better for clarithromycin-resistant bacteria (89% vs. 29%).154 A pooled analysis of studies evaluating sequential therapy confirmed its superior efficacy especially with macrolide-resistant bacterial strains.155 Such results are encouraging, although most experience with this treatment is geographically limited to Mediterranean countries. However, there is no reason to expect different efficacy in other regions.131,154 Although used more than a decade ago, dual regimens consisting of a single antibiotic (amoxicillin or clarithromycin) and a PPI are no longer recommended because eradication is significantly less than with three drug regimens.150 Bismuth-based therapy, which combines a bismuth salt, metronidazole 500 mg and tetracycline 500 mg each given four times a day, and daily acid suppression (usually a PPI every day) for two weeks was actually one of the first therapies used to treat H. pylori. Although it remains effective (more than 80% eradication), the number of daily pills and associated frequent minor side effects negatively affect tolerability and compliance. For that reason, in the United States it is usually reserved as a second-line or retreatment regimen.129,131,148,150 A combination capsule that contains bismuth subcitrate 140 mg, metronidazole 125 mg, and tetracycline 125 mg is available in the United States and Canada, simplifying bismuth-based treatment. In a comparative study, patients treated with three combination capsules four times daily and PPI twice daily for 10 days had comparable H. pylori eradication with those treated with traditional PPI triple therapy (88% vs. 83%).156 Short course (1 to 7 days)

bismuth-based treatment157 has been evaluated, but consistent long-term cure of infection has not been confirmed, so abbreviated treatment cannot be recommended.150 Initial treatment of H. pylori infection fails in up to 25% of patients as a result of an infection with antibioticresistant organisms, poor compliance with medication, and patient demographics such as younger age, smoking, prior antibiotic use, and underlying condition (functional dyspepsia vs. peptic ulcer).149,158,159 A review of various retreatment regimens (Table 50-4) reported eradication rates of 46%, 70%, 80%, and 76 % percent for PPI dual therapies, PPI triple therapies, ranitidine bismuth citrate–based triple therapy and bismuth-based therapy, respectively.160 Ranitidine bismuth citrate is no longer available in the United States. When two new antibiotics are used during retreatment, cure of infection appears to be superior compared with when only one new antimicrobial is used. One more recently recommended “rescue therapy” includes a PPI, levofloxacin 250 mg, and amoxicillin 1 g, all given twice daily for 10 days. This combination cures infection in up to 80% of patients who have failed one or more prior treatment attempts. Less well studied, but reported to be 85% effective when used as retreatment, is a combination of PPI and amoxicillin 1 g, each twice daily, along with rifabutin 300 mg every day for 10 days. A lower dose of rifabutin (150 mg) appears to be less effective. Successful retreatment with regimens substituting furazolidone for metronidazole has also been reported, but furazolidone is no longer commercially available in the United States. As initial treatment for H. pylori, a 10- to 14-day course of standard PPI triple therapy described previously (PPI, amoxicillin and clarithromycin) is recommended but a 10-day sequential regimen would be an appropriate alter­ native, especially if clarithromycin-resistant infection is suspected (see following). If infection persists after this treatment, bacteria are likely resistant to clarithromycin. Therefore, retreatment should be with one of the PPI triple regimens noted earlier that incorporates a different combination of medications or a bismuth-based therapy for 14 days. Subsequent courses of treatment if necessary should also incorporate different antibiotic combinations when

Chapter 50  Helicobacter pylori Table 50-4  Rescue Treatment for Persistent Helicobacter pylori Infection* REGIMEN

DURATION

ERADICATION RATE

COMMENTS

Bismuth subsalicylate 525 mg, metronidazole 500 mg, tetracycline 500 mg (each four times daily)   plus PPI† or H2RA (twice daily)

14 days

70%

Inexpensive but complicated regimen, so compliance should be emphasized; less effective as retreatment than as initial therapy; full dose of metronidazole and two weeks of treatment appear necessary

PPI†, amoxicillin 1000 mg, levofloxacin 250 mg (each twice daily)

10-14 days

57%-91%

Limited data from the United States

PPI† amoxicillin 1000 mg, rifabutin 150 mg (each twice daily)

14 days

60%-80%

Expensive; adverse hematologic events and drug interactions possible. Limited data from the United States

*Note that not all of these regimens are currently approved by the U.S. Food and Drug Administration (FDA). † Lansoprazole 30 mg, pantoprazole 40 mg, rabeprazole 20 mg, omeprazole 20 mg, or esomeprazole 40 mg (esomeprazole can be taken once daily). H2RA, histamine H2-receptor antagonist; PPI, proton pump inhibitor. Adapted from references 131, 150, and 151.

possible to lessen the effect of acquired antimicrobial resistance. Although selection of a treatment regimen based on antibiotic sensitivity testing might improve subsequent treatment results, this is not routinely recommended. Primary resistance to antibiotics commonly used to treat H. pylori varies widely throughout the world.138,144,148 In the United States resistance to metronidazole can be detected in up to 40% of stains, whereas clarithromycin resistance is approximately 11%. Resistance to tetracycline and amoxicillin is unusual, generally less than 1%.148,161,162 In the United States, clarithromycin and metronidazole resistance increase with age and are more common in women than in men. Clarithromycin resistance is more common in the mid-Atlantic and northeast regions of the country. Metronidazole resistance is more common in Hispanics and Asians. Antibiotic resistance significantly affects the success of PPI triple regimens but is less important with bismuthbased regimens.138,149 A bacterial point mutation(s) that prevents reduction of metronidazole to its active metabolite is responsible for drug resistance.138,144 Resistance to metronidazole appears to be a relative condition that can be overcome in most instances by using a higher dose (500 mg) or combining the drug with a bismuth preparation. On the other hand, clarithromycin resistance appears to be an absolute situation that cannot be easily overcome by increasing the macrolide dose. One of three bacterial point mutations within its conserved loop of 23S strand of ribosomal RNA (A2143G, A2142G, and A2142C) can interfere with ribosomal macrolide binding and lead to clarithromycin resistance.138,144 The A2143G mutation appears to have the greatest negative effect on treatment and is likely the major reason for PPI triple therapy failure. Testing for specific mutations is not clinically available, so if clarithromycin resistance is suspected or confirmed by culture, nonmacrolide regimens or sequential therapy are appropriate treatment options. Table 50-4 summarizes rescue treatments for persistent infection. Failed attempts at eradication generally result in secondary antibiotic resistance.148 Therefore, one can assume when treatment with clarithromycin or metronidazole-containing regimens is unsuccessful, specific drug resistance has emerged which should influence any subsequent choice of therapy.138,144 Treatment-related side effects can occur in as many as 50% of patients taking one of the treatment regimens

described previously, but generally these are mild and do not require discontinuation of therapy. Some of the more common side effects include taste alteration and gastrointestinal (GI) upset with metronidazole and clarithromycin and allergic reactions and diarrhea with amoxicillin. In addition, tetracycline should not be prescribed to children or pregnant women. Side effects of H. pylori treatment have recently been extensively reviewed elsewhere.150,151

KEY REFERENCES

Amieva MR, El-Omar EM. Host-bacterial interactions in Helicobacter pylori infection. Gastroenterology 2008; 134:306-23. (Ref 5.) Blaser MJ, Atherton JC. Helicobacter pylori persistence: Biology and disease. J Clin Invest 2004; 113:321-33. (Ref 26.) Chey WD, Wong BC. American College of Gastroenterology guideline on the management of Helicobacter pylori infection. Am J Gastroenterol 2007; 102:1808-25. (Ref 131.) Correa P, Houghton J. Carcinogenesis of Helicobacter pylori. Gastroenterology 2007; 133:659-72. (Ref 119.) Ernst PB, Peura DA, Crowe SE. The translation of Helicobacter pylori basic research to patient care. Gastroenterology 2006; 130:188-206. (Ref 2.) Gillen D, McColl KE. Gastroduodenal disease, Helicobacter pylori, and genetic polymorphisms. Clin Gastroenterol Hepatol 2005; 3:1180-6. (Ref 6.) Jafri NS, Hornung CA, Howden CW. Meta-analysis: Sequential therapy appears superior to standard therapy for Helicobacter pylori infection in patients naive to treatment. Ann Intern Med 2008; 148:923-31. (Ref 155.) Jodlowski TZ, Lam S, Ashby CR. Emerging therapies for the treatment of Helicobacter pylori infections. Ann Pharmacother 2008; 42:162139. (Ref 151.) Malfertheiner PF, Megraud, C, O’Morain, F et al. Current concepts in the management of Helicobacter pylori infection: The Maastricht III Consensus Report. Gut 2007; 56:772-81. (Ref 129.) Megraud F, Lehours P. Helicobacter pylori detection and antimicrobial susceptibility testing. Clin Microbiol Rev 2007; 20:280-322. (Ref 138.) Naumann M, Crabtree JE. Helicobacter pylori–induced epithelial cell signalling in gastric carcinogenesis. Trends Microbiol 2004; 12:29-36. (Ref 46.) Suerbaum S, Michetti P. Helicobacter pylori infection. N Engl J Med 2002; 347:1175-86. (Ref 4.) Talley NJ, Vakil N. Guidelines for the management of dyspepsia. Am J Gastroenterol 2005; 100:2324-37. (Ref 132.) Vakil N, Megraud F. Eradication therapy for Helicobacter pylori. Gastroenterology 2007; 133:985-1001. (Ref 148.) Versalovic J. Helicobacter pylori. Pathology and diagnostic strategies. Am J Clin Pathol 2003; 119:403-12. (Ref 137.) Full references for this chapter can be found on www.expertconsult.com.

843

CHAPTE R

51 Gastritis and Gastropathies Edward L. Lee and Mark Feldman

CHAPTER OUTLINE Classification  845 Chronic Gastritis  846 Helicobacter pylori Gastritis  846 Environmental Metaplastic Atrophic Gastritis  846 Autoimmune Metaplastic Atrophic Gastritis  847 Carditis  848 Infectious Gastritides  849 Viruses  849 Bacteria  849 Fungi  851 Parasites  852 Granulomatous Gastritides  852 Sarcoidosis  853 Xanthogranulomatous Gastritis  853 Distinctive Gastritides  853 Collagenous Gastritis  853 Lymphocytic Gastritis  853 Eosinophilic Gastritis  854 Miscellaneous Forms of Gastritis  855 Gastritis in Inflammatory Bowel Disease (Crohn’s and Ulcerative Colitis)  855

Patients, clinicians, endoscopists, and pathologists have different concepts of what gastritis is. Some think of it as a symptom complex, others as a description of the endoscopic appearance of the stomach, and still others use the term to describe microscopic inflammation of the stomach. This third definition of gastritis is used in this chapter. There is not a close relationship between the presence of microscopic inflammation (histologic gastritis) and gastric symptoms (epigastric pain, nausea, vomiting, bleeding). The correlation between microscopic and gastroscopic abnormalities is also poor.1-2 In fact, most patients with histologic gastritis are asymptomatic and have normal gastroscopic findings. Certain disorders of the gastric mucosa including erosive processes and hyperplastic disorders may be associated with little or no inflammation (gastritis). These conditions collectively are referred to as reactive and hyperplastic gastropathies, respectively. By the earlier definition, a gastric biopsy must be obtained to be able to diagnose gastritis. Every biopsy represents an excellent opportunity for the clinician and pathologist to communicate to correlate clinical data, endoscopic findings, and pathology. Errors may occur when the pathologist attempts to diagnose biopsies without clinical input. It is important for the pathologist to become familiar with the range of normal gastric biopsy findings because many gastrointestinal biopsies obtained endoscopically show normal mucosa.3 Indications for gastroscopic biopsies include gastric erosion or ulcer, thick gastric fold(s), gastric polyp(s) or

Gastritis Cystica Profunda  855 Gastric Graft-versus-Host Disease  856 Allergic Gastritis  856 Reactive Gastropathies (Acute Erosive Gastritis)  856 Medications and Toxins  856 Alcohol  856 Portal Hypertensive Gastropathy  857 Cocaine  857 Stress  857 Radiation  857 Bile Reflux  857 Ischemia  858 Prolapse  858 Linear Erosions in a Hiatal Hernia (Cameron Lesions)  858 Aging Gastropathy  858 Hyperplastic Gastropathies  858 Ménétrier’s Disease and Hyperplastic, Hypersecretory Gastropathy  858 Zollinger-Ellison Syndrome  859 Differential Diagnosis of Gastritis and Gastropathy  859 Treatment and Prevention of Gastritis and Gastropathy  859

mass(es), and diagnosis of Helicobacter pylori infection. A set of five biopsies should be taken from patients in whom clinical or endoscopic findings are suspicious for one of the forms of chronic gastritis (discussed later). Preferred sites for this set of biopsies are shown in Figure 51-1. The location of the biopsy sites should be identified for the pathologist on an accessioning form.

CLASSIFICATION 2-7 There is no universally accepted classification of gastritis. The Sydney system was an attempt to unify terminology for endoscopic and histologic gastritis and gastropathy, and it was updated in 1995.4 However, the complexity of the Sydney system precluded widespread use. Failure to obtain adequate numbers of biopsies from various regions of the stomach (see Fig. 51-1) often prevents accurate classification and often precludes a thorough assessment of the distribution of gastritis.5 In this chapter we use a combination of classifications of gastritis by four experts: Rubin,2 Genta,4 Appelman,6 and Montgomery.7 The keystone of the mentioned classification is the fact that H. pylori and nonsteroidal anti-inflammatory drugs (NSAIDs) are the most common causes of gastritis and reactive gastropathies (acute erosive gastritis), respectively. The chapter outline provides an etiology-based classification of gastritis and gastropathies.

845

846

Section VI  Stomach and Duodenum CHRONIC GASTRITIS 7-9 Most forms of chronic gastritis are clinically silent. Their importance relates to the fact that these gastritides are risk factors for other conditions such as peptic ulcer disease, gastric polyps, and benign and malignant gastric neoplasms.8,9 Three types of chronic gastritis are recognized (Figs. 51-2 and 51-3). Biopsies from the antrum and the incisura are useful for diagnosing H. pylori infection with its diffuse antralpredominant gastritis. However, biopsies from the gastric body mucosa may be more diagnostic for H. pylori infection in some patients treated with proton pump inhibitors. Environmental metaplastic atrophic gastritis (also called multifocal atrophic gastritis, or MAG) is patchy and involves the antrum and body mucosa and sometimes, but not always, is associated with H. pylori infection.7 The diagnosis of autoimmune metaplastic atrophic gastritis, also called diffuse corporal atrophic gastritis (DCAG) and type A gastritis, can be confirmed with multiple biopsies from the gastric body that show atrophy and biopsies from the antrum that do not show atrophy. In most cases biopsies are obtained at the time of endoscopy.

HELICOBACTER PYLORI GASTRITIS 2,6,7,10-19

H. pylori gastritis (HPG) is caused by infection of the antral mucosa with H. pylori.2,6,7 In the United States H. pylori

gastritis is seen mainly in low socioeconomic and immigrant populations,7 and there is no increased risk of gastric cancer. Most patients with HPG are asymptomatic. In most cases the antrum appears normal to the endoscopist; some patients with active disease in the antrum may demonstrate red streaks. Radiographic differences between antral gastritis due to H. pylori and not due to H. pylori have been described; thickened gastric folds, especially in a polypoid configuration, and enlarged areae gastricae favor H. pylori as the cause, whereas antral erosions favor causes other than H. pylori.10 Histologically, a diffuse, chronic inflammatory infiltrate which includes numerous lymphocytes and plasma cells expands the lamina propria and epithelium (see Fig 51-3).7 The presence of acute inflammatory cells is best designated an active gastritis and not acute gastritis. Additional microscopic changes include injury to the surface and foveolar epithelium with loss of apical mucin and reactive nuclear changes and erosions.6,11 Lymphoid follicles with germinal centers are characteristic of an infection with H. pylori.3,12 H. pylori organisms lie in the superficial mucous layer along the mucosal surface and within the gastric pits. Although the organisms can be seen in routine hematoxylin and eosin–stained tissue when numerous organisms are present, special stains are useful when few organisms are present. Stains that may be used to highlight the organisms are Giemsa stain, Warthin-Starry silver stain, Gram stain, and immunocytochemical stains.13-15 Helicobacter heilmannii spiral bacteria are a less frequent cause of active gastritis.16-18 The organisms originally known as Gastrospirillum hominis are longer than H. pylori and have multiple spirals.16,17 A topographic study of H. pylori density and distribution and the com­parison of biopsy sites for the histopathologic diagnosis of H. pylori conclude that two antral biopsy specimens, one from the lesser and one from the greater curvature, have close to 100% sensitivity for detecting H. pylori infection (see Fig. 51-1).19 Biopsy specimens from the corpus increase the diagnostic yield if extensive intestinal metaplasia is present in the antrum.19

ENVIRONMENTAL METAPLASTIC ATROPHIC GASTRITIS 4,7,20-32

Figure 51-1.  Gastric biopsy protocol. Blue and black symbols represent sites from which gastric mucosal biopsies should be obtained. Biopsies from the antrum (greater and lesser curvature) and from the incisura are useful for diagnosing Helicobacter pylori infection. Biopsies from the gastric body (greater and lesser curvature) are useful for diagnosing autoimmune metaplastic atrophic gastritis. Biopsies from the antrum and body in combination are useful for diagnosing environmental metaplastic atrophic gastritis.

Environmental metaplastic atrophic gastritis (EMAG), also called mutifocal atrophic gastritis (MAG), is characterized by the involvement of the antrum and body with mucosal atrophy and intestinal metaplasia.4,7,20-22 Atrophic gastritis involving the body may be associated with pseudopyloric metaplasia, in which the mucosa resembles antral mucosa but stains for pepsinogen I (PGI), a proenzyme expressed in body mucosa.23 Gastroscopy may show a pale mucosa, shiny surface, and prominent submucosal vessels,24 and magnify-

Figure 51-2.  Topographic patterns of chronic, nonspecific gastritis. The darkest areas in the schematics of autoimmune metaplastic atrophic gastritis and environmental metaplastic atrophic gastritis represent areas of focal atrophy and intestinal metaplasia.

Diffuse corporal atrophic gastritis

Diffuse antral gastritis

Multifocal atrophic gastritis

Chapter 51  Gastritis and Gastropathies

A

B

C

D

E Figure 51-3.  Chronic nonspecific gastritides. A and B, Normal mucosal biopsy from the gastric body/fundus and antrum, respectively. C, Diffuse antral gastritis. The glands show an infiltrate of neutrophils, in addition to an increase in inflammatory cells in the lamina propria. This lesion is typically associated with Helicobacter pylori infection. D, Multifocal atrophic gastritis with intestinal metaplasia. Note several glands lined by goblet cells (arrow). The biopsy specimen is from the gastric body, and similar changes were present in the antrum. E, Diffuse corporal atrophic gastritis in a man with pernicious anemia. The gland in the lower left is lined by goblet cells. Nests of enterochromaffin-like cells are also visible (arrows).

ing endoscopy is much more sensitive in detecting atrophy.25 The pathogenesis of EMAG is multifactorial. H. pylori plays an important role and has been identified in about 85% of patients with EMAG. EMAG can occur early in life in H. pylori–infected individuals living in developing countries.23 Genetic and environmental factors, especially diet, are also important. Certain population groups are predisposed to EMAG including African Americans, Scandinavians, Asians, Hispanics, Central and South Americans, Japanese, and Chinese. In patients with EMAG, intestinal metaplasia is a risk factor for dysplasia and gastric cancer, usually the intestinal type (see Chapter 54).2,4,7,22,26-31 Inflammation in EMAG destroys gastric epithelial cells, and eventually the atrophic glands are replaced by metaplastic epithelium.4,7,22 In some

cases, especially in patients living in the Pacific basin, metaplastic gastric cells are ciliated, probably due to environmental factors that are more prominent in the Pacific than the Atlantic Ocean basins.32 Because criteria for gastric atrophy among pathologists are debated, intestinal metaplasia is the most reliable marker of atrophy. Intestinal metaplasia of the gastric mucosa can be classified into three types as described in Chapter 54, where their possible associations with the intestinal type of gastric cancer are discussed.

AUTOIMMUNE METAPLASTIC ATROPHIC GASTRITIS 6,7,33-55

Autoimmune metaplastic atrophic gastritis (AMAG), also called diffuse corporal atrophic gastritis (DCAG), is an auto-

847

848

Section VI  Stomach and Duodenum immune destruction of body/fundic glands. AMAG is relatively uncommon, accounting for less than 5% of all cases of chronic gastritis. Endoscopic features of AMAG include effacement of the gastric folds and a thin body/fundic mucosa. AMAG is the pathologic process in patients with pernicious anemia, an autoimmune disorder usually occurring in patients of northern European or Scandinavian background.33 Patients with AMAG exhibit achlorhydria or hypochlorhydria, hypergastrinemia secondary to low or absent gastric acid with antral G-cell hyperplasia, and low serum PGI concentrations, and they often have circulating antibodies to parietal cell antigens and to intrinsic factor.6,7,33 Incomplete (colonic) intestinal metaplasia (type III) may occur in AMAG and be a risk factor for gastric carcinoma in areas of the world that experience a higher incidence of gastric carcinoma than in the United States.34 Metaplastic intestinal Paneth cells in AMAG appear to secrete an antibacterial peptide of the alpha-defensin family, human defensin 5 (HD-5), a peptide not produced in the normal stomach.35 HD-5 could help the atrophic stomach against invasion by indigenous bacterial flora that overgrow in the anacidic stomach (see Chapter 49). Metaplastic pancreatic acinar cells are also a feature of autoimmune gastritis.36 Atrophic glands with extensive intestinal metaplasia are confined to the body/fundic mucosa. Early in the course of this disease, atrophy may be focal and the preserved islands of relatively normal oxyntic mucosa may appear polypoid endo­scopically or radiologically.37 Rarely, AMAG progresses to diffuse (complete) atrophy. Hypergastrinemia, a con­sequence of achlorhydria, is associated with an increase in enterochromaffin-like cell hyperplasia and gastric carcinoid tumors. Cases of gastric carcinoids and simul­ taneous gastric cancer have been described.38 Gastric car­ cinoid tumors are discussed further in Chapters 31 and 54. In one study from Italy, half of 150 patients with AMAG had antibodies to H. pylori and another 25% had H. pylori in their oxyntic mucosa in addition to having antibodies against H. pylori.39 Thus, H. pylori could have contributed to three quarters of the cases of AMAG. Recent studies suggest a role for H. pylori in the early pathogenesis of autoimmune gastritis; evidence of infection early in the course of the disease in individuals with parietal cell antibodies is frequent.40 If gastric atrophy and achlorhydria develop, the incidence of H. pylori infection then decreases. Among 267 H. pylori–infected patients with dyspepsia, 65 had AMAG. Compared with the 202 patients without AMAG, the atrophics were older, more likely to have antibodies against cagA and vacA, more likely to consume alcohol and coffee, more likely to be taking sedative medicines, and less likely to have anxiety.41 Whether H. pylori results in AMAG thus appears to depend on length of infection, as well as bacterial, dietary, and emotional factors. With regard to bacterial factors promoting atrophy, it appears that cagA+/vacA+ H. pylori are more likely to cause AMAG. These H. pylori are often the s1m1 vacA subtype that also express Lewis blood group antigens X and Y.42 Lewis antigens may help camouflage H. pylori because these antigens are also present on human gastric epithelial cells. It has been suggested that when antibodies to Lewis antigens from H. pylori develop, they cross-react with antigens on epithelial cells such as the H+,K+-ATPase on parietal cells, resulting in autoimmune chronic gastritis.43 Based on uncontrolled studies from Tokyo,44 eradication of H. pylori often leads to a decrease in the amount of gastric atrophy and intestinal metaplasia, whereas failed eradication attempts accomplish neither of these endpoints. Antibodies to parietal cell antigens, most notably the proton pump (H+,K+-ATPase) are frequently present in auto-

immune gastritis.45 These antibodies are frequently detected in patients with various autoimmune diseases including type 1 diabetes mellitus46 and thyroid diseases (Graves’, Hashimoto’s), explaining the association of these conditions with pernicious anemia. The risk of AMAG is increased three- to five-fold in type 1 diabetic individuals, and some authors have suggested screening type 1 diabetics with gastroscopy and mucosal biopsy.47 One in eight patients with chronic hepatitis C treated with interferon-α develops antibodies to parietal cells and to thyroid tissue, and these antibodies recede after therapy is stopped48; the clinical significance of these findings in the stomach is yet to be elucidated. A proportion of the CD4+ lymphocytes present in the chronic inflammatory infiltrate within the gastric mucosa proliferate in response to H+,K+-ATPase, and most CD4+ cells secrete Th1 cytokines such as tumor necrosis factor-α (TNF-α); provide help for B cell immunoglobulin production; and enhance perforin-mediated cytotoxicity, as well as Fas ligand–mediated apoptosis.45 These factors in combination may contribute to gland destruction in autoimmune gastritis. An interesting animal model of autoimmune gastritis has been developed in mice in which CD4+ T cells target the β subunit of the H+,K+-ATPase.49 The risk of gastric adenocarcinoma in patients with AMAG is unclear. One recent study suggested a cancer risk of slightly more than 1% per year,50 which would favor periodic endoscopic screening for individuals known to have AMAG. However, other investigators have found cancer much less often and have questioned the costeffectiveness of cancer screening by endoscopy in AMAG.51,52 The importance of incomplete intestinal metaplasia (type III) as a predictor of gastric cancer also has been questioned.53 Thus, at what intervals AMAG patients should be screened, if at all, remains a matter of debate.54 Molecular events involved in the sequence from AMAG to intestinal metaplasia are beginning to be clarified. For example, the expression of the intestinal transcription factor CDX2 precedes expression of other intestinal-specific genes such as CDX1, alkaline phosphatase, MUC2, HD-5, and sucrase-isomaltase55 and thus may be an early trigger of the metaplastic process that precedes dysplasia and carcinogenesis.

CARDITIS 56,57 There has been recent attention to inflammation of the small rim of cardiac glands at the proximal portion of the stomach.56 The pathogenesis of carditis is currently con­ troversial.57 Inflammation of this gland area has been attributed to H. pylori gastritis, EMAG, AMAG, gastroesopha­ geal reflux disease, and other factors. Likewise, atrophy in this area, often accompanied by intestinal metaplasia, has been proposed to be a precursor of adenocarcinoma of the gastroesophageal junction (see Chapters 42 and 44). Der and associates56 reported on 141 patients in whom the cardiac mucosa could be identified in endoscopic biopsies. In this endoscopy population, all biopsies exhibited acute and/or chronic carditis. Nearly 80% of them had no evidence of H. pylori infection on simultaneous biopsies from the gastric body and antrum. H. pylori was present in 20 patients, 17 of whom had pan-gastritis and 15 of whom had H. pylori carditis. The severity of chronic carditis was related directly to 24-hour acid exposure of the lower esophagus, whereas acute carditis was related to H. pylori infection.

Chapter 51  Gastritis and Gastropathies INFECTIOUS GASTRITIDES 58-123 VIRUSES Cytomegalovirus60-64

Cytomegalovirus (CMV) is a human herpesvirus that may affect the esophagus, stomach, small bowel, colon, rectum, anus, liver, and gallbladder. CMV infection may occur in an immunocompetent patient.57 However, gastrointestinal CMV infection usually occurs in the immunocompromised patient. Eosinophilic gastroenteritis with cytomeglovirus infection has been reported in an immunocompetent child.62 Patients with malignant disease, immunosuppression (especially due to steroid therapy), transplants, and acquired immunodeficiency syndrome (AIDS) may experience lifethreatening CMV infections. Patients with CMV infection of the stomach may experience epigastric pain, fever, and atypical lymphocytosis. Upper gastrointestinal tract radiographic studies may reveal a rigid and narrowed gastric antrum suggestive of an infiltrating antral neoplasm. Endoscopic studies may reveal a congested and edematous mucosa of the gastric antrum, covered with multiple ulcerations, suggestive of gastric malignancy, submucosal antral mass, or gastric ulcer (Fig. 51-4). A hypertrophic and/or polypoid type of gastritis resembling Ménétrier’s disease with a similar type of protein-losing gastropathy has been described.58,59 Examination of biopsy specimens shows inflammatory debris, chronic active gastritis, and enlarged cells with CMV inclusion bodies indicative of an active infection (see Fig. 51-4). “Owl-eye” intranuclear inclusions are the hallmark of CMV infection in routine hematoxylin and eosin histologic preparations and may be found in vascular endothelial cells, mucosal epithelial cells, and connective tissue stromal cells. Multiple, granular, basophilic, cytoplasmic inclusions may also be present. Usual treatment with intravenous ganciclovir or foscarnet is of uncertain value (see Chapter 33).

Other Herpesviruses65-70

Gastric involvement with herpes simplex and varicellazoster virus is rare. Infected individuals experience the infection at an early age, and the virus remains dormant until reactivation. Activation has been related to radiation therapy, chemotherapy, lymphoma, and cancer. The typical immunocompromised patient may experience nausea, vomiting, fever, chills, fatigue, cough, and weight loss. An

A

acute abdomen caused by varicella-zoster virus–induced gastritis after autologous peripheral blood stem cell transplantation in a patient with non-Hodgkin’s lymphoma has been reported.67 Barium-air double-contrast radiographs show a cobblestone pattern, shallow ulcerations with a ragged contour, and an interlacing network of crevices filled with barium that corresponds to areas of ulceration. Upper gastrointestinal endoscopy reveals multiple, small, raised, ulcerated plaques or linear, superficial ulcers in a crisscrossing pattern, giving the stomach a cobblestone appearance. Grossly, the ulcers are multiple, small, and of uniform size. Microscopically, cytologic smears and biopsy specimens show numerous single cells and clumps of cells, with ground-glass nuclei and eosinophilic intranuclear inclusion bodies surrounded by halos. Brush cytology and biopsies should be performed at the time of endoscopy. Brush cytology has the advantage of sampling a wider area of mucosa because biopsies may not be representative. Treatment with acyclovir is reasonable but of unproven value. Human herpesvirus 7, a cause of roseola, is frequently present in the gastric mucosa but does not appear to cause gastritis.68 Epstein-Barr virus (EBV) may cause an acute gastritis with lymphoid hyperplasia.69 There is little evidence that EBV causes chronic gastritis.70

Measles71

Rare cases of morbilliform gastritis with giant cells of the Warthin-Finkeldey type have been described.71

Enterovirus72

Recently, it has been proposed that some patients with chronic fatigue syndrome are chronically infected with a noncytopathic, noncytolyic enterovirus that can be detected by immunostaining or by reverse transcriptase–polymerase chain reaction (PCR) techniques using gastric biopsy samples.72 Confirmatory studies are awaited.

BACTERIA Helicobacter pylori (see earlier and Chapter 50)

Phlegmonous (Suppurative) and Emphysematous Gastritis73-82 Phlegmonous gastritis is a rare bacterial infection of the submucosa and muscularis propria of the stomach. Acute necrotizing gastritis (gangrene of the stomach) is a rare, often fatal disease that is now thought to be a variant of phlegmonous gastritis. It has been suggested that acute

B

Figure 51-4.  Cytomegalovirus (CMV) gastritis. A, Low-power histopathology of CMV gastritis. An acute inflammatory infiltrate is present in the lamina propria. Glandular destruction and reactive glands are present. Cystic glands are also evident. B, High-power view of the cystic area deep in the mucosa shown in A. Several cytomegalic cells with the typical intranuclear and intracytoplasmic inclusions of cytomegalovirus are present.

849

850

Section VI  Stomach and Duodenum necrotizing gastritis begins as phlegmonous gastritis, producing primary necrosis and gangrene. Acute necrotizing gastritis and phlegmonous gastritis have been associated with a recent large intake of alcohol; upper respiratory tract infection; AIDS and other immunocompromised states; and an infected peritoneojugular venous shunt. Fulminant and fatal gas gangrene of the stomach in a healthy, live liver donor has been reported.76 Patients typically present with acute upper abdominal pain, peritonitis, purulent ascitic fluid, fever, and hypotension. Preoperative diagnosis is possible with plain film, ultrasonography, or computed tomography (CT), and gastroscopy with or without biopsy and culture of gastric contents may establish the diagnosis. Grossly, the stomach wall appears thick and edematous with multiple perforations, and the mucosa may demonstrate a granular, green-black exudate. Microscopically, the edematous submucosa reveals an intense polymorphonuclear infiltrate and numerous gram-positive and gramnegative organisms, as well as vascular thrombosis. The mucosa may demonstrate extensive areas of necrosis. The mortality rate of phlegmonous gastritis is close to 70%, probably because it is so often misdiagnosed and because treatment is initiated too late. The definitive treatment is resection or drainage of the stomach, combined initially with large doses of systemic broad-spectrum antibiotics directed against the most common organisms such as streptococci, Escherichia coli, enterobacters, other gramnegative bacilli, and Staphylococcus aureus. Emphysematous gastritis is a variant of phlegmonous gastritis in which the infection in the gastric wall is due to gas-forming organisms such as Clostridium welchii. Emphysematous gastritis associated with invasive gastric mucormycosis has been reported.77 Predisposing factors are gastroduodenal surgery, ingestion of corrosive materials, gastroenteritis, or gastrointestinal infarction. Radiographic studies (plain films, CT) show gas bubbles conforming to the contour of the stomach, often in the form of cystic gas pockets.82

Mycobacteria83-85

Gastric infection with Mycobacterium tuberculosis is a rare entity that usually occurs in association with pulmonary tuberculosis. Patients typically present with abdominal pain, nausea and vomiting, gastrointestinal bleeding, fever, and weight loss. Gastric tuberculosis associated with anemia has been reported.85 Gastric tuberculosis may be associated with gastric outlet obstruction or with bleeding from a tuberculous gastric ulcer. Radiographic studies reveal an enlarged stomach with narrowed, deformed antrum with prepyloric ulcerations. Upper endoscopy demonstrates ulcers, masses, or gastric outlet obstruction. Grossly, the stomach may demonstrate multiple small mucosal erosions, ulcers, an infiltrating mass (hypertrophic) form, a sclerosing inflammatory form, acute miliary dissemination, and pyloric obstruction either by extension from peripyloric nodes or by invasion from other neighboring organs. Biopsies show necrotizing granulomas with the presence of acid-fast bacilli, best demonstrated with Kinyoun acid-fast stain. Treatment is discussed in Chapter 107. Although Mycobacterium avium complex (MAC) is a common opportunistic bacterial infection among patients with AIDS, the stomach is rarely involved. Gastric MAC may be associated with a chronic gastric ulcer refractory to conventional antiulcer therapy. Patients may present with fever, night sweats, anorexia, weight loss, diarrhea, abdominal pain, chylous ascites, severe gastrointestinal hemorrhage, or chronic gastric ulcer. Serial CT scans of the abdomen may show mesenteric lymphadenopathy.

Endoscopy may show a chronic gastric ulcer, a coarsely granular duodenal mucosa, or fine white duodenal nodules. Microscopically, the gastric mucosa demonstrates numerous foamy histiocytes containing many acid-fast bacilli. Treatment of MAC is difficult and is discussed in Chapter 33.

Actinomycosis86-87

Primary gastric actinomycosis is a rare, chronic, progressive, suppurative disease characterized by formation of multiple abscesses, draining sinuses, abundant granulation, and dense fibrous tissue. Abdominal actinomycosis is more common and has a predilection for the terminal ileum, cecum, and appendix. The presenting symptoms include fever, epigastric pain, epigastric swelling, abdominal wall abscess with fistula, and upper gastrointestinal bleeding. Radiographic studies frequently suggest a malignant tumor or an ulcer. Endoscopy is suggestive of a circumscribed and ulcerated gastric carcinoma. Grossly, the resected stomach demonstrates a large, ill-defined, ulcerated mass in the wall of the stomach that measures up to 4 cm. Microscopically, multiple abscesses show the infective agent, Actinomyces israelii, a gram-positive filamentous anaerobic bacterium that normally resides in the mouth. A biopsy of a mass containing pus or a biopsy of a draining sinus may reveal actinomycosis. If the disease is recognized only by histologic examination, the prognosis is good. Prolonged (6- to 12-month) high-dose antibiotic treatment with penicillin or amoxicillin is indicated.

Syphilis88-94

The incidence of syphilis in the United States increased 34% from 13.7 to 18.4 cases per 100,000 persons between 1981 and 1989. Several case reports and small series emphasize the importance of the gastroenterologist and pathologist remaining alert to the protean manifestations of syphilis and familiar with the histopathologic pattern of the disease. Gastric involvement in secondary or tertiary syphilis is rarely recognized clinically, and its diagnosis by examination of endoscopic biopsy specimens has been reported infrequently. The features of syphilis in the stomach should be recognized because they can provide a window of opportunity for effective antibiotic therapy before the disease progresses and causes permanent disability. Patients typically present with symptoms of peptic ulcer disease, and the most common gastric complaint is upper gastrointestinal tract bleeding. Other diseases that may mimic gastric syphilis include benign ulcer disease, gastric carcinoma, gastric lymphoma, tuberculosis, and gastric Crohn’s disease. Gastric syphilis in the setting of human immunodeficiency virus (HIV) has been reported.92 The acute gastritis of early secondary syphilis produces the earliest radiologically detectable sign of the disease. Radiographs show a nonspecific gastritis with diffusely thickened folds that may become nodular with or without detectable ulcers. Strictures in the mid-stomach (“hourglass” stomach) may be present (Fig. 51-5A). Endoscopy shows numerous shallow, irregular ulcers with overlying white exudate and surrounding erythema (see Fig. 51-5B). The surrounding mucosa also demonstrates a nodular appearance. Gastroscopy may also demonstrate prominent, edematous gastric folds. Grossly, the stomach may be thickened and contracted and may show multiple serpiginous ulcers. Partial gastrectomy specimens may show compact, thick, mucosal rugae and numerous small mucosal ulcers. Microscopically, biopsies show severe gastritis with dense plasma cell infiltrate in the lamina propria, varying numbers of neutrophils and lymphocytes, gland destruction, vasculitis, and granulomas.

Chapter 51  Gastritis and Gastropathies

B

A

C

Figure 51-5.  Gastric syphilis (luetic gastritis). Film from an upper gastrointestinal series (A) showing a stricture in the mid-stomach (hourglass stomach), with antral deformity. Endoscopic appearance before (B) and 4 weeks after (C) penicillin therapy in another patient with gastric syphilis.

Warthin-Starry silver stain or modified Steiner silver impregnation stain reveals numerous spirochetes. Serum Venereal Disease Research Laboratory (VDRL) and Treponema immunofluorescence studies may be positive, and the Treponema pallidum gene may be detected by the PCR. Treatment with penicillin is highly effective (see Fig. 51-5C).

Other Bacteria95,96

Because approximately 25% of patients with chronic gastritis have no current or past evidence of infection with H. pylori or other Helicobacters such as H. heilmannii, other bacteria have been sought. One gram-negative bacillus, Acinetobacter lwoffi, is a common commensal that is normally not pathogenic in humans but has been proposed to cause gastritis in a manner analogous to H. pylori.84 A case of transient gastritis caused by the gram-positive enterococcus has also been described.85

FUNGI Candidiasis97,98

Fungal contamination of gastric ulcers with Candida species is not uncommon. Data from some studies suggest that fungal colonization in patients with gastric ulcers and chronic gastritis have little clinical significance, whereas others suggest that fungal infection aggravates and per­ petuates gastric ulceration. Endoscopically, gastric ulcers associated with Candida albicans tend to be larger in diameter and are more often suspected to be malignant than typical gastric ulcers. Diffuse superficial erosions may be noted. Fungal colonization of the gastrointestinal tract is frequent in patients with underlying malignancy and in immu-

nocompromised patients who have been treated with antibiotics or glucocorticoids but may occur also in immunocompetent patients. Symptoms are nonspecific. Massive growth of yeast organisms in the gastric lumen (yeast bezoar) is a potential complication of gastric surgical procedures, usually for peptic ulcer disease. Candida infection of the stomach may occur in alcoholic patients who ingest corrosive chemicals such as concentrated sulfuric acid and thiocyanates. Radiologic studies show tiny aphthoid erosions, which represent the earliest detectable radiographic change in gastric candidiasis. Aphthoid ulcers progress to deep linear ulcers. Grossly, the gastric mucosa demonstrates tiny aphthous erosions; widespread punctate, linear ulcerations; or gastric ulcers. Microscopically, the layer of necrotic fibrinoid debris demonstrates yeasts or pseudohyphae. The organisms can be seen in the hematoxylin and eosin stain; however, special stains such as periodic acid–Schiff-diastase stain or Gomori methenamine silver stain may be required. Treatment is usually not necessary, but if symptomatic candidiasis is suspected, fluconazole is reasonable but of unproven efficacy.

Histoplasmosis99-101

Progressive disseminated histoplasmosis is rare, occurring most frequently in the very young or the older adult or in those with immunodeficiency. Although disseminated histoplasmosis can involve any portion of the gastrointe­ stinal tract, gastric involvement is rare. Hypertrophic gastric folds or a mass that mimics a gastric carcinoma may be associated with gastric histoplasmosis or disseminated histoplasmosis has been reported.100 Radiographic studies may demonstrate an annular infiltrating lesion of the

851

852

Section VI  Stomach and Duodenum stomach, and endoscopy may demonstrate enlarged and reddened gastric folds. Biopsy specimens show an intensive infiltration of macrophages containing Histoplasma cap­ sulatum. Gastric histoplasmosis has also been associated with a fatal hemorrhage from a gastric ulcer. Treatment with intravenous amphotericin B is appropriate.

Phycomycosis102-104

Gastric phycomycosis (also called zygomycosis or mucormycosis) is a rare and highly lethal fungal infection. Phy­ comycosis usually affects the paranasal sinuses, central nervous system, or lungs and is rarely confined to the gastrointestinal tract. Risk factors include malnutrition, immunosuppression, antibiotic therapy, and acidosis, usually diabetic ketoacidosis. A nosocomial outbreak of gastric mucormycosis due to contamination of wooden tongue depressors by Rhizopus microspores has been reported. Most patients presented with upper gastrointe­ stinal bleeding.103 Gastric phycomycosis can be classified as invasive or noninvasive (colonization). The former is characterized by deep invasion of the stomach wall and by blood vessel involvement with the fungus. Abdominal pain is the most frequent presenting complaint. In the noninvasive type, the fungus colonizes the superficial mucosa without causing an inflammatory response. Grossly, surgical specimens from affected patients reveal hemorrhagic necrosis involving the mucosa and gastric wall. Microscopically, nonseptate 10- to 20-µm hyphae branched at right angles are present in the tissue and they infiltrate into blood vessel walls. Treatment is resection of the affected necrotic portion of the stomach. Unfortunately, invasive gastric phycomycosis is almost always fatal. 105-106

Aspergillosis

Acute aspergillus gastritis in a case of fatal aplastic anemia and a fatal case of pseudo-membranous gastritis due to this organism have been reported.

PARASITES (see also Chapters 109 and 110) Cryptosporidiosis107-110

Cryptosporidiosis may rarely involve the stomach. Gastric outlet obstruction and antral stricture due to cryptosporidiosis have been reported in patients with AIDS and diarrhea. A case of cryptosporidiosis-associated erosive gastritis in a patient with HIV infection also has been reported. Also, cryptosporidiosis associated with the immunocompromised state and small cell lung cancer has been reported.110

kis. Hundreds of cases of anisakiasis have been diagnosed in Japan, and the number of reported cases in the United States has also increased. The parasite may migrate into the wall of the stomach, small intestine, or colon. Typically, patients present with sporadic epigastric pain or have no symptoms at all. Gastric perforation due to chronic gastric aniskiasis has been reported.120 Misdiagnosis is common. Some patients may experience a mild peripheral eosinophilia. Endoscopy may show firm, yellowish submucosal masses with erosions.119 Radiographic studies may reveal notched-shadow defects suggestive of a gastric tumor. Grossly, the stomach demonstrates multiple erosive foci with hemorrhage and small 5- to 10-mm gastric lesions in the stomach wall. Microscopically, sections of the stomach show a marked eosinophilic granulomatous inflammatory process with intramural abscesses and granulation tissue. The eosinophilic abscess may contain a small worm measuring 0.3 mm in diameter, which can be identified as the larval form. The diagnosis may be confirmed by a serodiagnostic test for human anisakiasis on the patient’s serum when the larvae may no longer be detectable by endoscopy.

Ascariasis121,122

Although gastric ascariasis is rare, patients have experienced chronic, intermittent gastric outlet obstruction caused by roundworms (Ascaris lumbricoides) inhabiting the stomach. Gastric ascariasis has also been associated with upper gastrointestinal hemorrhage with endoscopic examination showing several Ascaris worms in the stomach and duodenum.

Hookworm123

Endoscopic discovery and capture of Necator americanus in the stomach has been reported.123

GRANULOMATOUS GASTRITIDES A variety of granulomatous diseases affect the stomach. Crohn’s disease is the most common (Fig. 51-6) and is discussed later and also in Chapter 111. The differential diagnosis of granulomatous gastritis also includes sarcoidosis, as well as rarer conditions such as xanthogranulomatous

Strongyloidiasis111-115

The stomach is rarely affected by Strongyloides stercoralis. However, the organisms may colonize the intact gastric mucosa and may be associated with a bleeding peptic ulcer. S. stercoralis hyperinfection has been associated with cimetidine therapy in an immunosuppressed patient and was diagnosed by endoscopic gastric biopsy. Diagnosis can be made by endoscopic biopsy, examination of stools, examination of duodenal aspirate, and examination of peripheral smear with elevated eosinophil count. A histologic diagnosis of strongyloidiasis must be taken into consideration when examining gastric and duodenal biopsies in immunocompromised patients.113 Disseminated strongyloidiasis can be rapidly fatal. Treatment is discussed in Chapter 110.

Anisakiasis116-120

Invasive anisakiasis may occur after the ingestion of raw marine fish containing nematode larvae of the genus Anisa-

Figure 51-6.  Histopathology of granulomatous gastritis in a patient with Crohn’s disease. A noncaseating granuloma is present within the lamina propria. (Hematoxylin and eosin, ×200.)

Chapter 51  Gastritis and Gastropathies gastritis, foreign bodies,105 lymphoma,106 Whipple’s disease (see Chapter 106),107 Langerhans cell histiocytosis (gastric eosinophilic granuloma),108 granulomatous vasculitis109 (Churg-Strauss syndrome), and chronic granulomatous disease of childhood.110 An isolated, idiopathic granulo­ matous gastritis also occurs.111 Some of these latter cases may evolve to Crohn’s disease or sarcoidosis over time. Other cases of “idiopathic” granulomatous gastritis appear to be due to H. pylori infection and may resolve, albeit slowly, following appropriate antibiotic therapy.112,113 Idiopathic granulo­matous gastritis can be associated with gastric cancer.114

SARCOIDOSIS 124-137

Gastrointestinal manifestations of sarcoidosis are uncommon (see Chapter 35). Sarcoidosis is a systemic disease, frequently involving the lungs, lymph nodes, skin, and eyes. Diagnosis of sarcoidosis of the stomach cannot be made with confidence in the absence of disease in other organs. The stomach (usually the antrum) is the most common part of the gastrointestinal tract affected in sarcoidosis, being involved in 10% of cases.137 Affected patients, usually in the third to fifth decades of life, typically present with epigastric pain, nausea, vomiting, and weight loss. Occasionally they present with massive hemorrhage. Gastric sarcoidosis may cause pyloric outlet obstruction, achlorhydria, and pernicious anemia. Radiographically, gastric sarcoidosis may mimic the diffuse form of gastric carcinoma (“linitis plastica”) or Ménétrier’s disease. Endoscopy may reveal a narrow distal half of the stomach with multiple prepyloric ulcers or erosions, atrophy, thick gastric folds with a diffuse cobblestone appearance, or normal mucosa associated with microscopic granulomas. Surgical specimens of patients with gastric sarcoidosis show a thickened stomach wall with foci of erosions and ulcers. Microscopically, mucosal biopsies show multiple noncaseating granulomas. However, the presence of noncaseating granulomas in gastrointestinal tissue is a nonspecific finding and special stains should be performed to rule out infections, especially tuberculosis. In some cases it may be difficult to differentiate gastric sarcoidosis from gastric Crohn’s disease or from isolated, idiopathic granulomatous gastritis. Glucocorticoid therapy is the cornerstone of treatment for gastric sarcoidosis. Subtotal gastric resection is reserved for patients with obstruction and severe hemorrhage.

XANTHOGRANULOMATOUS GASTRITIS 138-139

Xanthogranulomatous gastritis is characterized by inflammation of the gastric wall by foamy histiocytes, inflam­ matory cells, multinucleated giant cells, and fibrosis. The destructive inflammatory process may extend into adjacent organs and simulate a neoplasm. Xanthogranulomatous gastritis has been associated with xanthogranulomatous cholecystitis. Xanthogranulomatous gastritis with pseu­ dosarcomatous changes mimicking a neoplasm has been reported.139

DISTINCTIVE GASTRITIDES COLLAGENOUS GASTRITIS 140-148

Subepithelial fibrosis has been reported in the colon (collagenous colitis), small bowel (collagenous sprue), and stomach (collagenous gastritis). Collagenous gastritis is a rare form of gastritis, and fewer than 20 cases have been

1

2 Figure 51-7.  Histopathology of collagenous gastritis (top, Hematoxylin and eosin, ×200; bottom, Masson trichrome, ×400). The subepithelial thickening of the collagen band is apparent. (From Wang HL, Shah AG, Yerian LM, et al. Collagenous gastritis: An unusual association with profound weight loss. Arch Pathol Lab Med 2004; 128:229.)

reported in the literature. Collagenous gastritis may be associated with collagenous colitis, lymphocytic colitis, and celiac disease. Patients may experience intermittent epigastric abdom­ inal pain, hematemesis, hematochezia, anemia, diarrhea, hypotension, or weight loss. Collagenous gastritis in a Korean child who had been receiving treatment for refractory iron deficiency anemia has been reported.144 Upper gastrointestinal barium radiography may demonstrate an abnormal mucosal surface with a mosaic-like pattern in the body of the stomach, corresponding to mucosal nodularity. Endoscopy may reveal multiple diffusely scattered, discrete submucosal hemorrhages; erosions; and nodularity of the body of the stomach along the greater curvature. Biopsy specimens from the body and antrum of the stomach reveal a patchy, chronic, superficial gastritis, focal atrophy, and focal deposition of collagen in the subepi­ thelial region of the lamina propria, which measures from 20 to 75 mm thick (Fig. 51-7). Tiny erosions of the surface epithelium are present, and the inflammatory infiltrate consists of mainly plasma cells, intraepithelial lymphocytes and eosinophils, together with marked hypertrophy of muscularis mucosae.144 Little is known about the etiology, natural history, and proper treatment of this rare condition.

LYMPHOCYTIC GASTRITIS 149-157

Lymphocytic gastritis is characterized by a dense lym­ phocytic infiltration of surface and pit gastric epithelium (Fig. 51-8A). Lymphocytic gastritis is related to an endoscopic form of gastropathy known as varioliform gastritis.

853

854

Section VI  Stomach and Duodenum

A

B

Figure 51-8.  Histopathologic examples of two distinctive forms of gastritis: lymphocytic gastritis (A) and eosinophilic gastritis (B). A, High-power view of the antral mucosa shows numerous dark-staining mononuclear cells with striking intraepithelial lymphocytosis. B, Numerous eosinophils are noted within the lamina propria and within the walls and lumina of the gastric glands. The patient also had peripheral blood eosinophilia. (Hematoxylin and eosin.) (Courtesy Pamela Jensen, MD, Dallas, Tex.)

Lymphocytic gastritis is also seen in H. pylori infection and in celiac disease. Recent findings provide compelling evidence that lymphocytic gastritis may occur as a manifestation of celiac disease, and thus the lymphocytic infiltration of celiac disease may affect gastric epithelial mucous cells. Lymphocytic gastritis in untreated celiac disease may be associated with functional changes such as increased permeability. Gastric biopsies from 10 of 22 patients with diarrhea or malabsorption and small bowel changes characteristic of celiac disease showed striking lymphocytic gastritis. Following institution of a gluten-free diet, lymphocytic gastritis resolves after approximately two years. Lymphocytic gastritis has also been attributed to an atypical host immune response to H. pylori. H. pylori eradication treatment in patients with lymphocytic gastritis causes significant improvement in the gastric intraepithelial lymphocytic infiltrate, corpus inflammation, and dyspeptic symptoms. H. pylori may be the cause of some cases of protein-losing hypertrophic lymphocytic gastritis. The disease may resolve clinically, endoscopically, and pathologically with therapeutic eradication of H. pylori in some patients. The relationship between lymphocytic gastritis and gastric lymphoid hyperplasia, which also is associated with H. pylori, is not yet clear.154 Patients with gastric lymphoma have a significantly increased prevalence of lymphocytic gastritis due to H. pylori. Because intraepithelial lymphocytes are speculated to have a role in the regulation of normal mucosal inflammatory reaction, they may also participate in the pathogenesis of mucosal lymphoma. In a 10-year follow-up study of lymphocytic gastritis, the patients with lymphocytic gastritis also appeared to have a significant increase in the grade of intestinal metaplasia in the corpus mucosa. In another study, lymphocytic gastritis was more prevalent in patients with gastric adenocarcinoma (16 of 30 cases [12.3%]) than in unselected patients undergoing endoscopy (0.8% to 2.5%). Lymphocytic gastritis with anasarca and venous thrombosis at the confluence of the splenic and mesenteric veins has been reported.153 Endoscopy in lymphocytic gastritis shows thick mucosal folds, nodularity, and aphthous erosion, historically known as varioliform gastritis. Gastric biopsies show expansion of the lamina propria by an infiltrate of plasma cells, lymphocytes, and rare neutrophils. These findings

may be seen in the antral mucosa only, body mucosa only, or in antral as well as body mucosa. The surface and superficial pit epithelium shows a marked intraepithelial infiltrate with T lymphocytes, with flattening of the epithelium and loss of apical mucin secretion. Quantification of epithelial lymphocytes revealed 46.5 lymphocytes per 100 epithelial cells in lymphocytic gastritis, compared with 3.5 lymphocytes per 100 cells in normal controls and 5.1 lymphocytes per 100 cells in disease controls including patients with H. pylori gastritis. The immunohistochemistry profile of lymphocytic gastritis in celiac disease and H. pylori infection and the interplay between infection and inflammation has been reported. Only intraepithelial lymphocytes positive for CD3 and CD8 were increased significantly in celiac disease patients with or without H. pylori infection.155

EOSINOPHILIC GASTRITIS 158-164

Eosinophilic gastrointestinal disorders are characterized by marked tissue eosinophilia in the absence of known causes for eosinophilia or other gastrointestinal disorders. These disorders include eosinophilic esophagitis, eosinophilic gastritis, eosinophilic enteritis, and eosinophilic colitis. Eosinophilic gastroenteritis is a rare condition of unknown etiology characterized by peripheral eosinophilia, eosinophilic infiltration of the gastrointestinal tract, and gastrointestinal symptomatology. It is discussed in detail in Chapter 27. The gastric mucosa is frequently involved, and thus eosinophilic gastritis is one of the manifestations of eosinophilic gastroenteritis. Eosinophilic gastroenteritis is classified according to the layer of gastrointestinal tract involved (i.e., mucosal layer disease, muscle layer disease, and subserosal disease). Mucosal involvement may result in abdominal pain, nausea, vomiting, diarrhea, weight loss, anemia, protein-losing enteropathy, intestinal perforation, and iron deficiency anemia secondary to gastrointestinal blood loss. Patients with muscular layer disease generally have obstructive symptoms, and patients with subserosal eosinophilic infiltration develop eosinophilic ascites. Patients with gastric involvement frequently present with pyloric obstruction. Radiographic studies may demonstrate thickened mucosal folds, nodularity, or ulcerations. Endoscopy may reveal normal-appearing mucosa or hyperemic edematous mucosa with surface erosions or prominent gastric folds.

Chapter 51  Gastritis and Gastropathies Eosinophilic gastritis simulating gastric car­cinoma has been reported.161 Gastric mucosal biopsies are critical to the diagnosis and show marked eosinophilic infiltration, eosinophilic pit abscesses, necrosis with numerous neutrophils, and epithelial regeneration (see Fig. 51-8B). Abnormal eosinophilic infiltration is defined as at least 20 eosinophils per highpower field either diffusely or multifocally. A full-thickness surgical biopsy is necessary for the diagnosis of muscle layer disease. As discussed in Chapter 27, patients with disabling symptoms can be effectively treated with glucocorticoids (after other systemic disorders associated with peripheral eosinophilia have been excluded) or possibly with oral sodium cromoglycate. Surgical intervention may be required in patients with obstructive complications or refractory disease. Collagenous colitis associated with eosinophilic gastritis in a four-year-old girl has been reported.

MISCELLANEOUS FORMS OF GASTRITIS GASTRITIS IN INFLAMMATORY BOWEL DISEASE (CROHN’S AND ULCERATIVE COLITIS) 165-177

Crohn’s disease is the most common disease associated with granulomatous gastritis.111 Crohn’s disease involving the stomach is uncommon, however, and almost always occurs together with lower intestinal disease (see Chapter 111). Cases may be isolated to the stomach or the stomach and duodenum. The diagnosis of isolated Crohn’s disease of the stomach should be made with caution,165 and close followup is indicated for the subsequent development of Crohn’s disease elsewhere in the gastrointestinal tract or of other granulomatous diseases such as sarcoidosis. Symptoms are nonspecific and include nausea and vomiting, epigastric pain, anorexia, and weight loss. Radiologic studies show antral fold thickening, antral narrowing, shallow ulcers (aphthae), or deeper ulcers. Involvement of the stomach from adjacent ileal or colonic disease segments is best visualized by radiologic examination. Endoscopy allows better visualization of mucosal defects and is characterized by reddened mucosa, irregularly shaped ulcers, and erosions in a disrupted mucosal pattern. Nodular lesions occur and often reveal the presence of erosions on the top of nodules. An atypical cobblestone pattern may be associated with the nodules surrounded by fissure-like ulceration. In contrast with peptic ulcers, the ulcerations and erosions of Crohn’s disease are frequently serpiginous or longitudinal, rarely round or oval. Ulcerations or erosions associated with Crohn’s disease of the stomach typically are most commonly located in the antrum and the prepyloric region. The microscopic features of surgical specimens of gastric Crohn’s disease can be, but are not always, similar to those in the ileum or colon (see Chapter 111). They include granulomas, transmural chronic inflammation, ulcers, and marked submucosal fibrosis (see Fig. 51-6). Granulomas may be present in endoscopically normal antral mucosa. In the past few years it has been recognized that although Crohn’s patients may have granulomatous gastritis, H. pylori–associated gastritis (HAG) or focal active antral gastritis is even more common.166-175 The majority of pediatric inflammatory bowel disease (IBD) patients (Crohn’s or ulcerative colitis [UC]) have HAG and/or focal active gastritis, although the latter is more prevalent in Crohn’s patients

than in UC patients.174 The focal active gastritis is not due to H. pylori and is accompanied by macrophages in the center of the focal lesion and mast cells at its periphery.169 It is as yet unclear whether the type of gastritis identified in pediatric IBD patients can reliably distinguish Crohn’s colitis from UC (except for granulomatous gastritis, which favors Crohn’s)168 and even less is certain in adult IBD patients. Therapy of gastritis in Crohn’s disease should be driven by gastric symptoms and not solely by demonstration of gastritis on mucosal biopsy. Double-blinded, randomized, controlled clinical trials of pharmacologic agents are lacking in gastric and gastroduodenal Crohn’s disease. Proton pump inhibitors should be the first therapy for symptomatic patients.176,177 The effectiveness of glucocorticoids, immunosuppressive medications, and anti–TNF-α drugs such as infliximab has not been clearly demonstrated. Gastric outlet obstruction refractory to medical therapy can be treated by gastroenterostomy, ideally laparoscopically. Treatment of Crohn’s disease is discussed in more detail in Chapter 111.

GASTRITIS CYSTICA PROFUNDA 178-182

Gastritis cystica profunda is a rare complication of partial gastrectomy with gastrojejunostomy for benign peptic ulcer disease and typically occurs at the site of the gastroenterostomy. Gastritis cystica profunda may also develop in the unoperated stomach, and chronic atrophic gastritis may be a risk factor for it. Radiography and endoscopy typically demonstrate multiple exophytic gastric masses that simulate a malignancy. Endoscopic ultrasound may assist in the diagnosis. Grossly, the gastric mucosal surface demonstrates multiple nodules and exophytic masses. On section, the gastric wall is thick and multiple cysts are present. Microscopically, the mucosa is characterized by foveolar hyperplasia, and cystic glands extend through a disrupted muscularis mucosae into the submucosa and, rarely, the muscularis propria (Fig. 51-9). Gastritis cystica profunda may be associated with gastric stump adenocarcinoma. Removal of this lesion by snare polypectomy after submucosal injection to elevate the lesion has been reported.179 The disease may also coexist with gastric inverted hyperplastic polyp, and the latter may in fact be a variant of gastritis cystica profunda.182 Gastritis cystica profunda, if

Figure 51-9.  Gastritis cystica profunda. Note the cystic dilatation of numerous gastric glands that extend through the muscularis mucosae (arrow), simulating a gastric carcinoma.

855

856

Section VI  Stomach and Duodenum present, should lead to a thorough examination for a synchronous or metachronous gastric cancer, although exact recommendations for surveillance interval are not clear.

GASTRIC GRAFT-VERSUS-HOST DISEASE 183-187 (see also Chapter 34)

Graft-versus-host disease (GVHD) may affect any portion of the gastrointestinal tract; therefore, upper as well as lower gastrointestinal tract biopsies may provide diagnostic information not evidenced in biopsy from a single site. GVHD most often occurs after allogeneic bone marrow transplantation and rarely after solid transplantation. GVHD occurs in an acute and chronic form. Acute GVHD occurs between post-transplant days 21 and 100, whereas chronic GVHD occurs after day 100. The gastrointestinal tract is commonly affected in acute GVHD, especially the small and large intestines, and to a lesser extent the stomach and esophagus. Gastric GVHD is characterized by nausea, vomiting, and upper abdominal pain without diarrhea. Stomach biopsies may be necessary to diagnose GVHD in patients with upper gastrointestinal symptoms but no diarrhea and normal rectal biopsy specimens. The basic pathologic lesion consists of necrosis of single cells (apoptotic bodies) in the crypts of the large and small intestinal mucosa and in the neck region of the gastric mucosa. The necrosis consists of an intraepithelial vacuole filled with karyorrhectic debris and fragments of cytoplasm.

ALLERGIC GASTRITIS 188-190

Whether allergies to certain foods can lead to gastritis and whether the gastritis correlates with upper gastrointestinal symptoms are unclear (see Chapter 9). Children with food allergy as diagnosed by an open elimination challenge test have no higher incidence of gastritis than children without food allergy.188 An exception may be infants who are allergic to cow’s milk protein, in whom hematemesis and endoscopic signs of gastritis are common.189

REACTIVE GASTROPATHIES (ACUTE EROSIVE GASTRITIS) The gastric mucosa may be damaged by a variety of agents or factors that do not produce a significant inflammatory infiltrate. Because of the paucity of inflammatory cells, the mentioned lesions are best referred to as reactive gastropathies, as opposed to acute erosive gastritis. The gastric mucosa in patients who experience a reactive gastropathy demonstrates a spectrum of hemorrhages, erosions, and ulcers. Erosions and ulcers are frequently multiple, and the base of these lesions often stains dark brown owing to exposure to acid. Grossly, most erosions and acute ulcers appear as welldefined hemorrhagic lesions 1 to 2 mm in diameter. If the insult is severe, the mucosa between the lesions is intensely hemorrhagic. Microscopically, an erosion demonstrates superficial lamina propria necrosis. An acute ulcer is an area of necrosis that extends to the muscularis mucosa. Foveolar hyperplasia, the reactive epithelial changes secondary to regeneration of the mucosa (Fig. 51-10), is often associated with glands with atypical nuclei that can be misdiagnosed as dysplasia or carcinoma. The diagnosis of neoplasia in a background of necrosis, cellular debris, and granulation tissue should be made with utmost caution. The biopsy procedure itself may induce tissue hemorrhage; thus, subepithelial hemorrhage should involve more

Figure 51-10.  Histopathology of foveolar hyperplasia. The gastric pits show an elongated, corkscrew appearance.

than one fourth of a biopsy specimen to be considered significant.17

MEDICATIONS AND TOXINS 191-201

Aspirin (even in low daily or less-than-daily doses) and NSAIDs (including cyclooxygenase [COX]-2 selective inhibitors) are the most common causes of reactive gastropathy (see Chapter 52). Oral iron therapy may rarely cause mild endoscopic abnormalities consisting of erythema, small areas of subepithelial hemorrhage, and erosions.191 Oral potassium chloride may also be associated with endoscopic erosions.192 Endoscopic petechiae, erosions, and erythema have been associated with long-term fluoride ingestion.195 Bisphosphonates for osteoporosis or Paget’s disease can also cause gastric erosions, although their clinical significance is uncertain.196,197 These drugs exacerbate gastric damage from NSAIDs such as naproxen as well. Various intensive cancer chemotherapy drugs given to children with leukemia, lymphoma, or solid tumors are associated with a hemorrhagic or erosive gastropathy and histologic evidence of inflammation, but cause and effect have not been clearly established in these ill individuals.198 Reactive gastric epithelial atypia and gastric ulceration may be associated with hepatic arterial infusion chemotherapy for metastatic disease to the liver.199,200 The marked epithelial atypia that results may erroneously be interpreted as carcinoma and lead to unnecessary surgery. Toxin ingestion of heavy metals such as mercury sulfate poisoning may cause an erosive or ulcerative gastropathy with hematemesis.201 Corrosive gastric injuries are discussed in Chapter 25.

ALCOHOL 202-209

After acute alcohol ingestion, subepithelial hemorrhages are seen frequently at endoscopy, typically without prominent mucosal inflammation on biopsy specimens (Fig. 51-11).202 Gastric biopsy specimens obtained from patients with chronic alcoholism have shown a higher prevalence of chronic antral gastritis due to H. pylori, with almost complete normalization of histologic findings after treatment.203,204 The combined effects of alcohol and the NSAID ibuprofen were associated with more gastric mucosal damage by endoscopic assessment than with either agent alone. The combination of alcohol and aspirin also caused more damage in the stomach than either agent alone, though not to a signifi-

Chapter 51  Gastritis and Gastropathies ing the gastric antrum.212 GAVE does not respond as readily as PHG to measures that reduce portal pressure.212

COCAINE 215-217

Gastrointestinal hemorrhage due to erosion throughout the gastric fundus, duodenal bulb has been reported with Gastrointestinal hemorrhage or pyloric cocaine is well described.215-217

diffuse exudative body, antrum, and crack cocaine use. perforation due to

STRESS 218

Erosions of the gastric mucosa may occur rapidly after major physical or thermal trauma, shock, sepsis, or head injury. These are often referred to as stress ulcers and are discussed in Chapters 52 and 53.

RADIATION 218-219 (see also Chapter 39)

Figure 51-11.  Histopathology of alcoholic gastropathy. Hemorrhage is confined to the superficial portion of the mucosa, and there is a paucity of inflammatory cells.

cant degree.205 Alcohol appeared to be an acute triggering factor in 35% of patients admitted to an intensive care unit for massive upper gastrointestinal bleeding in Sweden.206 Chronic alcohol ingestion was related to an increased risk of chronic atrophic gastritis and hypochlorhydria in a study from Poland.207 Recent alcohol ingestion was also found to be a risk factor for gastric erosions and ulcers in cirrhotic patients referred for upper endoscopy.208 Other alcohols, besides ethanol, can injure the stomach. Even the topical application of isopropyl alcohol (rubbing alcohol) used to cool a child with fever has resulted in hemorrhagic gastropathy with hematemesis.209

PORTAL HYPERTENSIVE GASTROPATHY 210-214

As discussed in more detail in Chapters 19 and 90, gastric mucosal lesions are common in portal hypertension, occurring in up to 65% of cirrhotic patients, and represent an important cause of gastrointestinal blood loss.210 Portal hypertensive gastropathy was found in 33% of children after surgery for biliary atresia. Risk factors for portal hypertensive gastropathy were frequent endoscopic treatment of gastroesophageal varices, liver dysfunction, and hypersplenism.213 Biopsies show vascular ectasia and congestion in the mucosal layer without a significant degree of inflammatory infiltrate. Portal hypertensive gastropathy (PHG) is a risk factor not only for upper gastrointestinal bleeding but also for gastroduodenal erosions and ulcers in cirrhotic patients.208 It has been suggested that perturbations in the tissue levels of TNF-α, prostaglandins, endothelin, and nitric oxide/peroxynitrite participate in the vascular congestion and mucosal damage characteristic of PHG.211 As discussed in Chapter 90, lowering portal pressure pharmacologically or by creation of a portosystemic shunt effectively treats PHG and reduces bleeding. Some patients with cirrhosis and portal hypertension have gastric antral vascular ectasia (GAVE), which can bleed and is sometimes difficult to distinguish from PHG involv-

Radiation effects on the stomach depend on the cell kinetics of the gastric mucosa, as well as the dose of the radiation. The gastric mucosal response to radiation is unique, however, in that the most radiosensitive epithelial cells are the differentiated cells (parietal and chief cells) rather than the germinative cells in the mucous neck region. Radiation injury to the stomach can be classified into acute (less than six months) and chronic (more than one year) phases. It is thought that the tolerance level for radiation-induced gastric ulceration is approximately 4500 cGy. After a gastric dose of 5500 cGy or more, 50% of patients will develop clinical evidence of gastric ulcer formation.218 Radiation-induced gastric ulcers are usually solitary, from 0.5 to 2 cm in diameter, and located in the antrum. Massive hemorrhagic gastropathy requiring endoscopically administered therapy to control the bleeding has been reported.219

BILE REFLUX 220-231

Bile reflux into the stomach is common after partial gastrectomy with anastomosis to the duodenum (Billroth I) or jejunum (Billroth II) and after truncal vagotomy and pyloroplasty for peptic ulcer (see Chapter 53). It has even been reported after parietal cell vagotomy.220 Bile reflux gastropathy also may occur after cholecystectomy or sphincteroplasty, which allows the continuous exposure of bile to the duodenum with the potential for duodenogastric reflux. Occasionally, bile reflux gastropathy is observed in adult or pediatric patients who have not had surgery.221,223 Interleukins, particularly IL-8 and perhaps IL-6, may participate in the gastric damage.224,225 Bile reflux contributes to muscosal lesions in the stomach and may facilitate H. pylori colonization in the corpus region.222 Endoscopy shows swelling, redness, erosions, and bile staining of the gastric mucosa. Biopsy specimens show foveolar hyperplasia, dilated cystic glands, atypical glands that may be misdiagnosed as dysplasia or carcinoma, and a paucity of acute and chronic inflammatory cells. Gastric atrophy may result and increase the risk of carcinoma in the gastric stump (see Chapter 54). In fact, bile reflux into the unoperated stomach has been proposed to be a risk factor for intestinal metaplasia in the distal stomach, at the gastroesophageal junction (cardia), and in the distal esophagus (Barrett’s esophagus).226-228 Unfortunately, bile-diverting procedures performed because of severe bile gastropathy do not reverse gastric atrophy or intestinal metaplasia.229 It may be worthwhile, at the time of the original gastric surgery for gastric cancer or peptic ulcer, to construct a 30-cm Roux-en-Y limb176 or perform a 10- to 12-cm isoperistaltic jejunal interposition230 to prevent bile gastropathy and subsequent metaplastic changes.

857

858

Section VI  Stomach and Duodenum In selected, previously unoperated patients with primary bile gastropathy, surgery using a Roux-en-Y choledochojejunostomy without gastric resection has been successful.221 In unoperated patients with bile gastropathy following cholecystectomy, the proton pump inhibitor rabeprazole and sucralfate were equally effective in relieving symptoms and improving gastroscopic evidence of mucosal damage as compared with observation alone (no placebo was given).231 At present, lacking definitive studies, medical therapy should precede surgical therapy for bile gastropathy occurring in the unoperated stomach such as it does spontaneously or after cholecystectomy or biliary sphincterotomy.

ISCHEMIA 232-237

Chronic ischemic gastropathy may occur secondary to chronic mesenteric insufficiency and can be reversed after a revascularization operation.191,192 Chronic ischemic gastropathy, as well as chronic ischemic gastric ulcers, may also occur in association with atheromatous embolization.193,194 Athletes involved in intense physical activity, especially long-distance running, may experience recurrent ischemic gastropathy and chronic gastrointestinal bleeding with anemia.195,196

PROLAPSE 238-239

The mucosa of the gastric cardia may prolapse into the esophageal lumen during retching and vomiting.238 The resulting mechanical injury to the cardia has been proposed to be a cause of upper gastrointestinal hemorrhage, but this association has been questioned.239 Esophagoscopy may demonstrate the prolapsed gastric mucosa. The congested mucosa may show erosions and superficial ulcerations.

LINEAR EROSIONS IN A HIATAL HERNIA (CAMERON LESIONS) 240-241

Linear gastric erosions in a hiatal hernia are discussed in Chapters 19, 24, and 52.

AGING GASTROPATHY 242,243

Normal aging is associated with impaired gastric mucosal defense against injury in animals and humans.242 Two proteins have been implicated in aging gastropathy: PTEN (phosphatase and tensin homolog deleted on chromosome 10), which is overexpressed with aging, and survivin, which is underexpressed.243

HYPERPLASTIC GASTROPATHIES 244-251 MÉNÉTRIER’S DISEASE AND HYPERPLASTIC, HYPERSECRETORY GASTROPATHY

Hyperplastic gastropathy is a rare condition characterized by giant gastric folds associated with epithelial hyperplasia. Two clinical syndromes have been identified: Ménétrier’s disease and a variant of it referred to as hyperplastic, hypersecretory gastropathy, and Zollinger-Ellison syndrome, which is discussed in Chapter 32. Figure 51-12A and B demonstrates enlarged gastric folds in these conditions. Ménétrier’s disease is typically associated with proteinlosing gastropathy (see Chapter 28) and with hypochlor­ hydria, whereas the hyperplastic, hypersecretory variant is associated with increased or normal acid secretion and parietal and chief cell hyperplasia, with or without excessive gastric protein loss. Other more common conditions can also cause enlarged gastric folds or protein-losing gastropathy including gastric

neoplasm (lymphoma, carcinoma), granulomatous gastri­ tides, gastric varices, infectious gastritis (particularly H. pylori and CMV58), eosinophilic gastritis, and ZollingerEllison syndrome. The enlarged gastric folds in Ménétrier’s disease are due to foveolar cell hyperplasia, edema, and variable degrees of inflammation. Patients may present with weight loss, epigastric pain, vomiting, anorexia, dyspepsia, hematemesis, and positive fecal occult blood tests. Most patients with a clinical syndrome associated with hyperplastic gastritis showed histology typical for the syndrome; however, clinical-histologic concordance was not absolute.247 The mechanism responsible for the low gastric acid secretion is unclear, but it could be related to overexpression of transforming growth factor-α (TGF-α), a ligand for epidermal growth factor receptor (EGFR), a receptor tyrosine kinase.244 Although mucus hypersecretion is often seen in Ménétrier’s disease, abnormalities in mucins have not been consistent in a few cases that have been examined.245,246 Ménétrier’s disease may be associated with hypertrophic lymphocytic gastritis and a carcinoid-like syndrome due to increased gastric mucosal production of prostaglandin E2 concentrations. Ménétrier’s disease may be self-limited and may completely resolve in patients younger than 10 years of age and when it occurs in the postpartum period. The cause of Ménétrier’s disease of childhood may be infection with CMV and activation of TGF-α. The risk of developing carcinoma in association with Ménétrier’s disease is an open question, and some authors question whether there is a significant risk. Of the 200 cases of Ménétrier’s disease reported in the literature, 30 (15%) have been associated with carcinoma. The mucosa of patients with Ménétrier’s disease demonstrates irregular hypertrophic folds that involve the entire gastric body. The mucosa also demonstrates a swollen, spongy appearance subdivided by creases, creating a picture similar to cerebral convolutions. A polypoid variant of Ménétrier’s disease that resembles multiple hyperplastic gastric polyps has been described (see Chapter 54). Gastric resections from patients with Ménétrier’s disease typically show large polypoid gastric folds or large cerebriform gastric folds with antral sparing (see Fig. 51-12C). In the absence of a gastrectomy, a full-thickness gastric mucosal biopsy is required to adequately assess the gastric histology in patients with hyperplastic gastropathy. The predominant microscopic feature of Ménétrier’s disease and hyperplastic, hypersecretory gastropathy is foveolar hyperplasia with cystic dilation (see Fig. 51-12D). The parietal and chief cells may be decreased and replaced by mucous glands. Inflammation in hyperplastic gastropathies is variable and may be absent. The etiology of Ménétrier’s disease is unknown, although some cases have undoubtedly been infections with CMV or H. pylori. Genetic factors have recently been emphasized after the report of the disorder in identical twin men who presented at ages 29 and 35, respectively.247 Hyperplasia of surface mucous cells may be due to enhanced EGFR signaling in the gastric mucosa due to local overproduction of TGF-α.244 Ideal treatment of hyperplastic gastropathy is unclear because the condition is rare and controlled trials are lacking. Spontaneous resolution may occur, especially in children. It is likely that some cases, particularly in children, were actually cases of CMV gastritis (see earlier discussion). Ganciclovir has been used successfully in children with Ménétrier’s disease associated with CMV gastritis.249 H. pylori infection should be treated, if present, and the entire syndrome may resolve.250 Symptoms may improve with antisecretory agents (histamine-2 [H2] receptor antagonists,

Chapter 51  Gastritis and Gastropathies

A

B

C

D

Figure 51-12.  Radiologic and histopathologic examples of hyperplastic gastropathy with giant gastric folds. A, Zollinger-Ellison syndrome (barium radiograph). B, Ménétrier’s disease (barium radiograph). C, Total gastrectomy specimen in a patient with Ménétrier’s disease (right: body, revealing hyperplastic mucosa and cerebriform rugal folds; left: antrum, with relative sparing). D, Histopathology of Ménétrier’s disease showing enlarged folds with foveolar hyperplasia, cystically dilated glands, and minimal gastritis.

anticholinergic agents, proton pump inhibitors), especially if the patient has Zollinger-Ellison syndrome or normo­ gastrinemic hyperplastic, hypersecretory gastropathy. It has been suggested that H2 blockers and anticholinergics reduce gastric protein loss by strengthening intercellular tight junctions. Some patients with Ménétrier’s disease have responded to glucocorticoids, octreotide, antifibrinolytic agents, or monoclonal antibody against the EGFR.251 Partial or total gastric resection is reserved for severe complications such as refractory or recurrent bleeding, obstruction, severe hypoproteinemia, or cancer development.

ZOLLINGER-ELLISON SYNDROME (see Chapter 32) DIFFERENTIAL DIAGNOSIS OF GASTRITIS AND GASTROPATHY The most important disorders that can simulate gastritis and gastropathy are gastric polyps (non-neoplastic and neoplastic) and gastric neoplasms such as adenocarcinoma and lymphoma (see Chapter 54).252,253 Although CT criteria have been useful in distinguishing benign gastritis or gastropathy

from gastric malignancy,254 endoscopy and gastric biopsy with review by an expert pathologist are the most useful diagnostic procedures. B cell clonality using advanced PCR technology can also help distinguish gastric marginal zone lymphomas from chronic gastritis.255 At the other end of the spectrum, many patients with gastritis have a normal endoscopic appearance,256 so the differential diagnosis of gastritis also includes functional dyspepsia (see Chapter 13), in which case the gastric biopsy is usually normal.

TREATMENT AND PREVENTION OF GASTRITIS AND GASTROPATHY 257 The treatment of these disorders depends on the underlying etiology (if one can be identified). In countries where the incidence of H. pylori infection is declining, the prevalence of chronic gastritis will decline as well (see Chapter 50). It has been shown in a case-control study performed in a region of southeastern China with a very high prevalence of chronic gastritis and gastric cancer that ingestion of green tea reduced the risk of gastritis and gastric cancer by close to 50%.257

859

860

Section VI  Stomach and Duodenum KEY REFERENCES

Carlson AP, Chan JWH, Ketai LH, Demarest GB. Emphysematous gas­ tritis in a severely burned patient: Case report and literature review. J Trauma 2007; 62:765-7. (Ref 82.) Chia JKS, Chia AY. Chronic fatigue syndrome is associated with chronic enterovirus infection of the stomach. J Clin Pathol 2008; 61:43-8. (Ref 72.) Coffey RJ, Washington MK, Corless CL, Heinrich MC. Ménétrier disease and gastrointestinal stromal tumors: Hyperproliferative disorders of the stomach. J Clin Invest 2007; 117:70-80. (Ref 244.) D’Elios MM, Bergman MP, Azzurri A, et al. H+,K+-ATPase (proton pump) is the target autoantigen of Th1-type cytotoxic T cells in autoimmune gastritis. Gastroenterology 2001; 120:377-86. (Ref 45.) DeBlock CEM, DeLeeuw IH, VanGaal LF. Autoimmune gastritis in type 1 diabetes: A clinically oriented review. J Clin Endocrinol Metab 2008; 93:363-71. (Ref 47.) Fox JG, Wang TC. Inflammation, atrophy, and gastric cancer. J Clin Invest 2007; 117:60-9. (Ref 9.) Hummel M, Oeschger S, Barth TFE, et al. Wotherspoon criteria combined with B cell clonality analysis by advanced polymerase chain reaction technology discriminates covert gastric marginal zone lymphoma from chronic gastritis. Gut 2006; 55:782-7. (Ref 255.) Insko EK, Levine MS, Birnbaum BA, et al. Benign and malignant lesions of the stomach: Evaluation of CT criteria for differentiation. Radiology 2003; 228:166-71. (Ref 254.)

Kaur G, Raj SM. A study of the concordance between endoscopic gastritis and histological gastritis in an area with a low background prevalence of Helicobacter pylori infection. Singapore Med J 2002; 43:90-2. (Ref 1.) Marshall JK, Thabane M, James C. Randomized active and placebocontrolled endoscopy study of a novel protected formulation of oral alendronate. Dig Dis Sci 2006; 51:864-8. (Ref 196.) Redeen S, Peterson F, Jonsson KA, et al. Relationship of gastroscopic features to histological findings in gastritis and Helicobacter pylori infection in a general population sample. Endoscopy 2003; 35:94650. (Ref 256.) Ricuarte O, Gutierrez O, Cardona H, et al. Atrophic gastritis in young children and adolescents. J Clin Pathol 2005; 58:1189-93. (Ref 23.) Rubio CA, Nesi G, Zampi GC, et al. Gastric ciliated metaplasia. A study of 3406 gastrectomy specimens from dwellers of the Atlantic and the Pacific basins. J Clin Pathol 2005; 58:605-10. (Ref 32.) Tarnawski A, Pai R, Deng X et al. Aging gastropathy-novel mechanisms: Hypoxia, up-regulation of multifunctional phosphatase PTEN, and proapoptotic factors. Gastroenterology 2007; 133:1938-47. (Ref 243.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

52 Peptic Ulcer Disease Nimish Vakil

CHAPTER OUTLINE Epidemiology  861 Population-Based Studies  861 Time Trends  861 Risk Factors  862 Helicobacter pylori Infection  862 Aspirin and Nonsteroidal Anti-inflammatory Drugs  862 Other Ulcerogenic Drugs  862 Other Risk Factors  862 Pathogenesis  863 Gastric Acid Secretion  863 Helicobacter pylori Infection  863 Gastric and Duodenal Mucosal Defense Mechanisms  863 Helicobacter pylori-Negative, NSAID-Negative Ulcers  864 Other Causes of Ulcer Disease in the Stomach and Duodenum  864

An ulcer in the gastrointestinal (GI) tract may be defined as a break in the lining of the mucosa, with appreciable depth at endoscopy or histologic evidence of involvement of the submucosa. Erosions are breaks in the surface epithelium that do not have perceptible depth. The term peptic ulcer disease is used broadly to include ulcerations and erosions in the stomach and duodenum from a number of causes. This is because pepsin, which is proteolytic in acidic solution, plays a major role in causing the mucosal breaks regardless of the cause of the inciting agent (e.g., Helicobacter pylori, aspirin, or a nonsteroidal anti-inflammatory drug [NSAID]). Decades of research focused on the role of acid secretion and the effects of stress, personality type, and genetics in the pathogenesis of ulcer disease. The discovery of histamine-2 (H2) receptors1 and development of H2receptor antagonist drugs, and the subsequent development of proton pump inhibitor drugs led to major changes in the management of peptic ulcer disease. The discovery of H. pylori and its treatment led to dramatic changes in the prevalence and recurrence of peptic ulcer disease, transforming peptic ulcer from a chronic recurrent disease to a curable one.2 H. pylori infection remains an important cause of peptic ulceration in the developing world. In the developed world, the use of NSAIDs has emerged as a leading cause of peptic ulcer disease, especially in the aging population in whom these drugs are often prescribed. Through all of these developments, the role of acid and pepsin in the genesis and perpetuation of mucosal injury remains a unifying aspect of the pathogenesis of peptic ulcer disease.

EPIDEMIOLOGY The epidemiology of peptic ulcer disease has undergone a remarkable change in the past century. The incidence of

Gastrinoma with or without Multiple Endocrine Neoplasia Syndrome, Type I  864 Systemic Mastocytosis  864 Miscellaneous Disorders  864 Clinical Features  864 Diagnosis  864 Endoscopy  864 Contrast Radiography  865 Diagnostic and Management Strategies for Patients with Suspected Peptic Ulcer Disease  866 Complications  867 Hemorrhage  867 Penetration and Perforation  867 Obstruction  868 Treatment  868

duodenal ulcer and gastric ulcer has declined in parallel with the decline in H. pylori prevalence, likely as a result of improved sanitary conditions and a safer food and water supply. The risk of developing peptic ulcer disease and the risk of dying from peptic ulcer disease increased in successive cohorts born between 1840 and 1890 and then declined thereafter.3 A peak in the incidence of gastric ulcer in the first half of the 19th century and a subsequent peak in the incidence of duodenal ulcer in the second half of the 19th century remain unexplained, although a number of theories have been proposed, among them the widespread adoption of smoking after the commercial manufacture of cigarettes in a setting of widespread H. pylori infection.

POPULATION-BASED STUDIES

In northern Sweden, a random sample of 1001 people underwent upper GI endoscopy after filling out symptom questionnaires.4 The prevalence of peptic ulcer disease in this sample was 4.1%, with 20 gastric ulcers and 21 duo­ denal ulcers. In a prospective Danish study of 2416 subjects, the 11-year cumulative incidence of peptic ulcer was 2.9%: 1.6% for duodenal ulcer, 1.3% for gastric ulcer, and 0.04% for combined ulcers.5 In countries with a high prevalence of H. pylori infection, the ratio between duodenal and gastric ulcers may be quite different. In a case-control study from Shanghai, China, in which the prevalence of H. pylori infection was 76%, recurrent or new peptic ulcers occurred in 3.6% of the population over 2 years and 85% of the ulcers were duodenal.6

TIME TRENDS

There has been a significant decline in mortality from peptic ulcer disease over time in most age groups.7,8 A notable exception is older adults, in whom peptic ulcer bleeding remains a life-threatening condition. A study based on the U.S. National Discharge survey reported that from 1992 to 1999, the annual rate of hospitalization for peptic ulcer

861

862

Section VI  Stomach and Duodenum disease declined 20%, from 205/100,000 population to 165/100,000 population. Mortality also declined 22%, from 7.7/100,000 to 6/100,000, respectively. Sales of acidinhibitory drugs correlated with a decrease in peptic ulcer disease hospitalizations and mortality,9 although this correlation does not prove causality. The observed decline in ulcer-related mortality was probably related to improvements in the general health of the population and the availability of effective treatment for peptic ulcer disease. In an analysis of the Canadian Institute of Health Information database, the prevalence of acute nonvariceal upper GI bleeding (largely ulcer-related bleeding) declined by 31%, from 77/100,000 population to 53/100,000 population over the 10-year period from 1993 to 2003.10 The need for surgical intervention also declined but the mortality rate remained unchanged. In the United Kingdom, overall hospital admission rates for peptic ulcer disease have also declined, as has mortality, but admission rates for peptic ulcer hemorrhage and perforation have increased.11 Peptic ulcer bleeding is most often seen in older adults, with 68% of patients presenting older than age 60, and 27% older than age 80.12

RISK FACTORS The principal risk factors of peptic ulcer disease are H. pylori infection and NSAID use (Fig. 52-1). However, some patients with peptic ulcer disease have neither of these risk factors.

HELICOBACTER PYLORI INFECTION

(see Chapter 50) H. pylori is a gram-negative bacillus that is uniquely adapted to life in the stomach. It is a major cause of peptic ulcer disease and accounts for a large proportion of peptic ulcers in countries where H. pylori infection is highly prevalent such as in Asia. In the United States and in western Europe,

None known

None known

ZES, other NSAID use

H. pylori infection

ZES, other NSAID use

H. pylori infection

the original estimates that H. pylori infection was the cause of 90% or more of duodenal ulcers, and 60% or more of all gastric ulcers have been lowered by the declining prevalence of H. pylori infection.13 It is estimated that close to 70% of duodenal ulcers are related to H. pylori in Western populations. For example, in an analysis of patients who participated in H. pylori eradication trials in the United States, the proportion of patients with peptic ulcer disease and H. pylori infection was 73%.14 In Rochester, New York, the prevalence of H. pylori infection in patients with duodenal ulcer disease was 61%.15

ASPIRIN AND NONSTEROIDAL ANTIINFLAMMATORY DRUGS (see also Chapter 51)

Aspirin is increasingly used in the prevention of cardiovascular disease.16,17 Aspirin and clopidogrel are frequently used in combination in patients who have had ischemic cardiac events or in patients who have recently had a stent placed in the coronary arteries.18 NSAIDs are used by approximately 11% of the U.S. population on a regular basis.19 This frequency is likely to increase as the population ages. Regular use of NSAIDs increases the odds of gastrointestinal bleeding five- to six-fold compared with persons not taking NSAIDs.20 Serious ulcer-related complications occur in 1% to 4% of NSAID users and NSAID-related complications are thought to account for 100,000 admissions every year.21 NSAID users who also take aspirin are at an especially high risk for complications. In a population-based study from Denmark, the odds ratio for GI bleeding if a person was taking low-dose aspirin was 2.6, and this risk increased to 5.6 in patients who were also taking an NSAID.22 In Spain, the death rate attributed to NSAIDaspirin use was 15.3/100,000. Up to one third of all NSAIDaspirin–related deaths in that study were attributable to low-dose aspirin use.23 H. pylori and NSAIDs may have a synergistic role in causing peptic ulcer disease. In a meta-analysis, the odds ratios for the development of a peptic ulcer in patients with H. pylori infection or NSAID use were 4.05 and 2.99, respectively, but the odds ratio increased significantly to 15.4 if both factors were present.24 The risk factors for peptic ulcer disease among patients taking NSAIDS and their risk ratios are listed in the next chapter (see Table 53-1).

OTHER ULCEROGENIC DRUGS

Deep ulcers and perforations of the stomach and duodenum have been described in cocaine and methamphetamine users, presumably due to mucosal ischemia.25 Bisphosphonates have also been associated with gastroduodenal ulceration,26 although esophageal injury with bisphosphonates in clinically more of a concern (see Chapter 45). There is little if any risk for peptic ulcer disease in patients taking glucocorticoids.27 In combination with NSAIDS, however, glucocorticoids increase the risk of peptic ulcer disease above the risk with NSAIDs alone.28

OTHER RISK FACTORS

Duodenal Gastric Figure 52-1.  Pie charts depicting conditions associated with peptic ulcer disease. The percentages shown are rough approximations based on studies from Western countries. The relative contributions of Helicobacter pylori infection and NSAID use to peptic ulcer vary considerably among different populations and, within populations, vary with age and socioeconomic status. Also, the separation depicted in this figure is somewhat artificial because NSAID use and H. pylori infection often coexist. NSAID, non-steroidal anti-inflammatory drug; ZES, Zollinger-Ellison syndrome.

Smoking has been implicated in the pathogenesis of peptic ulcer disease for decades, but its importance as a risk factor has declined after the discovery of H. pylori. A populationbased study evaluated the risk factors for peptic ulcer disease in 2416 Danish adults who were interviewed between 1982 and 1994.29 As expected, H. pylori seropositivity was a significant risk factor for ulcer disease; smoking increased the risk of peptic ulcer only in H. pylori–infected subjects. A large body of literature suggests that smoking

Chapter 52  Peptic Ulcer Disease may predispose to peptic ulcer disease, but H. pylori infection remains a confounder that was not addressed in earlier studies. It is noteworthy that cigarette smoking does not increase the risk of recurrent ulceration once H. pylori has been eradicated, suggesting that smoking may only play a role in infected subjects.30 The role of alcohol remains uncertain. Alcoholic beverages stimulate gastric acid production. Moreover, direct application of high concentrations of alcohol to the gastric mucosa causes demonstrable mucosal injury. In the Danish study alluded to previously,29 intake of spirits increased the risk of peptic ulcer disease in H. pylori– infected patients. With regard to diet, ulcer prevalence rates differ considerably in the north of India, where the principal cereal in the diet is wheat, and in the south, where rice is the predominant cereal. However, many other potential confounders were not accounted for in these populations.31 An association between the ingestion of spicy foods and peptic ulcer disease is weak, at best.32 Emotional stress was proposed as a major cause of ulcer disease or as a precipitant of ulcer complications,33 and much was written about its relationship to peptic ulcer disease prior to the description of H. pylori. Personality types and psychological profiles have been proposed to be linked to ulcer, but much of this literature is confounded by the lack of information on H. pylori infection. Welldocumented descriptions of an increase in ulcer disease after natural calamities such as earthquakes suggest that emotional stress among those not physically injured may play a role in triggering overt manifestations of peptic ulcer disease, especially in individuals who may be otherwise predisposed to ulcer (e.g., patients infected with H. pylori).33,34

PATHOGENESIS GASTRIC ACID SECRETION

Gastric acid and pepsin have long been considered the principal inciting agents in the pathogenesis of peptic ulcer disease.35 The control of gastric acid secretion in health and disease has been reviewed in detail36 and is discussed in Chapter 49.

HELICOBACTER PYLORI INFECTION

(see also Chapters 50 and 51) Most patients who have a chronic infection with H. pylori have a pan-gastritis in which the body and the antrum are equally involved. Gastritis results in inhibition of acid secretion, and various mechanisms have been proposed.37,38 These include direct inhibition of the parietal cell by lipopolysaccharides or toxins secreted by H. pylori or an indirect effect through stimulation of cytokines caused by the inflammation.1 Due to the decreased output of acid, these individuals with pan-gastritis do not usually develop ulcer disease related to H. pylori infection.39 The reduced acid production in H. pylori–infected individuals has been suggested as a factor that may protect against the development of reflux disease and its complications such as Barrett’s esophagus (see Chapters 43 and 44). Another pattern of H. pylori infection affects approximately 10% to 20% of patients with chronic infection and consists of an antrum-predominant gastritis (see Chapters 50 and 51). In these individuals, through a series of steps

not well understood but possibly involving reduced somatostatin concentrations in the antrum, basal and mealstimulated acid secretion often increases. The increased acid output from the stomach results in increased acid delivery to the duodenum that can result in gastric metaplasia in the duodenal bulb. Some investigators believe that the metaplastic epithelium then becomes infected with H. pylori from the stomach, resulting in focal duodenitis, sometimes followed by ulcer formation. In contrast with patients with duodenal ulcer disease who often have increased acid secretion, patients with gastric ulcer typically have normal or decreased acid production, suggesting that the mechanism for the development of ulceration is a failure in the gastric mucosal protective mechanisms, described in more detail below. Acid antisecretory medication heals ulcers in patients with gastric ulcer and hypochlorhydria because acid changes the balance in favor of mucosal defense factors that restore mucosal integrity (see Chapter 53).

GASTRIC AND DUODENAL MUCOSAL DEFENSE MECHANISMS

As described in Chapter 49 the gastric mucosa has multiple defense mechanisms to protect it from digestion by acid and pepsin.40 These include the gastric surface epithelium,41-43 the mucus/phospholipid and bicarbonate barrier (mucus layer),44,45 epithelial cell renewal and regeneration,46,47 mucosal blood flow and the alkaline tide, and prostaglandin production.48-51 Prostaglandins also stimulate mucus, bicarbonate, and phospholipid production. Many prostaglandins also increase mucosal blood flow and stimulate epithelial regenerative processes. Inhibition of these protective effects by NSAIDs increases the likelihood of injury to the epithelium and decreases its ability to respond and regenerate. Cyclooxygenase-1 (COX-1) and COX-2 are the enzymes responsible for the synthesis of prostaglandins. COX-1 is expressed in the stomach and is responsible for the maintenance of the integrity of gastric epithelium and the mucous barrier. COX-2 is not expressed in the healthy stomach but is rapidly expressed in response to the cytokines generated by inflammatory processes. Conventional NSAIDs such as ibuprofen inhibit the COX-1 and the COX-2 enzymes. It is believed that COX-1 inhibition reduces prostaglandin synthesis, which leads to a reduction in mucosal blood flow, hypoxia, and a reduction in mucosal defense. Neutrophilendothelial interactions then occur as a result of the vascular disturbances and neutrophil activation. In experimental studies, COX-1 inhibition alone is not sufficient to cause ulceration.52 The inhibition of COX-1 up-regulates the expression of COX-2 that suppresses the neutrophil endothelial interaction that is stimulated by COX-1 inhibition. Inhibition of COX-1 and COX-2 enzymes is therefore important in the generation of gastric injury. Some studies have suggested that a prostaglandinindependent mechanism may also contribute to damage by NSAIDs. One such mechanism is increased leukotriene production that occurs because arachidonic acid metabolism shifts to the alternative 5-lipooxygenase pathway when the COX-1 pathway is inhibited.40 The duodenum, like the stomach, secretes bicarbonate that neutralizes acid arriving into the duodenum.53 Decreased duodenal bicarbonate secretion has been reported in patients with duodenal ulcer.54 A protective role has also been postulated for pancreatic juice but the results have been conflicting.

863

864

Section VI  Stomach and Duodenum HELICOBACTER PYLORI-NEGATIVE, NSAID-NEGATIVE ULCERS

As the prevalence of H. pylori infection declines in the United States, a growing proportion of patients with peptic ulcers who have no evidence of H. pylori infection and have no history of aspirin or NSAID use have emerged. Eradication of H. pylori generally cures peptic ulcer disease and prevents recurrence, but a small number of patients do not heal their ulcers or suffer an ulcer recurrence despite successful eradication of H. pylori. The exact prevalence of H. pylori–negative, NSAID-negative (idiopathic) peptic ulcers is unknown. The pathogenesis of H. pylori–negative, NSAIDnegative (idiopathic) ulcers remains uncertain. One theory that has been proposed is that some idiopathic ulcers may occur from H. pylori colonization of the duodenum.55 These patients may test negative with conventional endoscopic tests for H. pylori because these tests focus on detecting H. pylori in gastric biopsies. Crohn’s disease, lymphoma, and other disorders that cause ulcers in the stomach or duodenum (see next section) should be distinguished from H. pylori–negative, NSAIDaspirin–negative ulcers.

OTHER CAUSES OF ULCER DISEASE IN THE STOMACH AND DUODENUM GASTRINOMA WITH OR WITHOUT MULTIPLE ENDOCRINE NEOPLASIA SYNDROME, TYPE I

(see Chapter 32) Gastrin-secreting tumors are an important cause of acid hypersecretion due to the sustained drive for acid secretion. This disorder may result in multiple ulcerations in the stomach or duodenum that are refractory to conventional treatment and are often associated with a chronic diarrhea. One fourth of all patients with Zollinger-Ellison syndrome have an autosomal dominant disorder characterized by pancreatic endocrine tumors, hyperparathyroidism, and pituitary adenomas.56

SYSTEMIC MASTOCYTOSIS (see Chapter 32)

This is an uncommon condition in which multiple ulcers may occur in the stomach or duodenum with infiltration of the mucosa with mast cells that can be recognized histologically.57 Secretion of histamine by the mast cells is thought to result in the excessive stimulation of acid production through the histamine receptor.

MISCELLANEOUS DISORDERS

Ulcerations in the upper GI tract may be a manifestation of Crohn’s disease, which can cause ulcerations anywhere in the gastrointestinal tract (see Chapter 111). Associations between peptic ulcers and α1-antitrypsin deficiency, chronic lung disease, and chronic renal failure have been described. Many of these studies predate the routine evaluation for H. pylori in patients with peptic ulcer disease, and some of the observed associations may be questioned. For example, a recent study of α1-antitrypsin–deficient patients with duodenal ulcer found that all the patients were infected with H. pylori.58 Another study reported a higher than expected rate of ulcer recurrence in patients with renal failure following successful eradication of H. pylori.59

CLINICAL FEATURES The “classic” symptoms attributed to peptic ulcer disease included a burning pain the epigastrium relieved by antac-

ids. These assumptions were challenged by the advent of endoscopic examinations that made it clear that the same symptoms were often present in patients with no visible abnormalities at endoscopy (nonulcer dyspepsia). In patients with NSAID-related ulcer disease, pain is often absent. Thus, classic symptoms are neither sensitive nor specific for ulcer disease. The physical examination is usually normal in patients with uncomplicated peptic ulcer disease. Epigastric tenderness is neither sensitive nor specific for ulcer. A systematic review evaluated, for the patient presenting with upper abdominal symptoms, the utility of the clinical examination and various computer models in predicting if peptic ulcer disease is present.60 Some studies evaluated the accuracy of gastroenterologists, some evaluated primary care physicians, and one evaluated both.61-66 The studies included 4684 patients of whom 802 (17%) had peptic ulcer disease (Fig. 52-2). The positive likelihood ratio of accurately diagnosing peptic ulcer by clinical features was 2.9 for gastroenterologists, which was somewhat better than the performance of primary care physicians (likelihood ratio, 2.2). The negative likelihood ratio for accurately excluding peptic ulcer by clinical features was similar in gastroenterologists and primary care phy­sicians (0.62). Computer models for the prediction of peptic ulcer disease were also not highly accurate, with a positive likelihood ratio of 1.9 and a negative likelihood ratio of 0.34.

DIAGNOSIS ENDOSCOPY

Endoscopy is the current reference standard for diagnosis of peptic ulcer disease. Its main limitation is its high cost in some countries such as the United States. The decision to perform endoscopy in a patient suspected of having peptic ulcer disease is based on a number of factors. Patients presenting with complications of peptic ulcer disease such as bleeding need endoscopic evaluation to allow an accurate diagnosis and for the administration of endoscopic therapy (see Chapters 19 and 53). The presence of “alarm” features such as weight loss or recurrent vomiting may prompt concern for malignancy (Table 52-1). Although endoscopy is regarded as the standard for a diagnosis of peptic ulcer disease (Figs. 52-3 and 52-4), small ulcers may be missed at endoscopy.67 The proportion of ulcers that may be missed at endoscopy is uncertain because comparative studies of contrast radiography and endoscopy were performed many years ago and technologic innovations in the quality and resolution of endoscopes have

Table 52-1  Alarm Features in Patients with Suspected Peptic Ulcer Disease* Age older than 55 years with new-onset dyspepsia Family history of upper gastrointestinal cancer Gastrointestinal bleeding, acute or chronic, including unexplained iron deficiency Jaundice Left supraclavicular lymphadenopathy (Virchow’s node) Palpable abdominal mass Persistent vomiting Progressive dysphagia Unintended weight loss *These features should prompt an upper endoscopy and often other testing to establish a definitive diagnosis.

Chapter 52  Peptic Ulcer Disease Primary care physicians Heikkinen et al Hansen et al DDSG

Positive likelihood ratio (95% CI)

Negative likelihood ratio (95% CI)

2.1 (1.4-3.2) 2.1 (1.7-2.6) 2.2 (1.6-3.0)

0.76 (0.61-0.91) 0.58 (0.45-0.72) 0.54 (0.37-0.73)

Gastroenterologists Bytzer et al Fjosne et al DDSG Barenys et al

1.9 (1.6-2.3) 3.9 (3.3-4.5) 3.4 (2.3-4.8) 2.8 (2.2-3.6)

0.50 (0.41-0.60) 0.47 (0.40-0.53) 0.54 (0.38-0.70) 0.45 (0.37-0.54)

Computer models Barenys et al Bytzer et al Johannessen et al Mann et al Holdstock et al Numans et al

3.0 (2.5-3.8) 1.4 (1.2-1.5) 2.0 (1.8-2.2) 1.6 (1.2-1.9) 1.7 (1.6-1.9) 2.2 (1.9-2.5)

0.28 (0.22-0.36) 0.62 (0.46-0.80) 0.19 (0.12-0.30) 0.37 (0.18-0.72) 0.58 (0.29-0.49) 0.31 (0.22-0.44)

Combined

2.2 (1.9-2.6)

0.45 (0.38-0.53)

0.1 1.0 10.0 Positive likelihood ratio (95% CI)

0.1 1.0 10.0 Negative likelihood ratio (95% CI)

Figure 52-2.  Positive and negative likelihood ratios of different approaches to diagnosing peptic ulcer disease. Each square represents an individual study. The size of the square is a measure of the size of the study and the horizontal line through the square indicates a graphical representation of the 95% CI of that study. For the combined analysis, the diamond and vertical dashed line indicate the pooled positive or negative likelihood ratio, with the left and right ends of the diamond indicating the pooled 95% CI. CI, confidence interval; DDSG, Danish Dyspepsia Study Group. (Reproduced from Moayyedi P, Talley N, Fennerty B, Vakil N. Can the clinical history distinguish between organic and functional dyspepsia? JAMA 2006; 295:1566-76.)

Figure 52-4.  Endoscopic view of a duodenal ulcer in a patient with a positive rapid urease test for Helicobacter pylori. There was no history of nonsteroidal anti-inflammatory drug use.

Figure 52-3.  Endoscopic view of a clean based antral gastric ulcer in a patient taking a non­steroidal anti-inflammatory drug. Tests for infection with Helicobacter pylori were negative.

ulcers at initial endoscopy can be found to be malignant at a subsequent examination, and cancers so detected are typically at an early stage with the possibility of a curative resection.

CONTRAST RADIOGRAPHY

improved the yield of endoscopy. Biopsies should be taken from the edges of a gastric ulcer because of the risk of malignancy. In patients with gastric ulcers it has been customary to repeat the endoscopy after approximately eight weeks of treatment to ensure that the ulcer has healed (see Chapter 53). A small number of chronic, nonhealing gastric ulcers can be identified as being malignant at the repeat endoscopy even though the initial biopsies had revealed no evidence of a malignancy. Although conflicting reports on the utility and cost-effectiveness of performing a repeat endoscopy after approximately eight weeks in patients with benign-appearing gastric ulcers have appeared in the literature, recent studies support the use of a repeat endoscopy in this setting.68,69 Up to 4% of apparently benign gastric

Contrast radiography of the upper GI tract, also referred to as the barium meal or an upper GI series, can often demonstrate a peptic ulcer (Fig. 52-5). In historical studies, contrast radiography has performed as well as endoscopy in the diagnosis of ulcer disease when the contrast examinations are performed and interpreted by experts.67 Much has changed since then. Contrast radiography of the upper GI tract is now frequently performed by technicians, and the availability of personnel trained in the interpretation of barium studies has decreased. A further disadvantage to contrast radiography is radiation exposure, which can be substantial. Endoscopy offers the obvious advantage of allowing mucosal biopsy, which can be used to diagnose H. pylori infection and to rule out malignancy in chronic gastric ulcers. Contrast radiography therefore has limited utility in modern practice.

865

866

Section VI  Stomach and Duodenum Dyspepsia not due to GERD not associated with NSAID use

Figure 52-5.  Radiograph from an upper gastrointestinal series showing a benign gastric ulcer. Note the smooth, symmetrical folds radiating to the ulcer crater, which appears to project outside the lumen of the stomach. (Courtesy Mark Feldman, MD, Dallas, Tex.)

DIAGNOSTIC AND MANAGEMENT STRATEGIES FOR PATIENTS WITH SUSPECTED PEPTIC ULCER DISEASE Upper abdominal symptoms that raise the question of peptic ulcer disease are common in clinical practice and account for 2% to 5% of visits in family practice.70 These symptoms, consisting of pain or discomfort in the upper abdomen are frequently referred to as dyspepsia. Due to the high cost of subjecting all dyspeptic individuals to endoscopy, noninvasive alternative strategies have been proposed as an initial step in the management of suspected peptic ulcer disease (Fig. 52-6).71 Economic models suggested that these noninvasive strategies could reduce the costs of managing peptic ulcer disease substantially.72 Subsequent clinical trials demonstrated the clinical efficacy of the approach and the initial strategy proposed was to perform noninvasive testing for infection, followed by antimicrobial treatment directed against H. pylori if the patient tested positive.73,74 This “test and treat” strategy offers a potentially curative treatment for patients who have H. pylori related ulcer disease and may cure a small proportion of patients with nonulcer dys­pepsia. As the prevalence of H. pylori infection declines in Western populations, other empirical strategies can be considered. Because acid inhibition plays a significant role in the management of patients with upper GI symptoms, empirical treatment with a proton pump inhibitor has been proposed as an alternative to routine endoscopy. Cost-models have suggested that when the prevalence of H. pylori infection in the population falls to the 10% to 15% range, an empirical course of therapy with a proton pump inhibitor may be a reasonable initial strategy.75 A cost simulation model in the United States76 supported the empirical management of patients, as recommended in the American Gastroenterological Association dyspepsia guidelines (see Fig. 52-6).71 A randomized controlled clinical trial compared empirical testing and treatment for H. pylori with empirical proton pump inhibitor therapy and concluded that both were equally cost-effective in the initial man­agement of dyspeptic patients, with the choice of which strategy to be taken left to a discussion between patient and physician.77

Age >55 or alarm features present (See Table 52-1.)

Age ≤55 No alarm features (See Table 52-1.)

EGD

Test for Helicobacter pylori

Negative

Positive

PPI trial 4-6 weeks

Treat for H. pylori

Fails

Fails

PPI trial 4 weeks Fails

Reassurance Reassess diagnosis

Consider EGD Figure 52-6.  American Gastroenterological Association guideline for the management of dyspepsia. This is the current management approach for patients with suspected peptic ulcer disease. EGD, esophagogastroduodenoscopy; GERD, gastroesophageal reflux disease; NSAID, nonsteroidal anti-inflammatory drug; PPI, proton pump inhibitor. (Adapted from Talley NJ. American Gastroenterological Association medical position statement: evaluation of dyspepsia. Gastroenterology 2005; 129:1753-5.)

The incidence of upper GI malignancies rises with age, and thus current empirical management strategies are generally reserved for younger patients with upper abdo­ minal symptoms. The age beyond which endoscopy should become routine is controversial and to a substantial degree depends on the epidemiology of gastric cancer in the population under consideration. In Western populations, upper GI cancer is rare in young individuals and therefore an age cutoff of 50 or 55 years is used.78 Older patients presenting with new-onset upper abdominal symptoms suggestive of peptic ulcer disease should be referred for endoscopy. Guidelines in the United Kingdom have not set an age cutoff for early endoscopy, and empirical therapy is preferred at all ages. In Asia and eastern Europe, an earlier age cutoff may be reasonable because the risk of gastric cancer is substantially higher than in Western nations and setting the age threshold too high may lead to a delayed diagnosis of a treatable cancer.79 The principal concern of any empirical treatment strategy for treating dyspepsia is that an underlying malignancy will be missed. To increase the odds of detecting an underlying malignancy, an age-based strategy has been

Chapter 52  Peptic Ulcer Disease Study A

0.67 (0.22, 0.96)

Study B

0.71 (0.44, 0.90)

Study C

0.56 (0.45, 0.66)

Study D

0.83 (0.61, 0.95)

Study E

0.50 (0.16, 0.84)

Studies combined

0.67 (0.54, 0.83) 0.1

A

0.2

0.5

1

Sensitivity (95% confidence interval)

Study A

0.95 (0.93, 0.97)

Study B

0.68 (0.66, 0.69)

Study C

0.86 (0.85, 0.86)

Study D

0.39 (0.37, 0.41)

Study E

0.59 (0.57, 0.62)

Studies combined

0.66 (0.55, 0.79) 0.2

B

0.5 Specificity (95% confidence interval)

proposed in which younger dyspeptic patients are treated empirically and older dyspeptic patients receive early endoscopy. Although there is some controversy about the age at which routine endoscopy for dyspepsia should begin, an age cutoff of 50 to 55 years is proposed in recent guidelines.80 A systematic review and meta-analysis of alarm symptoms in predicting malignancy (as opposed to peptic ulcer and nonulcer dyspepsia) has raised questions about their accuracy.81 The sensitivity of alarm symptoms for serious underlying pathology in 15 studies including 57,363 patients ranged from 0% to 83%, with considerable heterogeneity between studies (Fig. 52-7). Specificity of alarm symptoms for malignancy varied from 40% to 83%. Although current management schemes still list alarm features as an indication for early endoscopy in patients with undiagnosed upper abdominal pain (see Fig. 52-4), further research is necessary in this area. The issue of evaluation of undiagnosed dyspepsia is also discussed in Chapter 13.

COMPLICATIONS HEMORRHAGE

Hemorrhage from a peptic ulcer occurs when the ulcer crater erodes a blood vessel. It is a major cause of morbidity

1

Figure 52-7.  Meta-analysis plots for the utility of clinical alarm features in the diagnosis of upper gastrointestinal malignancies. A, Sensitivity of the presence of any alarm feature in detecting upper gastrointestinal (GI) malignancy in five studies (A-E). When the studies were combined, the sensitivity was only 67%. B, Specificity of the presence of any alarm feature in detecting upper GI malignancy. The combined specificity was only 66%. (Reproduced from Vakil N, Moayyedi P, Fennerty MB, Talley NJ. Limited value of alarm features in the diagnosis of upper gastrointestinal malignancy: Systematic review and meta-analysis. Gastroenterology 2006; 131:390-401.)

in the United States with an annual cost that exceeds 2 billion dollars.82,83 GI hemorrhage is an important cause of morbidity and mortality in older adutls.84 The typical presentation is with melena or hematemesis although a small proportion of patients with bleeding may present with hematochezia. Bleeding peptic ulcers are discussed in more detail in Chapters 19 and 53. Strategies to prevent NSAIDrelated bleeding in patients at high risk have been developed85 and are also discussed in more detail in Chapter 53, as is the treatment of acute ulcers occurring in critically ill patients in intensive care units (stress ulcers).

PENETRATION AND PERFORATION 86-93

As an ulcer deepens it can burrow into adjoining structures (penetration) or rupture into the peritoneal cavity (perforation). Ulcer-related penetration and perforation are less common than upper GI bleeding as a complication of peptic ulcer disease and are more common in older adults. Ulcers located posteriorally in the duodenum may penetrate the pancreas, whereas ulcers located more anteriorally located ulcers may penetrate the liver or biliary tract.88-90 There are no typical presenting signs that identify a penetrating ulcer, but pain radiating to the back is often cited as a symptom of a posterior penetrating ulcer. Computed tomography (CT) may be helpful in establishing a diagnosis.91 Free ulcer perforation with peritonitis is a medical emergency. The patient typically presents with severe gener­

867

868

Section VI  Stomach and Duodenum alized abdominal pain and signs of peritonitis (see Chapters 10 and 37). CT usually establishes the diagnosis.92 Perforation has a high mortality, particularly in older adults. A population-based study from three Danish counties reported that among 2061 patients hospitalized with peptic ulcer perforation, 38% were current NSAID users.93 The 30-day mortality associated with ulcer perforation in this population was 25% overall, and 35% among current NSAID users.

OBSTRUCTION

Chronic scarring, usually in the antrum of the stomach, can lead to a disorder characterized by recurrent vomiting and a narrowing of the H. pylori channel, which is usually visualized endoscopically or on contrast radiography. The symptoms may be insidious, manifesting as reflux disease that is difficult to control or dyspepsia until the narrowing becomes pronounced when vomiting and early satiety become prominent features. Benign pyloric stenosis due to recurrent peptic ulcer disease was a common condition in the past but is rather rare now. It is therefore important to rule out a neoplastic process, which is a more common cause of gastric outlet obstruction than benign pyloric stenosis due to ulcer disease.94

TREATMENT Treatment of peptic ulcer disease and its complications is discussed in Chapter 53.

KEY REFERENCES

Aro P, Storskrubb T, Ronkainen J, et al. Peptic ulcer disease in a general adult population: The Kalixanda study: A random population-based study. Am J Epidemiol 2006; 163:1025-34. (Ref 4.) Black J, Duncan W, Durant D. Definition and antagonism of histamine H2 receptors. Nature 1972; 236:365-90. (Ref 1.) El-Omar EM. Mechanisms of increased acid secretion after eradication of Helicobacter pylori infection. Gut 2006; 55:144-6. (Ref 37.)

Laine L, Takeuchi K, Tarnawski A. Gastric mucosal defence and cytoprotection: Bench to bedside. Gastroenterology 2008; 135:41-60. (Ref 40.) Lewis JD, Bilker WB, Brensinger C, et al. Hospitalization and mortality rates from peptic ulcer disease and GI bleeding in the 1990s: Relationship to sales of nonsteroidal anti-inflammatory drugs and acid suppression medications. Am J Gastroenterol 2002; 97:25409. (Ref 9.) Marshall B, Warren J. Unidentified curved bacilli in the stomachs of patients with gastritis and peptic ulcer. Lancet 1984; 1:1311-15. (Ref 2.) Moayyedi P, Talley N, Fennerty B, Vakil N. Can the clinical history distinguish between organic and functional dyspepsia? JAMA 2006; 295:1566-76. (Ref 60.) Ofman JJ, Etchason J, Fullerton S, et al. Management strategies for Helicobacter pylori–seropositive patients with dyspepsia (clinical and economic consequences). Ann Intern Med 1997; 126:280-91. (Ref 72.) Rosenstock S, Jorgensen T, Bonnevie O, Andersen L. Risk factors for peptic ulcer disease: A population based prospective cohort study comprising 2416 Danish adults. Gut 2003; 52:186-93. (Ref 29.) Schubert M, Peura D. Control of gastric acid secretion in health and disease. Gastroenterology 2008; 134:1842-60. (Ref 36.) Sonnenberg A. Time trends of ulcer mortality in non-European countries. Am J Gastroenterol 2007; 102:1101-7. (Ref 7.) Spiegel BM, Vakil NB, Ofman JJ. Dyspepsia management in primary care (a decision analysis of competing strategies) Gastroenterology 2002; 122:1270-85. (Ref 75.) Talley NJ, Vakil N. Practice Parameters Committee of the American College of Gastroenterology. Guidelines for the management of dyspepsia. Am J Gastroenterol 2005; 100:2324-37. (Ref 80.) Targownik L, Nabalamba A. Trends in management and outcomes of acute nonvariceal upper gastrointestinal bleeding: 1993-2003. Clin Gastroenterol Hepatol 2006; 4:1459-66. (Ref 10.) Vakil N, Moayyedi P, Fennerty BM, Talley N. Limited value of alarm features in the diagnosis of upper gastrointestinal malignancy: Systematic review and meta-analysis. Gastroenterology 2006; 131:390401. (Ref 81.) Wilcox C, Allison J, Benzuly K, et al. Consensus Development Conference on the Use of Nonsteroidal Anti-Inflammatory Agents, Including Cyclooxygenase-2 Enzyme Inhibitors and Aspirin. Clin Gastroenterol Hepatol 2006; 4:1082-9. (Ref 21.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

53 Treatment of Peptic Ulcer Disease Francis K. L. Chan and James Y. W. Lau

CHAPTER OUTLINE Overview  869 Antisecretory and Acid-Neutralizing Agents  869 Mucosa-Protective Agents  871 Ulcers Associated with Helicobacter pylori Infection  872 Duodenal Ulcer  872 Gastric Ulcer  872 Role of Maintenance Therapy  872 Ulcers Associated with Nonsteroidal Anti-inflammatory Drugs  872 Active Ulcers  872 Ulcer Prophylaxis  873

Chapter 52 reviews the epidemiology, clinical features, diagnosis, and complications of peptic ulcer disease. This chapter focuses on its treatment.

OVERVIEW For more than a century, peptic ulcer was considered a chronic, incurable disorder characterized by frequent exacerbations and remissions. The discovery of the link between Helicobacter pylori and peptic ulcer by Marshall and Warren1 in the mid-1980s revolutionalized the concept and treatment of peptic ulcer. Now there is overwhelming evidence to support H. pylori infection as the most important cause of duodenal and gastric ulcers worldwide. Curing the infection not only heals peptic ulcer but also prevents ulcer relapse.2 Peptic ulcer is, in fact, a curable infectious disease. Beside H. pylori–related peptic ulcers, use of nonsteroidal anti-inflammatory drugs (NSAIDs) and low-dose aspirin is another major cause of peptic ulcer complications particularly among older adults.3 Co-therapy with antiulcer drugs and the replacement of conventional nonselective NSAIDs with NSAIDs selective for cyclooxygenase-2 (COX-2 selective NSAIDs) have become alternative treatments for patients who are at risk for peptic ulcer disease. Data suggest that COX-2 selective NSAIDs and some nonselective NSAIDs increase the risk of serious cardiothrombotic events. Prescribing NSAIDs therefore requires a careful assessment of individual patients’ gastrointestinal (GI) and cardio­ vascular risks. As discussed in the preceding chapter, due to the declining prevalence of H. pylori infection, the proportion of patients with H. pylori–negative, NSAID-negative idiopathic ulcers is growing. In the United States, the reported propor-

The Role of Helicobacter pylori Infection in Ulcer Disease Associated with NSAID Use  876 Recommendations for the Prevention of NSAID-Induced Ulcer Complications  876 Refractory Ulcers  877 Stress-Related Mucosal Injury  877 Treatment of Complications of Peptic Ulcer Disease  878 Hemorrhage  878 Perforation  883 Obstruction  884

tion of H. pylori–negative, NSAID-negative idiopathic ulcers is between 20% and 30%.4,5 It has been argued that as the incidence of H. pylori–related ulcers falls, a greater proportion of H. pylori–negative, NSAID-negative idiopathic ulcers will be seen.6 Of interest, a prospective cohort study has demonstrated a four-fold rise in the absolute incidence of idiopathic bleeding ulcers, and the risk of recurrent ulcer bleeding in these patients is high.7 Thus, long-term prophylaxis with antisecretory drugs for idiopathic bleeding ulcers is advisable (see later), although this recommendation is not evidence based.

ANTISECRETORY AND ACID-NEUTRALIZING AGENTS

Before the discovery of H. pylori as a causal factor in peptic ulcer disease, drugs designed to reduce gastric acidity were the mainstays of treatment. Antisecretory or antacid therapy is not routinely required for patients with uncomplicated H. pylori ulcers in whom the bacterium is successfully eradicated, but these classes of drugs do play an important role in promoting healing of large ulcers, preventing early recurrent bleeding after endoscopic therapy for bleeding ulcers, and reducing the risk of ulcer relapse associated with NSAIDs. Specific therapies for peptic ulcer are discussed in the following sections.

Antacids

Mechanisms of Action When Peterson and coworkers8 showed that a liquid antacid preparation of magnesium-aluminum hydroxide (approximately 1000 mmol HCl neutralizing capacity per day) was more effective than placebo for hastening the healing of duodenal ulcer, it was thought antacids promoted ulcer healing by neutralizing gastric acid. However, later studies showed that far smaller doses of antacids (neutralizing

869

870

Section VI  Stomach and Duodenum capacity 120 mmol HCl per day) had virtually identical efficacy for healing peptic ulcerations.9 The precise mechanisms by which antacids hasten the healing of peptic ulcerations are not clear. Although antacids have been shown to be superior to placebo in healing peptic ulcers, their efficacy in ulcer healing is limited. In one study, the healing rates for gastric ulcer after 6 weeks were 67% in the antacids group and 25% in the placebo group.9 Adverse Effects For the magnesium-containing agents, the most common side effect is diarrhea. In contrast, antacids that contain aluminum hydroxide primarily, and those that contain calcium, may cause constipation. All antacids must be used with caution, if at all, in patients who have chronic kidney disease. In such patients, magnesium-containing agents can cause hypermagnesemia; the use of calcium carbonate can cause hypercalcemia and alkalosis and further renal impairment (milk-alkali syndrome), and aluminum hydroxide antacids can cause aluminum neurotoxicity.10

Histamine-2 Receptor Antagonists

Mechanisms of Action Four histamine-2 (H2) receptor antagonists are available— cimetidine (Tagamet), ranitidine (Zantac), famotidine (Pepcid), and nizatidine (Axid). These compounds are competitive inhibitors of histamine-stimulated acid secretion, although famotidine appears to have some component of noncompetitive inhibition as well.11 All four agents suppress basal acid output as well as acid output stimulated by meals (see Chapter 49). Pharmacokinetics The H2 receptor antagonists are well absorbed after oral dosing, and their absorption is not affected by food. Peak blood levels are achieved within 1 to 3 hours after an oral dose. These drugs cross the blood-brain barrier and the placenta.12,13 After oral administration, cimetidine, ranitidine, and famotidine undergo first-pass hepatic metabolism, which reduces their bioavailability by 35% to 60%. In contrast, nizatidine does not undergo first-pass metabolism, and its bioavailability approaches 100% with oral dosing. When administered in the evening, the drugs are especially effective in suppressing nocturnal acid output.14 All four H2 receptor antagonists are eliminated by a combination of renal excretion and hepatic metabolism. After intravenous administration, in contrast, all four agents are eliminated principally through renal excretion. For cimetidine and famotidine it is recommended that the doses be cut in half in patients whose creatinine clearance is 15 to 30 mL/minute. For nizatidine and ranitidine, the dose should be halved if the creatinine clearance is less than 50 mL/minute. Dialysis does not remove substantial amounts of the H2 receptor antagonists, so dose adjustments for dialysis are not necessary. Liver failure has been found to prolong the half-life of cimetidine, but dose reductions are generally not needed for patients with hepatic failure unless it is accompanied by chronic kidney disease.11 Tolerance to the antisecretory effects of H2 receptor antagonists appears to develop quickly and frequently.15 The mechanisms that mediate tolerance to the antisecretory effects of H2 receptor antagonists are not entirely clear. Adverse Effects The H2 receptor antagonists are a remarkably safe and welltolerated group of agents. The overall incidence of side effects is less than 4%, and serious side effects are decidedly uncommon. One meta-analysis of randomized clinical trials

concluded that the overall rate of adverse effects reported for the H2 blockers did not differ significantly from that for placebo.16 Nevertheless, a number of untoward effects have been described, primarily in anecdotal reports and uncontrolled series. Cimetidine has weak antiandrogenic activity that occasionally can cause gynecomastia and impotence.17 Myelosuppression is an uncommon, presumably idiosyncratic side effect of the H2 receptor antagonists. In one large series of patients with bone marrow transplants, however, ranitidine was implicated as a possible cause of myelosuppression in 5%.18 The contribution of ranitidine to the bone marrow suppression in such patients is not clear, but pending further data, it seems prudent to avoid the use of H2 receptor antagonists in bone marrow transplant recipients. Drug Interactions Both cimetidine and ranitidine bind to the hepatic cytochrome P-450 (CYP) mixed-function oxidase system, and this binding can inhibit the elimination of other drugs that are metabolized through the same system, including theophylline, phenytoin, lidocaine, quinidine, and warfarin.19 Famotidine and nizatidine have no significant avidity for the CYP system, and these agents do not appear to have any important drug interactions.

Proton Pump Inhibitors

Mechanisms of Action The proton pump inhibitors (PPIs) are a class of drugs that decrease gastric acid secretion through inhibition of H+,K+-ATPase, the proton pump of the parietal cell (see Chapter 49). Five PPIs are used widely as antisecretory agents—omeprazole (Prilosec), esomeprazole (Nexium; the S optical isomer of omeprazole), lansoprazole (Prevacid), pantoprazole (Protonix), and rabeprazole (Aciphex). These agents are prodrugs that must be activated by acid to inhibit the H+,K+-ATPase. However, PPIs as prodrugs are acid-labile compounds that must be protected from degradation by stomach acid after oral administration.20 Pharmacokinetics The PPIs are well absorbed after oral dosing. Absorption of the enteric-coated agents may be erratic, and peak serum concentrations are not achieved until 2 to 5 hours after oral administration. Although the plasma half-life of the PPIs is short (about 2 hours), the duration of acid inhibition is long (about 24 hours) as a result of covalent binding of the active metabolite to the H+,K+-ATPase. All PPIs undergo significant hepatic metabolism. Because there is no direct toxicity from PPIs, dose adjustments are not required even in patients with significant renal or hepatic impairment. However, there are significant genetic polymorphisms for one of the CYP isoenzymes involved in PPI metabolism, CYP2C19. Approximately 3% of white persons and 15% of Asians are deficient in CYP2C19. This polymorphism has been shown to substantially raise plasma levels of omeprazole, lansoprazole, and pantoprazole but not those of rabeprazole.21,22 As a result of their requirement for concentration and activation in acidic compartments, the PPIs bind predominantly to those proton pumps that are actively secreting acid. Thus, the efficacy of the PPIs for inhibiting acid secretion is limited if they are administered during the fasting state, when only approximately 5% of the stomach’s proton pumps are active. With meal stimulation, in contrast, 60% to 70% of the proton pumps actively secrete acid. Thus, the PPIs are most effective if they are administered immediately before meals. For once-daily dosing, it is recom-

Chapter 53  Treatment of Peptic Ulcer Disease mended that the PPIs be taken immediately before breakfast.23 Unlike H2 receptor antagonists, tolerance to the antisecretory effects of PPI therapy has not been seen during shortterm investigations. Adverse Effects The PPIs are a remarkably safe and well-tolerated group of agents. The most commonly reported side effects are headache and diarrhea, yet the rate at which patients experience these symptoms does not differ significantly from that for patients treated with placebo.24 PPIs and other antisecretory agents cause hypergas­ trinemia by inhibiting gastric acid secretion (see Chapter 49). In addition to stimulating acid secretion, gastrin has been shown to have trophic effects on the GI entero­ chromaffin-like (ECL) cells. Female rats in which protracted hypergastrinemia has been induced by PPIs develop ECL cell hyperplasia and gastric carcinoid tumors.25 However, there are no reports of gastric carcinoid tumors attributable to PPIs in humans. Even in patients with Zollinger-Ellison syndrome who have severe hypergas­ trinemia, carcinoid tumors are uncommon and occur predominantly in patients with multiple endocrine neoplasia type I (MEN-I).26 Some data suggest that the long-term administration of PPIs to patients who are infected with H. pylori might accelerate the development of gastric atrophy.27 However, these early observations have not been conformed by subsequent studies.28 The U.S. Food and Drug Administration (FDA) advisory group concluded that the available data did not establish such an effect, and did not recommend routine testing for and treatment of H. pylori before initiation of PPI therapy.29 Data from observational studies have found that PPIs increase the risk of osteoporosis-related fracture. The strength of the association increases with increasing duration and dose of PPI therapy.30,31 The mechanisms underlying such an association are unknown. Drug Interactions The elevation of gastric pH induced by the PPIs can affect the absorption of a number of medications. However, this antisecretory action rarely has clinically important effects on drug pharmacokinetics, except when the PPIs are given with ketoconazole or digoxin.32 Ketoconazole requires stomach acid for absorption, and this drug may not be absorbed effectively after PPIs have inhibited gastric acid secretion. Conversely, an elevated gastric pH facilitates the absorption of digoxin, resulting in higher plasma levels of this agent. If a patient requires both PPI and antifungal therapy, it is recommended that an agent other than ketoconazole be chosen. For patients treated concomitantly with PPIs and digoxin, clinicians should consider monitoring plasma digoxin levels. Because the PPIs are metabolized by the CYP system, there is potential for them to alter the metabolism of other drugs that are eliminated by CYP enzymes. Among the available PPIs, omeprazole appears to have the greatest potential for such drug interactions and has been shown to delay the clearance of warfarin, diazepam, and phenytoin.33 Lansoprazole, pantoprazole, and rabeprazole do not appear to interact significantly with drugs metabolized by the CYP system. Even with omeprazole, however, clinically important drug interactions are uncommon. Evidence is accumulating that several PPIs may inhibit the activation of clopidogrel to its active metabolite, thus impairing the antiplatelet effect of clopidogrel, with adverse cardiovascular outcomes.

MUCOSA-PROTECTIVE AGENTS Sucralfate

Mechanisms of Action Sucralfate (Carafate) has demonstrated efficacy (similar to that of the H2 receptor antagonists) in healing duodenal ulcer when given in a dose of 1 g four times daily.34 The drug has demonstrated efficacy in the treatment of gastric ulcer as well, but sucralfate has not been approved by the FDA for this indication. Sucralfate is a complex metal salt of sulfated sucrose. Although the sucralfate molecule contains aluminum hydroxide, the agent has little acidneutralizing capacity. When exposed to gastric acid, the sulfate anions can bind electrostatically to positively charged proteins in damaged tissue, thereby forming a protective barrier that may prevent further acid-peptic attack. Other proposed beneficial effects of sucralfate are enhancement of mucosal prostaglandin levels, stimulation of mucus and bicarbonate secretion, binding of mucosa-irritating bile salts, binding of epidermal growth factors, and promotion of angiogenesis.34 Pharmacokinetics Less than 5% of the sucralfate administered is absorbed owing to its poor solubility.34 The drug is excreted in the feces. The high aluminum content causes a small but significant rise in serum and urine aluminum levels within 2 days. In patients with normal renal function, the minor amounts of aluminum absorption with short-term therapy are of no clinical significance. Toxicity and Drug Interactions Because of the lack of systemic absorption, sucralfate appears to have no systemic toxicity. The effect on the disposition of aluminum in the body has not been adequately studied in patients with chronic kidney disease. Sucralfate is best avoided in this population. The drug can bind to a number of medications, including phenytoin and warfarin, reducing their absorption. Important drug interactions appear to be rare, however, and can be avoided entirely if sucralfate is administered at a time separate from other medications.34

Bismuth

Mechanisms of Action Two colloidal preparations of bismuth have been most commonly used, colloidal bismuth subcitrate and bismuth subsalicylate (e.g., Pepto-Bismol). These agents have some efficacy in healing peptic ulcers, but the mechanisms underlying this therapeutic effect are not clear.35 The bismuth forms complexes with mucus that appear to coat ulcer craters. Effects on increasing mucosal prostaglandin synthesis and bicarbonate secretion also have been proposed, and bismuth has documented antimicrobial actions against H. pylori. Bismuth has been approved by the FDA for use in combination with other agents for the treatment of H. pylori infection (see Chapter 50). Pharmacokinetics Bismuth is largely unabsorbed and is excreted in the feces. Colonic bacteria convert bismuth subcitrate and bismuth subsalicylate to bismuth sulfide, which turns the stools black. Trace amounts of bismuth are absorbed in the upper GI tract. Absorbed bismuth is slowly excreted in the urine for three months or longer.35 Toxicity Short-term, standard-dose therapy with bismuth appears to carry little risk of toxicity. However, there is the potential

871

872

Section VI  Stomach and Duodenum for bismuth neurotoxicity if the agent is given for extended periods in high dosage, especially in patients with chronic kidney disease.35

Prostaglandin E Analogs

Mechanisms of Action Endogenous prostaglandins, including prostaglandin E2 (PGE2), regulate mucosal blood flow, epithelial cell proliferation, epithelial restitution, mucosal immunocyte function, mucus and bicarbonate secretion, and basal acid secretion.36 There is substantial evidence that the ulcerogenic effect of an NSAID correlates well with its ability to suppress prostaglandin synthesis.37 Misoprostol, a prostaglandin E1 analog, is the only prostaglandin analog approved by the FDA for the prevention of NSAID-induced ulcer disease. The drug not only enhances mucosal defense mechanisms but also inhibits gastric acid secretion. After binding to the prostaglandin receptor on the parietal cell, misoprostol inhibits gastric acid secretion in a dosedependent manner that is mediated through inhibition of histamine-stimulated cyclic adenosine monophosphate (cAMP) production.38 It has been shown that misoprostol significantly reduces nocturnal, basal, and meal-stimulated acid secretion at a standard therapeutic dose, although the effect is not as potent as that of other classes of antisecretory agents.39 Pharmacokinetics Misoprostol is well absorbed after oral administration. The plasma concentration peaks at about 30 minutes, with a serum half-life of approximately 1.5 hours. The drug has no effect on hepatic cytochrome P-450. Misoprostol metabolites are excreted in the urine, but dose reduction is unnecessary in patients with chronic kidney disease.40 Toxicity Dose-related diarrhea is the most common side effect, occurring in up to 30% of patients and limiting the usefulness of misoprostol.41 Diarrhea is related to prostaglandin-induced increases in intestinal water and electrolyte secretion or acceleration of intestinal transit time. Administration of misoprostol with food may reduce diarrhea. Prostaglandins stimulate uterine smooth muscle. Misoprostol is therefore contraindicated in women who may be pregnant.

ULCERS ASSOCIATED WITH HELICOBACTER PYLORI INFECTION The discovery of H. pylori and its role in peptic ulcer disease has revolutionized the approach to management. Before this discovery, annual ulcer recurrence rates were as high as 80%, often requiring long-term maintenance therapy for ulcer prevention. Now it is well established that curing H. pylori infection not only heals peptic ulcer but also prevents relapse. The following sections outline the management of peptic ulcers associated with H. pylori infection. The choice of diagnostic tests and treatment regimens for H. pylori infection is discussed in Chapter 50.

DUODENAL ULCER

Because H. pylori infection accounts for 70% or more of duodenal ulcers, one must test for the infection using one of the noninvasive tests recommended in Chapter 50. If the diagnosis of ulcer disease is made endoscopically, gastric biopsy specimens should be taken to detect H. pylori infection. If H. pylori infection is documented, the patient should

be treated with one of the regimens recommended in Chapter 50, irrespective of whether he or she has a history of NSAID use. There is good evidence that a course of H. pylori eradication therapy is sufficient to heal complicated and uncomplicated duodenal ulcers such that additional antisecretory therapy is usually not required. In a meta-analysis of 52 trials, it was found that the eradication of H. pylori alone was superior to use of an ulcer-healing drug (relative risk of ulcer, 0.66) and to no treatment (relative risk, 0.37).42 Follow-up endoscopic examination to ensure healing and testing to document H. pylori eradication after antibiotic therapy are not recommended routinely in patients with uncomplicated ulcers. However, noninvasive tests such as the urea breath test and fecal antigen test can be used to confirm H. pylori eradication in patients with ulcer complications.

GASTRIC ULCER

If H. pylori infection is documented, the patient should be treated with one of the regimens recommended in Chapter 50 regardless of whether he or she has a history of NSAID use. Whether antisecretory therapy is required after a course of H. pylori eradication therapy is controversial. It has been shown that 1 week of antibacterial therapy without acid suppression effectively heals gastric ulcers.42 In a metaanalysis of ulcer healing trials, treatment with H. pylori eradication therapy was not significantly different from treatment with an ulcer healing drug.43 For patients with large (>1.5 cm) or complicated gastric ulcers, however, additional antisecretory therapy has been shown to promote ulcer healing.44,45 Routine follow-up endoscopy is recommended to document ulcer healing, to exclude malignancy, and to confirm successful H. pylori eradication (see also Chapter 52).

ROLE OF MAINTENANCE THERAPY

After the eradication of H. pylori infection, there is little evidence that maintenance therapy with antisecretory agents is required, even for patients with complicated peptic ulcers.46,47 A meta-analysis showed that H. pylori eradication therapy was superior to no treatment in preventing recurrence of duodenal ulcer (relative risk, 0.19) or gastric ulcer (relative risk, 0.31).42 In another meta-analysis of H. pylori eradication therapy versus maintenance antisecretory therapy in prevention of recurrent ulcer bleeding, rebleeding occurred in 1.6% of the H. pylori eradication therapy group and 5.6% of the maintenance therapy group (odds ratio, 0.25; 95% confidence interval [CI], 0.08 to 0.76).48 Although some prospective trials reported that patients with duodenal ulcer had asymptomatic ulcer recurrences after eradication of H. pylori,49 these asymptomatic ulcers probably had little clinical significance.

ULCERS ASSOCIATED WITH NONSTEROIDAL ANTI-INFLAMMATORY DRUGS ACTIVE ULCERS Histamine-2 Receptor Antagonists There are limited data on the efficacy of H2 receptor antagonists in healing NSAID-associated ulcers. Current evidence suggests that conventional doses of H2 receptor antagonists effectively heal duodenal ulcers but are ineffective for gastric ulcers. In a multicenter study, the effects of ranitidine on ulcer healing were compared in a group of patients who had stopped NSAID therapy and another group who continued NSAID therapy. Gastric ulcers healed in 63% of

Chapter 53  Treatment of Peptic Ulcer Disease those still taking NSAIDs compared with 95% of those who had stopped. At 12 weeks, 79% of gastric ulcers and 92% of duodenal ulcers were healed in the group continuing NSAIDs, whereas all ulcers healed in those who had stopped taking NSAIDs.50 The ability of H2 receptor antagonists given in conventional doses to heal NSAID-associated ulcers also depends on the size of the ulcers. One early study reported that when NSAIDs were continued, 90% of gastric ulcers smaller than 5 mm healed after 8 weeks of cimetidine, whereas only 25% of ulcers larger than 5 mm healed.51

Proton Pump Inhibitors

Several large-scale studies have investigated the efficacy of PPIs for healing of NSAID-associated ulcers.52-55 Current evidence indicates that PPIs are superior to standard-dose H2 receptor antagonist therapy in healing NSAID-associated ulcers. In a large-scale randomized comparison of two doses of esomeprazole, 20 and 40 mg, and ranitidine, 150 mg twice daily, in patients who continue to take NSAIDs, ulcer healing at eight weeks was found in 85.7% of patients given esomeprazole 40 mg daily, in 84.8% of those given esomeprazole 20 mg daily, and in 76.3% of those given ranitidine 150 mg twice daily.52 In another study of 350 patients with NSAID-associated gastric ulcers who continued to use NSAIDs, ulcer healing at eight weeks was found in 69% of patients given lansoprazole 15 mg daily, in 73% of those given lansoprazole 30 mg daily, but in only 53% of those given ranitidine 150 mg twice daily.55

Misoprostol

In a randomized placebo-controlled trial in which patients continued NSAID therapy, misoprostol resulted in healing of gastric and duodenal ulcers in 67% of patients at eight weeks, compared with 26% of patients treated with placebo.56 However, misoprostol is not as effective as PPIs in healing NSAID-associated ulcers. One large-scale, randomized trial compared misoprostol 200 µg four times daily with omeprazole 20 or 40 mg daily in patients who continued NSAID treatment.53 After eight weeks, duodenal ulcers healed in 89% of patients receiving either dose of omeprazole and in 77% of those receiving misoprostol. Gastric ulcers healed in 80% of those receiving 40 mg of omeprazole, in 87% of those receiving 20 mg of omeprazole, and in 73% of those receiving misoprostol.

Sucralfrate

In a single-blind endoscopic study, sucralfate was significantly less effective than omeprazole in healing NSAIDassociated gastroduodenal ulcers.57

Adverse Role of Cyclooxygenase-2 Inhibitors in Ulcer Healing

There is good evidence that COX-2 inhibitors induce less gastric mucosal injury than conventional NSAIDs. However, animal experiments have consistently shown that COX-2, but not COX-1, is up-regulated in gastric ulcer.58,59 The administration of COX-2 inhibitors actually retards the healing of rodent gastric ulcers.60-62 These results suggest that prostaglandins generated by COX-2 contribute to restoring the integrity of gastric mucosa. A double-blind randomized trial of celecoxib on the healing of bleeding gastric ulcer found that after eight weeks, the ulcer healing rate was 65.7% in the celecoxib group and 80% in the placebo group.63 This finding indicates that treatment with COX-2 inhibitors such as celecoxib delays the healing of complicated gastric ulcers.

Recommendations

For patients in whom ulcers develop in association with the use of NSAIDs, it is recommended that NSAIDs should be discontinued if possible. Current evidence indicates that PPIs are more effective than H2 receptor antagonists, sucralfate, and misoprostol in healing NSAID-associated ulcers when continuous NSAID treatment is required. When NSAIDs can be discontinued, an H2 receptor antagonist is an effective alternative. Treatment with COX-2 inhibitors in patients with active ulcers who continue to require antiinflammatory therapy is not recommended. In the Maastricht III Consensus Report, eradication of H. pylori is advisable in patients who plan to start long-term NSAID therapy.2 The influence of H. pylori infection on the healing and relapse of NSAID-associated ulcer is discussed later.

ULCER PROPHYLAXIS

For many years an ulcer visible at endoscopy has been extensively used as a surrogate endpoint to assess the efficacy of prophylactic agents in preventing complications of NSAID-induced ulcers. An “endoscopic ulcer” has been arbitrarily defined as a circumscribed mucosal defect having a diameter of 5 mm or more with a perceivable depth.63 However, many studies have loosened this criterion to include flat mucosal breaks with a diameter of 3 mm or more as ulcers. The distinction between small ulcers and erosions is arbitrary and is prone to interobserver bias. The clinical relevance of these minor endoscopic lesions is uncertain. Although endoscopic findings roughly correlate with clinical outcomes in subjects at low to average risk for complications, current evidence indicates that the results of endoscopic studies cannot be generalized to high-risk patients.64 Because there are few prospective outcome trials to evaluate the true efficacy of prophylactic agents, clinical judgment relies on data largely using endoscopic endpoints.

Antacids

Antacids have no proven efficacy in the prevention of NSAID-induced ulcers. However, many clinicians still prescribe antacids as co-therapy for patients taking NSAIDs, both for symptom relief and prevention of ulcers. A casecontrol study showed that NSAID users receiving prophylaxis with antacids and H2 receptor antagonists had a more than two-fold increased risk of ulcer complications compared with those not taking these prophylactic agents.65 This finding was attributed to the possibility that antacids might have masked the dyspeptic symptoms, thereby creating a false sense of protection and raising the risk of silent ulcer complications. Co-prescription of antacids in patients taking NSAIDs who are at risk for ulcer should be discouraged.

Histamine-2 Receptor Antagonists

Several endoscopic studies investigated the efficacy of standard-dose H2 receptor antagonist therapy for the pre­ vention of NSAID-induced ulcers.66 A systematic review of randomized trials showed that H2 receptor antagonists significantly reduce the risk of endoscopic duodenal ulcers but not gastric ulcers.96 In contrast, it has been shown that double-dose famotidine significantly reduced the risk of both endoscopic duodenal and gastric ulcers. In one study, NSAID-related gastric ulcers developed in 20% of patients receiving placebo, in 13% of those receiving 40 mg of famotidine once daily, and in only 8% of those receiving 40 mg of famotidine twice daily.67 In another study, however, the gastric ulcer rates at 24 weeks were 41% and 19% in the groups receiving placebo and double-dose famotidine, respectively.68 The large discrepancy (8% vs.

873

874

Section VI  Stomach and Duodenum 19%) in the efficacy of double-dose famotidine between these two studies raises doubt about the true efficacy of this agent in preventing NSAID-induced gastric injury. To date there is no clinical outcome study to assess whether highdose H2 receptor antagonist therapy prevents NSAIDinduced ulcer complications.

Misoprostol

More than 20 randomized controlled trials assessed the efficacy of misoprostol in preventing NSAID-induced ulcers. A meta-analysis of the randomized trials indicated that all doses of misoprostol (400 to 800 µg per day) reduce the risk of NSAID-induced endoscopic ulcers.66 Only 800 µg per day of misoprostol is documented to reduce ulcer complications, however. In a large-scale, randomized double-blind trial in patients with rheumatoid arthritis who received NSAIDs, misoprostol (200 µg four times daily) significantly lowered the rate of GI complications, by 40% (0.95% in the placebo group vs. 0.57% in the misoprostol group). However, up to 30% of misoprostol-treated patients experienced GI upset, thus limiting its clinical use.69 Subsequent endoscopic studies suggested that lower doses of misoprostol, such as 200 µg two or three times per day, prevented NSAIDinduced endoscopic ulcers with fewer side effects.70 However, there is evidence that low-dose misoprostol therapy fails to prevent ulcer complications.71 Misoprostol has been found to be superior to H2 receptor antagonists for the prevention of NSAID-induced gastric ulcers. In one study, ranitidine (150 mg twice daily) was compared with misoprostol (200 µg four times daily) in long-term NSAID users.72 After four to eight weeks, about 1% of patients in each group demonstrated endoscopic duodenal ulcers. In contrast, gastric ulcers occurred in 5.7% of patients receiving ranitidine, compared with 0.6% of those receiving misoprostol.

Proton Pump Inhibitors

A systematic review of randomized controlled trials of PPIs for prophylaxis against NSAID-induced endoscopic ulcers found that PPIs significantly reduce the risk of endoscopic duodenal and gastric ulcers.66 The efficacy of PPIs has been compared with that of H2 receptor antagonists and misoprostol in patients who continued to receive NSAIDs. Two studies compared omeprazole 20 mg once daily with standard-dose ranitidine (150 mg twice daily) and halfdose misoprostol (200 µg twice daily) for six months.53,54 Omeprazole was found to be more effective than standarddose ranitidine but only comparable with half-dose misoprostol in preventing gastric ulcers. However, it should be noted that the superiority of omeprazole in preventing NSAID-related ulcer was due to a significant reduction of duodenal ulcers. A post hoc analysis revealed that most of the added protection attributable to omeprazole occurred among those with H. pylori infection. Another study compared high-dose misoprostol (200 µg four times daily) with two doses of lansoprazole (15 and 30 mg daily) for the prevention of ulcers in long-term NSAID users without H. pylori infection and with a history of gastric ulcer.73 Misoprostol was more effective than the two doses of lansoprazole in preventing gastric ulcer, but there was no practical advantage of misoprostol over lansoprazole because of the high withdrawal rate in the misoprostol group. In a headto-head endoscopic ulcer prevention study comparing two doses of pantoprazole with 20 mg/day of omeprazole in patients with rheumatoid arthritis receiving NSAIDs, the six-month probabilities of remaining ulcer free were 91%, 95%, and 93% for pantoprazole 20 mg, pantoprazole 40 mg, and omeprazole 20 mg, respectively.74

Two identical multicenter randomized clinical trials (RCTs) have been reported together. They compared esomeprazole (20 or 40 mg) with placebo in the prevention of ulcers in patients taking NSAIDs or COX-2 inhibitors over a six-month period. Patients in both studies were H. pylori negative, older than 60, and had a history of gastric or duodenal ulcer. Overall, the rates of ulcers were 17.0%, 5.2%, and 4.6% in the groups receiving placebo, esomeprazole 20 mg, and esomeprazole 40 mg, respectively.75 Whether PPIs can reduce the risk of NSAID-associated ulcer bleeding is largely based on observational studies and one randomized trial in high-risk patients. A large-scale case control study found that PPI therapy was associated with a significant reduction in risk of upper GI bleeding among chronic NSAID users (relative risk, 0.13; 95% CI, 0.09 to 0.19).76 One randomized trial compared long-term (six-month) omeprazole therapy with one week of H. pylori eradication therapy for the prevention of recurrent ulcer bleeding in H. pylori–infected patients with a recent history of NSAID-related ulcer bleeding who continued to use naproxen.77 Recurrent ulcer bleeding was seen in 18.8% of patients undergoing eradication therapy, compared with 4.4% of patients receiving omeprazole. In a randomized comparison of diclofenac plus omeprazole versus celecoxib for secondary prevention of ulcer bleeding in patients who either were H. pylori negative or had undergone H. pylori eradication,78 a similar proportion had recurrent bleeding in six months (6.4% in the combination therapy group compared with 4.9% of patients in the celecoxib group). These results indicate that omeprazole reduces but does not eliminate the risk of ulcer bleeding associated with NSAID use in very-high-risk patients. However, the following two important issues remain unresolved: first, the actual risk reduction achieved by PPI is unknown because of the lack of a placebo group, and second, there are no data on the efficacy of PPIs in preventing ulcer complications in low- or moderate-risk users of NSAIDs.

Role of Cyclooxygenase-2 Inhibitors in Ulcer Prevention

Consistent with the notion that inhibition of COX-2 spares the gastric mucosa, clinical trials using endoscopic ulcer as the endpoint have consistently shown that COX-2 inhibitors induced fewer ulcers than do conventional NSAIDs. Five COX-2 inhibitors have been evaluated in clinical trials: the sulfonamides celecoxib [Celebrex] and valdecoxib [Bextra] (parecoxib is a prodrug of valdecoxib), the methylsulfones rofecoxib [Vioxx] and etoricoxib, and the phenylacetic acid derivative lumiracoxib. Four large-scale clinical outcome studies—the Celecoxib Long-Term Arthritis Safety Study (CLASS),79 the Vioxx Gastrointestinal Outcomes Research Study (VIGOR),80 the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET),81 and the Multinational Etoricoxib and Diclofenac Arthritis Long-term programme (MEDAL)82—evaluated the gastrointestinal safety of celecoxib, rofecoxib, lumiracoxib, and etoricoxib, respectively. In the CLASS trial, patients with osteoarthritis or rheumatoid arthritis were randomized to receive celecoxib versus diclofenac or ibuprofen in two substudies of identical design. On primary analysis, there was no significant difference in the incidence of ulcer complications between the celecoxib group and the nonselective NSAIDs group.79 Whether the failure of CLASS was due to flaws in the study design remains controversial. The VIGOR study compared rofecoxib with naproxen in patients with rheumatoid arthritis.80 Unlike CLASS, VIGOR demonstrated that treatment with rofecoxib significantly reduced clinical GI events

Chapter 53  Treatment of Peptic Ulcer Disease (combined endpoint of ulcer complications and symptomatic ulcers) by about 50%, compared with treatment with naproxen. Patients requiring low-dose aspirin were excluded from the VIGOR study, whereas 20% of patients in CLASS received low-dose aspirin concomitantly. The TARGET study compared lumiracoxib with naproxen or ibuprofen in patients with osteoarthritis using two substudies of identical design.81 Randomization was stratified for low-dose aspirin use and age. TARGET showed that treatment with lumiracoxib significantly reduced the incidence of ulcer complications compared with nonselective NSAIDs in the subgroup of patients not taking low-dose aspirin. Current evidence indicates that low-dose aspirin negates the GI mucosa-sparing effect of COX-2 inhibitors. The MEDAL program was a prespecified pooled analysis of data from three prospective trials. A total of 34,701 arthritic patients were treated with 60 or 90 mg of etoricoxib or 150 mg of diclofenac daily. Unlike the previous large-scale studies, the primary endpoint of this study was cardiothrombotic events. Therefore, patients on low-dose aspirin were enrolled and encouraged to receive PPI co-therapy. There was no between-group difference in terms of complicated GI events including bleeding, perforation, or obstruction. However, the overall incidence of uncomplicated GI events was significantly less with etoricoxib than with diclofenac. The reduction in uncomplicated GI events with etoricoxib is maintained in patients treated with PPIs and is also observed with regular low-dose aspirin use.82 In a systematic review of randomized trials of COX-2 inhibitors, COX-2 inhibitors produced significantly fewer gastroduodenal ulcers (relative risk, 0.26; 95% CI, 0.23 to 0.30) and ulcer complications (relative risk, 0.39; 95% CI, 0.31 to 0.5), as well as fewer withdrawals caused by GI symptoms when compared with nonselective NSAIDs.83 Current evidence indicates that COX-2 inhibitors prob­ ably are as effective as a combination of nonselective NSAIDs combined with a PPI in patients at risk for ulcers. In a double-blind randomized outcome trial of celecoxib and the combination of omeprazole and diclofenac in patients with a recent history of ulcer bleeding, approximately 5% of patients in the two treatment groups still had recurrent ulcer bleeding at 6 months.78 Although the two treatments were comparable in terms of the incidence of ulcer bleeding, a subsequent follow-up endoscopic study showed that 20% to 25% of patients receiving either treatment developed recurrent ulcers at 6 months.84 These findings suggest that neither treatment could eliminate the risk of recurrent bleeding in very-high-risk patients. Recently, a double-blind randomized trial compared celecoxib alone with combination of celecoxib and esomeprazole in patients with a history of NSAID-associated ulcer bleeding. All patients had a negative test for H. pylori infection before randomization. After a median follow-up of 13 months, 8.9% of the celecoxib-alone group had recurrent ulcer bleeding compared with none of the combined therapy group (P = 0.0004).85

Cardiovascular Risk of COX-2 Inhibitors and Nonselective NSAIDs

Despite the improved gastric safety profile of COX-2 inhibitors, the cardiovascular risk associated with this new class of NSAIDs has been the subject of much concern. In the VIGOR study, the incidence of acute myocardial events, although low, was four times higher among patients receiving rofecoxib than among patients receiving naproxen.80 Whether the observed difference in infarction rates between the two treatments was related to an antiplatelet property

of naproxen or to a thrombotic effect of rofecoxib was hotly debated. Further data regarding the cardiovascular hazard of COX-2 inhibitors were derived from two long-term studies of colon polyp prevention using rofecoxib (Adenomatous Polyp Prevention on Vioxx [APPROVE] study)86 and celecoxib (Adenoma Prevention with Celecoxib [APC] study).87 In the APPROVE study, interim data at 18 months indicated that patients who received 25 mg rofecoxib a day had double the risk of serious cardiovascular events compared with patients who received placebo.86 In September 2004, rofecoxib was voluntarily withdrawn from worldwide markets in light of this unexpected finding. In APC study, interim data at 33 months showed that the occurrence of serious cardiovascular events was significantly higher for celecoxib at the very high dose of 400 mg twice a day (hazard ratio, 1.9; 95% CI, 1 to 3.3).87 In addition, a randomized, placebo-controlled trial of parecoxib and valdecoxib in patients who had undergone coronary artery bypass surgery found an almost four-fold increased risk of myocardial infarction.88 Do COX-2 inhibitors as a class increase the risk of myocardial infarction? Results of these three placebo-controlled trials indicate that this is the case. Both polyp prevention trials, however, investigated by design supratherapeutic doses of rofecoxib and celecoxib for extended time periods.86,87 In the TARGET study, rates of myocardial infarction with lumiracoxib were lower than with ibuprofen but higher than with naproxen. Neither result was statistically significant because the trial was underpowered to detect a difference in cardiovascular outcomes between treatment groups.84 The MEDAL program was a prespecified pooled analysis of cardiothrombotic events from three trials in which patients with osteoarthritis or rheumatoid arthritis were randomly assigned to etoricoxib (60 mg or 90 mg daily) or diclofenac (150 mg daily). After an average treatment of 18 months, rates of cardiothrombotic events were similar between the two treatment groups.89 Emerging evidence suggests that not only COX-2 inhibitors but also nonselective NSAIDs, with the exception of full-dose naproxen (1000 mg a day), increase cardiothrombotic risk. In a meta-analysis of randomized trials of COX-2 inhibitors (data mostly derived from rofecoxib and celecoxib), all COX-2 inhibitors increased the cardiothrombotic risk compared with placebo (risk ratio, 1.42; 95% CI, 1.13 to 1.78). This was largely attributable to an increased risk of myocardial infarction, with little difference in other vascular outcomes. A dose-dependent increase in cardiothrombotic events was observed with celecoxib. Importantly, there was no significant difference in cardiothrombotic risk between COX-2 inhibitors and nonselective NSAIDs. Naproxen (500 mg twice daily) was the only exception.90,91 In a meta-analysis of observational studies, high-dose rofecoxib (>25 mg a day), diclofenac, and indomethacin were associated with an increase in cardiothrombotic events, whereas celecoxib did not significantly increase the cardiothrombotic risk, though an increased risk could not be excluded with doses greater than 200 mg a day.92 In February 2005, the FDA issued recommendations proposing new serious labeling warnings for valdecoxib and celecoxib with respect to increased cardiovascular risks. In April of the same year, the FDA took the further step of asking the manufacturer to remove valdecoxib from the market. All sponsors of marketed prescription NSAIDs have been asked to revise the labeling for their products to include a boxed warning highlighting the potential for increased risk of cardiovascular events in addition to the potential life-threatening GI bleeding associated with their use. Manufacturers of over-the-counter NSAIDs are also

875

876

Section VI  Stomach and Duodenum being asked to revise their labeling to provide more specific information about the potential cardiovascular and GI risks associated with their individual products and to remind patients of the importance of limited dose and duration of treatment of these products in accordance with the package instructions.93 A large-scale, randomized head-to-head comparison of COX-2 inhibitors and nonselective NSAIDs using predefined cardiothrombotic events as the primary endpoint is under way.

THE ROLE OF HELICOBACTER PYLORI INFECTION IN ULCER DISEASE ASSOCIATED WITH NSAID USE

Whether H. pylori infection influences the risk of ulcer in patients receiving NSAIDs is one of the controversial issues in peptic ulcer research. Factors such as the choice of H. pylori diagnostic test, a history of ulcer complication, concomitant use of antisecretory agents, prior exposure to NSAIDs, and the use of low-dose aspirin affect the outcome.94 A meta-analysis showed that H. pylori raised the risk of ulcer bleeding more than 6-fold in patients receiving long-term NSAIDs, whereas H. pylori and NSAIDs alone raised the risk 1.79-fold and 4.85-fold, respectively.95 An updated meta-analysis showed similar findings.96 Among patients who are about to start NSAID therapy, eradication of H. pylori reduces the subsequent risk of ulcer development.97,98 Two systematic reviews have consistently shown that eradication of H. pylori is superior to placebo in preventing peptic ulcers among NSAID users.96,99 However, eradication of H. pylori infection alone is not sufficient for the prevention of ulcer bleeding in NSAID users with high ulcer risk.77 It has been suggested that the eradication of H. pylori might retard healing of gastric ulcers.100 This was not confirmed by a prospective randomized trial using ulcer healing as the predefined endpoint.101 Currently, there is no evidence that curing H. pylori infection has any clinically important negative effect on the healing of NSAID-related ulcers. There is growing evidence that H. pylori increases the ulcer risk in patients receiving low-dose aspirin. Among patients with H. pylori infection and a history of ulcer bleeding who continued to use low-dose aspirin, a randomized trial found that successful eradication of H. pylori alone substantially reduced the risk of recurrent bleeding in six months.77 However, a later low-dose aspirin study suggested that co-therapy with a PPI after eradication of H. pylori was still required because of a high failure rate

of H. pylori eradication and because concomitant NSAID use is not uncommon in clinical practice.102

RECOMMENDATIONS FOR THE PREVENTION OF NSAID-INDUCED ULCER COMPLICATIONS Assessment of Gastrointestinal Risk

Before the cardiovascular hazards of COX-2 inhibitors and nonselective NSAIDs was a concern, prevention of NSAIDinduced ulcer complications had been based on assessment of individual patients’ GI risk factors (Table 53-1). In clinical practice, patients receiving NSAIDs can be stratified according to their levels of GI risk, as follows (Table 53-2): • Low risk: absence of risk factors • Moderate risk: presence of one or two risk factors • High risk: history of ulcer complications, multiple (three or more) risk factors, or concomitant use of lowdose aspirin, glucocorticoids, or anticoagulant therapy Because H. pylori infection raises the risk of ulcer complications in NSAID users, patients with a history of ulcer who require NSAIDs should be tested for H. pylori, and if present, the infection should be eradicated. The Maastricht III Consensus Guidelines also consider it advisable to test and treat for H. pylori infection in patients who are about to start regular NSAID therapy.2 Low-Risk Patients Patients without risk factors are at very low risk of ulcer complications with NSAID use (1% per year). Rational use of NSAIDs, including avoidance of high doses of NSAIDs and use of a less ulcerogenic NSAID (e.g., ibu­profen, diclo­ fenac) at the lowest effective dose is a cost-effective approach. Table 53-1  Risk Ratios for the Various Risk Factors for Ulcer Complications Induced by NSAIDs* risk factor

risk ratio

History of complicated ulcer Use of multiple NSAIDs (including aspirin), cyclooxygenase-2 [COX-2] inhibitor) High doses of NSAIDs Use of an anticoagulant History of uncomplicated ulcer Age > 70 years Helicobacter pylori infection Use of a glucocorticoid

13.5 9 7 6.4 6.1 5.6 3.5 2.2

*Not all NSAIDs pose the same risk. NSAIDs, nonsteroidal anti-inflammatory drugs.

Table 53-2  Recommendations for Reducing the Risk of Ulcers Associated with Nonsteroidal Anti-inflammatory Drugs (NSAIDs) as a Function of Gastrointestinal and Cardiovascular Risk GI Risk

Low CV risk High CV risk†

LOW*

MODERATE*

HIGH*

Use the least ulcerogenic NSAID at the lowest effective dose Naproxen plus either a PPI or misoprostol

NSAID plus either a PPI or misoprostol Naproxen plus either a PPI or misoprostol

COX-2 inhibitor plus a PPI, or misoprostol Avoid NSAIDs or COX-2 inhibitors; use alternative therapy.

*Low GI risk denotes absence of any risk factors (see Table 53-1); moderate GI risk denotes presence of one or two risk factors; high GI risk denotes presence of three or more risk factors, prior complicated ulcer, or concomitant use of low-dose aspirin or anticoagulants. All patients with a history of ulcer who require NSAIDs should be tested for H. pylori, and if infection is present, eradication therapy should be given (see Chapter 50). † High CV risk denotes the requirement for prophylactic low-dose aspirin for primary or secondary prevention of serious cardiovascular events. COX-2, cyclooxygenase-2; CV, cardiovascular; GI, gastrointestinal; PPI, proton pump inhibitor.

Chapter 53  Treatment of Peptic Ulcer Disease Moderate-Risk Patients Moderate-risk patients, a group that by definition includes the older adult population, account for the majority of cases of NSAID-induced ulcer complications. These patients should receive co-therapy with antiulcer agents (PPIs or misoprostol). Alternatively, substitution of a COX-2 inhibitor alone maybe as effective as the combination therapy, although no study has focused on the moderaterisk group. High-Risk Patients In general, NSAIDs should be avoided in these patients, not only because of the high risk of ulcer complications but also owing to the serious consequences of ulcer complications in the presence of comorbidities. Glucocorticoid therapy can be considered if short-term anti-inflammatory therapy is required for acute, self-limiting arthritis (e.g., gout), because glucocorticoids alone do not increase the risk of ulcer. If regular anti-inflammatory therapy is required for chronic arthritis, the combination of a COX-2 inhibitor and either misoprostol or a PPI probably offers the best GI protection, although this approach remains to be examined in prospective trials.85

Approach to Patients with High Cardiovascular Risk

After the withdrawal of rofecoxib, rational prescription of NSAIDs has become a clinical challenge. Not only GI but also cardiovascular risk factor must be assessed for the individual patient (see Table 53-2). For patients who do not have a history of coronary heart disease or ischemic stroke, identifying those with significant cardiovascular risk is not that straightforward. The American Heart Association (AHA) recommends that aspirin should be considered in all apparently healthy men and women whose 10-year risk for a cardiovascular event is 10% or above.103 We consider patients with arthritis to have significant cardiovascular risk if they are already on aspirin for secondary prophylaxis or if they require aspirin for primary prophylaxis according to the AHA guidelines. Among patients with low cardiovascular risk, prescription of NSAIDs can be based on the presence of GI risk factors only. Patients known to have high cardiovascular risk should receive low-dose aspirin irrespective of NSAID use. Because of the potential cardiovascular hazards of COX-2 inhibitors and some nonselective NSAIDs, patients with high cardiovascular risk should avoid using these drugs if possible. Ibuprofen has been found to attenuate the cardioprotective effect of aspirin, possibly through competitive binding to platelet COX-1.104,105 Concomitant use of ibuprofen and low-dose aspirin therefore should be avoided. If anti-inflammatory analgesics are deemed necessary in patients at high cardiovascular risk, evidence suggests that full-dose naproxen (500 mg twice daily) does not increase the cardiothrombotic risk. However, it remains uncertain whether the cardioprotective effect of naproxen will persist at lower doses or when naproxen is co-prescribed with low-dose aspirin. Although naproxen has an antiplatelet effect, we do not recommend using naproxen as a substitute for low-dose aspirin in patients with high cardiovascular risk. This is because naproxen has a weak antiplatelet effect and patients take NSAIDs only intermittently for pain relief. One major drawback of concomitant use of NSAIDs and low-dose aspirin is that the combination will markedly increase the risk of ulcer complications over that incurred with NSAIDs alone (see Table 53-1; Chapter 52). Thus, co-therapy with a PPI or misoprostol is necessary even if patients do not have other GI risk factors (see Table 53-2).

REFRACTORY ULCERS Most peptic ulcers heal within eight weeks of initiation of antisecretory therapy. Nevertheless, there is a small, but considerable minority of patients whose ulcers persist despite conventional treatment. Such ulcers can be considered refractory. There is no standardized definition for refractory peptic ulcer, making comparisons among studies difficult. For the patient whose peptic ulcer does not heal despite a trial of conventional therapy, the clinician should ask the following questions: 1. Has the patient complied with the prescribed treatment? 2. Has the patient received an H2 receptor antagonist or a PPI? 3. Is there H. pylori infection? If antibiotic therapy already has been prescribed, the patient should be tested to confirm that the infection has indeed been eradicated. If no attempt has been made to seek and eradicate H. pylori infection, it should be made now. False-negative test results for H. pylori should be considered if the patient is undergoing acid-suppressive therapy. 4. Is the patient still taking an NSAID? NSAID use may be surreptitious. A careful history regarding the use of over-the-counter NSAIDs including low-dose aspirin should be obtained, and NSAIDs should be stopped if possible. 5. Does the patient smoke cigarettes? If so, he or she should be counseled strongly to discontinue cigarettes. 6. Has the duration of ulcer treatment been adequate? Large ulcerations require a longer duration of therapy than small ulcers to heal. A large ulceration (e.g., >2 cm) probably should not be considered refractory until it has persisted beyond 12 weeks of antisecretory therapy. 7. Is there evidence of a hypersecretory condition? A family history of gastrinoma or multiple endocrine neoplasia type 1 or a personal history of chronic diarrhea, hypercalcemia due to hyperparathyroidism, or ulcers involving the postbulbar duodenum suggests a diagnosis of Zollinger-Ellison syndrome or other hypersecretory state (see Chapter 32). 8. Is the ulcer penetrating the pancreas, liver, or other organ? 9. Is the ulcer indeed peptic? Primary or metastatic neoplasms, infections (e.g., cytomegalovirus), cocaine use, eosinophilic gastroenteritis, and inflammatory bowel diseases (e.g., Crohn’s disease) can cause ulcerations of the stomach and duodenum that can mimic peptic ulcers (see Chapter 52). These disorders should be considered and excluded appropriately.

STRESS-RELATED MUCOSAL INJURY Stress-related mucosal injury is an illness of the critically ill who are typically cared for in intensive care units. Fortunately, only a small proportion of patients with stressrelated mucosal lesions have clinically overt bleeding. In a prospective study of more than 2000 patients admitted to intensive care units, only 1.5% experienced clinically important bleeding.106 Respiratory failure and coagulopathy were strong, independent risk factors for stress-related hemorrhage. Important bleeding occurred in 3.7% of the 847 patients who had one or both of these risk factors, whereas

877

878

Section VI  Stomach and Duodenum only 0.1% of 1405 patients without respiratory failure or coagulopathy experienced such bleeding. Thus, routine uses of stress ulcer prophylaxis in the intensive care unit is not recommended unless the patient has a coagulopathy or is receiving mechanical ventilation. The approach to treatment of actively bleeding stress ulcers is similar to that of bleeding peptic ulcers (see later). Endoscopic control of hemorrhage followed by adjuvant therapy with PPI infusion is the preferred treatment option. However, the risk of recurrent bleeding is high, and bleeding carries a high mortality due to multiorgan failure. Although intravenous H2 receptor antagonists and oral or intragastric sucralfate are widely used to prevent stress ulcers in critically ill patients, the true efficacy of these treatments remains controversial. Moreover, few studies have evaluated PPIs for stress ulcer prophylaxis. The majority of studies have demonstrated that enteral or intravenous administration of PPIs to critically ill patients elevates intragastric pH and consistently maintains pH at 4 or higher.107 To date, no large-scale, clinical outcome study has been conducted to assess the efficacy of PPIs for this condition, despite their widespread use. An early, widely quoted metaanalysis conducted by Cook and associates108 concluded that H2 receptor antagonists were more effective than placebo in reducing the incidence of clinically important GI bleeding. However, in a later meta-analysis performed by Messori and colleagues,109 trials of ranitidine showed no difference from placebo; moreover, the data available on sucralfate did not allow any conclusion to be drawn. The Messori meta-analysis differed from the Cook meta-analysis in two important aspects. First, Cook and associates108 included trials that used either overt bleeding or clinically important GI bleeding as an endpoint, whereas Messori and colleagues included only trials that used clinically important GI bleeding as an endpoint. Second, in assessing the effectiveness of H2 receptor antagonists, Cook’s group included five trials that used cimetidine and three trials with negative results that used ranitidine, whereas Messori’s group included only the trials of ranitidine. A separate analysis showed that cimetidine but not ranitidine signi­ficantly reduced the rate of bleeding. The differences in outcome between cimetidine and ranitidine were probably due to chance rather than to a genuine difference between these two H2 receptor antagonists. Despite these discrepancies, there was either a rising trend108 or a significant increase109 in the incidence of nosocomial pneumonia with ranitidine compared with sucralfate. Overall, the findings are based on small numbers of patients with variable study design, and thus, firm conclusions cannot be drawn. Currently, neither the FDA nor the European Medicines Evaluation Agency has given approval to the routine use of H2 receptor antagonists, sucralfate, or PPIs for stress ulcer prophylaxis.

TREATMENT OF COMPLICATIONS OF PEPTIC ULCER DISEASE HEMORRHAGE 100,111 Initial Management (see Chapter 19)

Consensus groups have recommended a multidisciplinary approach to the care of patients presenting with upper GI bleeding.112,113 A team should include both medical and surgical gastroenterologists with access to skills in endoscopic hemostasis, which has become the mainstay of treatment. Patients identified as being at high risk of continued or recurrent bleeding should be admitted to an intensive care unit.

Patients with acute upper GI bleeding should be assessed promptly on presentation. Resuscitation and volume restoration should take priority and should precede endoscopy. Features of liver cirrhosis should call attention to the possibility of bleeding from esophagogastric varices rather than an ulcer. This distinction has prognostic as well as management implications. Variceal hemorrhage carries a higher death rate. The possibility of variceal hemorrhage calls for specific measures prior to endoscopy, such as the use of vasoactive drugs (e.g., octreotide) and antibiotics (e.g., cefotaxime) as prophylaxis against spontaneous bacterial peritonitis (see Chapters 90 and 91). The routine use of a nasogastric tube to obtain a gastric aspirate and to lavage the stomach cannot be recommended because in the majority of cases, the diagnosis of bleeding from an upper GI source is obvious. The insertion of a nasogastric tube is associated with a small risk of pulmonary aspiration, particularly in patients with compromised airways. Gastric lavage prior to endoscopy is often unnecessary and ineffective because blood and blood clots tend to pool at the fundus during endoscopic examination with the patient in the left lateral position, whereas bleeding lesions are usually located along the lesser curvature, angular notch, distal stomach, or duodenal bulb. The finding of fresh blood in an “in-and-out” nasogastric intubation and aspiration indicates ongoing and often massive bleeding and may predict a poor outcome.

Risk Stratification

Bleeding stops in about 80% of patients presenting with acute upper GI bleeding. The remaining 20% constitute a high-risk group with substantial morbidity and mortality. It is therefore important to identify and direct appropriate care to patients at risk of continued or recurrent bleeding. For practical purposes, the management distinctions to be made are whether the patient is in need of urgent endoscopy and whether the patient is likely to have recurrent bleeding after initial endoscopic control. Some of the clinical predictors of increased risk for continued or recurrent bleeding are older age, shock, comorbid illnesses, low hemoglobin value, need for transfusion, and the finding of fresh blood in emesis or on rectal examination.106 Patients with signs of ongoing bleeding need urgent endoscopy with a view to securing hemostasis. Older adult patients tolerate blood loss poorly and are likely to have organ dysfunction consequent to bleeding; in such patients, the basis for early intervention should be more liberal. Other clinical predictors associated with higher mortality include the onset of bleeding in patients already hospitalized for other reasons. Several derived risk scores have been developed to aid physicians in clinical decisions.114-116 The Rockall and Baylor scores are composite systems consisting of two components, the pre-endoscopy and postendoscopy scores.114,115 The Rockall scoring system (see Chapter 19) was derived from data gathered from the National United Kingdom Audit. A score of 0 to 2 indicates an excellent prognosis, whereas a score of 9 or more is associated with a high risk of death. The Blatchford score, on the other hand, uses clinical parameters only and is calculated from patients’ hemoglobin and blood urea concentrations, pulse and systolic blood pressure on admission, the presence or absence of melena or syncope, as well as of evidence of cardiac or hepatic failure.116 Endoscopic stigmata of bleeding not only pinpoint the source of bleeding but are themselves prognostic. The commonly used nomenclature is a version modified from Forrest and Finlayson’s117 original description, as follows (Table 53-3):

Chapter 53  Treatment of Peptic Ulcer Disease Table 53-3  Frequency and Prognosis of Various Endoscopic Stigmata of Hemorrhage in Patients with Bleeding Peptic Ulcer* ENDOSCOPIC CHARACTERISTIC†

FREQUENCY (%)

FURTHER BLEEDING (%)

SURGERY (%)

MORTALITY (%)

42 20 17 17 18

5 10 22 43 55

0.5 6 10 34 35

2 3 7 11 11

Clean base (type III) Flat pigmentation (type IIc) Adherent clot (type IIb) Nonbleeding visible vessel (type IIa) Active bleeding (type I)

*Percentages are average figures taken from multiple studies118 and therefore do not add up to 100%. † Classification by Forrest et al117 is shown in parentheses (see text). Adapted from Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med 1994; 331:717.

Figure 53-1.  Endoscopic appearance of a duodenal bulbar ulcer with a fresh adherent clot (Forrest type IIb).

Type I: Active bleeding: Ia: Spurting hemorrhage Ib: Oozing hemorrhage Type II: Stigmata of recent hemorrhage: IIa: Nonbleeding visible vessel IIb: Adherent clot IIc: Flat pigmentation Type III: Clean-base ulcers Reported prevalences of these endoscopic stigmata and their respective risks of recurrent bleeding have varied widely. This variation may be attributed to differences in visual interpretation among endoscopists and to varying definitions of recurrent bleeding. In one review,118 the rate of recurrent bleeding was less than 5% in patients with a clean ulcer base (type III) and increased to 10% in patients with a flat pigmentation (IIc), to 22% in those with an adherent clot (IIb), to 43% in those with a nonbleeding visible vessel (IIa), and to 55% in those with active bleeding, either spurting or oozing (type I) (see Table 53-3). Actively bleeding ulcers and ulcers with nonbleeding visible vessels (“protuberant discoloration”) warrant endoscopic therapy.119 Endoscopic therapy of ulcers with “adherent clots” has been controversial (Fig. 53-1). The definition of adherent clot varies with the vigor in endoscopic washing. Some endoscopists use targeted irrigation from a thermal probe. Some go to the extent of mechanical removal using a

Figure 53-2.  Endoscopic appearance of a gastric ulcer at the angularis with a flat spot (Forrest type IIc).

polypectomy snare. Two randomized controlled studies and a meta-analysis compared medical therapy with endoscopic treatment in patients with ulcers harboring “adherent clots” and concluded that clot removal followed by endoscopic treatment to the vessel underneath would reduce the risk of recurrent bleeding from around 30% to 5%.120-122 It is not often possible to distinguish a clot from a vessel. Indeed, it is logical to believe that for every clot there is an underlying artery. Johnston123 introduced the term sentinel clot, which is often used synonymously with visible vessel. An ulcer stops bleeding when an eroded artery is plugged by a clot, which varies in color. The “sentinel clot” can be contiguous with a larger overlying clot. With time, the ulcer heals, initially leaving a flat pigment to the ulcer base, and the vessel eventually disappears from the ulcer floor. This evolution of a bleeding vessel usually takes less than 72 hours. Ulcers with a flat spot or a clean base do not warrant endoscopic therapy (Fig. 53-2).

Endoscopic Therapy (see also Chapter 19) Early endoscopy is generally defined as endoscopic examination performed within 24 hours of the patient’s admission. In patients with signs of active bleeding, it is common sense that urgent endoscopy establishes diagnosis and offers possible intervention. Such an approach in high-risk patients is generally believed to improve outcome. Rando­

879

880

Section VI  Stomach and Duodenum mized controlled trials demonstrated that early endoscopy in patients at low risk enabled their early hospital discharge and resource utilization or even management as outpatients.124-126 Many low-risk patients can be scheduled for endoscopy the morning after admission. A small but significant portion of patients with major bleeding require urgent endoscopy and therapy. Two meta-analyses of endoscopic therapy showed signi­ ficant reductions in the rates of further bleeding, surgery, and, importantly, mortality.127,128 Endoscopic therapy can be divided into injection, thermal, and mech­anical methods. Injection Methods Endoscopic injection of diluted epinephrine into a bleeding peptic ulcer works by volume tamponade and local vasoconstriction, as blanching and edema of the mucosa are observed. It is an easy technique to learn and diluted epinephrine is non-tissue damaging and therefore safe to use. Diluted epinephrine, however, does not induce vessel thrombosis. Recurrent bleeding after injection with diluted epinephrine alone occurs in 20% to 30% of patients. In theory, the addition of a second agent to cause vessel thrombosis would further reduce the rate of recurrent bleeding. Various sclerosants (e.g., sodium tetradecyl sulfate, polidocanol, absolute alcohol) have been applied to the vessel after initial hemostasis with epinephrine. The addition of a sclerosant has not shown to further reduce rebleeding.129 Sclerosants damage tissue in a dose-dependent manner. Cases of gastric necrosis, some of them fatal, have been reported after sclerosant injection.130 Thrombin and fibrin, derived from both bovine and human sources, have been used as injection agents. A largescale European multicenter trial demonstrated a statistically significant lower rate of recurrent bleeding associated with repeated injections of fibrin sealant (a mixture of fibrin and thrombin) at scheduled daily endoscopic examinations in comparison with a single injection of epinephrine plus polidocanol.131 The close surveillance rather than action of the fibrin sealant per se might have accounted for the difference. A single injection of fibrin sealant was not superior to epinephrine-polidocanol injection. There are also concerns about transmission of viral agents and anaphylaxis with the use of products derived from pooled plasma. Other agents for endoscopic injection therapy are normal saline and hypertonic saline. No single solution is superior to another for hemostasis. Diluted epinephrine is the only remaining agent that is widely used for injection therapy because of its safety, both local and systemic, as well as its low cost and easy availability. Thermal Methods Thermal methods of endoscopic therapy are divided into contact and noncontact methods. Noncontact methods refer to the former use of laser photocoagulation and the current use of argon plasma coagulation. Laser therapy is no longer used because laser units are bulky and difficult to transport. In canine mesenteric artery models, Johnston and colleagues132 compared laser photocoagulation with contact thermal probes in hemostasis. The use of 3.2-mm contact probes consistently sealed arteries up to 2 mm in size. Laser probes were much less effective. The researchers introduced the term coaptive thermocoagulation and emphasized the need for compression of vessel walls. The two walls of an artery are pressed together by firm tamponade. This in itself stops blood flow and reduces the “heat-sink” effect. Heat energy is then generated, welding the arterial lumen. The commonly used contact thermal probes are the heater probe, which has a polytetrafluoroethylene (Teflon)-coated copper

tip with three water ports for targeted irrigation, and bipolar probes. Firm tamponade is the key to successful application of contact probes. At least in animal experiments, thermal methods are superior to injection therapy in achieving hemostasis. Comparative clinical trials of injection and thermal methods did not show any difference in clinical outcomes. Mechanical Methods Because surgical plication of the bleeding artery is considered the most definitive treatment to achieve hemostasis, mechanical methods such as the endoscopic application of a hemoclip come closer to what would otherwise be done at surgery. The tangential application of hemoclips in treating posterior duodenal bulbar ulcerations or their use with the endoscope in a retroflexed position for treatment of high lesser curvature ulcers can be technically difficult. In a meta-analysis that included 15 randomized studies that compared injection, thermocoagulation, and hemoclipping, successful application of hemoclips was superior to injection alone but comparable to thermocoagulation in producing definitive hemostasis.133 Combination Methods A combination therapy incorporating advantages of injection and thermal methods may represent a better approach than either method alone. Preinjection with diluted epinephrine allows a clear view of the bleeding vessel, making accurate thermocoagulation or application of a second modality possible. The benefit of combination therapy has been confirmed in two meta-analyses.134,135 In the first meta-analysis, the addition of a second modality significantly reduced the rate of recurrent bleeding from 18.4% to 10.6% and that of emergency surgery from 11.3% to 7.6%. The mortality rate decreased significantly from 5.1% to 2.6%. Eleven studies used injected substances such as a sclerosant, tissue adhesive, or thrombin; two added hemoclips; and three evaluated the added use of thermal devices. Findings of the meta-analysis suggest that a second modality should be added after injection of diluted epinephrine to bleeding peptic ulcers. The meta-analysis also confirmed that the rate of significant complications such as perforation and gastric wall necrosis was higher in the combined therapy group (6 of 558 patients) than in the epinephrine-alone group (1 of 560 patients). Furthermore, the improvement in prognosis seems to be more evident in ulcers with active bleeding (Forrest type I ulcers). Despite the large volume of published literature, the best endoscopic therapy for bleeding peptic ulcers remains undefined. It is becoming clear that injection with diluted epinephrine alone is inadequate, especially in ulcers with active bleeding. The most widely adopted method is probably the combination therapy of preinjection with diluted epinephrine followed either by thermocoagulation using a 3.2-mm contact probe or the use of hemoclip. Thermocoagulation or hemoclipping on its own may suffice if the bleeding lesion can be seen clearly for its accurate application. The critical determinant of the efficacy of endoscopic therapy is the size of the eroded artery in the ulcer base. Swain and associates136 studied gastrectomy specimens in patients who required emergency gastrectomy for bleeding gastric ulcers. The researchers suggested that bleeding from arteries larger than 1 mm could not be stopped by existing methods of hemostasis. Most studies on predictors of persistent or recurrent bleeding from ulcers find ulcer size of larger than 2 cm and ulcer location either high on

Chapter 53  Treatment of Peptic Ulcer Disease the lesser curvature of the stomach or in the superior or posterior duodenal bulb to be associated with poorer outcomes. These are the classic locations for ulcers that erode into major artery complexes, such as the left gastric artery and the gastroduodenal artery, respectively. Surgery remains the only definitive method of securing bleeding in these patients.

Antisecretory Therapy

The rationale for antisecretory therapy is based on the fact that both pepsin activity and platelet aggregation are pH dependent. An ulcer stops bleeding when a fibrin or platelet plug blocks the rent in a bleeding artery. When gastric pH exceeds 4, pepsin is inactivated, preventing enzymatic digestion of blood clots. A gastric pH of 6 or greater is critical for clot stability and hemostasis. Labenz and associates137 studied gastric pH in patients with peptic ulcers receiving either a high dose of omeprazole (intravenous bolus 80 mg, followed by 8 mg/hr) or a high-dose ranitidine infusion (intravenous bolus 50 mg, followed by 0.25 mg/kg/hr). The gastric pH was less than 6 just 0.1% of the time in patients with either gastric or duodenal ulcers treated by high-dose omeprazole, much less than with ranitidine (20% in duodenal ulcers and 46% in gastric ulcers). In another study that measured gastric pH over three days, the use of histamine receptor antagonists given either in high-dose intravenous infusion or in bolus form led to progressive loss of antisecretory effect over days two and three because of tolerance. To achieve a gastric pH consistently above 6, a high-dose proton pump infusion is required. The use of H2 receptor antagonists in the management of bleeding peptic ulcers has been evaluated in numerous clinical trials and summarized in meta-analyses. Patients with duodenal ulcer bleeding, who typically have a higher gastric acid output, do not benefit from the use of H2 receptor antagonists. A recent meta-analysis of 30 rando­ mized trials concluded that the use of H2 receptor antagonists would be of benefit only in patients with gastric ulcers (absolute risk reductions of 7.2%, 6.7%, and 3.2% in the rates of recurrent bleeding, surgery, and death, respectively).138 Strong evidence for the use of PPIs in patients with bleeding peptic ulcers comes from a clinical trial reported by Lau and associates,139 in which early endoscopy was used to triage patients with bleeding peptic ulcers; only those at high risk of recurrent bleeding (i.e., those who had actively bleeding ulcers or ulcers with nonbleeding visible vessels) were enrolled. After endoscopic thermocoagulation of the ulcers, patients were randomly assigned to receive a highdose omeprazole infusion or a placebo for 72 hours. The rate of recurrent bleeding in those who received the PPI infusion was 6.7% at day 30, compared with 22.5% in those who received placebo. In addition, the trial showed significant reductions in the need for further intervention, transfusion, and hospitalization as well as a trend in reducing the death rate in patients who received omeprazole. Similar benefits with intravenous esomeprazole given after successful endoscopic therapy in patients at high risk for recurrent ulcer bleeding have recently been reported.140 A Cochrane Systematic Review later concluded that the use of PPI therapy significantly reduces rates of recurrent bleeding and surgery but not mortality.141 In a subgroup analysis including studies that allowed initial endoscopic control, a significant reduction in mortality among Asians was seen in association with the use of a PPI. This supports the use of PPI as an adjunct to endoscopic therapy. The optimal dose to use and the routine of PPI administration

continue to be controversial. The authors advocate the use of early endoscopic triage with a strategy to treat actively bleeding ulcers and ulcers with nonbleeding vessels, followed by adjunctive use of a high-dose intravenous infusion of a PPI. Preemptive use of an intravenous PPI infusion prior to endoscopy was studied in a large-scale randomized study.142 Patients with overt signs of upper GI bleeding were randomized to receive either a high dose PPI infusion or placebo. In the cohort, 60% were found to be bleeding from a peptic ulcer during endoscopy. The study demonstrated that early PPI infusion downstaged bleeding stigmata in ulcers and thereby reduced the need for endoscopic therapy. In the PPI group there were fewer ulcers with active bleeding or with major stigmata observed the next morning during endoscopy. PPI infusion starts ulcer healing, and significantly more clean-based ulcers are seen the next day. The study has cost-saving implications with less endoscopic therapy required with the use of intravenous PPI. In patients awaiting endoscopy it is reasonable to start PPI therapy.

Emergency Surgery

Indications Effective endoscopic intervention and improved pharmacotherapy have greatly reduced the need for emergency ulcer surgery. Not so long ago, surgery was the only reliable means of stopping bleeding. The National United Kingdom Audit performed more than a decade ago revealed an operative rate of 12% among 2071 patients with bleeding peptic ulcers and an associated mortality rate of 24%.143 Endoscopic intervention had not been used in 78% of these patients. In the current literature, surgery is often defined as an outcome in clinical trials of endoscopic therapy. Despite its diminished role in the management of bleeding peptic ulcer, surgery remains important. A common indication for emergency surgery is failure of endoscopic therapy. The usual scenarios are as follows: (1) spurting hemorrhage could not be stopped by endoscopic means; (2) the bleeding point could not be seen because of heavy active bleeding; and (3) recurrent bleeding appeared after initial endoscopic control (although as discussed later it is not entirely clear how many endoscopic attempts should be made before endoscopic therapy is deemed to have failed). Timing The timing of surgery for ulcer bleeding has been a subject of intense debate. In the 1980s, when endoscopic therapy was not available, Morris and colleagues144 published the only prospective randomized study that compared early surgery with delayed surgery, if needed, in 140 patients with bleeding ulcers. In patients younger than 60 years, there was no death in either group, but the more aggressive early surgery policy led to an unacceptably high operation rate (52% compared with only 5% for the delayed surgery group). For those older than 60 years, the operation rate was 62% in the early group and 27% in the delayed group. There were three deaths in 48 patients (6%) in the early surgery group and seven deaths in 52 patients (13%) in the delayed group. On intention-to-treat analysis, the difference did not reach statistical significance. According to treatmentreceived analysis, difference in mortality in patients with gastric ulcers was statistically significant (0 deaths in 19 of the early group vs. 5 in 21 of the delayed group, P < 0.01). The trial has been criticized for its small sample size and the use of subgroup analysis. In patients assigned to delayed surgery, ongoing bleeding was allowed before surgical intervention. Nevertheless, the study clearly demon-

881

882

Section VI  Stomach and Duodenum strated that early intervention reduced blood loss in older adult patients and improved outcome. Endoscopic therapy has replaced surgery as the first treatment in bleeding ulcers. After initial endoscopic control, most clinicians would adopt an expectant policy. Most of the affected patients are elderly and have comorbid illnesses. The prospect of long-term cure with a powerful PPI and H. pylori eradication provides incentives for clinicians to adopt a conservative stance. There remains a subgroup of patients predicted to be at risk for recurrent bleeding on the basis of ulcer characteristics at endoscopy. Such characteristics include larger ulcers located at the high lesser curvature and posterior duodenal bulb. The team approach allows experienced gastroenterologists and GI surgeons to confer after initial endoscopic control of bleeding. It is possible that some of these “difficult” ulcers should be selected for safer early elective surgery. A clinical trial to answer this question is unlikely. Choice of Operation The type of operation to be undertaken at emergency surgery for ulcer bleeding is controversial. Some surgeons maintain that oversewing of ulcers alone combined with acidsuppression therapy is safer than “definitive” surgery using either gastrectomy or vagotomy. H. pylori eradication and PPIs have provided incentives for surgeons to perform the minimum. Two randomized studies that compared minimal with definitive surgery have been published.145,146 A United Kingdom multicenter study compared minimal surgery (oversewing the vessel or ulcer excision alone plus intravenous H2 receptor antagonist therapy) with a definitive ulcer surgery (vagotomy and pyloroplasty or partial gastrectomy) in patients with bleeding gastric or duodenal ulcerations. Of the 62 patients assigned to conservative treatment, 7 experienced rebleeding, of whom 6 died. Of the 67 patients who received conventional ulcer surgery, 4 had rebleeding and none died; in all rebleeding cases, vagotomy and oversewing of the ulcer had been performed. The overall mortality in this high-risk group of patients was similar in the two groups: 26% after minimal surgery and 19% after conventional surgery. The trial was terminated because of the high rate of fatal rebleeding in the conservative surgery group in comparison with the conventional surgery group.145 In the French Association of Surgical Research trial, patients with duodenal ulcers were randomly assigned to either oversewing plus vagotomy and drainage or partial gastrectomy.146 After oversewing and vagotomy, recurrent postoperative bleeding occurred in 10 of 60 patients (17%), in 6 of whom conversion to a Billroth II gastrectomy was required. Five of these 6 patients experienced duodenal stump dehiscence. In the group of 60 assigned to undergo partial gastrectomy, only 2 patients (3%) had rebleeding, both of whom recovered after conservative treatment. Of the 60 patients assigned to partial gastrectomy, Billroth I reconstruction was performed in 18, Billroth I reconstruction plus vagotomy in 6, Billroth II reconstruction in 20, and Billroth II reconstruction plus vagotomy in 16 patients. No duodenal leak occurred in 24 patients after Billroth I reconstruction. Among the 36 patients who received Billroth II reconstruction, duodenal stump leaks occurred in 8 (22%). The rate of duodenal stump leak in the overall gastrectomy group was therefore 8 in 60 (13%). When the results were analyzed on an intention-to-treat basis, and data from patients with duodenal leaks after reoperations for rebleeding in the oversewing and vagotomy groups were included, the duodenal leak rates were similar in the two groups (7 of 58 vs. 8 of 60, respectively). The researchers concluded that a more

aggressive approach would be warranted in the surgical treatment of duodenal ulcers. The two randomized studies just reviewed emphasize that simple oversewing with or without vagotomy is associated with a higher rate of recurrent bleeding. In patients with recurrent bleeding, the mortality is exceptionally high. Exclusion of an ulcer (see later) or, in the case of gastric ulcers, ulcer excision is important in preventing recurrent bleeding. It is inappropriate, however, to recommend partial gastrectomy in all cases. The decision to perform a gastrectomy has to be balanced against the risk of duodenal stump dehiscence. The choice of resection is determined to a large extent by anatomic and pathologic considerations. The choice of the proper surgical procedure for the individual patient with massive gastric or duodenal ulcer bleeding also rests heavily on the experience and the judgment of the surgeon. Expertise in the surgical management of bleeding ulcers remains an important integral feature of an upper GI bleeding team. Difficult Ulcers Effective endoscopic therapy has selected a group of difficult ulcers for surgery. Anatomic factors that predict failure with endoscopic therapy are size larger than 2 cm and location at the posterior duodenal bulb or lesser curvature of the stomach. An ulcer at the former location often erodes into the gastroduodenal artery complex, and an ulcer at the latter location often erodes into the main left gastric artery or its first-generation branches. The size of the artery well exceeds the limits of endoscopic hemostasis. These ulcers represent challenges to GI surgeons, and expertise is required in dealing with them. At surgery, a bulbar duodenal ulcer can be accessed via a longitudinal pylorotomy extending into the duodenum. Berne and Rosoff147 identified the confluence of several branches of the gastroduodenal artery in the vicinity of a bleeding posterior duodenal ulcer. Ligations above and below the bleeding artery are insufficient to ensure hemostasis. Berne and Rosoff suggested a U stitch in the center after ligations above and below. Many surgeons perform plications at four quadrants, and a few figure-of-eight stitches at varying angles along the course of the artery are often required. The longitudinal pylorotomy is then closed vertically as a Mikulicz-Heineke type of pyloroplasty.147 Whether the procedure should be completed with a truncal vagotomy is unclear, because powerful PPI therapy is now available. Recurrent bleeding occurs in 5% to 17% of cases after a vagotomy and pyloroplasty, often with a fatal outcome. To avoid this complication, many surgeons argue for excluding the duodenal ulcer by closing the duodenal stump distal to the ulcer. Some surgeons advocate end-toend gastroduodenostomy (a Billroth I type reconstruction), in which the gastric remnant is advanced over the ulcer crater and sutured to the normal duodenal mucosa distal to the ulcer. No attempt is made to dissect the posterior duodenal wall distal to the inferior border of the ulcer. Often the duodenum retracts distally, and the stump can be closed by suturing of the divided anterior duodenal wall onto the distal lip of the ulcer (Nissen’s method). Other techniques in dealing with a difficult duodenal stump include a side catheter duodenostomy and the technique of Roux-en-Y jejunoduodenal anastomosis. Ulcers located at the incisura angularis and lesser curvature can erode into the left gastric artery or its first branches. In larger chronic ulcers, ulcer resection is often necessary. A high gastric ulcer presents a special problem. Oversewing of the bleeding point through an anterior gastrotomy is a simple option for old and frail patients. For lower-risk

Chapter 53  Treatment of Peptic Ulcer Disease patients, a Pauchet operation or a sleeve resection technique can be used. The stomach is transected from the greater curvature side. On the lesser curvature side, a tongue of gastric tissue that is based distally and stretches proximally toward the cardia is excised, usually freehand, to include the ulcer. A Roux-en-Y reconstruction is often required, with a large portion of stomach excised. Closure of duodenal stump is not a concern here because the duodenum is not diseased.

surgery for duodenal ulcer perforation are sterile; this is especially true when surgery is undertaken soon after perforation because the initial peritonitis is chemical in etiology. Nevertheless, acid-suppressive agents raise gastric pH and allow bacterial overgrowth in the stomach. The yield from culture increases with time as secondary peritonitis sets in. Coliforms, streptococci (usually anaerobic), staphylococci, and Candida species are the common organisms isolated. In late disease with abscess formations, anaerobes can often be isolated.

Surgery Versus Endoscopic Retreatment After Recurrent Bleeding

Medical Management

Recurrent ulcer bleeding is a major adverse prognostic factor for morbidity and mortality. Physicians often perform a second endoscopic examination to confirm recurrent bleeding and to re-treat the bleeding ulcer. The avoidance of salvage surgery may be desirable in older adult patients. There is, however, a concern that patients with rebleeding could be worse off after yet another failed endoscopic attempt and episodes of hypotension. The choice between endoscopic re-treatment and surgery for recurrent bleeding after initial endoscopic control was addressed by Lau and colleagues148 in a randomized trial. In a cohort of 1169 patients with bleeding peptic ulcers treated by epinephrine injection followed by thermocoagulation, recurrent bleeding occurred in 8.7%. Ninety-two patients (mean age 65 years, 76% men) were randomly assigned to undergo either endoscopic re-treatment or surgery. Using intention-to-treat analysis, the endoscopic re-treatment and surgery groups did not significantly differ in mortality at 30 days (10% for re-treatment vs. 18% for surgery), duration of hospitalization (median 10 vs. 11 days, respectively), need for intensive care or length of stay in an intensive care unit (5 vs. 10 patients, respectively; median of 59 days for both), or units of blood transfused (median 8 vs. 7 units, respectively). Patients who underwent surgery were significantly more likely to have complications (16 vs. 7, respectively). Endoscopic re-treatment was able to control bleeding in three quarters of the patients. In those for whom endoscopic re-treatment failed, salvage surgery carried substantial mortality. In a regression analysis of a small subgroup of patients, ulcers 2 cm or larger and hypotension at rebleeding were two independent factors predicting failure with endoscopic re-treatment. Findings of this trial suggest that a selective approach can be adopted on the basis of the characteristics of the ulcer.148 Large chronic ulcers should probably be treated with expedited surgery at the time of rebleeding. Early elective surgery may have been more appropriate in these chronic ulcers after initial endoscopic control. Angiographic embolization of bleeding arteries to peptic ulcer is a nonoperative option. In a nonrandomized comparison to surgery, angiographic embolization carried a similar rate of recurrent bleeding (29% vs. 23%), need for further intervention (16% vs. 31%), and death (26% vs. 21%).149

PERFORATION

A short period of resuscitation after ulcer perforation is diagnosed or suspected is often desirable, with restoration of fluids, electrolytes, and, if needed, blood. A nasogastric tube and a urinary catheter should be in place. Pain should be relieved with opiates after a presumptive diagnosis is made. As described in Chapter 37, intravenous broadspectrum antibiotics should be administered parenterally, even though gastric juice contains few organisms. A majority of cultures of peritoneal fluid collected at the time of

Nonoperative management of ulcer perforation has been advocated. The regimen involves nasogastric suctioning, parenteral antibiotics, and intravenous fluids. Crofts and associates150 randomly assigned patients with the presumptive diagnosis of perforated ulcers to either conservative treatment or surgery. Of 40 patients assigned to conservative treatment, 11 showed no improvement in 12 hours and underwent operation. Three of these 11 patients were found to have perforated carcinomas, 2 gastric and 1 sigmoid. Morbidity and mortality rates were similar in the medical and surgical groups. This trial may have included lower-risk patients or patients who presented early, because the 5% overall mortality is low. Findings of the study highlight common objections to the use of nonoperative management error in diagnosis, uncertainty of site of perforation, and the possibility of a perforated gastric tumor. Older adult patients should be operated on early for the following reasons: (1) malignancy is more likely; (2) atrophy of the greater omentum makes spontaneous sealing less likely; (3) such patients often withstand sepsis and organ dysfunction poorly; and (4) early surgery leads to a better outcome.150 In a patient in whom the perforation is considered confined on clinical assessment, it would be prudent to verify this judgment with a contrast radiographic study. If sealing is confirmed, conservative treatment may be reasonable.

Surgical Management

Surgery is the usual approach to an ulcer perforation. At the time of surgery the surgeon has to treat peritoneal contamination, the perforation, and the ulcer. The controversies in the operative management have been the need for definitive ulcer operation, the choice of which definitive operation to perform, and, more recently, the choice between laparoscopic and open suture repair. Treatment also differs for duodenal and gastric perforations. For example, simple closure of a perforated duodenal or a juxtapyloric ulcer with the use of an omental patch is widely practiced. Whether to perform definitive ulcer surgery at the time of ulcer perforation was an argument that predated the era of H. pylori and PPIs. Relapse of ulcer disease with modern medical therapy is now uncommon and therefore is no longer a factor in the consideration as to the type of surgery to perform. There is now evidence that H. pylori eradication reduces relapse of ulceration after patch repair.151 In a randomized study enrolling patients with perforated duodenal ulcers after omental patch repair, patients received either a PPI alone for four weeks or quadruple anti–H. pylori therapy. After a year, ulcer relapse had occurred in 38% in the PPI group compared with only 5% in the H. pylori eradication group. Thus, ulcer relapse is uncommon after simple closure of a perforation and H. pylori eradication. These data would support the performance of simple closure alone in perforated ulcers.151 Gastric ulcers account for about 20% of perforated peptic ulcers. Epidemiologic data suggest a rising proportion of gastric ulcers among perforated ulcers, especially in older

883

884

Section VI  Stomach and Duodenum adult patients who use NSAIDs. Patients with perforated gastric ulcers are more likely to be older and to have significant comorbid illnesses, making their prognosis less favorable. As with perforated duodenal ulcers, there has been debate regarding the choice of surgery for perforated gastric ulcers. No prospective controlled trials comparing simple closure and definitive treatment of a perforated gastric ulcer have been performed. An attempt should be made to determine the type of gastric ulcer. The optimal treatment of an angular notch ulcer along the lesser curvature of the stomach should entail an antrectomy followed by a Billroth type I gastroduodenostomy. High lesser curvature ulcers near the posterior wall can be managed with the Pauchet procedure as mentioned. Simple closure should be considered in prepyloric ulcers, especially when they are small. The role of a vagotomy or indeed gastric resection in these ulcers is unclear. Although the role of H. pylori in perforated prepyloric or antral gastric ulcers after their patch closure is not known, it is probably similar to the organism’s role in perforated duodenal ulcer. The advocates for primary resection in perforated gastric ulcers argue that mortality rates after gastrectomy are not increased and that the rate of postoperative ulcer-related complications is reduced. The arguments for primary resection include the possibility that the ulcer is malignant. Malignancy is seen in approximately 6% of perforated gastric ulcers.152 In a retrospective series comprising 287 perforated gastric ulcers, death occurred in 21.5% of patients who underwent patch closure alone and in 24.3% of those who underwent gastrectomy.153 A patch closure is a far more popular operation for perforated gastric ulcers than resection. Three randomized trials that compared laparoscopic repair with open repair in perforated ulcers favored the use of laparoscopic technique.154-156 In one study, Siu and associates155 randomly assigned 121 patients without other ulcer complications to undergo one of the two repairs. Laparoscopic repair was quicker to perform and led to less postoperative pain, smaller analgesic requirement, and shorter hospitalization. The same researchers later reported the routine use of laparoscopic repair in a cohort of 172 patients with perforated ulcers, about 80% of which were duodenal ulcers (Fig. 53-3).157 Conversion to open surgery occurred in 37 patients (around 1 in 5) because of perforations larger than 10 mm, non-juxtapyloric gastric ulcer, or perforations that could not be identified at laparoscopy. Persistent leaks after patch repair occurred in 2 patients who required

laparotomy and gastrectomy. The overall mortality rate was 8.1%. Several series studied risk factors for mortality in patients undergoing surgery for perforated ulcers. Boey and asso­ ciates158 identified preoperative shock, major medical illnesses, and perforation for longer than 12 hours as important risk factors. Irvin159 identified two other risk factors—age older than 70 years and use of NSAIDs. Definitive surgery (vagotomy and gastrectomy) also led to a higher mortality in Irvin’s series. The deaths, however, occurred in older patients who had other, concomitant ulcer complications.

OBSTRUCTION Medical Management

Patients with obstructing peptic ulcers are often volume depleted. The loss of fluid, hydrogen ions, and chloride ions in the vomitus leads to hypochloremic, hypokalemic metabolic alkalosis. The patient should be volume resuscitated with normal saline followed by potassium replacement once urine output is adequate. In severely malnourished patients, parenteral nutrition should be considered. A nasojejunal tube can be often inserted at endoscopy to bypass the stenosis and then used for enteral feeding. If the patient also has a nasogastric tube, however, management of two tubes is difficult. Decompression of the stomach by a largebore, preferably Salem-sump nasogastric catheter is an integral part of the management. This serves the following purposes: relief of vomiting, monitoring of fluid loss, and decompression of the stomach so it can regain tone. A high-volume non–bile-stained aspirate distinguishes gastric outlet obstruction from a high small bowel obstruction. The use of an intravenous PPI effectively reduces gastric acid output, making fluid and electrolyte management easier. PPI therapy also initiates ulcer healing, ameliorates inflammatory edema, and assists in resolving obstruction, although randomized controlled outcome studies of this issue are lacking. Around half of patients improve during the period of nasogastric suctioning, volume resuscitation, rehydration, and acid suppression. Improvement is especially noticeable in patients with active ulceration and acute edema. Surgery is thus deferred until after an adequate trial of conservative management. Other factors that may influence the decision to proceed to surgery are chronicity, a history of previous ulcer complication, and the patient’s age and general medical condition. Many authorities argue for initial endoscopic dilation before surgery.

Endoscopic Management

Figure 53-3.  Laparoscopic view of a perforated duodenal ulcer (arrow) with fibrinous exudate on the adjacent peritoneum.

Endoscopic balloon dilation has been used successfully in patients with benign gastric outlet obstruction (Fig. 53-4). During endoscopic examination, the stenosis is traversed by means of a biliary-type guidewire with a flexible hydrophilic tip. A low-compliance balloon is then passed over the guidewire. The use of a balloon is preferred because its inflation produces an even radial force, which has a theoretical advantage over the longitudinal shearing force associated with the use of conventional dilators. The availability of through-the-scope balloons that can be passed via the small channel of an endoscope enables the dilation to be seen and monitored. The procedure is typically performed with fluoroscopic guidance. A regimen of gradual dilation over two or three sessions seems sensible. The largest diameter of stenosis at which symptoms occur is unclear. Many authorities recommend dilation to 15 mm, which is often associated with relief of symptoms. The presence of gastric atony also contributes to symptoms. The risk of

Chapter 53  Treatment of Peptic Ulcer Disease

A

B

perforation rises with the size of balloon. Almost all of the perforations in one series occurred after dilation with a 20-mm balloon.160 Several series describing the use of endoscopic balloon dilation in the treatment of benign gastric outlet obstruction have been published.160-162 They varied in case mix, methods of dilation, and duration of follow-up. Most reported immediate relief of obstruction in 78% to 100% of cases. Recurrences, however, are common. In a series of 41 patients with a strong ulcer diathesis, half of the patients experienced relapse with recurrent obstruction, hemorrhage, or active ulceration after a median follow-up of 39 months.160 In a later series involving a handful of patients in whom H. pylori status was better defined, balloon dilation followed by H. pylori eradication in those shown to be infected led to more sustained symptom relief.163 Results of these studies suggest that endoscopic balloon dilation produces better results in patients with acute inflammatory edema than those with chronic scarring and fibrosis.

Surgical Management

A variety of operations have been described for obstructing duodenal, pyloric channel, and prepyloric ulcers. They include truncal vagotomy together with either a drainage procedure (either gastrojejunostomy or pyloroplasty) or an antrectomy. In the unusual event of an obstructing prepyloric gastric ulcer, an antrectomy followed by a Billroth type I gastroduodenostomy is the procedure of choice. Surgical management principles for gastric outlet obstruction were established in an era when (1) this peptic complication was common, (2) recurrence was likely without definitive surgical therapy, (3) potent antisecretory therapy and endoscopic dilation techniques were not available, and (4) parenteral nutrition options were limited. Although patients often presented with substantial debility due to profound weight loss and electrolyte disturbances, expectant medical management of the outlet obstruction ran the risk of dangerously delaying surgery while the malnutrition progressed. Consequently, early operation often was advised. Today, gastric outlet obstruction from peptic ulcer disease is uncommon, ulcer recurrence is unlikely once H. pylori and NSAIDs have been eliminated, excellent antisecretory therapy can be offered, there are a number of endoscopic techniques for dilating stenosis, and total parenteral nutri-

Figure 53-4.  A through-the-scope dilation of an obstructed pylorus caused by an ulcer. The procedure was performed under fluoroscopic guidance. A dual-channel endoscope with a 3.7-mm therapeutic channel was used. A, The stricture was first traversed with a biliary-type guidewire (arrowhead). A through-the-scope balloon was passed over the guidewire across the stricture. A waist, representing the stricture (arrow), was observed (B) and was nearly abolished on balloon inflation (C).

C

tion is widely available (see Chapter 5). Therefore, an immediate decision regarding the need for surgery is generally not necessary for a patient who presents with gastric outlet obstruction. The problem can be managed with medical and endoscopic means in approximately 70% of cases, and only 30% eventually require one of the previously mentioned operations to bypass the gastric outlet obstruction.

KEY REFERENCES

Barkun A, Bardou M, Marshall JK. Nonvariceal Upper GI Bleeding Consensus Conference Group: Consensus recommendations for managing patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med 2003; 139:843-57. (Ref 113.) Chan FK, Hung LC, Suen BY, et al. Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis. N Engl J Med 2002; 347:2104-10. (Ref 78.) Chan FK, Wong VW, Suen BY, et al. Combination of a cyclo-oxygenase-2 inhibitor and a proton pump inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: A double-blind randomized trial. Lancet 2007; 369:1621-6. (Ref 85.) Kahi CJ, Jensen DM, Sung JJ, et al. Endoscopic therapy versus medical therapy for bleeding peptic ulcer with adherent clot: A meta-analysis. Gastroenterology 2005; 129:855-62. (Ref 122.) Kearney PM, Baigent C, Godwin J, et al. Do selective cyclo-oxygenase-2 inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomized trials. BMJ 2006; 332:1302-8. (Ref 90.) Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med 1994; 331:717-27. (Ref 118.) Lanas A, Garcia-Rodriguez LA, Arroyo MT, et al. Effect of antisecretory drugs and nitrates on the risk of ulcer bleeding associated with nonsteroidal anti-inflammatory drugs, antiplatelet agents, and anticoagulants. Am J Gastroenterol 2007; 102:507-15. (Ref 76.) Lau JY, Leung WK, Wu JC, et al. Omeprazole before endoscopy in patients with gastrointestinal bleeding. N Engl J Med 2007; 356:163140. (Ref 142.) Leontiadis GI, Sreedharan A, Dorward S, et al. Systematic reviews of the clinical effectiveness and cost-effectiveness of proton pump inhibitors in acute upper gastrointestinal bleeding. Health Technol Assess 2007; 11:1-164. (Ref 99.) McGettigan P, Henry D. Cardiovascular risk and inhibition of cyclooxygenase: A systematic review of the observational studies of selective and nonselective inhibitors of cyclooxygenase 2. JAMA 2006; 296:1633-44. (Ref 92.) Rockall TA, Logan RF, Devlin HB, Northfield TC. Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. Steering committee and members of the National Audit of Acute Upper Gastrointestinal Haemorrhage. BMJ 1995; 311:222-6. (Ref 110.)

885

886

Section VI  Stomach and Duodenum Saeed ZA, Winchester CB, Michaletz PA, et al. A scoring system to predict rebleeding after endoscopic therapy of nonvariceal upper gastrointestinal hemorrhage, with a comparison of heat probe and ethanol injection. Am J Gastroenterol 1993; 88:1842-9. (Ref 115.) Sung JJY, Tsoi KK, Lai LH, et al. Endoscopic clipping versus injection and thermocoagulation in the treatment of non-variceal upper gastrointestinal bleeding: A meta-analysis. Gut 2007; 56:1364-73. (Ref 133.) Sung JJY, Barkum A, Kuipers EJ, et al. Intravenous esomeprazole for prevention of recurrent peptic ulcer bleeding. Ann Int Med 2009; 150:455-64. (Ref 140.)

Vergara M, Calvet X, Gisbert JP. Epinephrine injection versus epinephrine injection and a second endoscopic method in high risk bleeding ulcers. Cochrane Database Syst Rev 2007; 2:CD005584. (Ref 135.) Vergara M, Catalan M, Gisbert JP, Calvet X. Meta-analysis: Role of Helicobacter pylori eradication in the prevention of peptic ulcer in NSAID users. Aliment Pharmacol Ther 2005; 21:1411-18. (Ref 96.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

54 Adenocarcinoma and Other Tumors of the Stomach Julian A. Abrams and Timothy C. Wang

CHAPTER OUTLINE Epidemiology  887 Etiology and Pathogenesis  888 Helicobacter pylori Infection  889 Dietary Factors  892 Cigarette Smoking  892 Alcohol  892 Obesity  892 Inherited Predisposition  892 Genetics  893 Premalignant Conditions  895 Chronic Atrophic Gastritis  895 Intestinal Metaplasia  895 Dysplasia  895 Gastric Polyps  896 Previous Gastrectomy  896 Peptic Ulcer Disease  896 Ménétrier’s Disease  897 Screening and Prevention  897 Screening and Surveillance  897 Prevention  897

Gastric tumors are defined as benign or malignant based on their potential to metastasize. Gastric adenocarcinoma makes up the majority of malignant gastric tumors, and will be referred to as gastric cancer in this chapter.

EPIDEMIOLOGY Gastric cancer remains the second leading cause of cancer mortality in the world,1 although the overall incidence is declining.2 The incidence of gastric cancer in Western countries has decreased dramatically over the past century.3 For example, gastric cancer mortality has decreased 86% since 1950 in the United States, and the incidence of gastric cancer has diminished four-fold since 1930 to approximately 8 cases per 100,000 people.4,5 As recently as 1930, gastric cancer was the leading cause of cancer mortality in the United States for men and the third leading cause for women.6 Gastric cancer is now the seventh leading cause of cancer mortality in the United States.5 It was estimated that in 2008, approximately 21,500 Americans would be diagnosed with gastric cancer and 10,880 would die of it.7 There is great geographic variation in gastric cancer incidence, with the highest incidence rates in the Far East (Fig. 54-1). Eastern Europe and Central and South America also have high incidence rates, and the lowest incidence

Clinical Features  898 Diagnosis  899 Endoscopy  899 Upper Gastrointestinal Series  900 Computed Tomography Gastrography  900 Serum Markers  900 Classification and Staging  900 Methods of Staging  901 Prognosis and Treatment  902 Surgical Therapy  902 Endoscopic Mucosal Resection and Submucosal Dissection  902 Chemotherapy  903 Chemoradiation  903 Intraperitoneal Chemotherapy  904 Unresectable Disease  904 Gastric Lymphoma  904 Gastric Carcinoid Tumors  904 Gastrointestinal Stromal Tumors  904 Miscellaneous Tumors  904

rates are observed in North America, North Africa, South Asia, and Australia.6 Although gastric cancer was common in industrialized countries in the past, the latest epidemiologic data indicate that more than 60% of new cases of gastric cancer are in developing countries, reflecting a more rapid decline in developed countries. In the United States, the median age of diagnosis is 71 years, with the highest proportion (28%) diagnosed between the ages of 75 and 84.8 In Japan, a country with a high incidence of gastric cancer, the mean age of diagnosis is roughly a decade earlier, perhaps reflecting lead-time bias due to widespread screening. The incidence of gastric cancer in men is approximately twice that in women (Table 54-1).3 The incidence of gastric cancer in blacks in the United States is nearly double that in whites.5 Native Americans and Hispanics also have a higher risk of development of gastric cancer than whites.9 Lower socioeconomic status is associated with a much higher incidence of gastric cancer.3 In the United States, the distribution of gastric cancer within the stomach is 39% in the proximal third, 17% in the middle third, 32% in the distal third, and 12% involving the entire stomach.10 In contrast to the pattern seen with noncardia tumors, the incidence rates of gastric cardia cancer are rising.2,11 Dietary, environmental, and genetic risk factors for gastric adenocarcinoma are listed in Table 54-2, some of which are or may be protective.

887

888

Section VI  Stomach and Duodenum

Figure 54-1.  Worldwide incidence of gastric cancer in men. (From Parkin DM. International variation. Oncogene 2004; 23:6239-40.)

10 % Blood urea nitrogen increase of >5 mg/dL Serum calcium 6 L

GALLSTONE PANCREATITIS (1982) Age >70 yr >18,000/mm3 >220 mg/dL >400 IU/L >250 IU/L >10% >2 mg/dL

Figure 58-1.  Acute interstitial pancreatitis. Contrast-enhanced computed tomography shows diffuse swelling of the pancreas (P) with peripancreatic inflammatory changes (arrows). The pancreas was well perfused without evidence of necrosis. G, gallbladder.

5 mEq/L >4 L

NA, not applicable.

with two of the following criteria12: (1) symptoms, such as epigastric pain, consistent with the disease; (2) a serum amylase or lipase greater than three times the upper limit of normal; or (3) radiologic imaging consistent with the diagnosis, usually using computed tomography (CT) or magnetic resonance imaging (MRI). Pancreatitis is classified as acute unless there are CT, MRI, or endoscopic retrograde cholangiopancreatography (ERCP) findings of chronic pancreatitis. Then pancreatitis is classified as chronic pancreatitis, and any episode of acute pancreatitis is considered an exacerbation of inflammation superimposed on chronic pancreatitis (see Chapter 59). Once the diagnosis is established, patients are classified as having mild or severe pancreatitis. Mild acute pancreatitis consists of interstitial (edematous) pancreatitis on imaging, minimal or no extrapancreatic organ dysfunction, and typically an uneventful recovery. Severe pancreatitis manifests as organ failure or local complications such as necrosis, abscess, or pseudocyst. The Atlanta criteria13 (Table 58-1) defines severity by the presence of organ failure or pancreatic necrosis on dynamic contrast–enhanced CT scan (Figs. 58-1 and 58-2). Other acceptable markers of severe pancreatitis include three or more of Ranson’s 11 criteria for non-gallstone pancreatitis (Table 58-2),14 and an Acute Physiology and Chronic Health Evaluation (APACHEII) score of greater than eight.15 It is important to use precise terms in describing the anatomic complications of acute pancreatitis. The ability to

G

Figure 58-2.  Acute pancreatic necrosis. Contrast-enhanced computed tomography demonstrates focal areas of decreased perfusion in the pancreatic parenchyma (arrows) with surrounding peripancreatic inflammation. The necrosis was estimated to involve less than 30% of the pancreas. G, gallbladder.

apply appropriate therapy depends on a clear understanding of these terms. An old term that should be used only sparingly is phlegmon. Although this term is often used by radiologists to describe an inflammatory mass, this term has carried different meaning to gastroenterologists, internists, and surgeons. Whereas patients with interstitial pancreatitis have a normally perfused gland, manifesting contrastenhanced CT as a normal bright appearance indicating flow

Chapter 58  Acute Pancreatitis throughout the gland, patients with necrotizing pancreatitis have greater than 30% of the gland not perfused, with low attenuation. Pancreatic necrosis consists of focal or diffuse nonviable pancreatic parenchyma and usually peripancreatic fat necrosis. Pancreatic necrosis can be sterile or infected. Peripancreatic necrosis describes necrotic fatty and stromal tissue around the pancreas. It is more important to surgeons because this is typically not appreciated on imaging. However, the presence of peripancreatic necrosis may delineate a more complicated course for patients with acute pancreatitis. An acute fluid collection is fluid located in or near the pancreas that lacks a definite wall and typically occurs early in the course of acute pancreatitis. On CT scan these collections appear as a low attenuation mass with poor margins and no capsule. It is very difficult to distinguish acute fluid collections in the pancreatic parenchyma from pancreatic necrosis. An acute fluid collection occurs in 30% to 50% of cases of acute pancreatitis and most resolve spontaneously.16 A pseudocyst is a fluid collection that persists for 4 to 6 weeks and becomes encapsulated by a wall of fibrous or granulation tissue. Pseudocysts are located adjacent to or off the body of the pancreas. At times these enzyme-rich fluid-filled sacks can be found distantly in the pelvis and chest. When a pseudocyst is located within the body of the pancreas, the cyst may contain necrotic pancreatic debris even when the pseudocyst is fluidappearing with low attenuation on CT. The term for a walled-off fluid-appearing pseudocyst-like structure involving the pancreas is walled off pancreatic necrosis (WOPN). A pancreatic abscess is a circumscribed intra-abdominal collection of pus occurring after an episode of acute pancreatitis or pancreatic trauma. It usually develops close to the pancreas and contains little pancreatic necrosis. Due to confusion of whether an abscess represents an infected pseudocyst or infected pancreatic necrosis, the term abscess should be used sparingly. Because of important differences in management, it is best to use the terms infected pseudocyst and infected necrosis. The term hemorrhagic pancreatitis should also be used with caution, and this term is not a synonym for necrotizing pancreatitis. Hemorrhage is more commonly associated with pseudoaneurysm, an erosion of peripancreatic blood vessels with hemoperitoneum. Unfortunately, hemorrhagic pancreatitis has more commonly been used to inappropriately describe necrotizing pancreatitis. Of all these terms, the most important distinction is that between pancreatic necrosis and pseudocyst. WOPN is pancreatic necrosis that has liquefied after five to six weeks.17 Similar to a pseudocyst, a wall develops. However, whereas a pseudocyst always contains fluid, pancreatic necrosis, even if walled off early, contains a significant amount of debris that only becomes liquefied after five to six weeks. No attempt should be made to drain WOPN early (less than four weeks) because the debris is typically thick, often with the consistency of rubber early in the course of the disease. After five to six weeks, WOPN can be treated similar to the fluid-filled pseudocyst and drained surgically, endoscopically, or percutaneously.

NATURAL HISTORY Acute pancreatitis appears to have two distinct stages. The first stage is related to the pathophysiology of the inflammatory cascade. This first phase usually lasts a week. During this phase, the severity of acute pancreatitis is related to extrapancreatic organ failure secondary to the patient’s sys-

temic inflammatory response elicited by acinar cell injury. Infectious complications are uncommon at this time. Fever, tachycardia, hypotension, respiratory distress, and leukocytosis are typically related to the systemic inflammatory response syndrome (SIRS). Multiple cytokines are involved, including platelet activating factor, tumor necrosis factor-α (TNF-α) and various interleukins (ILs) (see Chapter 2). During the first week the initial state of inflammation evolves dynamically with variable degrees of pancreatic and peripancreatic ischemia or edema to either resolution or to irreversible necrosis and liquefaction, or the development of fluid collections in and around the pancreas. The extent of the pancreatic and peripancreatic changes is usually proportional to the severity of extrapancreatic organ failure. However, organ failure may develop independent of pancreatic necrosis.17 Approximately 75% to 80%, of patients with acute pancreatitis have a resolution of the disease process (interstitial pancreatitis) and do not enter the second phase. However, in 25% of patients, a more protracted course develops, often related to the necrotizing process (necrotizing pancreatitis) lasting weeks to months. The mortality peak in the second phase is related to a combination of factors, including organ failure secondary to sterile necrosis, infected necrosis, or complications from surgical intervention.11,12,18-20 There are two peaks for mortality. Most studies in the United States and Europe reveal that about half the deaths occur within the first week or two, usually of multiorgan failure.19-21 Death can be very rapid. About one quarter of all deaths in Scotland occurred within 24 hours of admission and one third within 48 hours.21 After the second week of illness, patients succumb to pancreatic infection associated with multiorgan failure. Some studies in Europe report a very high late mortality rate from infection.22 Patients who are older and have comorbid illnesses have a substantially higher mortality rate than younger healthier patients. In those who survive their illness, severe pancreatic necrosis can scar the pancreas, resulting in a stricture of the main pancreatic duct with subsequent obstructive chronic pancreatitis and permanent diabetes and malabsorption.23

PATHOGENESIS The initial step in the pathogenesis of acute pancreatitis is conversion of trypsinogen to trypsin within acinar cells in sufficient quantities to overwhelm normal mechanisms to remove active trypsin (see Fig. 57-3). Trypsin, in turn, catalyzes conversion of proenzymes, including trypsinogen and inactive precursors of elastase, phospholipase A2 (PLA2), and carboxypeptidase, to active enzymes. Trypsin also may activate the complement and kinin systems. Active enzymes autodigest the pancreas and initiate a cycle of releasing more active enzymes. Normally small amounts of trypsinogen are spontaneously activated within the pancreas, but intrapancreatic mechanisms quickly remove activated trypsin. Pancreatic secretory trypsin inhibitor (PSTI, now called SPINK1) binds and inactivates about 20% of the trypsin activity. Other mechanisms for removing trypsin involve mesotrypsin, enzyme Y, and trypsin itself, which splits and inactivates trypsin. The pancreas also contains nonspecific antiproteases such as α1-antitrypsin and α2macroglobulin. Additional protective mechanisms are the sequestration of pancreatic enzymes within intracellular compartments of the acinar cell during synthesis and transport and the separation of digestive enzymes from lysosomal hydrolases as they pass through the Golgi apparatus,

961

962

Section VII  Pancreas which is important because cathepsin B activates trypsin from trypsinogen. Low intra-acinar calcium concentrations also prevent further autoactivation of trypsin. In experimental pancreatitis, activation of trypsin occurs within 10 minutes, and large amounts of trypsin24 and increased concentrations of trypsinogen activation peptide (TAP) accumulate within the pancreas.25,26 TAP is cleaved when trypsinogen is activated to trypsin, and concentrations of TAP in plasma, urine, and ascites correlate with the severity of the pancreatic inflammatory response, with the highest levels associated with acinar necrosis and intrapancreatic hemorrhage.27,28 Co-localization of pancreatic enzymes in lysosomes, followed by acinar cell injury, is an attractive hypothesis for the pathogenesis of acute pancreatitis, but the relevance of co-localization to the pathogenesis of acute pancreatitis is unclear. Activation of trypsinogen occurs before biochemical or morphologic injury to acinar cells, in association with co-localization of lysosomal enzymes, such as cathepsin B, and digestive enzymes, including trypsinogen within unstable vacuoles.27,28 Complete inhibition of pancreatic cathepsin B activity in vitro prevents trypsinogen activation induced by the cholecystokinin (CCK) analog cerulein,29 supporting the co-localization hypothesis. Thus, complete inhibition of cathepsin B may prevent or be a treatment for acute pancreatitis. However, enzyme co-localization may occur without inducing significant acinar cell injury.30 Two other features of experimental acute pancreatitis are early blockade of the secretion of pancreatic enzymes while enzyme synthesis continues and disruption of the paracellular barrier of acinar cells and intralobular pancreatic duct cells. The disruption facilitates the extravasation of pancreatic enzymes from acinar cells and from the duct lumen into interstitial spaces. This phenomenon may explain the rapid development of interstitial edema and the increase of pancreatic enzymes in the serum.31 As discussed in Chapter 57, the discovery of genetic mutations associated with hereditary pancreatitis also lends support to the hypothesis that intrapancreatic activation of pancreatic zymogens is central to the pathogenesis of acute pancreatitis.32-35 The mutant trypsin in hereditary pancreatitis (usually R122H or N29I mutation) causes trypsin to be resistant to lysis or causes premature trypsinogen activation (gain of function mutation) leading to autodigestion of the pancreas and episodes of acute pancreatitis.36,37 Mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene have also been implicated in pancreatitis (see Chapter 57). CFTR anion channel allows for chloride and bicarbonate secretion into the pancreatic ducts and thus allows flushing of the liberated enzymes and proenzymes into the duodenum. There are more than 1200 mutations that have been described for the CFTR gene. Some of these are considered severe and some mild. Homozygote severe mutations produce a viscid, concentrated, acidic pancreatic juice leading to ductal obstruction and pancreatic insufficiency in infanthood. Heterozygotes of minor or major mutations may lead to acute recurrent or chronic pancreatitis by altering acinar or ductal cell function (e.g., alteration of bicarbonate conductance). A third genetic abnormality associated with pancreatitis is mutation of the SPINK1 gene.38 As noted, SPINK1 protects the pancreatic acinar cell by inhibiting prematurely activated trypsin. Mutations of this gene presumably limit the activity of this protein, but the exact mechanism is unclear. The pathogenesis of gallstone-related pancreatitis is unknown (see Chapter 65). Factors that may initiate gallstone pancreatitis include reflux of bile into the pancreatic duct39,40 or obstruction of the pancreatic duct at the ampulla

from stone(s) or edema resulting from the passage of a stone.41 Reflux of bile into the pancreatic duct could occur when the distal bile and pancreatic ducts form a common channel and a gallstone becomes impacted in the duodenal papilla. Alternatively, bile could reflux into the pancreatic duct from the duodenum through an incompetent sphincter of Oddi injured by recent passage of a gallstone. Experimentally, reflux of bile into the pancreatic duct, particularly if infected or mixed with pancreatic enzymes, causes pancreatic injury. Mixtures of bile and pancreatic enzymes increase the permeability of the main pancreatic duct, which is associated with local parenchymal inflammation.42 The common channel theory is somewhat problematic because pancreatic duct pressure is invariably higher than bile duct pressure, making bile reflux into the pancreatic duct unlikely. Reflux of bile from the duodenum also is unlikely because pancreatitis does not occur in conditions with easily demonstrable reflux, such as after surgical sphincteroplasty or endoscopic sphincterotomy. A popular opinion for the mechanism of gallstone pancreatitis is that an impacted gallstone in the distal bile duct obstructs the pancreatic duct, which increases pancreatic pressure, thereby damaging ductal and acinar cells. Experiments in the opossum that support this theory are the observations that ligation of the pancreatic duct causes severe necrotizing pancreatitis41 and that decompression of the ductal system within three days prevents progression to acinar cell necrosis and severe inflammation.43

PATHOPHYSIOLOGY The pathophysiology of acute pancreatitis starts with local acinar injury that, if unchecked, leads to local inflammatory complications, a systemic response and sepsis. Pathophysiologic mechanisms include microcirculatory injury, leukocyte chemoattraction, release of pro- and anti-inflammatory cytokines, oxidative stress, leakage of pancreatic fluid into the region of the pancreas, and bacterial translocation to the pancreas and systemic circulation. The release of pancreatic enzymes damages the vascular endothelium, the interstitium, and acinar cells.43-45 Acinar injury leads to expression of endothelial adhesion molecules (e.g., VCAM-1), which further propagates the inflammatory response.46 Microcirculatory changes, including vasoconstriction, capillary stasis, decreased oxygen saturation, and progressive ischemia, occur early in experimental acute pancreatitis. These abnormalities increase vascular permeability and lead to edema of the gland (edematous or interstitial pancreatitis). Vascular injury could lead to local microcirculatory failure and amplification of the pancreatic injury. It is uncertain whether ischemia-reperfusion injury occurs in the pancreas.40 Reperfusion of damaged pancreatic tissue could lead to the release of free radicals and inflammatory cytokines into the circulation, which could cause further injury. In early stages of animal and human pancreatitis, activation of complement and the subsequent release of C5a play significant roles in the recruitment of macrophages and polymorphonuclear leukocytes.47-49 Active granulocytes and macrophages release proinflammatory cytokines in response to transcription factors such as nuclear factor κB (NF-κB). Proinflammatory cytokines include TNF, IL-1, IL-6, and IL-8, and platelet-activating factor (PAF). Proinflammatory cytokines frequently are followed by anti-inflammatory cytokines (IL-2, IL-10, IL-11) that attempt to down-regulate inflammation.48 Other mediators of inflammation include arachidonic acid metabolites

Chapter 58  Acute Pancreatitis (prostaglandins, PAF, and leukotrienes), nitric oxide, proteolytic and lipolytic enzymes, and reactive oxygen metabolites that overwhelm scavenging by endogenous antioxidant systems. These substances also interact with the pancreatic microcirculation to increase vascular permeability, which induces thrombosis and hemorrhage and leads to pancreatic necrosis. A recent study suggests that gene polymorphisms that affect acinar cell glutathione concentrations may lead to increased oxidant stress and more severe pancreatitis.50 Meanwhile, ischemia and severe inflammation of the gland can lead to disruption of the main and secondary pancreatic ducts, leading to local fluid accumulations within and surrounding the pancreas that can eventuate into pseudocysts.51,52 Some patients with severe pancreatic damage develop systemic complications, including fever, acute respiratory distress syndrome (ARDS), pleural effusions, renal failure, shock, myocardial depression, and metabolic complications. SIRS is common in patients with acute pancreatitis and is probably mediated by activated pancreatic enzymes (phospholipase, elastase, trypsin) and cytokines (TNF, PAF) released into the portal circulation from the inflamed pancreas.53 Cytokines reaching the liver activate hepatic Kupffer cells, which, in turn, induces hepatic expression and secretion of cytokines into the systemic circulation. These cause acute phase protein synthesis (C-reactive protein [CRP], IL-6) and may cause SIRS and damage to the kidneys, lungs, and other organs leading to multiorgan dysfunction and failure.54 ARDS may be induced by active phospholipase A (lecithinase), which digests lecithin, a major component of lung surfactant. Acute renal failure has been explained on the basis of hypovolemia and hypotension. Myocardial depression and shock are likely secondary to vasoactive peptides and a myocardial depressant factor. Metabolic complications include hypocalcemia, hyperlipidemia, hyperglycemia with or without ketoacidosis, and hypoglycemia. The pathogenesis of hypocalcemia is multifactorial and includes hypoalbuminemia (the most important cause), hypomagnesemia, calcium-soap formation, hormonal imbalances (e.g., involving parathyroid hormone, calcitonin, and glucagon), binding of calcium by free fatty acid–albumin complexes, intracellular translocation of calcium, and systemic exposure to endotoxin.55 Pancreatic infection (infected necrosis and infected pseudocyst) can occur from the hematogenous route or from translocation of bacteria from the colon into the lymphatics. Under normal circumstances bacterial translocation does not occur because there are complex immunologic and morphologic barriers. However, during acute pancreatitis, these barriers break down, which can result in local and systemic infection.56 Penetration of the gut barrier by enteric bacteria is likely due to gut ischemia secondary to hypovolemia and pancreatitis-induced arteriovenous shunting in the gut.57 Indeed, in canine experimental pancreatitis, luminal Escherichia coli translocate to mesenteric lymph nodes and distant sites.58 In feline experimental pancreatitis, enclosing the colon in impermeable bags prevents translocation of bacteria from the colon to the pancreas.59

PREDISPOSING CONDITIONS Many conditions predispose to acute pancreatitis to varying degrees (Table 58-3). This list will undoubtedly continue to grow, and the number of cases diagnosed as “idiopathic” will decrease as our understanding of the disease improves.

Table 58-3  Conditions That Predispose to Acute Pancreatitis Obstructive Gallstones Tumors Parasites Duodenal diverticula Annular pancreas Choledochocele Other Alcohol/other toxins/drugs Ethyl alcohol Scorpion venom Methyl alcohol Organophosphorous insecticides Drugs (see Table 58-4) Metabolic Hypertriglyceridemia Hypercalcemia Infectious Vascular Vasculitis Emboli to pancreatic blood vessels Hypotension Trauma Post-endoscopic retrograde cholangiopancreatography (ERCP) Postoperative Hereditary/familial/genetic Controversial Pancreas divisum Sphincter of Oddi dysfunction Miscellaneous Idiopathic

Gallstones and chronic alcohol abuse account for 70% of acute pancreatitis in the United States.

OBSTRUCTION Gallstones

The most common obstructive process leading to pancreatitis is gallstones, which cause approximately 40% of cases acute pancreatitis,60 although only 3% to 7% of patients with gallstones develop pancreatitis. Gallstone pancreatitis is more common in women than men because gallstones are more frequent in women.61 Acute pancreatitis occurs more frequently when stones are less than 5  mm in diameter (odds ratio, 4 to 5),62 because small stones are more likely than large stones to pass through the cystic duct and cause ampullary obstruction. Cholecystectomy and clearing the bile duct of stones prevents recurrence, confirming the cause-and-effect relationship.61

Biliary Sludge and Microlithiasis

Biliary sludge is a viscous suspension in gallbladder bile that may contain small ( 40 mm Hg) is the most common abnormality discovered, occurring in approximately 35% to 40% of patients. The argument in favor of this entity as a cause of acute pancreatitis is the many observational series that report that endoscopic pancreatic sphincterotomy or surgical sphincteroplasty reduces or eliminates recurrent attacks of pancreatitis.174 The arguments against SOD as a cause of acute pancreatitis are (1) the lack of any prospective controlled blinded trials in the treatment of this disorder; (2) the short duration of follow-up in the observational reports; and (3) the high risk of pancreatitis (25% to 35%) associated with ERCP, sphincter of Oddi manometry, and pancreatic sphincterotomy in patients with suspected SOD. Furthermore, there is a relative dearth of data determining the normal range of pancreatic sphincter pressure.174

MISCELLANEOUS

Pancreatitis has been rarely associated with Crohn’s disease.164 A recent case control study from Denmark found a 4-fold increase in acute pancreatitis in patients with Crohn’s and a 1.5-fold increase in patients with ulcerative colitis. This has been attributed by some to the use of drugs such as aminosalicylates/sulfasalazine, azathioprine, or 6-mercaptopurine (see Table 58-4). Theories to support a relationship between idiopathic inflammatory bowel disease (IBD) and pancreatitis are that pancreatitis is an extraintestinal manifestation of IBD, that duodenal Crohn’s disease can cause obstruction to the flow of pancreatic juice, that granulomatous disease can involve the pancreas, or that there is an autoimmune process affecting the pancreas. Celiac disease165 has also been described in association with pancreatitis, but the relationship remains uncertain. It has been suggested that abnormalities in the normal barrier of the small bowel seen in patients with celiac disease may allow excessive absorption of amylase from the intestinal lumen, leading to hyperamylasemia. In the setting of abdominal pain in a patient with celiac disease, it is not uncommon to find elevations in the serum amylase in the absence of acute pancreatitis.166 Pancreatitis has been seen in patients after severe burns.167 A relationship of smoking with acute pancreatitis has been suggested. A Swedish case control study showed that there is a fourfold increased rate of acute pancreatitis in heavy smokers compared with nonsmokers.168 Autoimmune pancreatitis (discussed in the next chapter in more detail) typically presents as a mass or fullness in the pancreas. It is most commonly seen in older men, with biliary obstruction and an elevation of the serum immunoglobulin (IgG4) level. Occasionally, patients will present with signs and symptoms of chronic pancreatitis, such as stricturing of the main pancreatic duct, diabetes, exocrine pancreatic insufficiency, or as acute pancreatitis. Investigators have more recently described patients with autoimmune recurrent pancreatitis, especially in younger women often without the classic elevation of serum IgG4.169

CLINICAL FEATURES It is difficult to diagnose acute pancreatitis by history and physical examination because clinical features are similar to those of many acute abdominal illnesses (Table 58-6).

Table 58-6  Differential Diagnosis of Acute Pancreatitis Biliary colic Acute cholecystitis Perforated hollow viscus (e.g., perforated peptic ulcer) Mesenteric ischemia or infarction Intestinal obstruction Inferior wall myocardial infarction Dissecting aortic aneurysm Ectopic pregnancy

HISTORY

Abdominal pain is present at the onset of most attacks of acute pancreatitis, but the timing of abdominal pain is variable. Biliary colic may herald or progress to acute pancreatitis. Pain in pancreatitis usually involves the entire upper abdomen. However, it may be epigastric, in the right upper quadrant, or, infrequently, confined to the left side. Pain in the lower abdomen may arise from the rapid spread of pancreatic exudation to the left colon. Onset of pain is rapid but not as abrupt as that of a perforated viscus. Usually it is at maximal intensity in 10 to 20 minutes. Occasionally, pain gradually increases and takes several hours to reach maximum intensity. Pain is steady and moderate to very severe. There is little pain relief with changing position. Frequently, pain is unbearable, steady, and boring. Band-like radiation of the pain to the back occurs in half of patients. Pain that lasts only a few hours and then disappears suggests a disease other than pancreatitis, such as biliary colic or peptic ulcer. Pain is absent in 5% to 10% of attacks, and a painless presentation may be a feature of serious fatal disease.4 Ninety percent of patients have nausea and vomiting. Vomiting may be severe, may last for hours, may be accompanied by retching, and may not alleviate pain. Vomiting may be related to severe pain or to inflammation involving the posterior gastric wall.

PHYSICAL EXAMINATION

Physical findings vary with the severity of an attack. Patients with mild pancreatitis may not appear acutely ill. Abdominal tenderness may be mild, and abdominal guarding is absent. In severe pancreatitis, patients look severely ill and often have abdominal distention, especially epigastric, which is due to gastric, small bowel, or colonic ileus. Almost all patients are tender in the upper abdomen, which may be elicited by gently shaking the abdomen or by gentle percussion. Guarding is more marked in the upper abdomen. Tenderness and guarding are less than expected, considering the intensity of discomfort. Abdominal rigidity, as occurs in diffuse peritonitis, is unusual but can be present, and differentiation from a perforated viscus may be impossible in these instances. Bowel sounds are reduced and may be absent. Additional abdominal findings may include ecchymosis in one or both flanks (Grey Turner’s sign; Fig. 58-3) or about the periumbilical area (Cullen’s sign), owing to extravasation of hemorrhagic pancreatic exudate to these areas. These signs occur in less than 1% of cases and are associated with a poor prognosis. Rarely there is a brawny erythema of the flanks caused by extravasation of pancreatic exudate to the abdominal wall. A palpable epigastric mass may appear during the disease from a pseudocyst or a large inflammatory mass. The general physical examination, particularly in severe pancreatitis, may uncover markedly abnormal vital signs if

969

970

Section VII  Pancreas 65). The pain of a perforated ulcer is sudden, becomes diffuse, and precipitates a rigid abdomen; movement aggravates pain. Nausea and vomiting occur but disappear soon after onset of pain (see Chapter 52). In mesenteric ischemia or infarction, the clinical setting often is an older person with cardiac dysrhythmia or arteriosclerotic disease who develops sudden pain out of proportion to physical findings, bloody diarrhea, nausea, and vomiting. Abdominal tenderness may be mild to moderate, and muscular rigidity may not be severe despite severe pain (see Chapter 114). In intestinal obstruction, pain is cyclical, abdominal distention is prominent, vomiting persists and may become feculent, and peristalsis is hyperactive and often audible (see Chapter 119). Other conditions that enter into the differential diagnosis of acute pancreatitis are listed in Table 58-6. Figure 58-3.  Grey Turner’s sign. Ecchymoses in the flanks in a 24-year-old man with a two-day history of upper abdominal pain secondary to mild acute alcoholic pancreatitis. (Courtesy of Nadia Habal, MD, Dallas, Tex.)

LABORATORY DIAGNOSIS PANCREATIC ENZYMES

there are third-space fluid losses and systemic toxicity. Commonly, the pulse is 100 to 150 beats/minute. Blood pressure can be initially higher than normal and then lower than normal with third-space losses and hypovolemia. Initially the temperature may be normal, but within one to three days it may increase to 101°F to 103°F owing to the severe retroperitoneal inflammatory process and the release of inflammatory mediators from the pancreas.175 Tachypnea and shallow respirations may be present if the subdiaphragmatic inflammatory exudate causes painful breathing. Dyspnea may accompany pleural effusions, atelectasis, ARDS, or congestive heart failure. Chest examination may reveal limited diaphragmatic excursion if abdominal pain causes splinting of the diaphragm, or dullness to percussion and decreased breath sounds at the lung bases if there is a pleural effusion. There may be disorientation, hallucinations, agitation, or coma,176 which may be due to alcohol withdrawal, hypotension, electrolyte imbalance such as hyponatremia, hypoxemia, fever, or toxic effects of pancreatic enzymes on the central nervous system. Conjunctival icterus may be present due to choledocho­ lithiasis (gallstone pancreatitis) or bile duct obstruction from edema of the head of the pancreas, or from coexistent liver disease. Uncommon findings include subcutaneous nodular fat necrosis,177 thrombophlebitis in the legs, and polyarthritis. Subcutaneous fat necroses are 0.5- to 2-cm tender red nodules that usually appear over the distal extremities but may occur over the scalp, trunk, or buttocks. They occasionally precede abdominal pain or occur without abdominal pain, but usually they appear during a clinical episode and disappear with clinical improvement. If they occur over a joint, they may be confused with arthritis. Some physical findings point to a specific cause of acute pancreatitis. Hepatomegaly, spider angiomas, and thickening of palmar sheaths favor alcoholic pancreatitis. Eruptive xanthomas and lipemia retinalis suggest hyperlipidemic pancreatitis. Parotid pain and swelling are features of mumps. Band keratopathy (an infiltration on the lateral margin of the cornea) occurs with hypercalcemia.

DIFFERENTIAL DIAGNOSIS

The abdominal pain of biliary colic may simulate acute pancreatitis. It is frequently severe and epigastric, but it lasts for several hours rather than several days (see Chapter

In general, the diagnosis of acute pancreatitis relies on at least a three-fold elevation of amylase or lipase in the blood.178

Serum Amylase

In healthy persons, the pancreas accounts for 40% to 45% of serum amylase activity, and the salivary glands account for the rest. Simple analytic techniques can separate pancreatic and salivary amylases. Because pancreatic diseases increase serum pancreatic (P) isoamylase, measurement of P-isoamylase can improve diagnostic accuracy. However, this test is rarely used. The total serum amylase test is most frequently ordered to diagnose acute pancreatitis because it can be measured quickly and cheaply. It rises within 6 to 12 hours of onset and is cleared fairly rapidly from the blood (half-life, 10 hours). Probably less than 25% of serum amylase is removed by the kidneys. It is uncertain what other processes clear amylase from the circulation. The serum amylase is usually increased on the first day of symptoms, and it remains elevated for three to five days in uncomplicated attacks. The sensitivity of the serum amylase level for detecting acute pancreatitis is difficult to assess because an elevated amylase is often used to make the diagnosis. In mild attacks, other tests to confirm pancreatic inflammation are either not sensitive enough (pancreatic imaging studies, other biochemical markers) or not necessary (surgery). A limitation of serum amylase is that it is not 100% sensitive or specific. With respect to sensitivity, which is greater than 85%, the serum amylase may be normal or minimally elevated in fatal pancreatitis,4 during a mild attack or an attack superimposed on chronic pancreatitis (because the pancreas has little acinar tissue), or during recovery from acute pancreatitis. Serum amylase also may be falsely normal in hypertriglyceridemia-associated pancreatitis179 because an amylase inhibitor may be associated with triglyceride elevations. In this case, serial dilution of serum often reveals an elevated serum amylase. Hyperamylasemia is not specific for pancreatitis because it occurs in many conditions other than acute pancreatitis. In fact, one half of all patients with an elevated serum amylase may not have pancreatic disease.180 In acute pancreatitis, the serum amylase concentration is usually more than two to three times the upper limit of normal; it is usually less than this with other causes of hyperamylasemia.177 However, this level is not an absolute discriminator.

Chapter 58  Acute Pancreatitis Thus, an increased serum amylase level supports rather than confirms the diagnosis of acute pancreatitis. In addition, there are some individuals who have persistent hyperamylasemia without clinical symptoms. This has been reported due to macroamylasemia or pancreatic hyperamylasemia on a familial basis.174 Nonpancreatic diseases that cause hyperamylasemia include pathologic processes in organs (e.g., salivary glands, fallopian tubes) that normally produce amylase. Furthermore, mass lesions such as papillary cystadenocarcinoma of the ovary, benign ovarian cyst, and carcinoma of the lung, cause hyperamylasemia because they produce and secrete salivary-type isoamylase. Transmural leakage of pancreatic-type isoamylase and peritoneal absorption probably explain hyperamylasemia in intestinal infarction and in perforated viscus. Renal failure increases serum amylase up to four to five times the upper limit of normal due to decreased renal clearance of this enzyme.181 Patients on hemodialysis tend to have higher serum amylase levels than those on peritoneal dialysis. In patients with chronic kidney disease, there is no clear correlation between the creatinine clearance and serum levels of amylase, and about one third of patients with marked renal insufficiency have normal pancreatic enzyme levels. Chronic elevations of serum amylase (without amylasuria) occur in macroamylasemia. In this condition, normal serum amylase is bound to an immunoglobulin or abnormal serum protein to form a complex that is too large to be filtered by renal glomeruli and thus has a prolonged serum half-life.182 Macroamylasemia may complicate the diagnosis of pancreatic disease, but it has no other clinical consequence. The urinary amylase-to-creatinine clearance ratio (ACCR) increases from approximately 3% to approximately 10% in acute pancreatitis.183 However, even moderate renal insufficiency interferes with the accuracy and specificity of the ACCR. Other than to diagnose macroamylasemia, which has a low ACCR, urinary amylase and the ACCR are not used clinically. Macroamylasemia can also be measured directly using serum. Deliberate contamination of urine with saliva, as in Munchausen’s syndrome, can increase the urine amylase, with the serum amylase being normal. This situation can be excluded by measuring salivary amylase in the urine.

Serum Lipase

The sensitivity of serum lipase for the diagnosis of acute pancreatitis is similar to that of serum amylase and is between 85% and 100%.174 Lipase may have greater specificity for pancreatitis than amylase as serum lipase is normal when serum amylase is elevated as in salivary gland dysfunction, tumors, gynecologic conditions, and macroamylasemia. Serum lipase always is elevated on the first day of illness and remains elevated longer than does the serum amylase.184 Consequently some suggest combining lipase with amylase as a test for acute pancreatitis. However, we and others have found that combining enzymes does not improve diagnostic accuracy. Specificity of lipase can suffer from some of the same problems as those of amylase. In the absence of pancreatitis, serum lipase may increase less than two-fold above normal in severe renal insufficiency.185 With intra-abdominal conditions that resemble acute pancreatitis,186 lipase increases to levels less than three-fold above normal, presumably by absorption through an inflamed or perforated intestine. Rarely, a nonpancreatic abdominal condition such as small bowel obstruction can raise the amylase and lipase above three times normal. Some believe that serum lipase measurement is preferable to that of serum amylase because it is as sensitive as amylase measurement and more specific,

whereas others find no clear advantage of one over the other.5

Other Pancreatic Enzymes

During acute pancreatic inflammation, pancreatic digestive enzymes other than amylase and lipase leak into the systemic circulation and have been used to diagnose acute pancreatitis. They include PLA2, trypsin/typsinogen, car­ boxylester lipase, carboxypeptidase A, colipase, elastase, and ribonuclease. None, alone or in combination, are better than serum amylase or lipase, and most are not available on a routine basis.

OTHER BLOOD AND URINE TESTS

Many nonenzymatic proteins are overexpressed in acute pancreatitis. Pancreatitis-associated protein (PAP), a heat shock protein, is undetectable in the normal pancreas but markedly increases in acute pancreatitis, with PAP detectable in serum. The sensitivity of serum PAP or pancreaticspecific protein (PSP) is no better than that of conventional tests,187 but PAP and PSP are as accurate as serum amylase for the detection of acute pancreatitis. The methemalbumin level in the circulation increases in acute pancreatitis, but it also increases in serious intraabdominal conditions such as intestinal infarction, limiting its usefulness.

STANDARD BLOOD TESTS

The white blood cell count frequently is elevated, often markedly so in severe pancreatitis. The blood glucose also may be high and associated with high levels of serum glucagon. Serum aspartate transaminase (AST), alanine aminotransferase (ALT), alkaline phosphatase, and bilirubin also may increase, particularly in gallstone pancreatitis. Presumably, calculi in the bile duct account for these abnormalities. However, pancreatic inflammation may partially obstruct the distal bile duct in acute pancreatitis of other causes and cause abnormalities in liver biochemical tests. Nevertheless, aminotransferases may help distinguish biliary from alcoholic pancreatitis (see later).188 The erythrocyte mean corpuscular volume (MCV) has been also shown to help differentiate alcoholic from nonalcoholic acute pancreatitis.189 Alcoholic patients tend to have a higher MCV due to the toxic effects of alcohol on erythrocyte formation in the bone marrow. Serum triglyceride levels increase in acute pancreatitis, but also with alcohol use, uncontrolled diabetes mellitus, or defective tri­glyceride metabolism.

DIAGNOSTIC IMAGING ABDOMINAL PLAIN FILM

Findings on a plain radiograph range from no abnormalities in mild disease to localized ileus of a segment of small intestine (“sentinel loop”) or the colon cut-off sign in more severe disease. In addition, an abdominal plain film helps exclude other causes of abdominal pain, such as bowel obstruction and perforation. Images of the hollow GI tract on an abdominal plain radiograph depend on the spread and location of pancreatic exudate. Gastric abnormalities are caused by exudate in the lesser sac producing anterior displacement of the stomach, with separation of the contour of the stomach from the transverse colon. Small intestinal abnormalities are due to exudate in proximity to small bowel mesentery and include ileus of one or more loops of jejunum (the sentinel loop), of the distal ileum or cecum,

971

972

Section VII  Pancreas or of the duodenum. Generalized ileus may occur in severe disease. Besides ileus, other abnormalities of the hollow GI tract may be present. The descending duodenum may be displaced and stretched by an enlarged head of the pancreas. In addition, spread of exudate to specific areas of the colon may produce spasm of that part of the colon and either no air distal to the spasm (the colon cut-off sign) or dilated colon proximal to the spasm. Head-predominant pancreatitis predisposes to spread of exudate to the proximal transverse colon, producing colonic spasm and a dilated ascending colon. Uniform pancreatic inflammation predisposes spread of exudate to the inferior border of the transverse colon and an irregular haustral pattern. Exudate from the pancreatic tail to the phrenicocolic ligament adjacent to the descending colon may cause spasm of the descending colon and a dilated transverse colon. Other findings on plain radiography of the abdomen may give clues to etiology or severity, including calcified gallstones (gallstone pancreatitis), pancreatic stones or calcification (chronic pancreatitis with a bout of acute inflammation), and ascites (severe pancreatitis). Gas in the retroperitoneum may suggest a pancreatic abscess.

CHEST RADIOGRAPHY

Abnormalities visible on the chest roentgenogram occur in 30% of patients with acute pancreatitis, including elevation of a hemidiaphragm, pleural effusion(s), basal or plate-like atelectasis secondary to limited respiratory excursion, and pulmonary infiltrates. Pleural effusions may be bilateral or confined to the left side; rarely they are only on the right side.190 Patients with acute pancreatitis found to have a pleural effusion and/or infiltrate on admission are more likely to have severe disease.191 During the first 7 to 10 days, there also may be signs of congestive heart failure or acute respiratory distress syndrome. Pericardial effusion is rare.

ABDOMINAL ULTRASONOGRAPHY

Abdominal ultrasonography is used during the first 24 hours of hospitalization to search for gallstones, dilation of the bile duct due to choledocholithiasis, and ascites. If the pancreas is seen (bowel gas obscures the pancreas 25% to 35% of the time), it is usually diffusely enlarged and hypoechoic. Less frequently there are focal hypoechoic areas. There also may be ultrasonographic evidence of chronic pancreatitis, such as intraductal or parenchymal calcification and dilation of the pancreatic duct. Ultrasound is not a good imaging test to evaluate extrapancreatic spread of pancreatic inflammation or necrosis within the pancreas and consequently is not useful to ascertain severity of pancreatitis. During the course of acute pancreatitis, ultrasound can be used to evaluate progression of a pseudocyst (discussed later). Due to overlying gas, the diagnosis of cholelithiasis may be obscured during the acute attack but may be found after bowel gas has receded.

ENDOSCOPIC ULTRASONOGRAPHY

Usually endoscopic ultrasonography (EUS) is not helpful early in acute pancreatitis. Imaging of the pancreas during an attack of acute pancreatitis and weeks following an episode reveal signals that are not normal (typically hypoechoic) and indistinguishable from chronic pancreatitis and malignancy. However, after a month, especially in patients with idiopathic interstitial pancreatitis, EUS may help determine the presence of small tumors, pancreas divisum, and bile duct stones.192 EUS is equal to MRCP and ERCP but far more sensitive than either abdominal ultraso-

nography or CT in detecting common duct stones.193 In a patient with biliary pancreatitis, whose serum bilirubin is rising in the setting of biliary sepsis, ERCP should not be delayed by first performing EUS (see later). Although there has been some concern that ERCP can worsen pancreatitis in such settings, ERCP appears to be safe in acute pancreatitis if needed. One caveat is that the contrast instillation into the pancreatic duct could introduce infection into necrotic areas of the pancreas. For this reason, EUS might be the best method of evaluating the bile duct in a patient with necrotizing pancreatitis.194

COMPUTED TOMOGRAPHY

CT is the most important imaging test for the diagnosis of acute pancreatitis and its intra-abdominal complications.195 The three main indications for a CT in acute pancreatitis are to exclude other serious intra-abdominal conditions, such as mesenteric infarction or a perforated ulcer; to stage the severity of acute pancreatitis; and to determine whether complications of pancreatitis are present, such as involvement of the GI tract or nearby blood vessels and organs, including liver, spleen, and kidney.196 Helical CT is the most common technique. If possible, scanning should occur after the patient receives oral contrast, followed by intravenous contrast to identify any areas of pancreatic necrosis. If there is normal perfusion of the pancreas, interstitial pancreatitis is said to be present (see Fig. 58-1). Pancreatic necrosis manifested as perfusion defects after intravenous contrast may not appear until 48 to 72 hours after onset of acute pancreatitis (see Fig. 58-2). Contraindications to using intravenous contrast are a patient’s history of severe allergy (respiratory distress or anaphylaxis) or significant renal impairment (serum creatinine greater than 2  mg/dL). If severe renal impairment requires dialysis, intravenous contrast medium may be used.197 Hives or less severe allergic reactions with previous administration of iodinated contrast material are not absolute contraindications, but a nonionic contrast agent should be used, and 200 mg of hydrocortisone should be administered intravenously every six hours for four doses starting before the scan and 50 mg of diphenhydramine (Benadryl) should be given intramuscularly 30 minutes before the scan.198 It has been suggested that intravenous contrast media early in the course of acute pancreatitis might increase pancreatic necrosis because iodinated contrast medium given at the onset of pancreatitis increases necrosis in experimental rat acute pancreatitis.199 However, it did not do so in the opossum.200 Data in humans are conflicting. Two retrospective studies suggested that early contrast-enhanced CT worsened pancreatitis199 but this was not corroborated by a third retrospective study.198 The severity of acute pancreatitis has been classified into five grades (A to E) based on findings on unenhanced CT199 (see following). Although the presence of gas in the pancreas suggests pancreatic infection with a gas-forming organism, this finding can also accompany sterile necrosis (Fig. 58-4) with microperforation of the gut or adjacent pseudocyst into the pancreas.201 Moreover, the great majority of pancreatic infections occur in the absence of gas on CT scan.

MAGNETIC RESONANCE IMAGING

MRI provides similar information regarding the severity of pancreatitis as does CT. MRI is as good as CT in detecting necrosis and fluid collections. MRI is better than CT, but equal to EUS and ERCP in detecting choledocholithiasis202 The MRCP contrast agent gadolinium, previously thought

Chapter 58  Acute Pancreatitis

G P

Figure 58-4.  Acute necrotizing pancreatitis. Contrast-enhanced computed tomography that shows the pancreas (P) is surrounded by peripancreatic inflammation that contains bubbles of air (arrows) due to sterile necrosis. The patient was not clinically ill, and therefore an abscess was not considered likely. G, gallbladder.

Conventional transabdominal ultrasonography should be performed in every patient with a first attack of acute pancreatitis to search for gallstones in the gallbladder, common duct stones, or signs of extrahepatic biliary tract obstruction. However, bile duct stones are frequently missed by transabdominal ultrasonography, and most stones pass during the acute attack. ERCP is limited to patients with severe acute pancreatitis due to gallstones with persistent bile duct obstruction and to those in whom the stone could not be removed during surgery. In most patients with biliary pancreatitis the bile duct can be imaged with an operative cholangiogram at the time of laparoscopic cholecystectomy performed during the same admission. Although EUS is the most accurate method of detecting common duct stones and has been recommended for evaluating the common duct prior to cholecystectomy, it is rarely needed in this setting. MRCP is another noninvasive test that is highly accurate in determining if common duct stones are present. If a common duct stone is found at surgery, it is either removed at operation or endoscopically after surgery. Laparoscopic exploration of the bile duct is as safe and effective as postoperative ERCP in clearing stones from the common duct.209

PREDICTORS OF SEVERITY to be safe in patients with renal failure,203 can cause nephrogenic systemic fibrosis (NSF), which has raised concern.204 MRI is less accessible and more expensive than CT. MRI also requires the patient to remain still during capture of images, which typically is longer than with spiral CT. The use of intravenous secretin prior to MRCP allows a better visualization of the pancreatic ducts. This has been shown to be particularly useful in the evaluation of patients with idiopathic pancreatitis and recurrent pancreatitis.205

DISTINGUISHING ALCOHOLIC FROM GALLSTONE PANCREATITIS Differentiation between alcoholic and gallstone pancreatitis is important because eliminating these causes may prevent further attacks of pancreatitis. Alcoholic pancreatitis occurs more frequently in men approximately 40 years old. The first clinical episode usually occurs after 5 to 10 years of heavy alcohol consumption. By contrast, biliary pancreatitis is more frequent in women, and the first clinical episode is often after the age of 40 years. Recurrent attacks of acute pancreatitis suggest an alcoholic etiology, but unrecognized gallstones may cause recurrent pancreatitis. Among patients with acute biliary pancreatitis discharged from hospital without cholecystectomy, 30% to 50% develop recurrent acute pancreatitis relatively soon after discharge (average time to recurrent pancreatitis, 108 days).206 Thus, removing the gallbladder in biliary pancreatitis is imperative. Laboratory tests may help distinguish between these two disorders. The specificity for gallstone pancreatitis of ALT concentration greater than 150 IU/L (approximately a threefold elevation) is 96%; the positive predictive value is 95%, but the sensitivity is only 48%.205 The AST concentration is nearly as useful as the ALT, but the total bilirubin and alkaline phosphatase concentrations are not as helpful to distinguish gallstone pancreatitis from alcoholic and other etiologies. There are differing reports as to whether a high serum lipase-to-amylase ratio can differentiate alcoholic from other causes of pancreatitis.207,208

Predicting severity of pancreatitis early in the course of disease is critical to maximize therapy and to prevent and minimize organ dysfunction and complications. Unfortunately the management of patients with acute pancreatitis is complicated by the inability to distinguish mild from severe disease during the early stages. The definition of the severity of acute pancreatitis early in the course of disease (during the first week) is typically based on clinical rather than anatomic parameters. At admission, several potential risk factors of severity and measurements that may reflect severity should be documented including age, body mass index, elevated hematocrit, elevated blood urea nitrogen (BUN), and pleural effusions or infiltrates on admission chest radiograph. The height of elevation of the serum amylase and lipase does not correlate with severity. Obese patients with pancreatitis have a higher incidence of local complications,210 respiratory failure,211 severe acute pancreatitis,212 and death from sterile necrosis213 than do nonobese patients. Initially at presentation and over the first 48 hours, patients should be classified temporarily as having severe acute pancreatitis (and managed as such initially) based on the presence of SIRS or organ failure. SIRS is defined by two or more of the following four criteria: pulse greater than 90 beats/minute; rectal temperature less than 36°C or more than 38°C; white blood count less than 4000 or more than 12,000/mm3; and respirations greater than 20/minute or Pco2 less than 32 mm Hg. The presence of SIRS at admission and persistence of SIRS to 48 hours increases the morbidity and mortality rate. In one study, 25% of patients with persistent SIRS died from acute pancreatitis, 8% with transient SIRS, and less than 1% without SIRS.214 Although severity is now defined by the presence of organ failure or anatomic complications of acute pancreatitis, such as pancreatic necrosis, prospective systems using clinical criteria have been developed to determine severity in patients with acute pancreatitis. These systems include Ranson criteria (see Table 58-2) and APACHE score.13,14 Unfortunately these scoring systems (discussed following) are cumbersome, requiring multiple measurements. Addi-

973

974

Section VII  Pancreas tionally, the systems are not accurate until 48 hours after presentation.

SCORING SYSTEMS Ranson’s Score

Ranson and colleagues identified 11 signs that had prognostic significance during the first 48 hours. The original list14 was analyzed in patients who primarily suffered from alcoholic pancreatitis and was then modified 8 years later for those with gallstone pancreatitis (see Table 58-2).215 Higher Ranson’s scores predict more severe disease. In mild pancreatitis (scores < 2), the mortality is 2.5% and in severe pancreatitis (scores > 3) the mortality is 62%.216 Also, the higher the Ranson’s score the higher the incidence of systemic complications, necrosis, and infected necrosis. These lists continue to remain in wide use in both the United States and Europe.217 The Ranson criteria have several drawbacks. First, the list is cumbersome and there are two lists to follow depending on suspected etiology (see Table 58-2). Second, an accurate Ranson’s score takes 48 hours to compute and the criteria have not been validated beyond the 48-hour time limit. Third, not all laboratories measure all the parameters in routine blood tests (e.g., serum lactate dehydrogenase [LDH]). Fourth, the overall sensitivity of the Ranson criteria (using three signs as the cutoff) for diagnosing severe disease is only 40% to 88% and the specificity is only 43% to 90%. The positive predictive value is approximately 50% and the negative predictive value around 90%.218 Therefore, the best use of Ranson’s score is to exclude severe disease. The Imrie or Glasgow score219 is a slightly simplified list (eight criteria) that is used commonly in the United Kingdom. It has similar drawbacks to the Ranson score.

APACHE-II Scores

APACHE-II is another commonly used scoring system in the United States to predict severity. It has the advantage of being able to be used on a daily basis and has similar positive and negative predictive values as the Ranson score at 48 hours after admission. The APACHE-II system assigns points for 12 physiologic variables, for age, and for chronic health status, in generating a total point score. The 12 variables are temperature; heart rate; respiratory rate; mean arterial blood pressure; oxygenation; arterial pH; serum potassium, sodium, and creatinine; hematocrit; white blood cell (WBC); and Glasgow Coma Scale. APACHE-II scores on admission and within 48 hours help distinguish mild from severe pancreatitis and to predict death.218,220 Most patients survive if APACHE-II scores are 9 or less during the first 48 hours. However, patients with APACHE-II scores of 13 or more have a high likelihood of dying. At admission, sensitivity is 34% to 70%, and specificity is 76% to 98%. At 48 hours, sensitivity remains less than 50%, but specificity is close to 90% to 100%.220 Strong drawbacks are its complexity, its low sensitivity on admission, and the fact that at 48 hours the score is no better than other scoring systems.221 Like the Ranson criteria, the APACHE-II score has its highest value in predicting mild disease.

BISAP

The problem with scoring systems is that they are cumbersome, using multiple variables. As described above, accuracy in predicting morbidity and/or mortality of the most commonly used scoring systems, Ranson and APACHE, is typically not achieved until 48 hours. By this time, it is usually apparent that the patient has developed severe disease manifested by organ failure. In order to develop a simple scoring system for patients with acute pancreatitis

that would be useful within the first 12 hours from admission, the Pancreas Center at Brigham and Women’s Hospital performed a series of studies retrospectively and prospec­ tively.221a,221b The studies were performed on a large data­ base including almost 37,000 patients and more than 200 hospitals. After careful analysis, including a validation study, they determined that a simple system that included 5 variables could accurately determine severity early in the course of the disease. The scoring system, referred to as BISAP (Bedside Index for Severity in Acute Pancreatitis), also uses the first letter of each parameter for 1 point. The BISAP score provides a single point for 5 parameters: blood urea nitrogen (BUN) greater than 25 mg/dL, impaired mental status, systemic inflammatory response syndrome, age greater than 60, and/or the presence of a pleural effusion, for a possible total of 5 points. A BISAP score greater than 3 is associated with a seven- to twelve-fold increase in developing organ failure.221b Accurate, yet much easier to use, this new simple scoring system appears to be useful in the early identification of patients who are at risk of developing complications and mortality.221a

Blood Urea Nitrogen

Several prognostic scoring systems, including the Ranson criteria and BISAP, incorporate blood urea nitrogen (BUN) for the prediction of mortality in patients with acute pancreatitis. Hemoconcentration, as described above, has been shown to be an accurate predictor of necrosis and organ failure. Both BUN and the hematocrit or hemoglobin are routine laboratory tests that may provide information on changes in intravascular volume status. Either test may be used in monitoring the early response to initial fluid resuscitation. Wu and colleagues221c recently performed a large observational cohort study on data from 69 U.S. hospitals and found that BUN may be superior to hemoglobin (not hematocrit). For every 5  mg/dL increase in BUN during the first 24 hours, the age- and genderadjusted odds ratio for mortality increased by 2.2. Of multiple routine laboratory tests examined, BUN yielded the highest accuracy at 24 hours and 48 hours. Although further study is needed, this paper suggests that following serial BUN measurements would be the most valuable single routine laboratory test for predicting mortality in acute pancreatitis.

ORGAN FAILURE

There is considerable interest among pancreatologists in using organ failure to predict severity. The Atlanta criteria defined which organ systems are of importance: pulmonary, renal, and cardiovascular. However, these criteria did not attempt to quantitate or prognosticate using organ failure. It has been appreciated that multiorgan failure or persistent single organ failure has a greater associated mortality than transient single organ failure. Multisystem organ failure is defined as two or more organs failing on the same day, rather than one organ failing on one day and another failing on the subsequent day. Patients with multisystem organ failure or persistent organ failure have a much higher mortality rate (approaching 50%) compared with patients with single and transient organ failure.222 Persistent organ failure is defined as lasting greater than 24 hours regardless of intervention. Survival among patients with organ failure at admission has also been shown to correlate with the duration of organ failure. When organ failure is corrected within 48 hours, mortality is close to zero. When organ failure persists for more than 48 hours, mortality is 36%.223 The Marshall Scoring System224 for organ failure is commonly used by intensivists for patients admitted to an intensive

Chapter 58  Acute Pancreatitis car unit. Data have not yet been generated using this system to prognosticate mortality in acute pancreatitis. Studies are needed to determine if this scoring system improves on the Ranson and APACHE scoring systems.

PERITONEAL LAVAGE

Percutaneous recovery of any volume of peritoneal fluid with a dark color or recovery of at least 20 mL of free intraperitoneal fluid of dark color portends a significant mortality.225 The sensitivity of peritoneal lavage is 36% to 72%, and the specificity is greater than 80% to 100%.226 An advantage is that it can be used any time, but it has not gained wide acceptance because it is invasive.

severity, such as CRP, TAP is not a surrogate marker of inflammation. Normally trypsinogen is cleaved to trypsin in the intestinal lumen by the enzyme enterokinase. Premature intrapancreatic activation during acute pancreatitis results in the release of TAP. The degree of pancreatic necrosis and systemic inflammatory response or sepsis is directly related to TAP concentration. Elevated urinary TAP (>30  nmol/L) correlates with disease severity. The test can be applied within 12 hours of admission. The positive predictive value of an elevated TAP is 80% and the negative predictive value approaches 100%.

Procalcitonin

Because the degree of elevation of serum amylase and lipase does not distinguish mild from severe pancreatitis,219 other factors have been examined.

This propeptide is another acute-phase reactant that has been shown to differentiate mild from severe acute pancreatitis within the first 24 hours after symptom onset. A serum strip test has been developed that has a sensitivity of 86% and a specificity of 95% in detecting organ failure.234

Hematocrit

COMPUTED TOMOGRAPHY

LABORATORY MARKERS

A high hematocrit on admission, or one that fails to decrease after 24 hours of rehydration is thought to be a sign of hemoconcentration due to retroperitoneal fluid loss and thus a marker of severe disease.227 One study showed that a hematocrit greater than 44% had a sensitivity of 72% on admission and of 94% after 24 hours in detecting organ failure. The negative predictive value at 24 hours was 96%. Although one study from Germany found no correlation between admission hematocrit and organ failure, most investigators have found hematocrit to be important in the management of patients with acute pancreatitis.228 An elevated hematocrit (>44%) is a predictor for the development of necrosis. The hematocrit should be observed at admission for prognostic purposes and followed prospectively to assist in guiding the rate of intravenous hydration.12

C-Reactive Protein

CRP is an acute-phase reactant produced by the liver and is used extensively in Europe as a marker of severe pancreatitis. CRP is inexpensive to measure and readily available. The sensitivity for detecting severe disease is 60% to 100% (using cutoffs of 100 to 210 mg/L, or 10 to 21 mg/dL) and the specificity is 75% to 100%.229

Interleukin-6

IL-6 is an acute-phase-reactant cytokine that is produced by a variety of cells and induces hepatic synthesis of CRP. Several studies have shown that it is a reasonably good marker to differentiate mild from severe disease, but the test is not readily available.230

Polymorphonuclear Leukocyte Elastase

Polymorphonuclear leukocyte elastase rises very early (before CRP) in acute pancreatitis. High levels have been reported to differentiate severe from mild disease,231 but the test is not generally available.

Phospholipase A2

PLA2 is involved in the release of prostaglandins from cell membranes and degrades surfactant in the lung. It may play a role in the pulmonary dysfunction associated with acute pancreatitis. Levels of catalytic type II PLA2 have been reported to differentiate between mild and severe disease within 24 hours of admission.232

Urinary Trypsinogen Activation Peptide

Urinary TAP may serve as an early predictor of severity in patients with acute pancreatitis.233 Unlike other markers of

The finding of extensive fluid collections or extensive necrosis on CT has been correlated with severe disease. Balthazar reported that 5 of 37 (13.5%) patients who had grade D or E findings on CT died, as opposed to none of 51 who had grades B or C findings (Table 58-7).119 Using the CT severity index (CTSI score) (see Table 58-7), among those with a score of 0 to 6, 3 of 77 (3.8%) died, as compared with 2 of 11 (18%) with scores of 7 to 10. The CT grading scores correlate better with local complications (pseudocysts and abscesses) than with mortality. Among the 37 patients with a grade D or E score, 54% developed a local complication, whereas only 2 of 51 (3.9%) with grades A through C developed this problem.211 Thus, the data do not confirm that the CTSI is any more predictive than the grades A through E score. There is controversy in the literature as to whether the extent of necrosis on CT predicts organ failure. Two studies

Table 58-7  Computed Tomography (CT) Grading System of Balthazar and CT Severity Index (CTSI) Balthazar Grades Grade A

Normal pancreas consistent with mild pancreatitis Grade B Focal or diffuse enlargement of the gland, including contour irregularities and inhomogeneous attenuation but without peripancreatic inflammation Grade C Grade B plus peripancreatic inflammation Grade D Grade C plus associated single fluid collection Grade E Grade C plus two or more peripancreatic fluid collections or gas in the pancreas or retroperitoneum CTSI = Balthazar Grade Score Plus Necrosis Score* Balthazar grade score: A = 0 B = 1 C = 2 D = 3 E = 4 Necrosis score: Absence of necrosis = 0 Necrosis of up to 33% of pancreas = 2 Necrosis of 33% to 50% = 4 Necrosis of >50% = 6 *Highest attainable score = 10 (Balthazar grade E + necrosis >50%).

975

976

Section VII  Pancreas failed to show any correlation between these two events,223,235 whereas a third study found a strong correlation.236

CHEST RADIOGRAPHY

A pleural effusion documented within 72 hours of admission190,193 by chest radiography (or CT) correlates with severe disease.

TREATMENT (Fig. 58-5) GENERAL CONSIDERATIONS

Patients with acute pancreatitis require adequate intravenous hydration and adequate analgesia to eliminate or markedly reduce pain. The patient is usually on nothing

Early course: 0-72 hours Is there organ failure? No

Yes

Admission to medical/surgical floor NPO, IV hydration (250-400 cc/hr) Nasal oxygen Frequent evaluation of oxygen saturation Hematocrit daily Serum electrolytes daily Pain control

Admission to an ICU Same orders as for floor admission Central line placement Evaluate need for assisted ventilation Assess for bile duct obstruction If bilirubin rising, consider urgent ERCP

Later course: >72 hours Evidence of severe disease or organ failure? No

Yes

Early refeeding Evaluate for etiology If GS, early cholecystectomy If ETOH, address psychosocial issues If high serum TGs, medical therapy

Interstitial pancreatitis on CT without peripancreatic necrosis: Continue supportive care Observation

To ICU if patient not already there Observe for biliary sepsis; if present, consider emergency ERCP Enteral feedings (NJ or NG) CT to evaluate for necrosis

Pancreatic/peripancreatic necrosis on CT: Continue supportive care Enteral feedings If infection suspected, consider antibiotics

Late course: 7-28 days Patient improving? Yes

No

Consider oral refeeding

If on antibiotics, consider FNA of pancreas for culture and change of antibiotics If not on antibiotics and FNA negative, keep off antibiotics Beyond 28 days Patient improving?

Yes Consider refeeding If patient cannot tolerate feedings, consider necrosectomy

No Consider necrosectomy by endoscopic, radiologic, or surgical means

Figure 58-5.  Algorithm for the management of acute pancreatitis at various stages in its course. CT, computed tomography; ERCP, endoscopic retrograde cholangiopancreatography; ETOH, ethyl alcohol; FNA, fine-needle aspiration; GS, gallstones; ICU, intensive care unit; IV, intravenous; NG, nasogastric; NJ, nasojejunal; NPO, nil per os (nothing by mouth); TGs, triglycerides.

Chapter 58  Acute Pancreatitis per mouth until any nausea and vomiting have subsided. Abdominal pain can be treated with opiate analgesics, often by a patient-controlled anesthesia pump. Opiate dosing is monitored carefully and adjusted on a daily basis according to ongoing needs. Although morphine has been reported to increase sphincter of Oddi tone and increase serum amylase,237 its use to treat the pain of pancreatitis has not been shown to adversely effect outcome. Nasogastric intubation is not used routinely because it is not beneficial in mild pancreatitis. It is used only to treat gastric or intestinal ileus or intractable nausea and vomiting. Similarly, proton pump inhibitors or H2-receptor blocking agents69 are not beneficial and not used. The patient should be carefully monitored for any signs of early organ failure such as hypotension, pulmonary, or renal insufficiency by closely following vital signs and urinary output. Tachypnea should not be assumed to be due to abdominal pain; monitoring oxygen saturation and, if needed, blood gases is advised and oxygen supplementation is mandatory if there is hypoxemia. It cannot be overemphasized that any patient who exhibits signs of early organ dysfunction should be immediately transferred to intensive care monitoring as deterioration can be rapid and fatal.

FLUID RESUSCITATION

As the inflammatory process progresses early in the course of the disease, there is an extravasation of protein-rich intravascular fluid into the peritoneal cavity resulting in hemoconcentration. The decreased perfusion pressure into the pancreas leads to microcirculatory changes that lead to pancreatic necrosis. An admission hematocrit of more than 47% and a failure of the admission hematocrit to decrease at 24 hours have been shown to be predictors of necrotizing pancreatitis.238 The relationship of hematocrit to severity of pancreatitis implies that the opposite is also true. Early vigorous intravenous hydration for the purpose of intravascular resuscitation is of foremost importance. The goal is to decrease the hematocrit. Laboratory and clinical studies with intravenous dextran to promote hemodilution have suggested efficacy in preventing severe disease.239 Too often patients with acute pancreatitis are given suboptimal intravenous hydration. One of the markers of severity of pancreatitis defined by Ranson and colleagues is intravascular losses (“fluid sequestration”). Ranson and colleagues found that a sequestration of more than 6 L of fluids during the first 48 hours was an independent predictor of disease severity in nongallstone pancreatitis.14 If this amount of fluid (6 L) is added to the minimal intravenous fluid requirements of a 70-kg person during the first 48 hours, intravenous hydration should be at least 250 to 300 mL/hour for 48 hours. The rate of volume replacement is likely to be more important during the first 24 hours, when a rising hematocrit has been shown to correlate closely with severe disease. A study from the Mayo Clinic showed that patients with severe acute pancreatitis who do not receive at least one third of their initial 72-hour cumulative intravenous fluid volume during the first 24 hours after emergency department presentation are at risk for greater mortality than those who are initially resuscitated more aggressively.240 Maintaining adequate intravascular volume in patients with severe disease may require 5 to 10 L of fluid such as isotonic saline daily for the first several days (200 to 400 mL/ hour). Respiratory distress often suggests development of ARDS independent of intravascular volume status. However, in a patient with unclear cardiac output, a Swan-Ganz catheter can be useful to gauge fluid resuscitation and to avoid congestive heart failure.

RESPIRATORY CARE

Hypoxemia (oxygen saturation 80  mEq/L) in 69 patients with normal or equivocal histology, mean bicarbonate concentration was 70, 63, and 50  mEq/L in those with mild, moderate, and severe histologic chronic pancreatitis, respectively. The overall sensitivity of hormonal stimulation testing in this study was 67%, with a specificity of 90% and overall accuracy of 81%. When the analysis was restricted to the 29 patients with moderate or severe histologic changes of chronic pancreatitis, the sensitivity of hormonal stimulation testing rose to 79%. In this same group of 29 patients, the sensitivity of ERCP was 66%. In comparisons with ERCP, direct hormonal stimulation tests appear to be on average somewhat more sensitive for the diagnosis of chronic pancreatitis. The values for sensitivity of pancreatic function testing range from 74% to 97%, with specificity ranging from 80% to 90%.174,175,180-185 In these studies the two tests agree in about three quarters of patients, although some studies note higher rates of concordance. Most studies also note a general correlation between increasing structural abnormalities and progressive abnormalities of hormone stimulation test results, although the relationship is not exact. Most of these studies also identify patients with discordant results—patients with abnormal ERCPs and normal hormonal stimulation test results as well as those with normal ERCPs and abnormal hormonal stimulation test results. In four studies the percentage of patients with an abnormal hormonal stimulation test result and a normal ERCP ranged from 3% to 20%.180-184 Two small studies have followed such patients whose diagnosis was based solely on an abnormal hormonal stimulation test result, and both found development of chronic pancreatitis on follow-up in 90% of patients.184,186 These data point out that direct pancreatic function testing appears to be able to identify a group of patients with chronic pancreatitis who have functional abnormalities of stimulated secretion but who do not (yet) have ERCP-identifiable structural abnormalities. Conversely, most of these studies also report patients with normal hormonal stimulation test results and abnormal ERCPs. This group of patients is generally less common,

averaging less than 10% in several studies.180-186 Long-term follow-up in a small group of such patients noted development of chronic pancreatitis in 0% to 26%.184,186 These studies point out that in situations in which results of the two tests disagree, hormonal stimulation testing appears to be somewhat more sensitive and specific than ERCP. Some experts have suggested that the pancreas has such reserve that 30% to 50% damage to the gland is necessary before direct pancreatic function tests yield reliably positive results. Despite their theoretical advantages, direct pancreatic function tests have a number of limitations.175,176 First, they have not been well standardized across institutions offering the test. Second, they are available at only a very few referral centers and so are not available to the majority of clinicians seeing patients with chronic pancreatitis. Third, it can be difficult for patients to tolerate unsedated placement of an oroduodenal tube for the hour or more required for the test. Fourth, accurate measurement of bicarbonate concentrations or enzyme output may be challenging. False-positive results have been reported in patients who have undergone Billroth II gastrectomy, in patients with diabetes, celiac disease, and cirrhosis, and in patients recovering from a recent attack of acute pancreatitis. A direct pancreatic function test is most useful in patients with presumed chronic pancreatitis in whom easily identifiable structural and functional abnormalities have not been demonstrated on more widely available diagnostic modalities such as CT (e.g., a patient with small-duct disease). This type of test is most useful in ruling out chronic pancreatitis in patients who present with a chronic abdominal pain syndrome suggestive of chronic pancreatitis, saving these patients the label of chronic pancreatitis with its negative repercussions and the risk of such diagnostic tests as ERCP. There are variations of direct pancreatic function tests that may be easier for patients to tolerate (by sedating them) and might be able to be made more widely available. One proposed variation is to collect pancreatic secretions at the time of ERCP by placement of a catheter directly in the pancreatic duct (the so-called intraductal secretin test). This test typically samples pancreatic output for only 15 minutes, to minimize the likelihood of ERCP-induced pancreatitis. It is not standardized and does not appear to be as accurate as standard direct pancreatic function testing,175,187,188 probably because of the rather brief collection time. Another variation of pancreatic function testing is to use sedation and a standard upper endoscope to take the place of the usual oroduodenal tube, with analysis of bicarbonate output using the regular hospital laboratory. This variation attempts to bypass the difficulties limiting the widespread application of standard direct pancreatic function tests, such as passage of the collection tube in unanesthetized patients and need for a dedicated laboratory to measure the bicarbonate concentration by back-titration. Unlike the intraductal secretin test, this endoscopic variation appears to be nearly, although not quite, as accurate as standard direct pancreatic function testing.189-191 The initial descriptions of this test used a 60-minute collection with timed aspirates of duodenal fluid every 15 minutes. This is a long time to keep a patient sedated and an endoscopy room occupied, although it may be possible to shorten the test and only collect samples at 30 and 45 minutes after secretin injection192 with only moderate loss of sensitivity. Another variation of the test is to measure lipase output rather than bicarbonate output, with secretin193 or CCK194 as the secretagogue. These variations appear to be less accurate than standard direct pancreatic function testing. Like traditional pancreatic function testing, endoscopic-based pancreatic function tests have been compared to alternative diagnostic

999

1000

Section VII  Pancreas tests like ERCP,195 with an overall agreement in 86% of patients and a high negative predictive value for endoscopic pancreatic function testing. The value of pancreatic function tests, whether traditional or endoscopic, lies in their high sensitivity and consequent ability to rule out chronic pancreatitis.196

Indirect Tests

The desire to develop indirect tests of pancreatic function is an outgrowth of the complexity, discomfort, and limited availability of direct pancreatic function testing. Indirect tests can generally measure pancreatic enzymes in blood or stool. Tests that measure the effect of pancreatic enzymes on an orally administered substrate with collection of metabolites in blood, breath, or urine (see Chapter 56) are of historical interest only. Serum Trypsinogen Serum trypsinogen (often called serum trypsin) can be measured in blood and provides a rough estimation of pancreatic function. Very low levels of serum trypsinogen (4 mm)   Pancreatic duct irregularity   Cavities 10 mm   Intraductal filling defects   Calculi/pancreatic calcification   Ductal obstruction (stricture)   Severe duct dilatation or irregularity   Contiguous organ invasion

Equivocal

Mild-moderate

Severe

Adapted from Sarner M, Cotton PB: Classification of pancreatitis. Gut 1984; 25:756.

Chapter 59  Chronic Pancreatitis In one more recent study comparing transabdominal ultrasonography with CT, ERCP, and EUS, the accuracy of ultrasonography was 56%.206 In this study, some abnormality (such as changes in parenchymal echotexture) was noted on ultrasonography in 40% of patients who had a normal pancreas as defined by the other diagnostic tests. A large screening study of transabdominal ultrasonography in Japan encompassing 130,000 examinations found increased echogenicity, mild dilation of the pancreatic duct, small cystic cavities, and even ductal calcification in the absence of clinical features of chronic pancreatitis.207 The majority of these abnormalities could not be attributed to chronic pancreatitis and were instead attributed to aging. These studies would suggest that there is a large spectrum of ultrasonographic findings in normal individuals and that it can be difficult to distinguish normal (or age-related) variability from chronic pancreatitis if the visualized changes are mild. Thus, transabdominal ultrasonography is often not useful in the evaluation of patients with suspected chronic pancreatitis. The finding of a normal pancreas or moderate to marked changes of advanced chronic pancreatitis is generally definitive. Mild changes of chronic pancreatitis are less specific and must be interpreted in light of the clinical history and the patient’s age. Ultrasonography can be useful in screening for complications of chronic pancreatitis (e.g., pseudocyst or bile duct obstruction) and in evaluating for other conditions that might mimic the symptoms of chronic pancreatitis (i.e., biliary tract disease).

Computed Tomography

The overall sensitivity of CT for chronic pancreatitis is between 75% and 90%, with a specificity of 85% or more.208-210 CT is able to image the pancreas in essentially all patients and hence has a substantial advantage over ultrasonography. Table 59-4 outlines the diagnostic abnormalities seen on CT in chronic pancreatitis. Most studies of diagnostic CT in chronic pancreatitis have not used stateof-the-art CT technology. One study using a 64 multi-row detector CT demonstrated the ability to accurately image the pancreatic duct and an overall sensitivity or more than 80% and specificity of near 100% when compared with MRCP or ERCP.211 It is almost certain that modern multidetector scanners using a pancreas protocol have better sensitivity than these older studies suggest, although the magnitude of the greater sensitivity is not known. Like all diagnostic tests, CT is most accurate in advanced chronic pancreatitis after substantial structural changes have developed (Fig. 59-4). Although CT is more expensive than ultrasonography and exposes the patient to ionizing radiation, it is more sensitive and more specific.

Magnetic Resonance Imaging

MRI, coupled with MRCP is as accurate, and probably more so, than CT in patients with chronic pancreatitis.208-212 MRCP results agree with ERCP results in about 90% of cases.212-213 Agreement between MRCP and ERCP is less common in areas where the pancreatic duct is small (tail of pancreas and side branches) or when the ductal changes are more subtle.214 Improved visualization of the pancreatic duct can be achieved by administering secretin.212-216 In addition, signal intensity and arterial enhancement ratios can be obtained, using gadolinium as a contrast agent, which may improve the ability to image the gland.212,216 Finally, a qualitative or semi-quantitative assessment of fluid output from the pancreas to the duodenum can be made during MRCP after secretin injection, which may allow additional insights into pancreatic function.217-219 The accuracy of this type of pancreatic function test has been compared with the intra-

Figure 59-4.  Computed tomography scan demonstrating several large, densely calcified stones (arrows) within a markedly dilated pancreatic duct in long-standing “big-duct” chronic pancreatitis.

ductal secretin test217 and with low-sensitivity tests such as fecal elastase,220 but not with a more formal hormonalstimulation pancreatic function test. Some analyses suggest that just measuring volume after secretin stimulation, instead of bicarbonate concentration, is too inaccurate to be clinically useful.221 Advancements in MR image analysis will continue to improve the image quality of MRCP, which in the future will equal ERCP in accuracy. Like ERCP, however, MRCP will be inaccurate in patients without significant ductal abnormalities. Although MRI is widely available, not all centers have the capacity to perform highquality MRCP.

Endoscopic Retrograde Cholangiopancreatography

Pancreatography has been considered the most specific and sensitive test of pancreatic structure, and many consider it the de facto gold standard. It also has the advantage over all previously discussed tests in that therapy (e.g., pancreatic duct stenting or stone extraction) may be administered during its performance. The disadvantage, however, is that ERCP is the riskiest diagnostic test, with complications occurring in at least 5% of patients (in as many as 20% of certain subgroups) and a mortality rate of 0.1% to 0.5%. In most studies in patients with chronic pancreatitis, the sensitivity of ERCP is between 70% and 90%, with a specificity of 80% to 100%.173,174,181-185,222,223 Thus, chronic pancreatitis can exist in the absence of any visible changes within the pancreatic duct.177,222-224 The diagnostic features of chronic pancreatitis on ERCP are listed in Table 59-5. These were developed at an international consensus conference held more than 20 years ago.225 The diagnosis is based on abnormalities seen in the main pancreatic duct and the side branches. ERCP is highly sensitive and specific in patients with advanced disease. The appearance of a massively dilated pancreatic duct with alternating strictures (the chain-of-lakes appearance) is characteristic of the most advanced chronic pancreatitis (Fig. 59-5). Less dramatic pancreatographic changes are less definitive and specific (Fig. 59-6).222,223

1001

1002

Section VII  Pancreas

Figure 59-5.  Endoscopic retrograde pancreatogram showing a markedly dilated pancreatic duct with alternating strictures and dilatation. This “chain-of-lakes” appearance is diagnostic of chronic pancreatitis.

Table 59-5  Cambridge Grading of Chronic Pancreatitis on Endoscopic Retrograde Pancreatography GRADE

MAIN PANCREATIC DUCT

SIDE BRANCHES

Normal Equivocal Mild Moderate Severe

Normal Normal Normal Abnormal Abnormal with at least 1 of the following:   Large cavity (>10 mm)   Obstruction   Filling defects   Severe dilatation or irregularity

Normal 192 ng/mL) Mucinous epithelial cells or malignant cells Resection

Thin, clear

Communication with pancreatic duct Fluid characteristics

Thin, clear to dark

Amylase level in fluid CEA level in fluid Cytology of fluid or wall

High Low Inflammatory cells

Therapy

Drainage

Microcystic, honeycomb appearance Uncommon

Low Low Cuboidal cells without mucin Observation

CEA, carcinoembryonic antigen; CT, computed tomography; EUS, endoscopic ultrasonography.

Pseudoaneurysm

Figure 59-9.  Computed tomography scan demonstrating a cystic neoplasm seen in the tail of the pancreas in an older adult woman with no history of pancreatic disease. The loculations and mural nodules seen within the cavity (arrow) are suggestive of a cystic neoplasm, rather than a pseudocyst (see Chapter 60).

pseudocyst, pseudoaneurysm, and portal or splenic vein thrombosis.

Pseudocyst

Bleeding may occur from the wall of a pseudocyst. Bleeding occurs from small vessels (venous, capillary, or arteriole) in the wall, which can lead to expansion of the pseudocyst and further rupture of these small vessels.323 Blood may remain in the pseudocyst or may reach the gut via a spontaneous pseudocyst decompression into the GI lumen or into the pancreatic duct (hemosuccus pancreaticus). Bleeding from small vessels in the wall of the pseudocyst is generally of low volume.

Pseudoaneurysms form as a consequence of enzymatic and pressure digestion of the muscular wall of an artery by a pseudocyst. The pseudoaneurysm may rupture either into the pseudocyst (converting the pseudocyst into a larger pseudoaneurysm) or directly into an adjacent viscus, peritoneal cavity, or pancreatic duct. Pseudoaneurysmal bleeding may complicate 5% to 10% of all cases of chronic pancreatitis with pseudocysts, although pseudoaneurysms may be seen in up to 21% of patients with chronic pancreatitis undergoing angiography.323 Pseudoaneurysms are also seen after pancreatic surgery. Many visceral arteries may be involved, the splenic artery being most common, followed by gastroduodenal or pancreaticoduodenal arteries. Once bleeding occurs, the mortality is at least 40% being related both to the severity of the blood loss and to the presence of coexisting conditions. Although death from a pseudocyst is rare, more than half the overall mortality of pseudocysts is due to hemorrhage. Bleeding from a pseudoaneurysm may be slow and intermittent or acute and massive. Common presentations are abdominal pain (due to the enlargement of the pseudocyst), unexplained anemia, and overt GI bleeding (if the blood can reach the gut lumen through the pseudocyst or through the pancreatic duct). In many cases, an initial self-limited bleed occurs (so-called sentinel bleed), followed hours or days later by a massive exsanguinating hemorrhage. The initial self-limited nature of the bleed may be due to transient tamponade of the bleeding within the confines of the pseudocyst. The presence of unexplained blood loss or any amount of GI bleeding in a patient with pancreatitis or a known pseudocyst should immediately raise the possibility of a pseudoaneurysm. If a pseudoaneurysm is suspected in the setting of upper GI blood loss, an urgent upper endoscopy should be undertaken. If no obvious bleeding site is seen, pseudoaneurysm formation should be considered. Rarely, blood may be seen issuing from the ampulla (hemosuccus pancreaticus), but the absence of this finding does not rule out pseudoaneurysm. The next step in the evaluation should be an emergency CT scan with intravenous contrast. The finding of highdensity material within a pseudocyst on the initial noncon-

Chapter 59  Chronic Pancreatitis

Figure 59-10.  Computed tomography scan demonstrating a pseudocyst containing a pseudoaneurysm (arrow) that is opacified following intravenous injection of contrast agent.

trast images is highly suggestive of a pseudoaneurysm as is a circular opacifying structure within the low-attenuation pseudocyst after the intravenous administration of contrast agent (Fig. 59-10). It is prudent to avoid oral administration of a contrast agent so that it will not interfere with angiography if required. In most centers, such a CT finding is followed immediately by angiography to define and embolize the pseudoaneurysm. Once a pseudoaneurysm has been identified, it should be treated whether or not it has caused bleeding. Angiographic embolization or stent-graft placement has largely replaced primary surgery, which is reserved for cases in which these therapies have failed.324

Variceal Bleeding from Splenic Vein Thrombosis

Variceal bleeding may complicate chronic pancreatitis because of either associated alcoholic cirrhosis or thrombosis of the splenic (and, less commonly, portal) vein (see Chapters 19 and 90). Thrombosis of the splenic vein is most common and produces a segmental or left-sided portal hypertension.323 Decompression of splenic venous outflow occurs through the short gastric veins to the coronary vein, producing prominent variceal channels in the gastric cardia and fundus. Depending on the venous anatomy, esophageal varices may also be produced, although they are generally smaller than the gastric varices. The natural history of gastric varices in this setting is not known, but the overall risk of bleeding is less than with esophageal varices due to cirrhosis. In one analysis, the risk of gastric variceal bleeding was only 4%.325 Therapy is therefore not required in the absence of bleeding. Should bleeding occur, splenectomy is curative. Endoscopic control of bleeding is sometimes possible with gastric varices, utilizing cyanoacrylate injection or other techniques (see Chapter 19).

BILE DUCT OBSTRUCTION

The distal bile duct is enclosed within the posterior portion of the head of the pancreas. Inflammatory and fibrotic conditions of the head of the pancreas, as well as pseudocysts in this location, can compress this intrapancreatic bile duct, leading to abnormal liver chemistry values, jaundice, biliary pain, or cholangitis. Symptomatic bile duct obstruction occurs in about 10% of patients. The ductal stricture can be suspected from cholestatic liver chemistry values, CT or ultrasonography findings of biliary ductal dilation, or

Figure 59-11.  A retrograde cholangiogram showing a smooth stricture of the bile duct (arrows) as it passes through the head of the pancreas in a patient with chronic pancreatitis.

both. ERCP characteristically demonstrates a long tapered stenosis of the distal bile duct (Fig. 59-11). The occurrence of cholangitis is an absolute indication for therapy. The presence of abnormal liver chemistry values or jaundice is not so straightforward because those most affected are alcoholic patients, and alcoholic (and other intrinsic) liver disease can also produce substantial abnormalities in liver chemistry values. The clinical, biochemical, and even radiologic features are not always sufficient to distinguish biliary stenosis from intrinsic liver disease.326 It is for this reason that liver biopsy is often necessary to enable the choice of therapy. The mere presence of a stenosis of the intrapancreatic bile duct, in the absence of symptoms or progressive abnormalities in liver chemistry values, can usually be followed conservatively. If there is a concern about the development of secondary biliary cirrhosis, a liver biopsy should be performed. In patients with increasing jaundice or biliary pain, in the absence of alternative explanations (i.e., intrinsic liver disease), therapy should be considered. Definitive therapy of bile duct obstruction usually requires surgical biliary bypass with choledochojejunostomy or choledochoduodenostomy. One study suggested that hepatic fibrosis due to chronic biliary obstruction may actually decrease after successful surgical decompression.327 Although endoscopic stent therapy for biliary obstruction due to chronic pancreatitis is generally temporarily effective (see Chapter 61), the long-term success is relatively low.328 Placement of one or more plastic stents to treat bile obstruction is relatively simple, but the long-term management is complicated by the need for multiple stent exchanges over many months to years, and stent migration and obstruction are common. The use of multiple parallel stents shows a better response than a single stent.329 Long-term endoscopic stent therapy in alcoholic patients is particularly difficult and is associated with high complication rates owing to missed appointments for scheduled stent exchanges.330 The use of uncoated permanent metallic stents is discouraged because of the high rate of ultimate stent occlusion, although the use of fully coated, removable metal stents may provide additional options in the future.

1013

1014

Section VII  Pancreas The development of a bile duct stenosis in a patient with chronic pancreatitis may also signal the development of a pancreatic malignancy. EUS and ERCP are useful in this setting to attempt to differentiate benign from malignant strictures.

DUODENAL OBSTRUCTION

Approximately 5% of patients with chronic pancreatitis experience symptomatic duodenal stenosis. The majority of such patients have chronic alcoholic pancreatitis. Fibrosis in the head of the pancreas, often associated with an inflammatory mass, is the most common explanation. Symptoms include nausea, vomiting, weight loss, and abdominal pain. Coexistent obstruction of the bile duct may occur. The diagnosis is best made with CT using oral contrast or an upper GI barium study, because the extent of duodenal stenosis is often underestimated at the time of endoscopy. Because the degree of stenosis may improve with resolution of some of the inflammation, a trial of medical therapy may be worthwhile. Surgical therapy is required for those in whom conservative management fails. The simplest approach is a bypass with a gastrojejunostomy, which may be performed with laparoscopic techniques. This may be coupled with drainage of the bile duct and/or pancreatic duct (lateral pancreaticojejunostomy). Resection of the head of the pancreas with a Frey, Beger, Berne, or Whipple procedure (see earlier) may be considered in centers with this expertise in select patients with a large inflammatory mass of the head of the pancreas, or in those in whom malignancy is also being considered.

PANCREATIC FISTULA External Fistulas

External pancreatic fistulas occur most commonly as a consequence of surgical or percutaneous therapy for chronic pancreatitis or pseudocyst. It has been estimated that perhaps half of such fistulas heal with complete bowel rest and parenteral hyperalimentation. The most common complications are abscess and bleeding. There is some evidence that the addition of octreotide, in a dosage of 100 µg subcutaneously every eight hours, can hasten closure of such fistulas. Successful medical treatment, even with octreotide, can take many weeks. The placement of an endoscopic stent across the site of ductal disruption is effective at closing the fistula rapidly (see Chapter 61). Up to 75% of pancreaticocutaneous fistulas may be effectively treated with endoscopic techniques,331,332 although this approach may need to be coupled with percutaneous drainage of intra-abdominal fluid collections. In patients in whom endoscopic therapy fails or cannot be performed, surgical treatment can involve pancreatic resection (if the fistula is in the tail) or a fistulojejunostomy, in which the fistula tract is “capped” with a defunctionalized limb of jejunum.

Internal Fistulas

Internal pancreatic fistulas occur mainly in the setting of chronic pancreatitis after rupture of a pseudocyst. The fluid may track to the peritoneal cavity (pancreatic ascites) or into the pleural space (pancreatic pleural effusion). Affected patients may not complain of symptoms of chronic pancreatitis but may instead note abdominal distention or shortness of breath, respectively. Although such fistulas invariably occur in advanced chronic pancreatitis (particularly alcoholic) there may not be a clear-cut history of recent pancreatitis. The diagnosis can be established through documentation of high levels of amylase within the respective fluid, typically more than 4000 U/L.

Conservative treatment, consisting of complete bowel rest, parenteral hyperalimentation, paracentesis or thoracentesis, and octreotide, is effective in some internal pancreatic fistulas.333 If the leak is in the body or head of the pancreas, a pancreatic duct stent covering the fistula site is highly effective (see Fig. 61-8).331,332 In some cases, merely bridging the ampulla with a short pancreatic duct stent may be enough to heal the fistula. Endoscopic therapy is less effective but still worthwhile if the leak is from the tail but is ineffective if the leak is present upstream from a complete blockage of the pancreatic duct (excluded pancreatic tail syndrome). In this situation, resection or surgical drainage of the pseudocyst is required, and ERCP or MRCP is used preoperatively to delineate the ductal anatomy for surgical planning.

MALIGNANCY

The risk of pancreatic cancer is higher with all forms of chronic pancreatitis (see Chapter 60). The lifetime risk for pancreatic cancer in patients with chronic pancreatitis is about 4%.334 The risk of pancreatic cancer is highest in patients with hereditary pancreatitis, and particularly those who smoke (see Chapter 57). At present, there is no completely reliable way to differentiate chronic pancreatitis alone from chronic pancreatitis complicated by adenocarcinoma. The symptoms and signs may be similar (abdominal pain, weight loss, jaundice). In the absence of widespread metastases, imaging studies such as CT, ultrasonography, and even ERCP may be unable to establish the diagnosis. The role of EUS is evolving, but finding a small hypoechoic tumor within a diseased gland with preexisting altered echotexture can be difficult. However, EUS is superior to CT for detection of coexistent malignancy particularly when the lesion is small. EUS also has the substantial advantage of allowing directed tissue biopsy of any suspicions lesions. Tumor markers may also be helpful in attempting to differentiate chronic pancreatitis from cancer. CA 19-9, the tumor marker most commonly used for pancreatic adenocarcinoma, is elevated in the serum in 70% to 80% of patients with adenocarcinoma of the pancreas.335 Biliary obstruction and cholangitis can also raise CA 19-9 levels. The use of any of these techniques for surveillance is not cost-effective in the general population of patients with chronic pancreatitis, although they may be useful in families with hereditary pancreatic cancer. In some patients, laparotomy is required to determine the presence or absence of coexistent pancreatic carcinoma. In those with a benign pseudotumor who undergo resection to rule out malignancy, a variant of autoimmune chronic pancreatitis is often found.95,97 Several reports have also called attention to the development of extrapancreatic cancer in association with chronic pancreatitis, which are more common than pancreatic malignancy in these patients. These cancers, particularly those of the upper digestive tract and lungs, are probably related to the effect of concomitant tobacco use.20,334 The incidence of extrapancreatic carcinomas in these reports varies between 4% and 12%.

DYSMOTILITY

Gastroparesis and antroduodenal dysmotility are seen in patients with chronic pancreatitis,336,337 possibly as a consequence of perigastric inflammation, hormonal changes associated with chronic pancreatitis (e.g., increases in plasma CCK), or a side effect of narcotic analgesics. Gastroparesis is clinically important because it may produce symptoms occasionally indistinguishable from those of the disease and

Chapter 59  Chronic Pancreatitis may interfere with the effective delivery of pancreatic enzymes.337 Gastroparesis should be considered in patients with early satiety, nausea, vomiting, and weight loss.

KEY REFERENCES

Anapurthy R, Pasricha PJ. Pain and chronic pancreatitis. Is it the plumbing or the wiring? Curr Gastroenterol Rep 2008; 10:101-6. (Ref 144.) Baillie J. Pancreatic pseudocysts (part I). Gastrointest Endosc 2004; 59:873-9. (Ref 311.) Baillie J. Pancreatic pseudocysts (part II). Gastrointest Endosc 2004; 60:105-13. (Ref 312.) Blondet JJ, Carlson AM, Kobayashi T, et al. The role of total pancreatectomy and islet autotransplantation for chronic pancreatitis. Surg Clin North Am 2007; 87:1477-501. (Ref 296.) Cahen DL, Gouma DJ, Nio Y, et al. Endoscopic versus surgical drainage of the pancreatic duct in chronic pancreatitis. N Engl J Med 2007; 356:676-84. (Ref 281.) Chowdhury RS, Forsmark CE. Review article: Pancreatic function testing. Aliment Pharmacol Ther 2003; 17:733-50. (Ref 175.) Collins D, Penman I, Mishra G, Draganov P. EUS-guided celiac block and neurolysis. Endoscopy 2006; 38:935-9. (Ref 299.) Dite P, Ruizicka M, Zboril V, Novotmy I. A prospective, randomized trial comparing endoscopic and surgical therapy for chronic pancreatitis. Endoscopy 2003; 35:553-8. (Ref 280.)

Frulloni L, Lunardi C, Simone R, et al. Identification of a novel antibody associated with autoimmune pancreatitis. N Engl J Med 2009; 361: 2135-42. (Ref 87a.) Gardner TB, Chari ST. Autoimmune pancreatitis. Gastroenterol Clin North Am 2008; 37:439-60. (Ref 24.) Gleeson FC, Topazian M. Endoscopic retrograde cholangiopancreatography and endoscopic ultrasonography for diagnosis of chronic pancreatitis. Curr Gastroenterol Rep 2007; 9:123-9. (Ref 222.) Khalid A, Brugge W. ACG Practice Guidelines for the diagnosis and management of neoplastic cysts. Am J Gastroenterol 2007; 102:233949. (Ref 321.) Van der Gaag MA, Gouma DJ, Van Gulik TM, et al. Review article: Surgical management of chronic pancreatitis. Aliment Pharmacol Ther 2007; 26:221-32. (Ref 283.) Wilcox CM, Varadarajulu S. Endoscopic therapy for chronic pancreatitis: An evidence-based review. Curr Gastroneterol Rep 2006; 8:10410. (Ref 272.) Witt J, Apte MV, Keim V, Wilson JS. Chronic pancreatitis: Challenges and advances in pathogenesis, genetics, diagnosis, and therapy. Gastroenterology 2007; 132:1557-73. (Ref 36.) Full references for this chapter can be found on www.expertconsult.com.

1015

CHAPTE R

60 Tumors of the Pancreas Ramon E. Jimenez and Carlos Fernández-del Castillo

CHAPTER OUTLINE Pancreatic Cancer  1017 Epidemiology  1017 Pathology  1018 Clinical Features  1020 Diagnosis  1020 Staging  1023 Treatment  1024

PANCREATIC CANCER Pancreatic cancer is the second most common gastrointestinal malignancy in the United States, and approximately 37,680 new cases are expected to occur in 2008.1 Despite its relatively low incidence compared with other malignancies, it represents the fourth leading cause of cancer death in men and women (34,290 deaths expected in 2008). Overall, pancreatic cancer carries an unfavorable prognosis. For all stages combined, the one- and five-year relative survival rates are 24% and 5%, respectively.1 Early detection, accurate preoperative staging, and better treatment options remain a challenge.

EPIDEMIOLOGY Incidence

Pancreatic cancer is very rare before the age of 45 years, but its occurrence rises sharply thereafter. It affects men more than women (ratio of 1.3 : 1), and is more common in blacks. The incidence in black men is 14.8 per 100,000, compared with 8.8 in the general population.2

Populations at Risk

Genetic as well as environmental factors have been found to be associated with the development of pancreatic cancer (Fig. 60-1). Table 60-1 summarizes some of the genetic syndromes associated with an increased risk of pancreatic cancer. One of the most prominent of these syndromes is hereditary pancreatitis, even though it accounts for only a small fraction of pancreatic cancer cases (see Chapter 57). Affected patients have an abnormal trypsin gene that is transmitted as an autosomal dominant trait; their risk for development of pancreatic cancer by age 70 years is estimated at 40%.3 Patients with other, nonhereditary forms of chronic pancreatitis also have a higher likelihood of pancreatic cancer (see Chapter 59). A multinational study found this risk to be 2% per decade, independent of the type of pancreatitis.4 Individuals with familial Peutz-Jeghers syndrome, discussed in Chapter 122, also have an increased risk of pancreatic cancer, with an impressive 132-fold increased risk over the general population.5 Germline mutations in p16 are

Cystic Tumors of the Pancreas  1027 Differential Diagnosis  1027 Diagnostic Imaging  1028 Mucinous Cystic Neoplasms  1029 Serous Cystadenomas  1029 Intraductal Papillary Mucinous Neoplasms  1030 Solid Pseudopapillary Tumors  1032 Other Nonendocrine Pancreatic Tumors  1033

observed in kindreds with the familial atypical molemalignant melanoma (FAMMM) syndrome, and these individuals are at risk for pancreatic cancer and melanoma.6 Lastly, patients with BRCA2 gene mutations (which pre­ dispose to hereditary breast cancer in women and men), also have a familial predisposition to pancreatic cancer.7 Other unknown genetic abnormalities are yet to be identified. In several population studies, 7% to 8% of patients with pancreatic cancer have a first-degree relative with the disease.8 Patients with three or more first-degree relatives with the disease, have a 32-fold increased risk over the general population; those with two first degree relatives have 6-fold increased risk; and those with one first-degree relative have a 2.3-fold increased risk.9 There are no specific recommendations for screening patients at risk for pancreatic cancer, because available techniques lack sensitivity for detection of very small lesions. The timing and frequency of such screening are also uncertain. The American Gastroenterological Association (AGA) suggests that screening should begin at age 35 in patients with hereditary pancreatitis. Patients with a history of familial pancreatic cancer should begin screening 10 years before the age at which pancreatic cancer has been first diagnosed in their relatives. The AGA also states that such screening is probably best done with spiral computed tomography (CT) and endoscopic ultrasonography (EUS).10,11

Environmental Factors

The most important environmental factor in pancreatic cancer, and possibly the only one that has been firmly established, is cigarette smoking.12 Multiple cohort and casecontrol studies have found that the relative risk for smokers of developing pancreatic cancer is at least 1.5.13-15 The risk may be particularly elevated in smokers who have homozygous deletions of the gene for glutathione S-transferase T1 (GSTT1), which is a carcinogen metabolizing enzyme.16 Furthermore, the risk rises with the amount of cigarette consumption, and the excess risk level returns to baseline by 15 years after cessation of the habit.15 The second most important environmental factor associated with pancreatic cancer appears to be dietary influences. A high intake of fat or meat has been linked to the development of this neoplasm,17 and a protective effect is ascribed to fresh fruits and vegetables.18 Reduced serum levels of

1017

1018

Section VII  Pancreas Table 60-1  Historical Features and Germline Genetic Alterations Associated with an Increased Risk of Pancreatic Cancer history

GENE

RELATIVE RISK

individual RISK BY AGE 70

None Breast cancer

None BRCA2 BRCA1 p16 (CDKN2A) Unknown PRSS1 STK11/LKB1 MLH1, MSH2, others FANC-C, FANC-G, others Palladin

1 3.5-10× 2× 20-34× 32× 50-80× 132× Unknown Unknown Unknown

0.5% 5% 1% 10%-17% 16% 25%-40% 30%-60% 90% >95% 50%-70% 55% 10%-20% 10% 10% 7%-10% A, RFLP

ABCB11 ABCG5/G8 ARDB3 APOA1 APOB APOC1 AR CCK1R CETP CYP7A1 ESR2 LRPAP1

Apolipoprotein B

Apolipoprotein C1 Androgen receptor Cholecystokinin 1 receptor Cholesterol ester transfer protein

Cytochrome P450 7A1 Estrogen receptor 2 LRP-associated protein 1

+ (China) + (Greece) + (India)

− − − − − − − − − − −

− − − + − − + − + − −

+ (India) + (Greece) + (India) + (Finland)

−

+

+ (China, Poland)

+ (Germany Serbia, Chile, China) + (Germany) + (China, India)

−

−

+

−

Common polygenic

Familial oligogenic

Rare monogenic

Gallbladder hypomotility ↑ Biliary cholesterol secretion secondary to ↑ reverse cholesterol transport ↑ Biliary cholesterol secretion secondary to hepatic VLDL synthesis and ↑ intestinal cholesterol absorption ↑ APOC1 remnant-like particle cholesterol Gallbladder hypomotility Gallbladder and small intestinal hypomotility ↑ Hepatic cholesterol uptake from HDL catabolism ↑ Bile salt synthesis ↑ Cholesterol synthesis ↑ Hepatic cholesterol uptake from chylomicron remnants via LRP

↑ Biliary cholesterol secretion

↓ Biliary bile salt secretion

↓ Biliary phospholipid secretion

POTENTIAL MECHANISMS

*Countries where inheritance patterns were reported are shown in parentheses. HDL, high density lipoprotein; LRP, low-density lipoprotein receptor-related protein; OMIM, Online Mendelian Inheritance in Man; RFLP, restriction fragment length polymorphism; rs, restriction site; SNP, single nucleotide polymorphism; VLDL, very-low-density lipoprotein. Reproduced with slight modifications and with permission from Lammert F, Sauerbruch T. Pathogenesis of gallstone formation: Updated inventory of human lithogenic genes. In: Carey MC, Dité P, Gabryelewicz A, et al, editors. Future Perspectives in Gastroenterology (Falk Symposium 161). Dordrecht, Germany: Springer; 2008. pp 99-107.

Promoter SNP -204A>C c.1092+3607(CA)n Intron 5 insertion/deletion (rs11267919)

RFLP c.172(CAG)n RFLP RFLP

c.2488C>T, c.4154G>A

Multiple

ABCB4

ATP binding cassette transporter B4 ATP binding cassette transporter B11 ATP binding cassette transporters G5/G8 β3 Adrenergic receptor Apolipoprotein A1

GENE VARIANTS

GENE SYMBOL

PROTEIN

Inheritance Pattern*

Table 65-1  Human Cholesterol Gallstone (LITH) Genes and Gene Products That Have Been Identified as of 2008

Chapter 65  Gallstone Disease 1103

1104

Section VIII  Biliary Tract contains an excess of unconjugated bilirubin, analogous to the saturation of bile with cholesterol in patients with cholesterol stones.169 Also, both types of pigment stones are composed primarily of bile pigment and contain a matrix of mucin glycoproteins. In black stones, however, the pigment is predominantly an insoluble, highly cross-linked polymer of calcium bilirubinate, whereas in brown stones, the main pigment is monomeric calcium bilirubinate. The two types of pigment stones also differ in radiodensity, location within the biliary system, and geographic distribution.

BLACK PIGMENT STONES

Black pigment stones are formed in uninfected gallbladders, particularly in patients with chronic hemolytic anemia (e.g., β-thalassemia, hereditary spherocytosis, and sickle cell disease) and liver cirrhosis. The unconjugated bilirubin produced in increased amounts precipitates as calcium bilirubinate to form stones.170 This type of stone is composed of either pure calcium bilirubinate or polymer-like complexes consisting of unconjugated bilirubin, calcium bilirubinate, calcium, and copper. Mucin glycoproteins account for as much as 20% of the weight of black stones. A regular crystalline structure is not present. Under normal physiologic conditions, unconjugated bilirubin is not secreted into bile. Bilirubin glucuronides are hydrolyzed by endogenous β-glucuronidase, and unconjugated bilirubin constitutes less than 1% of total bile pigment, mostly because the activity of the enzyme is inhibited by β-glucaro-1,4-lactone in the biliary system.171,172 The unifying predisposing factor in the formation of black pigment stones is the hypersecretion of bilirubin conjugates (especially monoglucuronides) into bile. In the presence of hemolysis, secretion of these bilirubin conjugates increases ten-fold. Unconjugated monohydrogenated bilirubin is formed by the action of endogenous β-glucuronidase, which coprecipitates with calcium as a result of supersaturation. A 1% hydrolysis rate could give rise to high concentrations of unconjugated bilirubin that often greatly exceed the solubility of bilirubin in bile. A defect in acidification of bile also may be induced by gallbladder inflammation or the reduced buffering capacity of sialic acid and sulfate moieties in the mucin gel. The reduction in buffering capacity facilitates the supersaturation of calcium carbonate and phosphate that would not occur at a more acidic pH. Gallbladder motility defects are not observed in patients with black pigment stones, as inferred from in vitro experiments of human gallbladder muscles.

BROWN PIGMENT STONES

Brown pigment stones are composed mainly of calcium salts of unconjugated bilirubin, with varying amounts of cholesterol, fatty acids, pigment fraction, and mucin glycoproteins, as well as small amounts of bile salts, phospholipids, and residues. Brown pigment stones may be easily distinguished grossly from black pigment stones by their reddish brown to dark brown color and lack of brightness. Their shape is irregular or molded and occasionally spherical. Most of the stones are muddy in consistency, and some show facet formation. Brown pigment stones are either smooth or rough without any surface luster and are soft, fragile, and light in comparison with other gallstones. The cut surface is generally a stratified structure (lamellation) or is amorphous without the radiating crystalline structure seen in cholesterol stones. Almost invariably, brown pigment stones have a lamellated cross-sectional surface with calcium bilirubinate-rich layers alternating with calcium palmitate-rich layers. Brown pigment stones are formed not only in the gallbladder, but also commonly in other portions of the biliary

tree, especially in intrahepatic bile ducts. The formation of brown pigment stones requires the presence of structural or functional stasis of bile associated with biliary infection, especially with Escherichia coli.173 These stones are more common in areas such as Asia, where Clonorchis sinensis and roundworm infestations are prevalent, and parasitic elements have been considered to be kernels of brown pigment stone formation (see Chapters 68 and 82).174 Bile stasis predisposes to the bacterial infection as well as the accumulation of mucins and bacterial cytoskeletons in the bile ducts. Bile stasis may be caused by bile duct stenosis and bacterial infection caused by infestation by parasites such as Clonorchis sinensis, roundworms, and their ova.175 Additionally, bacterial infection and colonization in bile ducts by enteric bacteria are found commonly in patients with brown pigment stones. As the incidence of biliary infections has decreased in Asian populations prone to development of brown pigment stones, the ratio of cholesterol stones to pigment stones also has changed in these populations. The percentage of brown pigment stones in Japan has fallen from 60% to 24% since the 1950s, and similar changes have been reported from other Asian countries.176-178 Enteric bacteria produce β-glucuronidase, phospholipase A1, and conjugated bile acid hydrolase. Activity of β-glucuronidase results in the production of unconjugated bilirubin from bilirubin glucuronide; phospholipase A1 liberates palmitic and stearic acids from phospholipids; and bile acid hydrolases produce unconjugated bile salts from glycine or taurine-conjugated bile salts. Partially ionized saturated fatty acids, unconjugated bilirubin, and unconjugated bile salts may precipitate as calcium salts. Mucin gel can trap these complex precipitates and facilitate their growth into macroscopic stones. Figure 65-6 shows the postulated mechanisms underlying the formation of brown pigment stones. Under normal physiologic conditions, bilirubin in bile exists mainly as bilirubin glucuronide, which is soluble in aqueous media. Bile also contains β-glucuronidase of tissue origin, the activity of which is inhibited by glucaro-1,4-lactone, which is also formed in the liver. If infection with E. coli occurs, the concentration of bacterial β-glucuronidase increases significantly and exceeds the inhibitory power of glucaro-1,4-lactone. As a result, bilirubin glucuronide is hydrolyzed to produce unconjugated bilirubin and glucuronic acid; the former is water-insoluble and combines with calcium to form calcium bilirubin at its carboxyl radical, leading to the formation of brown pigment gallstones.

NATURAL HISTORY The natural history of gallstones typically is described in two separate groups of patients: those who have symptoms and those who are asymptomatic. Necropsy studies clearly show that the vast majority of patients with gallstones are asymptomatic and remain so. Ascertaining the true frequency of complications in persons with asymptomatic stones (as well as those with symptomatic stones) is critical to providing rational, cost-effective recommendations regarding therapy (see later). Unfortunately, the information available on the natural history of gallstones has been sparse and somewhat varied.179-181

ASYMPTOMATIC STONES

The study that changed our understanding of the course and appropriate therapy of gallstone disease was performed by Gracie and Ransohoff.179 They monitored 123 University of

Chapter 65  Gallstone Disease Phospholipids

Bilirubin glucuronides

Conjugated bile salts

b-Glucaro–1,4– lactone Bacterial b-glucuronidase Glucuronic acid

(–)

Phospholipase A1

Bile salt hydrolase

Endogenous b-glucuronidase

Free bilirubin

Free fatty acids

Free bile salts

Brown pigment stones

Calcium

Dead bacteria and parasites

Michigan faculty members for 15 years after they had been found to have gallstones on routine screening ultrasonography. At 5, 10, and 15 years of follow-up, 10%, 15%, and 18% of the patients, respectively, had become symptomatic, and none had experienced serious complications. The investigators suggested that the rate at which biliary pain develops in persons with asymptomatic gallstones is about 2% per year for five years and then decreases over time. Biliary complications developed in only three patients in this study, and all complications were preceded by episodes of biliary pain. Studies have suggested that biliary pain, not a biliary complication, is the initial manifesting symptom in 90% of people with previously asymptomatic gallstones.179 Therefore, in patients with asymptomatic stones, the frequency of complications is low, and prophylactic cholecystectomy is not necessary. Subsequent studies have reported slightly higher rates of biliary pain and complications in patients with initially asymptomatic gallstones,180 but only one was a long-term and prospective study.181 The Group for Epidemiology and Prevention of Cholelithiasis (GREPCO) in Rome reported the courses of 151 subjects with gallstones, 118 of whom were asymptomatic on entering the study. In those who were initially asymptomatic, the frequency of biliary pain was 12% at 2 years, 17% at 4 years, and 26% at 10 years, and the cumulative rate of biliary complication was 3% at 10 years.181

Mucin gel

Figure 65-6.  Proposed mechanisms for the pathogenesis of brown pigment stones. Under normal physiologic conditions, unconjugated bilirubin is not secreted into bile. Although modest hydrolysis of bilirubin glucuronides by endogenous b-glucuronidase occurs, unconjugated bilirubin constitutes less than 1% of total bile pigment, mostly because the activity of b-glucuronidase is inhibited by b-glucaro-1,4-lactone in the biliary system. The presence of excess bacterial β-glucuronidase, however, overcomes the inhibitory (-) effect of b-glucaro-1,4-lactone, which results in hydrolysis of bilirubin glucuronide into free bilirubin and glucuronic acid. Free bilirubinate combines with calcium to yield water-insoluble calcium bilirubinate. In addition, phospholipase A1 liberates free fatty acids such as palmitic acid and stearic acid from phospholipids, and bile salt hydrolases produce unconjugated bile salts from glycine or taurine-conjugated bile salts. Dead bacteria and parasites could act as nuclei that accelerate the precipitation of calcium bilirubinate. The mucin gel in the gallbladder can trap these complex precipitates and facilitate their growth into macroscopic stones.

fore, cholecystectomy should be offered to patients only after biliary symptoms develop. Depending on the patient, a reasonable alternative approach may be to observe the pattern of pain before deciding on therapy because up to 30% of patients with one episode of biliary pain do not have a recurrent episode. This approach is particularly useful in patients with a high operative risk.

STONES IN PATIENTS WITH DIABETES MELLITUS

Diabetic patients with incidental cholelithiasis were long considered to have an increased risk of serious complications even when the gallstones were asymptomatic. Subsequent studies have shown that the natural history of gallstones in diabetic patients follows the same pattern observed in nondiabetic persons. A prospective study of patients with insulin-resistant diabetes mellitus showed that after five years of follow-up, symptoms had developed in 15% of the asymptomatic patients.185 This frequency is roughly the same as that reported for nondiabetic patients. Moreover, the complication and mortality rates were comparable to those in studies of nondiabetic patients with gallstones. Therefore, prophylactic cholecystectomy is generally not recommended in patients with insulin-resistant diabetes mellitus and asymptomatic gallstones.

SYMPTOMATIC STONES

The natural history of symptomatic gallstones has a more aggressive course than that of asymptomatic stones. The U.S. National Cooperative Gallstone Study showed that in persons who had an episode of uncomplicated biliary pain in the year before entering the study, the rate of recurrent biliary pain was 38% per year.182 Other investigators have reported a rate of recurrent biliary pain as high as 50% per year in persons with symptomatic gallstones.183 As noted earlier, biliary complications also are more likely to develop in persons with symptomatic gallstones. The risk of biliary complications is estimated to be 1% to 2% per year and is believed to remain relatively constant over time.184 There-

DIAGNOSIS AND CLINICAL DISODERS The clinical manifestations of gallstones are shown schematically in Figure 65-7 and summarized in more detail in Table 65-2.186-190 Biliary pancreatitis is discussed in Chapter 58. Although the standard approach to asymptomatic gallstones is observation, some patients with asymptomatic gallstones may be at increased risk of complications and may require special consideration. An increased risk of cholangiocarcinoma and gallbladder carcinoma has been associated with certain disorders of

1105

1106

Section VIII  Biliary Tract Stone intermittently obstructing cystic duct, causing intermittent biliary pain (20%)

Stone impacted in cystic duct, causing acute cholecystitis (10%) 3

2

4

Asymptomatic stone (75%)

** * *

1

*

7

*

Long-standing cholelithiasis, resulting in gallbladder carcinoma (100% because patients with acute cholecystitis generally have had prior episodes of biliary pain.)

Stone in the cystic duct compressing or fistulizing into the bile duct, causing Mirizzi’s syndrome (10 mg/ dL suggests malignant obstruction or coexisting hemolysis A transient “spike” in serum aminotransferase or amylase (or lipase) levels suggests the passage of a stone

Leukocytosis in 80%, but remainder may have normal white blood cell count with or without band forms Serum bilirubin level >2 mg/ dL in 80% Serum alkaline phosphatase level is usually elevated Blood cultures are usually positive, especially during chills or fever spike; two organisms are grown in cultures from one half of patients

ERCP Endoscopic ultrasonography Magnetic resonance cholangiography Percutaneous THC

ERCP Percutaneous THC

Natural history is not well defined, but complications are more common and more severe than for asymptomatic stones in the gallbladder

High mortality rate if unrecognized, with death from septicemia Emergency decompression of the BD (usually by ERCP) improves survival dramatically

Stone removal at the time of ERCP followed in most cases by early laparoscopic cholecystectomy

Emergency ERCP with stone removal or at least biliary decompression Antibiotics to cover gramnegative and possibly anaerobic organisms and Enterococcus spp. Subsequent cholecystectomy

Physical findings

Laboratory findings

Usually normal Elevated serum bilirubin, alkaline phosphatase, or amylase levels suggest coexisting BD stones

Diagnostic studies (see Table 65-3 for details on imaging studies) Natural history

Ultrasonography Oral cholecystography Meltzer-Lyon test (see Chapter 67)

Treatment (see Chapters 66 and 70)

After the initial attack, 30% of patients have no further symptoms Symptoms develop in the remainder at a rate of 6% per year, and severe complications at a rate of 1% to 2% per year Elective laparoscopic cholecystectomy possibly with IOC ERCP for stone removal or BD exploration if IOC shows stones

50% of cases resolve spontaneously in 7-10 days without surgery Left untreated, 10% of cases are complicated by a localized perforation and 1% by a free perforation and peritonitis Laparoscopic cholecystectomy possibly with IOC if feasible; otherwise open cholecystectomy BD exploration or ERCP for stone removal if IOC shows stones

BD, bile duct; ERCP, endoscopic retrograde cholangiopancreatography; IOC, intraoperative cholangiography; MRC, magnetic resonance cholangiography; RUQ, right upper quadrant; THC, transhepatic cholangiography. *See Chapter 58 for a discussion of biliary pancreatitis.

1108

Section VIII  Biliary Tract Table 65-3  Imaging Studies of the Biliary Tract TECHNIQUE

CONDITION TESTED FOR

Ultrasonography

Cholelithiasis

Choledocholithiasis Acute cholecystitis

EUS

Choledocholithiasis

Oral cholecystography*

Cholelithiasis

Cholescintigraphy (hepatobiliary scintigraphy; hydroxyiminodiacetic acid or diisopropyl iminodiacetic acid scan)

Acute cholecystitis

ERCP

Choledocholithiasis

Cholelithiasis MRCP

Choledocholithiasis

CT

Complications of gallstones

FINDINGS/COMMENTS Stones manifest as mobile, dependent echogenic foci within the gallbladder lumen with acoustic shadowing Sludge appears as layering echogenic material without shadows Sensitivity rate >95% for stones >2 mm Specificity rate >95% for stones with acoustic shadows Rarely, a stone-filled gallbladder may be contracted and difficult to see, with a “wall-echoshadow” sign Best single test for stones in the gallbladder Stones are seen in BD in only ≈50% of cases but can be inferred from the finding of a dilated BD (>6 mm diameter), with or without gallstones, in another ≈25% of cases Can confirm, but not exclude, BD stones Ultrasonographic Murphy’s sign (focal gallbladder tenderness under the transducer) has a positive predictive value of >90% in detecting acute cholecystitis when stones are seen Pericholecystic fluid (in the absence of ascites) and gallbladder wall thickening to >4 mm (in the absence of hypoalbuminemia) are nonspecific findings but are suggestive of acute cholecystitis Highly accurate for excluding or confirming stones in the BD Concordance of EUS with the ERCP diagnosis ≈95%; many studies suggest slightly higher sensitivity rates for EUS than for ERCP Specificity rate ≈97% Positive predictive value ≈99%, negative predictive value ≈98%, accuracy rate ≈97% With experienced operators, EUS can be used in lieu of ERCP to exclude BD stones, particularly when the clinical suspicion is low or intermediate Considered for patients with a low-to-moderate clinical probability of choledocholithiasis Stones manifest as mobile filling defects in an opacified gallbladder Sensitivity and specificity rates exceed 90% when the gallbladder is opacified, but nonvisualization occurs in 25% of studies and can result from multiple causes other than stones Opacification of the gallbladder indicates patency of the cystic duct May be useful in the evaluation of acalculous gallbladder diseases such as cholesterolosis and adenomyomatosis (see Chapter 67) Assesses patency of the cystic duct Normal scan shows radioactivity in the gallbladder, BD, and small bowel within 30-60 minutes Positive result is defined as nonvisualization of the gallbladder with preserved hepatic excretion of radionuclide into the BD or small bowel Sensitivity rate is ≈95% and specificity rate is ≈90%, with false-positive results seen in fasted, critically ill patients With cholecystokinin stimulation, gallbladder “ejection fraction” can be determined and may help evaluate patients with acalculous biliary pain (see Chapter 67) Normal scan result virtually excludes acute cholecystitis ERCP is the standard diagnostic test for stones in the BD, with sensitivity and specificity rates of ≈95% Use of ERCP to extract stones (or at least to drain infected bile) is life-saving in severe cholangitis and reduces the need for BD exploration at the time of cholecystectomy Recommended for patients with a high clinical probability of choledocholithiasis When contrast agent flows retrograde into the gallbladder, stones appear as filling defects and can be detected with a sensitivity rate of ≈80%, but ultrasonography remains the mainstay for confirming cholelithiasis Rapid, noninvasive modality that provides detailed bile duct and pancreatic duct images equal to those of ERCP Sensitivity rate ≈93% and specificity rate ≈94%, comparable with those for ERCP Useful for examining nondilated ducts, particularly at the distal portion, which often is not well visualized by ultrasonography Adjacent structures such as liver and pancreas can be examined at the same time Recommended for patients with a low-to-moderate clinical probability of choledocholithiasis Not well suited for detecting uncomplicated stones, but excellent for detecting complications, such as abscess, perforation of the gallbladder or BD, and pancreatitis Spiral CT may prove useful as a noninvasive means of excluding BD stones; some studies suggest improved diagnostic accuracy when CT is combined with an oral cholecystographic contrast agent

BD, bile duct; CT, computed tomography; ERCP, endoscopic retrograde cholangiopancreatography; EUS, endoscopic ultrasonography; MRCP, magnetic resonance cholangiopancreatography. *Performed infrequently nowadays.

Chapter 65  Gallstone Disease

GB

Stone

A

Acoustic shadowing

GB

Stones

B

Acoustic shadowing

Figure 65-8.  A, Typical ultrasonographic appearance of cholelithiasis. A gallstone is present within the lumen of the gallbladder (GB), casting an acoustic shadow. With repositioning of the patient, the stones will move, thereby excluding the possibility of a gallbladder polyp. B, Cholelithiasis in the setting of acute cholecystitis. Multiple gallstones can be seen within the gallbladder lumen with associated acoustic shadowing. In addition, the gallbladder wall is thickened (arrowheads). (Courtesy of Julie Champine, MD, Dallas, Tex.)

additional advantage of being portable and thus available at the bedside of a critically ill patient.194 The diagnosis of gallstones relies on the detection of echogenic objects within the lumen of the gallbladder that produce an acoustic shadow (Fig. 65-8A). The stones are mobile and generally congregate in the dependent portion of the gallbladder. Modern ultrasonography is able to detect stones as small as 2 mm in diameter routinely. Smaller stones may be missed or may be confused with biliary sludge (layering echogenic material that does not cast acoustic shadows).197 The sensitivity of ultrasonography for the detection of gallstones in the gallbladder is more than 95% for stones larger than 2 mm.198 The specificity is greater than 95% when stones produce acoustic shadows. Rarely, advanced scarring and contraction of the gallbladder around gallstones make it impossible to locate the gallbladder or the stones; this finding should also raise the possibility of gallbladder cancer. The contracted gallbladder filled with stones may give a “double-arc shadow” or “wall-echo shadow” sign, with the gallbladder wall, echogenic stones, and acoustic shadowing seen in immediate proximity. If the

gallbladder cannot be identified ultrasonographically, then a complementary imaging modality such as oral cholecystography or abdominal computed tomography (CT) is warranted. Ultrasonography is the standard for the diagnosis of stones in the gallbladder but is distinctly less sensitive for the detection of stones in the bile duct (previously termed common bile duct).199 Because of the proximity of the distal bile duct to the duodenum, luminal bowel gas often interferes with the ultrasonographic image, and the entire length of the bile duct cannot be examined.200 As a result, only approximately 50% of bile duct stones are actually seen on ultrasonography.194 The presence of an obstructing bile duct stone, however, can be inferred when a dilated duct is found. Now that endoscopic retrograde cholangiopancreatography (ERCP) has uncovered a rising frequency of falsely negative ultrasonograms, the upper limit of normal of the diameter of the bile duct has declined from 10 mm to 6 mm. Even so, inferring choledocholithiasis from a dilated bile duct on ultrasonography has a sensitivity of only 75%. Finally, ultrasonography is quite useful for diagnosing acute cholecystitis.201 Pericholecystic fluid (in the absence of ascites) and gallbladder wall thickening to more than 4 mm (in the absence of hypoalbuminemia) are suggestive of acute cholecystitis (see Fig. 65-8B). Unfortunately, in the critical care setting, these nonspecific findings are seen frequently in patients with no other evidence of gallbladder disease.201 A more specific finding is the so-called sonographic Murphy’s sign, in which the ultrasonographer elicits focal gallbladder tenderness under the ultrasound transducer. Eliciting a sonographic Murphy’s sign is somewhat operator dependent and requires an alert patient. Presence of the sign has a positive predictive value of greater than 90% for detecting acute cholecystitis if gallstones are present.202 Because it provides accurate anatomic localization of biliary tract abnormalities, ultrasonography may help localize other abdominal diseases, such as abscesses or pseudocysts, that may be in the differential diagnosis.

Endoscopic Ultrasonography

Endoscopic ultrasonography (EUS) is highly accurate for detecting choledocholithiasis. Inherently more invasive and more expensive than standard ultrasonography, EUS has the advantage of being able to visualize the bile duct from within the gastrointestinal lumen and is reported to be comparable to ERCP in this respect. Intraluminal imaging provides several advantages over transabdominal ultrasonography, including closer proximity to the bile duct, higher resolution, and lack of interference by bowel gas or abdo­minal wall layers (Fig. 65-9). In several studies, EUS had a positive predictive value of 99%, a negative predictive value of 98%, and an accuracy rate of 97% for the diagnosis of bile duct stones compared with ERCP.203,204 If bile duct stones are found on EUS, endoscopic removal of the stones is necessary, and it can be argued that ERCP should be the initial study if choledocholithiasis is strongly suspected. Nonetheless, several studies that compared EUS with ERCP have found both techniques to be accurate for confirming or excluding choledocholithiasis, with EUS having advantages in both safety and cost.205-207 EUS also has been found to be superior to magnetic resonance cholangiopancreatography (MRCP) (or simply magnetic resonance cholangiography [MRC]) in detecting the presence or absence of bile duct stones (see later). The major benefit of EUS in patients with a clinical suspicion of choledo­cholithiasis is the ability to avoid unnecessary ERCP and sphincterotomy, which is not without risk. There-

1109

1110

Section VIII  Biliary Tract

GB

5 min.

10 min.

15 min.

20 min.

25 min.

30 min.

BD

BD PD PV

Distal BD Conf

Stone

Figure 65-9.  Endoscopic ultrasonography, with a radial sector scanning endoscope, demonstrating choledocholithiasis. The bile duct (BD) is shown extending to the level of the gallbladder (GB) (top) and distally (middle and bottom pictures). The greatest diameter of the BD is 12 mm (middle), and the duct tapers distally to a diameter of 7 mm (bottom). Within the distal BD a gallstone is clearly visualized (bottom). Note the proximity of adjacent structures to the bile duct and the ease with which these structures are resolved by endoscopic ultrasonography. Conf, confluence of the portal and splenic veins; PD, pancreatic duct; PV, portal vein.

fore, EUS is currently considered an appropriate modality for excluding bile duct stones, especially if the pretest probability of finding stones is low to intermediate.

Oral Cholecystography

Once the mainstay of imaging studies of the gallbladder, oral cholecystography (OCG) now has limited application as a secondary approach to identifying stones in the gall-

35 min. 45 min. 60 min. Figure 65-10.  Cholescintigraphy demonstrating an obstructed cystic duct characteristic of acute cholecystitis. The gamma-emitting radioisotope diisopropyl iminodiacetic acid (DISIDA) is injected intravenously, rapidly (at 5 min) taken up by the liver, and excreted into bile (at 20 min). Sequential images show the isotope quickly entering the duodenum (at 45 min) and passing distally in the small intestine without ever being concentrated in the gallbladder. The failure of the gallbladder to be visualized as a hot spot within 30 to 60 minutes constitutes a positive result and implies obstruction of the cystic duct.

bladder.194 The ease, reliability, and rapidity with which it can detect stones, along with the absence of ionizing radiation, have made ultrasonography the imaging study of choice. In unusual cases in which the gallbladder cannot be identified ultrasonographically (as when the gallbladder is contracted and full of stones),208 OCG is complementary to ultrasonography for demonstrating cholelithiasis. Additionally, when medical dissolution of stones or lithotripsy is being considered, visualization of the gallbladder on OCG excludes cystic duct obstruction (see Chapter 66).209 Because of the time required to complete the test (48 hours), OCG is not useful in patients with suspected acute cholecystitis or other complications of gallstone disease. On occasion, OCG may detect unsuspected disease of the gallbladder, such as adenomyomatosis or cholesterolosis (see Chapter 67).

Cholescintigraphy

Cholescintigraphy (hepatobiliary scintigraphy) is a radionuclide imaging test of the gallbladder and biliary tract that is most useful for evaluating patients with suspected acute cholecystitis. By demonstrating patency of the cystic duct, cholescintigraphy can exclude acute cholecystitis rapidly (within 90 minutes) from the differential diagnosis in a patient who presents with abdominal pain.210,211 The procedure can be performed on an emergency basis in a nonfasting patient after intravenous administration of gamma-emitting 99mTc-labeled hydroxyl iminodiacetic acid (HIDA) or diisopropyl iminodiacetic acid (DISIDA), which is taken up rapidly by the liver and secreted into bile. As shown in Figure 65-10, serial scans after injection normally should show radioactivity in the gallbladder, bile

Chapter 65  Gallstone Disease duct, and small intestine within 30 to 60 minutes.167 In the past, imaging of jaundiced patients with this technique was limited, but use of DISIDA may allow imaging of the biliary tree in a patient with a serum bilirubin value as high as 20 mg/dL. An abnormal or “positive” scan result is defined as nonvisualization of the gallbladder with preserved excretion into the bile duct or small intestine. The accuracy of the test for detecting acute cholecystitis is 92%, superior to that for ultrasonography. False-positive results occur primarily in fasting or critically ill patients, in whom gallbladder motility is decreased. The reduction in gallbladder motility leads to greater water resorption, which results in a gelatinous bile. In critically ill patients, cholestasis and hepatocyte dysfunction result in reduced clearance of radionuclide imaging agents. Although nonvisualization of the gallbladder because of cystic duct obstruction is the hallmark of acute cholecystitis, pericholecystic hepatic uptake of radionuclide is a useful secondary sign.212 In some patients (e.g., those with chronic cholecystitis, liver disease, or choledocholithiasis), imaging of the gallbladder on a radionuclide scan is delayed for several hours, and scanning must be repeated in four or more hours to confirm absence of acute cholecystitis. This delay in visualization of the gallbladder is problematic in the acutely ill patient but has largely been overcome with the administration of intravenous morphine sulfate to patients in whom the gallbladder fails to be visualized within 60 minutes. Morphine raises the pressure within the sphincter of Oddi, thereby leading to the preferential flow of bile into the gallbladder if the cystic duct is not obstructed. Another scan is obtained 30 minutes after injection of morphine, and if the gallbladder is visualized, cystic duct obstruction and, hence, acute cholecystitis, is excluded. The gallbladder may not be visualized in approximately half of critically ill patients even after injection of morphine, thereby leading to falsepositive cholescintigraphy results. Although primarily a tool for evaluating acutely ill patients with suspected acute cholecystitis, cholescintigraphy after administration of CCK may be useful in identi­ fying patients with chronic acalculous biliary pain who are likely to benefit from empirical cholecystectomy (see Chapter 67). An additional important role for cholescintigraphy is the noninvasive and clear detection of bile leakage from the cystic duct as a complication of cholecystectomy (see Chapter 66).213

Endoscopic Retrograde Cholangiopancreatography

Endoscopic retrograde cholangiopancreatography (ERCP) is one of the most effective modalities for detecting choledocholithiasis. The technique of this procedure is discussed in more detail in Chapter 70. Stones within the bile duct appear as filling defects and can be detected with a sensitivity of approximately 95% (Fig. 65-11). Care should be taken to avoid inadvertent injection of air into the biliary tract214 because bubbles may mimic gallstones. The specificity of ERCP for the detection of bile duct stones is approximately 95%. The therapeutic applications of ERCP have revolutionized the treatment of patients with choledocholithiasis215 and other bile duct disorders (see Chapter 70). Because ERCP is invasive and is associated with potential complications such as pancreatitis and postsphincterotomy bleeding, however, the role of ERCP has been challenged by other modalities that exclude choledocholithiasis more safely and accurately in patients in whom the clinical suspicion of choledocholithiasis is not high. As the use of EUS and MRC has increased, the role of ERCP in the diagnosis of choledocholithiasis has changed considerably. A National Institutes

Figure 65-11.  An endoscopic retrograde cholangiogram demonstrating dilatation of the bile duct to 15 mm with multiple filling defects representing choledocholithiasis (arrows). (Courtesy of Steve Burdick, MD, Dallas, Tex.)

GB

Stone

Figure 65-12.  Abdominal computed tomography demonstrating emphy­ sematous cholecystitis with associated cholelithiasis. Pockets of gas (yellow arrow), resulting from a secondary infection with gas-forming organisms, are present within the wall of the gallbladder (GB). (Courtesy of Julie Champine, MD, Dallas, Tex.)

of Health consensus conference has recommended the use of ERCP only when the clinical probability for choledocholithiasis is high (i.e., when the need for therapeutic intervention is likely). For diagnosis of choledocholithiasis alone, EUS and MRC are equal in accuracy to ERCP.216

Computed Tomographic Cholangiography and Magnetic Resonance Cholangiography

In patients with cholelithiasis or choledocholithiasis, CT has been used principally for detecting complications such as pericholecystic fluid in acute cholecystitis, gas in the gallbladder wall suggesting emphysematous cholecystitis, gallbladder perforation, and abscesses (Fig. 65-12). Spiral CT cholangiography (CTC) with use of an oral cholecysto-

1111

1112

Section VIII  Biliary Tract

GB

Stone

BD

obstruction. The term chronic cholecystitis to describe biliary pain should be avoided because it implies the presence of a chronic inflammatory infiltrate that may or may not be present in a given patient. Indeed, the severity and frequency of biliary pain and the pathologic changes in the gallbladder do not correlate significantly.222 The most common histologic changes observed in patients with biliary pain are mild fibrosis of the gallbladder wall with a chronic inflammatory cell infiltrate and an intact mucosa. Recurrent episodes of biliary pain also can be associated with a scarred, shrunken gallbladder and RokitanskyAschoff sinuses (intramural diverticula). Bacteria can be cultured from gallbladder bile or gallstones themselves in about 10% of patients with biliary pain, but bacterial infection is not believed to contribute to the symptoms (see also Chapter 67).

Clinical Features

Figure 65-13.  Magnetic resonance cholangiography demonstrating choledocholithiasis. Within the bile duct (BD) are two filling defects repre­ senting gallstones. GB, gallbladder. (Courtesy of Charles Owen, III, MD, Dallas, Tex.)

graphic contrast agent has been studied for the detection of choledocholithiasis.217,218 Although CTC is still inferior to ERCP imaging for detecting bile duct stones, it may reveal other surrounding pathologic abnormalities.217 MRC is highly useful for imaging the bile duct and detecting gallstones. This modality is especially useful for detecting abnormalities in the most distal extrahepatic portion of the bile duct when the duct is not dilated; this region often is not well visualized by transabdominal ultrasonography.195 With the advent of laparoscopic cholecystectomy, an easy, quick, and, preferably, noninvasive method of excluding bile duct stones is needed. MRC permits the construction of a three-dimensional image of the bile duct with a high sensitivity for detecting bile duct stones (Fig. 65-13).219,220 In a systematic review that compared MRC with diagnostic ERCP for the detection of choledocholithiasis, MRC had a sensitivity of 93% and a specificity of 94%.221 CTC and MRC are noninvasive but, unlike ERCP, have no therapeutic application. They are most useful for excluding choledocholithiasis (either preoperatively or postoperatively) in patients who undergo cholecystectomy and in whom the probability of bile duct stones is believed to be low. ERCP is now reserved for patients with a higher probability of bile duct stones and thus the need for therapeutic intervention.

BILIARY PAIN AND CHRONIC CHOLECYSTITIS

Biliary pain is the most common manifesting symptom of cholelithiasis, and about 75% of patients with symptomatic gallstone disease seek medical attention for episodic abdominal pain. In patients who present with a complication of gallstones, such as acute cholecystitis, a history of recurrent episodes of abdominal pain in the months preceding the complication can often be elicited.

Pathogenesis

Biliary pain (conventionally referred to as biliary “colic,” a misnomer) is caused by intermittent obstruction of the cystic duct by one or more gallstones. Biliary pain does not require that inflammation of the gallbladder accompany the

Biliary pain is visceral in nature and, thus, poorly localized.223 In a typical case, the patient experiences episodes of upper abdominal pain, usually in the epigastrium or right upper quadrant (RUQ) but sometimes in other abdominal locations. Ingestion of a meal often can precipitate pain, but more commonly no inciting event is apparent. The onset of biliary pain is more likely to occur during periods of weight reduction and marked physical inactivity such as prolonged bed rest than at other times. The term biliary colic, used in the past, is a misnomer because the pain is steady rather than intermittent as would be suggested by the word “colic.” The pain increases gradually over a period of 15 minutes to an hour and then remains at a plateau for an hour or more before slowly resolving. In one third of patients, the onset of pain may be more sudden, and on rare occasions, the pain may cease abruptly. Pain lasting more than six hours suggests acute cholecystitis rather than simple biliary pain. In order of decreasing frequency, biliary pain is felt maximally in the epigastrium, RUQ, left upper quadrant, and various parts of the precordium or lower abdomen. Therefore, the notion that pain not located in the RUQ is atypical of gallstone disease is incorrect. Radiation of the pain to the scapula, right shoulder, or lower abdomen occurs in one half of patients. Diaphoresis and nausea with some vomiting are common, although vomiting is not as protracted as in intestinal obstruction or acute pancreatitis. Like patients with other kinds of visceral pain, the patient with biliary pain is usually restless and active during an episode. Complaints of gas, bloating, flatulence, and dyspepsia, which are common in patients with gallstones, are probably not related to the stones themselves. These nonspecific symptoms are found with similar frequencies in persons without gallstones. Accordingly, patients with gallstones whose only symptoms are dyspepsia and other nonspecific upper gastrointestinal tract complaints are not candidates for cholecystectomy. Physical findings are usually normal, with only mild to moderate gallbladder tenderness during an attack and perhaps mild residual tenderness lasting several days after an attack.

Natural History

Biliary pain is cause for concern but not alarm. Approximately 30% of patients who have an attack of classic biliary pain will experience no additional attacks over the next 24 months. Therefore, a reasonable approach would be to offer cholecystectomy to patients with recurring episodes of biliary pain.224 In the remaining 70%, the frequency of recurrent attacks varies widely from patient to patient, but the pattern remains relatively constant for an individual

Chapter 65  Gallstone Disease patient over time. In patients monitored after an initial attack of biliary pain, symptoms sufficient to warrant cholecystectomy develop on average at a rate of approximately 6% per year. The cumulative risk that symptoms that require therapy will develop in asymptomatic persons with gallstones who are followed up for five years is 7.6%.225 The probability that a patient with a history of biliary pain will experience a complication of gallstones that requires urgent surgical intervention is only 1% to 2% per year.224

Diagnosis

In a patient with uncomplicated biliary pain, laboratory parameters are usually normal. Elevations of serum bilirubin, alkaline phosphatase, or amylase levels suggest coexisting choledocholithiasis. In general, the first, and often the only, imaging study recommended in patients with biliary pain is ultrasound of the RUQ. Ultrasonography is rapid, noninvasive, highly sensitive, and highly specific for detecting stones in the gallbladder. Despite the impressive diagnostic accuracy of ultrasonography, a clinically important stone is occasionally missed and the correct diagnosis delayed because of the large number of patients who undergo ultrasonography for any reason.195 Given the relatively benign natural history of biliary pain, patients with suspected gallstones but a negative ultrasonography result can safely be observed, further diagnostic testing being reserved for those in whom symptoms recur.226 Oral cholecystography is generally viewed as a secondary imaging study of the gallbladder and is reserved for patients in whom medical dissolution therapy or lithotripsy of gallstones is planned (see Chapter 66). In such cases, patency of the cystic duct must be confirmed by OCG before therapy. On rare occasions, OCG may demonstrate small floating gallstones that were missed by ultrasonography.

ciated with abdominal pain, RUQ tenderness, fever, and leukocytosis is the hallmark of acute cholecystitis. In approximately 90% of cases, the underlying cause is obstruction of the outlet of the gallbladder by a gallstone in the cystic duct, the gallbladder neck, or Hartman’s pouch.229 In the remaining 10% of cases, cholecystitis occurs in the absence of gallstones (acalculous cholecystitis; see Chapter 67). Acute cholecystitis caused by gallstones is a disease of young, otherwise healthy women and generally has a favorable prognosis, whereas acute acalculous cholecystitis occurs more commonly in critically ill elderly men and is associated with high morbidity and mortality rates.

Pathogenesis

Patients with recurrent, uncomplicated biliary pain and documented gallstones are generally treated with elective laparoscopic cholecystectomy, as discussed in Chapter 66. Acute biliary pain improves with administration of meperidine, with or without ketorolac or diclofenac. Aspirin taken prophylactically has been reported to prevent gallstone formation as well as acute attacks of biliary pain in patients with gallstones, but long-term use of NSAIDs does not prevent gallstone formation.227,228

Acute cholecystitis generally occurs when a stone becomes embedded in the cystic duct and causes chronic obstruction, rather than transient obstruction as in biliary pain.229 Stasis of bile within the gallbladder lumen results in damage of the gallbladder mucosa with consequent release of intracellular enzymes and activation of a cascade of inflammatory mediators. In animal studies, if the cystic duct is ligated, the usual result is gradual absorption of the gallbladder contents without the development of inflammation230; the additional instillation of a luminal irritant, such as concentrated bile or lysolecithin, or trauma from an indwelling catheter is required to cause acute cholecystitis in an obstructed gallbladder. Phospholipase A is believed to be released by gallstoneinduced mucosal trauma and converts lecithin to lysolecithin. Although normally absent from gallbladder bile, lysolecithin is present in the gallbladder contents of patients with acute cholecystitis.231 In animal models, installation of lysolecithin into the gallbladder produces acute cholecystitis associated with increased protein secretion, decreased water absorption, and evidence of white blood cell (WBC) invasion associated with elevated production of prostaglandins E and F1α. Administration of indomethacin, a cyclooxygenase inhibitor, has been shown to block this inflammatory response. Studies of human tissue obtained at cholecystectomy have demonstrated enhanced prostaglandin production in the inflamed gallbladder. Additionally, administration of intravenous indomethacin and oral ibuprofen to patients with acute cholecystitis has been shown to diminish both luminal pressure in the gallbladder and pain.231 Supporting evidence for the role of prostaglandins in the development of acute cholecystitis comes from a prospective study in which patients who presented with biliary pain were randomized to receive diclofenac, a prostaglandin synthetase inhibitor, or placebo.232 Ultimately, acute cholecystitis developed in 9 of 40 patients who received placebo, whereas episodes of biliary pain resolved in all 20 patients who received diclofenac. These data suggest a chain of events in which obstruction of the cystic duct in association with one or more intraluminal factors damages the gallbladder mucosa and stimulates prostaglandin synthetase. The resulting fluid secretion and inflammatory changes promote a cycle of further mucosal damage and inflammation.232 Enteric bacteria can be cultured from gallbladder bile in approximately one half of patients with acute cholecystitis.233 Bacteria are not believed, however, to trigger the actual onset of acute cholecystitis.

ACUTE CHOLECYSTITIS

Pathology

Differential Diagnosis

The differential diagnosis of recurrent, episodic upper abdominal symptoms includes reflux esophagitis, peptic ulcer, pancreatitis, renal colic, diverticulitis, carcinoma of the colon, irritable bowel syndrome, radiculopathy, and angina pectoris. Usually, a carefully taken history assists in narrowing the differential diagnosis. For example, relief of pain with food, antacids, or antisecretory drugs suggests peptic ulcer, whereas pain of a cramping nature suggests an intestinal disorder. The pain of angina pectoris usually is precipitated by exercise and does not last for hours, and the pain of renal stones usually is associated with abnormal findings on urinalysis. Like biliary pain, irritable bowel syndrome is common in young women, but pain is associated with altered bowel habits. The pain of shingles or a radiculopathy from osteoarthritis occasionally may resemble biliary pain.

Treatment

Acute cholecystitis is the most common complication of gallstone disease. Inflammation of the gallbladder wall asso-

If examined in the first few days of an attack of acute cholecystitis, the gallbladder usually is distended and contains

1113

1114

Section VIII  Biliary Tract a stone embedded in the cystic duct.234 After the gallbladder is opened, inflammatory exudate and, rarely, pus are present. Later in the attack, the bile pigments that are normally present are absorbed, having been replaced by thin mucoid fluid, pus, or blood. If the attack of acute cholecystitis is left untreated for a long period but the cystic duct remains obstructed, the lumen of the gallbladder may become distended with clear mucoid fluid, a condition known as hydrops of the gallbladder. Histologic changes range from mild acute inflammation with edema to necrosis and perforation of the gallbladder wall. Surprisingly, the severity of histologic changes correlates little with the patient’s symptoms.234 If the gallbladder is resected for acute cholecystitis and no stones are found, the specimen should be carefully examined histologically for evidence of vasculitis or cholesterol emboli because these systemic disorders may manifest as acalculous cholecystitis (see Chapter 35).

Clinical Features

Approximately 75% of patients with acute cholecystitis report prior attacks of biliary pain (see Table 65-2).235 Often, such a patient is alerted to the possibility that more than simple biliary pain is occurring by the prolonged duration of the pain. If biliary pain has been constant for more than six hours, uncomplicated biliary pain is increasingly unlikely, and acute cholecystitis should be suspected. As inflammation in the gallbladder wall progresses, poorly localized visceral pain gives way to moderately severe parietal pain that localizes to the RUQ.235 Less commonly the back or rarely the chest may be the site of maximal pain. Nausea with some vomiting is characteristic of acute cholecystitis, but these symptoms almost invariably follow, rather than precede, the onset of pain. Vomiting is not as persistent or as severe as that with intestinal obstruction or acute pancreatitis. In some patients, the symptoms of acute cholecystitis are nonspecific, with only a mild ache and anorexia, whereas other patients may present with toxic manifestations, including fever and severe RUQ pain with guarding and localized rebound tenderness. In contrast to uncomplicated biliary pain, the physical findings can, in many cases, suggest the diagnosis of acute cholecystitis. Fever is common, but body temperature is usually less than 102°F unless the gallbladder has become gangrenous or has perforated (Fig. 65-14). Mild jaundice is present in 20% of patients with acute cholecystitis and 40% of elderly patients. Serum bilirubin levels usually are less than 4 mg/dL.236 Bilirubin levels above this value suggest the possibility of bile duct stones, which may be found in 50% of jaundiced patients with acute cholecystitis. Another cause of pronounced jaundice in patients with acute cholecystitis is Mirizzi’s syndrome, which is associated with inflammatory obstruction of the common hepatic duct (see later). The abdominal examination often demonstrates right subcostal tenderness, with a palpable gallbladder in one third of patients. A palpable gallbladder is more common in patients having a first attack of acute cholecystitis because repeated attacks usually result in a scarred, fibrotic gallbladder that is unable to distend. For unclear reasons, the gallbladder is usually palpable lateral to its normal anatomic location. A relatively specific finding of acute cholecystitis is Murphy’s sign.235 During palpation in the right subcostal region, pain and inspiratory arrest may occur when the patient takes a deep breath that brings the inflamed gallbladder into

GB Fluid collection

Figure 65-14.  Ultrasonography demonstrating a complex fluid collection adjacent to the gallbladder (GB), consistent with gallbladder perforation. (Courtesy of Julie Champine, MD, Dallas, Tex.)

contact with the examiner’s hand. The presence of Murphy’s sign in the appropriate clinical setting is a reliable predictor of acute cholecystitis, although gallstones should still be confirmed by ultrasonography.

Natural History

The pain of untreated acute cholecystitis generally resolves in 7 to 10 days.237 Not uncommonly, symptoms remit within 48 hours of hospitalization. One study has shown that acute cholecystitis resolves without complications in approximately 83% of patients but results in gangrenous cholecystitis in 7%, gallbladder empyema in 6%, perforation in 3%, and emphysematous cholecystitis in fewer than 1%.238

Diagnosis

Perhaps because it is so common, acute cholecystitis is often at the top of the differential diagnosis of abdominal symptoms and is actually overdiagnosed when clinical criteria alone are considered. In a prospective series of 100 patients with RUQ pain and tenderness and suspected acute cholecystitis, this diagnosis was correct in only two thirds of cases. The clinician must therefore use laboratory and imaging studies to confirm the presence of acute chole­ cystitis, exclude complications such as gangrene and per­ foration, and look for alternative causes of the clinical findings. Table 65-3 details the most common laboratory findings in acute cholecystitis.237 Leukocytosis with a shift to immature neutrophils is common. Because a diagnosis of bile duct stones with cholangitis usually is in the differential diagnosis, attention is directed to results of liver biochemical tests.236 Even without any detectable bile duct obstruction, acute cholecystitis often causes mild elevations in serum aminotransferase and alkaline phosphatase levels. As noted earlier, the serum bilirubin level may also be mildly elevated (2 to 4 mg/dL), and even serum amylase and lipase values may be elevated nonspecifically. A serum bilirubin value greater than 4 mg/dL or amylase value greater than 1000 U/L usually indicates coexisting bile duct obstruction or acute pancreatitis and warrants further evaluation. When the level of leukocytosis exceeds 15,000 cells/mm3, particularly in the setting of worsening pain, high fever (temperature > 102°F), and chills, suppurative cholecystitis

Chapter 65  Gallstone Disease (empyema) or perforation should be suspected, and urgent surgical intervention may be required. Such advanced gallbladder disease may be present even if local and systemic manifestations are unimpressive. Ultrasonography is the single most useful imaging study in acutely ill patients with RUQ pain and tenderness. It accurately establishes the presence or absence of gallstones and serves as an extension of the physical examination. Presence of sonographic Murphy’s sign, defined as focal gallbladder tenderness under the transducer, has a positive predictive value greater than 90% for detecting acute cholecystitis if gallstones are also present, the operator is skillful, and the patient is alert.239 Additionally, ultraso­ nography can detect nonspecific findings suggestive of acute cholecystitis, such as pericholecystic fluid and gallbladder wall thickening greater than 4 mm. Both findings lose specificity for acute cholecystitis if the patient has ascites or hypoalbuminemia.195,240 Because the prevalence of gallstones is high in the population, many patients with nonbiliary tract diseases that manifest as acute abdominal pain (such as acute pancreatitis and complications of peptic ulcer) may have incidental and clinically irrelevant gallstones. The greatest usefulness of cholescintigraphy in these patients is its ability to exclude acute cholecystitis and allow the clinician to focus on nonbiliary causes of the patient’s acute abdominal pain.188 A normal cholescintigraphy scan result shows radioactivity in the gallbladder, bile duct, and small intestine within 30 to 60 minutes of injection of the isotope. With rare exceptions, a normal result excludes acute cholecystitis caused by gallstones. Several studies have suggested that the sensitivity and specificity of scintigraphy in the setting of acute cholecystitis are approximately 94% each. Its sensitivity and specificity are reduced considerably, however, in patients who have liver disease, are receiving parenteral nutrition, or are fasting. These conditions can lead to a false-positive scan result, defined as the absence of isotope in the gallbladder in a patient who does not have acute cholecystitis. If a positive scan result is defined as the absence of isotope in the gallbladder, then a false-negative scan result would be defined as filling of the gallbladder with isotope in the setting of acute cholecystitis, a situation that virtually never occurs. Therefore, scintigraphy should not be used as the initial imaging study in a patient with suspected cholecystitis but rather should be used as a secondary imaging study in patients who already are known to have gallstones and in whom a nonbiliary cause of acute abdominal pain is possible.241 In an effort to reduce the occurrence of false-positive cholescintigraphy scan results, morphine can be administered to the patient if the gallbladder has failed to be visualized after 60 minutes. As discussed earlier, by increasing pressure within the sphincter of Oddi, administration of morphine directs bile into the gallbladder unless the cystic duct is obstructed. Additional scans obtained 30 minutes after the injection of morphine occasionally show filling of the gallbladder with isotope, thereby excluding cystic duct obstruction. Unfortunately, despite the use of morphine augmentation, cholescintigraphy continues to have a falsepositive rate of 60% in critically ill patients. The greatest usefulness of abdominal CT in patients with acute cholecystitis is to detect complications such as emphysematous cholecystitis and perforation of the gallbladder. At the same time, CT can exclude other intraabdominal processes that may engender a similar clinical picture. For example, abdominal CT is highly sensitive for detecting pneumoperitoneum, acute pancreatitis, pancre-

atic pseudocysts, hepatic or intra-abdominal abscesses, appendicitis, and obstruction or perforation of a hollow viscus. An abdominal CT scan usually is not warranted in patients with obvious acute cholecystitis, but if the diagnosis is uncertain or the optimal timing of surgery is in doubt, CT may be invaluable.

Differential Diagnosis

The principal conditions to consider in the differential diagnosis of acute cholecystitis are appendicitis, acute pancreatitis, pyelonephritis or renal calculi, peptic ulcer, acute hepatitis, pneumonia, hepatic abscess or tumor, and gonococcal or chlamydial perihepatitis. These possibi­ lities should be considered before a cholecystectomy is recommended. Acute appendicitis is the disease most often confused with acute cholecystitis because the initial diagnostic impression is based largely on localized right-sided abdo­ minal tenderness, which may be lower than expected in cholecystitis or higher than expected in appendicitis. In general, fever, leukocytosis, and tenderness progress more inexorably in appendicitis. Abdominal CT usually can distinguish these two entities (see Chapter 116). Acute pancreatitis also may be difficult to distinguish from acute cholecystitis on the basis of the history and physical examination alone. Generally, vomiting is a more prominent feature of acute pancreatitis, and affected persons are more uncomfortable in a supine position. Hyperamylasemia is more profound in pancreatitis than cholecystitis, and an elevated serum lipase value is more specific (see Chapter 58). Diseases of the right kidney may produce pain and tenderness that mimic findings in acute cholecystitis, but urinalysis and ultrasonography can usually differentiate renal disease from cholecystitis. The pain of an uncomplicated peptic ulcer is usually chronic and seldom confused with that of acute cholecystitis, but a perforated ulcer, at least initially, may mimic severe acute cholecystitis. Signs of generalized peritonitis or pneumoperitoneum strongly suggest a perforated viscus, necessitating emergency laparotomy (see Chapters 10 and 52). Pneumonia with pleurisy may cause abdominal pain and tenderness, but the pleuritic nature of the pain and the chest radiograph findings should suggest the correct diagnosis. In some instances, acute hepatitis, especially when caused by alcohol, may manifest as severe RUQ pain and tenderness, fever, and leukocytosis and may be confused diagnostically with acute cholecystitis. In such cases, careful assessment of the liver biochemical values over time in combination with ultrasonography or cholescintigraphy may exclude a diagnosis of acute cholecystitis. Rarely, liver biopsy may be warranted (see Chapters 77 to 81 and 84). Gonococcal perihepatitis (Fitz-Hugh–Curtis syndrome) produces RUQ pain and tenderness, which often overshadow any pelvic complaints, as well as leukocytosis. Nevertheless, adnexal tenderness is found on physical examination, and a Gram stain of the cervical smear should show gonococci (see Chapter 37). Hepatic abscesses and tumors usually can be differentiated from acute cholecystitis on the basis of ultrasonographic findings. Prior undiagnosed gallbladder perforation may manifest with fever from a subhepatic abscess. Pseudolithiasis due to ceftriaxone therapy, most often in children, has caused symptoms resembling those of acute cholecystitis, although the gallbladder is his­ tologically normal (see Chapters 82 and 94).

1115

1116

Section VIII  Biliary Tract Treatment

The patient in whom acute cholecystitis is suspected should be hospitalized. The patient is often hypovolemic from vomiting and poor oral intake, and fluid and electrolytes should be administered intravenously. Oral feeding should be withheld and a nasogastric tube inserted if the patient has a distended abdomen or is vomiting persistently. In uncomplicated cases of acute cholecystitis, antibiotics need not be given. Antibiotics are warranted if the patient appears particularly toxic or a complication such as perforation of the gallbladder or emphysematous cholecystitis is suspected. Antibiotics that cover gram-negative enteric bacteria are effective. Coverage with a single agent such as cefoxitin is appropriate in mild cases, but more severely ill patients should receive broad-spectrum coverage with ampicillin and an aminoglycoside or with a thirdgeneration cephalosporin and metronidazole. Definitive therapy of acute cholecystitis consists of cholecystectomy. The safety and effectiveness of using a laparoscopic approach in the setting of acute cholecystitis have been demonstrated (see Chapter 66).242

CHOLEDOCHOLITHIASIS

Choledocholithiasis is defined as the occurrence of stones in the bile ducts. Like stones in the gallbladder, stones in the bile ducts may remain asymptomatic for years, and stones from the bile duct are known to pass silently into the duodenum, perhaps frequently. Unlike stones in the gallbladder, which usually become clinically evident as relatively benign episodes of recurrent biliary pain, stones in the bile duct, when they do cause symptoms, tend to manifest as life-threatening complications such as cholangitis and acute pancreatitis (see Chapter 58). Therefore, discovery of choledocholithiasis generally should be followed by some type of intervention to remove the stones (see Chapter 70).

Etiology

Gallstones may pass from the gallbladder into the bile duct or can form de novo in the duct. Generally, all gallstones from one patient, whether from the gallbladder or bile duct, are of one type, either cholesterol or pigment. Cholesterol stones form only in the gallbladder, and any cholesterol stones found in the bile duct must have migrated there from the gallbladder. Black pigment stones, which are associated with old age, hemolysis, alcoholism, and cirrhosis, also form in the gallbladder and only rarely migrate into the bile duct. The majority of pigment stones in the bile duct are the softer brown pigment stones. These stones form de novo in the bile duct as a result of bacterial action on phospholipid and bilirubin in bile (see earlier).243 They are often found proximal to biliary strictures and are frequently associated with cholangitis. Brown pigment stones are found in patients with hepatolithiasis and recurrent pyogenic cholangitis (see Chapter 68).244 Fifteen percent of patients with gallbladder stones also have bile duct stones. Conversely, of patients with ductal stones, 95% also have gallbladder stones.245 In patients who present with choledocholithiasis months or years after a cholecystectomy, determining whether the stones were overlooked at the earlier operation or have subsequently formed may be impossible. In fact, formation of pigment stones in the bile duct is also a late complication of endoscopic sphincterotomy.246 In a study of the long-term consequences of endoscopic sphincterotomy in more than 400 patients, the cumulative frequency of recurrent bile duct stones was 12%; all the recurrent stones were of the brown pigment type, irrespective of the chemical composition of

the original gallstones. This observation suggests that sphincterotomy permits chronic bacterial colonization of the bile duct that results in deconjugation of bilirubin and precipitation of pigment stones. Stones in the bile duct usually come to rest at the lower end of the ampulla of Vater. Obstruction of the bile duct raises bile pressure proximally and causes the ducts to dilate. Pressure in the bile duct is normally 10 to 15 cm H2O and rises to 25 to 40 cm H2O with complete obstruction. When pressure exceeds 15 cm H2O, bile flow decreases, and at 30 cm H2O, bile flow stops. The bile duct dilates to the point that dilatation can be detected on either ultrasonography or abdominal CT in approximately 75% of cases. In the patient who has had recurrent bouts of cholangitis, the bile duct may become fibrotic and thus unable to dilate. Moreover, dilatation of the duct is sometimes absent in patients with choledocholithiasis because the obstruction is low-grade and intermittent.

Clinical Features

The morbidity of choledocholithiasis stems principally from biliary obstruction, which raises biliary pressure and diminishes bile flow. The rate of onset of obstruction, its extent, and the amount of bacterial contamination of the bile are the major factors that determine the resulting symptoms. Acute obstruction usually causes biliary pain and jaundice, whereas obstruction that develops gradually over several months may manifest initially as pruritus or jaundice alone.247 If bacteria proliferate, life-threatening cholangitis may result (see later). The physical findings are usually normal if obstruction of the bile duct is intermittent. Mild to moderate jaundice may be noted when obstruction has been present for several days to a few weeks. Deep jaundice without pain, particularly with a palpable gallbladder (Courvoisier’s sign), suggests neoplastic obstruction of the bile duct, even when the patient has stones in the gallbladder. With longstanding obstruction, secondary biliary cirrhosis may result, leading to physical findings of chronic liver disease. As shown in Table 65-2, results of laboratory studies may be the only clue to the presence of choledocholithiasis.248 With bile duct obstruction, serum bilirubin and alkaline phosphatase levels both increase. Bilirubin accumulates in serum because of blocked excretion, whereas alkaline phosphatase levels rise because of increased synthesis of the enzyme by the canalicular epithelium. The rise in the alkaline phosphatase level is more rapid than and precedes the rise in bilirubin level.249 The absolute height of the serum bilirubin level is proportional to the extent of obstruction, but the height of the alkaline phosphatase level bears no relation to either the extent of obstruction or its cause. In cases of choledocholithiasis, the serum bilirubin level is typically in the range of 2 to 5 mg/dL205 and rarely exceeds 12 mg/dL. Transient “spikes” in serum aminotransferase or amylase levels suggest passage of a bile duct stone into the duodenum. The overall sensitivity of liver biochemical testing for detecting choledocholithiasis is reported to be 94%; serum levels of gamma glutamyl transpeptidase are elevated most commonly but may not be assessed in clinical practice.249

Natural History

Little information is available on the natural history of asymptomatic bile duct stones. In many patients such stones remain asymptomatic for months or years, but available evidence suggests that the natural history of asymptomatic bile duct stones is less benign than that of asymptomatic gallstones.247,250

Chapter 65  Gallstone Disease Diagnosis

Ultrasonography actually visualizes bile duct stones in only about 50% of cases,199 whereas dilatation of the bile duct to a diameter greater than 6 mm is seen in about 75% of cases. Ultrasonography can confirm, or at least suggest, the presence of bile duct stones but cannot exclude choledocho­ lithiasis definitively. EUS, although clearly more invasive than standard ultrasonography, has the advantage of visualizing the bile duct more accurately. In preliminary studies, EUS has excluded or confirmed choledocholithiasis with sensitivity and specificity rates of approximately 98% as compared with ERCP.203 ERCP is the standard method for the diagnosis and therapy of bile duct stones,251 with sensitivity and speci­ ficity rates of approximately 95%. When the clinical probabi­lity of choledocholithiasis is low, however, less invasive studies, such as EUS and MRCP, should be performed first.216 Percutaneous transhepatic cholangiography (percutaneous THC) is also an accurate test for confirming the presence of choledocholithiasis. The procedure is most readily accomplished when the intrahepatic bile ducts are dilated and now is performed primarily when ERCP is unavailable or has been technically unsuccessful. Laparoscopic ultrasonography may be used in the surgical suite immediately before mobilization of the gallbladder during cholecystectomy. Laparoscopic ultrasono­ graphy may be as accurate as surgical cholangiography in detecting bile duct stones and may thereby obviate the need for the latter.252

Differential Diagnosis

Symptoms caused by obstruction of the bile duct cannot be distinguished from those caused by obstruction of the cystic duct. Therefore, biliary pain is always in the differential diagnosis in patients with an intact gallbladder. The presence of jaundice or abnormal liver biochemical test results strongly points to the bile duct rather than the gallbladder as the source of the pain. In patients who present with jaundice, malignant obstruction of the bile duct or obstruction from a choledochal cyst may be indistinguishable clinically from choledocholithiasis (see Chapters 62 and 69). Acute passive congestion of the liver, associated with cardiac decompensation, may cause intense RUQ pain, tenderness, and even jaundice with serum bilirubin levels higher than 10 mg/dL (see Chapter 83); however, fever is usually absent, and the WBC count is normal or only slightly elevated. The patient typically has other obvious signs of cardiac decompensation. Constrictive pericarditis and cor pulmonale also may cause acute congestion of the liver with only subtle cardiac findings. Acute viral hepatitis rarely may cause severe RUQ pain with tenderness and fever. The WBC count, however, usually is not elevated, whereas serum alanine aminotransferase and aspartate aminotransferase levels are markedly elevated. Acquired immunodeficiency syndrome (AIDS)-associated cholangiopathy253 and papillary stenosis must be considered in human immunodeficiency virus–positive patients with RUQ pain and abnormal liver biochemical test results (see Chapter 33).

Treatment

Because of its propensity to result in serious complications such as cholangitis and acute pancreatitis, choledocho­ lithiasis warrants treatment in nearly all cases.254 The optimal therapy for a given patient depends on the severity

of symptoms, presence of coexisting medical problems, availability of local expertise, and presence or absence of the gallbladder. Bile duct stones discovered at the time of a laparoscopic cholecystectomy present a dilemma to the surgeon. Some surgeons may attempt laparoscopic exploration of the bile duct. In other cases, the operation can be converted to an open cholecystectomy with bile duct exploration, but this approach results in greater morbidity and a more prolonged hospital stay. Alternatively, the laparoscopic cholecystectomy can be carried out as planned, and the patient can return for ERCP with removal of the bile duct stones. Such an approach, if successful, cures the disease but runs the risk of necessitating a third procedure, namely a bile duct exploration, if the stones cannot be removed at ERCP. In general, the greater the expertise of the therapeutic endoscopist, the more inclined the surgeon should be to complete the laparoscopic cholecystectomy and have the bile duct stones removed endoscopically.254 In especially high-risk patients, endoscopic removal of bile duct stones may be performed without cholecystectomy. This approach is particularly appropriate for elderly patients with other severe illnesses.255 Cholecystectomy is required subsequently for recurrent symptoms in only 10% of patients. The surgical management and endoscopic treatment of gallstones are discussed in detail in Chapters 66 and 70, respectively.

CHOLANGITIS

Of all the common complications of gallstones, the most serious and the most lethal is acute bacterial cholangitis. Pus under pressure in the bile ducts leads to rapid spread of bacteria via the liver into the blood, with resulting septicemia. Moreover, the diagnosis of cholangitis is often problematic (especially in the critical early phase of the disease) because clinical features that point to the biliary tract as the source of sepsis are often absent.26 Table 65-2 delineates the symptoms, signs, and laboratory findings that can aid in an early diagnosis of cholangitis.

Etiology and Pathophysiology

In approximately 85% of cases, cholangitis is caused by a stone embedded in the bile duct, with resulting bile stasis.257 Other causes of bile duct obstruction that may result in cholangitis are neoplasms (see Chapters 60 and 69), biliary strictures (see Chapters 68 and 70), parasitic infections (see Chapters 68 and 82), and congenital abnormalities of the bile ducts (see Chapter 62). This discussion deals specifically with cholangitis caused by gallstones in the bile duct. Bile duct obstruction is necessary, but not sufficient, to cause cholangitis. Cholangitis is relatively common in patients with choledocholithiasis and nearly universal in patients with a post-traumatic bile duct stricture but is seen in only 15% of patients with neoplastic obstruction of the bile duct. It is most likely to result when a bile duct that already contains bacteria becomes obstructed, as is the case in most patients with choledocholithiasis and stricture but in few patients with neoplastic obstruction. Malignant obstruction is more often complete than obstruction by a stricture or a bile duct stone and less commonly permits the reflux of bacteria from duodenal contents into the bile ducts.258 The bacterial species most commonly cultured from the bile are E. coli, Klebsiella, Pseudomonas, Proteus, and enterococci. Anaerobic species such as Bacteroides fragilis and Clostridium perfringens are found in about 15% of appropriately cultured bile specimens. Anaerobes usually accompany aerobes, especially E. coli. The shaking chills

1117

1118

Section VIII  Biliary Tract and fever of cholangitis are caused by bacteremia from bile duct organisms. The degree of regurgitation of bacteria from bile into hepatic venous blood is directly proportional to the biliary pressure and, hence, the degree of obstruction.258 For this reason, decompression alone often effectively treats the illness.

Clinical Features

The hallmark of cholangitis is Charcot’s triad, consisting of RUQ pain, jaundice, and fever (see Table 65-2). The full triad is present in only 70% of patients.258 The pain of cholangitis may be surprisingly mild and transient but is often accompanied by chills and rigors. Elderly patients in particular may present solely with mental confusion, lethargy, and delirium. Altered mental status and hypo­ tension in combination with Charcot’s triad, known commonly as Reynolds’ pentad, occur in severe suppurative cholangitis. On physical examination, fever is almost universal, occurring in 95% of patients. RUQ tenderness is elicited in approximately 90% of patients, but jaundice is clinically detectable in only 80%. Notably, peritoneal signs are found in only 15% of patients. The combination of hypo­ tension and mental confusion indicates gram-negative septicemia. In overlooked cases of severe cholangitis, intrahepatic abscess may manifest as a late complication (see Chapter 82). Laboratory study results are often helpful in pointing to the biliary tract as the source of sepsis. In particular, the serum bilirubin level exceeds 2 mg/dL in 80% of patients. When the bilirubin level is normal initially, the diagnosis of cholangitis may not be suspected.249 The WBC count is elevated in 80% of patients. In many patients who have a normal WBC count, examination of the peripheral blood smear reveals a dramatic shift to immature neutrophil forms. The serum alkaline phosphatase level is usually elevated, and the serum amylase level may also be elevated if pancreatitis is also present. In the majority of cases, blood culture results are positive for enteric organisms, especially if culture specimens are obtained during chills and fever spikes. The organism found in the blood is invariably the same as that found in the bile.

Diagnosis

The principles of radiologic diagnosis of cholangitis are the same as those for choledocholithiasis. As shown in Table 65-3, stones in the bile duct are seen ultrasonographically in only about 50% of cases155 but can be inferred by detection of a dilated bile duct in about 75% of cases. Normal ultrasonography findings do not exclude the possibility of choledocholithiasis in a patient in whom the clinical presentation suggests cholangitis.241 An abdominal CT is an excellent test for excluding complications of gallstones such as acute pancreatitis and abscess, but a standard abdominal CT scan is not capable of excluding bile duct stones. EUS and MRC, as noted earlier, have a much higher accuracy rate than CT for detecting and excluding stones in the bile duct. ERCP is the standard test for the diagnosis of bile duct stones and cholangitis. Moreover, the ability of ERCP to establish drainage of infected bile under pressure can be life-saving. If ERCP is unsuccessful, percutaneous THC can be performed (see Chapter 70).

Treatment

In cases of suspected bacterial cholangitis, blood culture specimens should be obtained immediately and therapy

started with antibiotics effective against the likely causative organisms.259 In mild cases, initial therapy with a single drug, such as cefoxitin, 2.0 g intravenously every six to eight hours, is usually sufficient. In severe cases, more intensive therapy (e.g., gentamicin, ampicillin, and metronidazole or a broad-spectrum agent such as piperacillintazobactam 3.375 g intravenously every six hours or, if resistant organisms are suspected, meropenem 1 g intravenously every eight hours) is indicated. The patient’s condition should improve within 6 to 12 hours, and in most cases, the infection comes under control within 2 to 3 days, with defervescence, relief of discomfort, and a decline in the WBC count. In these cases, definitive therapy can be planned on an elective basis. If, however, after 6 to 12 hours of careful observation, the patient’s clinical status declines with worsening fever, pain, mental confusion, or hypotension, the bile duct must be decompressed immediately.259 If available, ERCP with stone extraction, or at least decompression of the bile duct with an intrabiliary stent, is the treatment of choice. Controlled studies in which ERCP and decompression of the bile duct were compared with emergency surgery and bile duct exploration have shown dramatically lower morbidity and mortality rates in patients treated endoscopically.254 The surgical treatment and endoscopic management of cholangitis are discussed in detail in Chapters 66 and 70, respectively.

UNCOMMON COMPLICATIONS Table 65-4 describes the clinical manifestations, diagnosis, and treatment of several uncommon complications of gallstone disease.

EMPHYSEMATOUS CHOLECYSTITIS

Patients who have emphysematous cholecystitis present with the same clinical manifestations as patients with uncomplicated acute cholecystitis, but in the former, gasforming organisms have secondarily infected the gallbladder wall. Pockets of gas are evident in the area of the gallbladder fossa on plain abdominal films, ultrasonog­ raphy, and abdominal CT (see Fig. 65-13).260 Emergency antibiotic therapy with anaerobic coverage and early cholecystectomy are warranted because the risk of gallbladder perforation is high. Emphysematous cholecystitis often occurs in diabetic persons or older men who do not have gallstones, in whom atherosclerosis of the cystic artery with resulting ischemia may be the initiating event (see Chapter 67).

CHOLECYSTOENTERIC FISTULA

A cholecystoenteric fistula occurs when a stone erodes through the gallbladder wall (usually the neck) and into a hollow viscus. The most common entry point into the bowel is the duodenum, followed in frequency by the hepatic flexure of the colon, the stomach, and the jejunum. Symptoms are initially similar to those of acute cholecystitis, although at times the stone may pass into the bowel and may be excreted without causing any symptoms.261 Because the biliary tract is decompressed, cholangitis is not common despite gross seeding of the gallbladder and bile ducts with bacteria. The diagnosis of a cholecystoenteric fistula is suspected from radiographic evidence of pneumobilia and may be confirmed by barium contrast studies of the upper or lower gastrointestinal tract; often the precise anatomic location of the fistula is not identified until surgery.

Chapter 65  Gallstone Disease Table 65-4  Uncommon Complications of Gallstone Disease COMPLICATION

PATHOGENESIS

CLINICAL FEATURES

DIAGNOSIS/TREATMENT

Emphysematous cholecystitis

Secondary infection of the gallbladder wall with gasforming organisms (Clostridium welchii, Escherichia coli, and anaerobic streptococci) More common in elderly, diabetic men; can occur without stones (see Chapter 67) Erosion of a (usually large) stone through the gallbladder wall into adjacent bowel, most often the duodenum, followed in frequency by the hepatic flexure, stomach, and jejunum

Symptoms and signs similar to those of severe acute cholecystitis

Mirizzi’s syndrome

Impacted stone in the gallbladder neck or cystic duct with extrinsic compression of the common hepatic duct from accompanying inflammation or fistula

Jaundice and right upper quadrant pain

Porcelain gallbladder

Intramural calcification of the gallbladder wall, usually in association with stones

No symptoms attributable to the calcified wall per se, but carcinoma of the gallbladder is a late complication in ≈20% (see Chapter 69)

Plain abdominal films may show gas in the gallbladder fossa Ultrasonography and CT are sensitive for confirming gas Treatment with intravenous antibiotics, including anaerobic coverage, and early cholecystectomy High morbidity and mortality rates Plain abdominal films may show gas in the biliary tree and/or a small bowel obstruction in gallstone ileus as well as a stone in the right lower quadrant, if the stone is calcified Contrast gastrointestinal series may demonstrate the fistula A fistula from a solitary stone that passes may close spontaneously Cholecystectomy and bowel closure are curative Gallstone ileus requires emergency laparotomy; the diagnosis is often delayed, with a resulting mortality rate of ≈20% ERCP demonstrates dilated intrahepatic ducts and extrinsic compression of the common hepatic duct and possible fistula Preoperative diagnosis is important to guide surgery and minimize the risk of BD injury Plain abdominal films or CT shows intramural calcification of the gallbladder wall Prophylactic cholecystectomy is indicated to prevent carcinoma

Cholecystoenteric fistula

Symptoms and signs similar to those of acute cholecystitis; although sometimes the fistula may be clinically silent Stones >25 mm, especially in elderly women, may produce a bowel obstruction, or “gallstone ileus”; the terminal ileum is the most common site of obstruction Gastric outlet obstruction (Bouveret’s syndrome) may occur rarely

BD, bile duct; CT, computed tomography; ERCP, endoscopic retrograde cholangiopancreatography.

If the gallstone exceeds 25 mm in diameter, it may manifest, especially in elderly women, as a small intestinal obstruction (gallstone ileus); the ileocecal area is the most common site of obstruction.262 In such cases, a plain abdominal film may show the pathognomonic features of pneumobilia, a dilated small bowel, and a large gallstone in the right lower quadrant. Unfortunately, the diagnosis of a gallstone ileus is often delayed, with a resulting mortality rate of approximately 20%. Bouveret’s syndrome is characterized by gastric outlet obstruction resulting from duodenal impaction of a large gallstone that has migrated through a cholecystoduodenal fistula.263

tion of gallstones but is mentioned here because of the remarkable tendency of carcinoma to develop as a late complication of gallbladder calcification (specifically, a gallbladder with focal rather than diffuse wall calcification).267 The diagnosis of a porcelain gallbladder can be made with a plain abdominal film or abdominal CT, which shows intramural calcification of the gallbladder wall. Prophy­ lactic cholecystectomy, preferably through a laparoscopic approach, is indicated to prevent the subsequent development of carcinoma, which may otherwise occur in up to 20% of cases (see Chapter 69).268

MIRIZZI’S SYNDROME

The authors wish to acknowledge the significant contributions of colleagues in the gallstone field. This work was supported in part by research grants DK54012 and DK73917 (D.Q.-H.W.) from the National Institutes of Health (U.S. Public Health Service).

Mirizzi’s syndrome is a rare complication in which a stone embedded in the neck of the gallbladder or cystic duct extrinsically compresses the common hepatic duct with resulting jaundice, bile duct obstruction, and, in some cases, a fistula.264,265 Typically the gallbladder contracted and contains stones. ERCP usually demonstrates the characteristic extrinsic compression of the common hepatic duct. Treatment is traditionally by an open cholecystectomy, although endoscopic stenting and laparoscopic cholecystectomy have been performed successfully. Preoperative diagnosis of Mirizzi’s syndrome is important so that bile duct injury can be avoided.266

PORCELAIN GALLBLADDER

Strictly speaking, porcelain gallbladder, defined as intramural calcification of the gallbladder wall, is not a complica-

ACKNOWLEDGMENTS

KEY REFERENCES

Buch S, Schafmayer C, Volzke H, et al. A genome-wide association scan identifies the hepatic cholesterol transporter ABCG8 as a suscepti­ bility factor for human gallstone disease. Nat Genet 2007; 39:995-9. (Ref 159.) Buhman KK, Accad M, Novak S, et al. Resistance to diet-induced hypercholesterolemia and gallstone formation in ACAT2-deficient mice. Nat Med 2000; 6:1341-7. (Ref 164.) Collins C, Maguire D, Ireland A, et al. A prospective study of common bile duct calculi in patients undergoing laparoscopic cholecystectomy: Natural history of choledocholithiasis revisited. Ann Surg 2004; 239:28-33. (Ref 250.)

1119

1120

Section VIII  Biliary Tract Konikoff FM, Chung DS, Donovan JM, et al. Filamentous, helical, and tubular microstructures during cholesterol crystallization from bile. Evidence that cholesterol does not nucleate classic monohydrate plates. J Clin Invest 1992; 90:1155-60. (Ref 74.) Maurer KJ, Ihrig MM, Rogers AB, et al. Identification of cholelithogenic enterohepatic helicobacter species and their role in murine cho­ lesterol gallstone formation. Gastroenterology 2005; 128:1023-33. (Ref 122.) Paigen B, Carey MC. Gallstones. New York: Oxford University Press; 2002. p 298. (Ref 3.) Portincasa P, Di Ciaula A, Wang HH, et al. Coordinate regulation of gallbladder motor function in the gut-liver axis. Hepatology 2008; 47:2112-26. (Ref 97.) Wang DQ, Carey MC. Complete mapping of crystallization pathways during cholesterol precipitation from model bile: Influence of physical-chemical variables of pathophysiologic relevance and identi­ fication of a stable liquid crystalline state in cold, dilute and hydrophilic bile salt-containing systems. J Lipid Res 1996; 37:606-30. (Ref 52.) Wang DQ, Paigen B, Carey MC. Phenotypic characterization of Lith genes that determine susceptibility to cholesterol cholelithiasis in inbred mice: Physical-chemistry of gallbladder bile. J Lipid Res 1997; 38:1395-411. (Ref 73.)

Wang DQ, Schmitz F, Kopin AS, et al. Targeted disruption of the murine cholecystokinin-1 receptor promotes intestinal cholesterol absorption and susceptibility to cholesterol cholelithiasis. J Clin Invest 2004; 114:521-8. (Ref 115.) Wang HH, Portincasa P, Wang DQ. Molecular pathophysiology and physical chemistry of cholesterol gallstones. Front Biosci 2008; 13:401-23. (Ref 4.) Wang HH, Afdhal NH, Gendler SJ, et al. Evidence that gallbladder epithelial mucin enhances cholesterol cholelithogenesis in MUC1 transgenic mice. Gastroenterology 2006; 131:210-22. (Ref 89.) Wang HH, Afdhal NH, Wang DQ. Estrogen receptor alpha, but not beta, plays a major role in 17beta-estradiol-induced murine cholesterol gallstones. Gastroenterology 2004; 127:239-49. (Ref 21.) Wang HH, Afdhal NH, Wang DQ. Overexpression of estrogen receptor alpha increases hepatic cholesterogenesis, leading to biliary hypersecretion in mice. J Lipid Res 2006; 47:778-86. (Ref 22.) Yu L, Hammer RE, Li-Hawkins J, et al. Disruption of Abcg5 and Abcg8 in mice reveals their crucial role in biliary cholesterol secretion. Proc Natl Acad Sci U S A 2002; 99:16237-42. (Ref 60.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

66 Treatment of Gallstone Disease Robert E. Glasgow and Sean J. Mulvihill

CHAPTER OUTLINE Medical Treatment  1121 Dissolution Therapy  1121 Extracorporeal Shock-Wave Lithotripsy  1122 Surgical Treatment  1124 Open Cholecystectomy  1125 Laparoscopic Cholecystectomy  1126 Choice of Treatment  1129 Indications for Treatment  1129 Asymptomatic Gallstones  1129 Biliary Pain  1130 Acute Cholecystitis  1130 Gallstone Pancreatitis  1132 Special Problems  1132

Many options are available for the treatment of patients with symptomatic gallstone disease. Improvements in endoscopic, radiologic, and chemical therapies for gallstones have enhanced the overall management of these patients. Nevertheless, surgery remains the most important therapeutic option. Laparoscopic cholecystectomy has become the standard method for the management of patients with biliary pain and complications of gallstone disease, such as acute cholecystitis, gallstone pancreatitis, and choledocholithiasis. In this chapter, treatment of patients with symptomatic and asymptomatic gallstones is discussed, with particular emphasis on indications for medical and surgical treatment, patient selection, and outcomes.

MEDICAL TREATMENT Medical treatment of gallstone disease was first proposed by Schiff in Italy in 1873.1 Dabney of Virginia first reported the effective treatment of gallstones with bile acids in 1876, an observation later confirmed by Rewbridge of Minnesota in 1937.2,3 Despite these initial reports, the use of medical dissolution treatment did not gain acceptance until large clinical series were reported in the 1970s. Contact dissolution of gallstones with solvents and percutaneous cholecystolithotomy techniques also have been reported, but these modalities have not proved superior to oral dissolution, shock-wave lithotripsy, or laparoscopic cholecystectomy and have been abandoned. The mainstay of current nonsurgical treatment of gallstone disease is oral dissolution with ursodeoxycholic acid, with or without extracorporeal shockwave lithotripsy.

Choledocholithiasis  1133 Choledocholithiasis Known Preoperatively  1134 Choledocholithiasis Identified during Cholecystectomy  1134 Choledocholithiasis Identified after Cholecystectomy  1134 Bile Duct Stricture  1134 Postcholecystectomy Syndrome  1135 Choledocholithiasis  1135 Cystic Duct Remnant  1137 Sphincter of Oddi Dysfunction  1137 Gallstones, Cholecystectomy, and Cancer  1137 Biliary Tract Cancer  1137 Colorectal Cancer  1137

Although nonsurgical treatment of gallstones has proved effective in carefully selected patients, only a limited number of patients are candidates for this treatment option. Nonsurgical treatments are effective only in patients with small, cholesterol gallstones. Significant admixtures of pigment or calcium salts make stones indissoluble. In addition, long-term success with medical treatment of gallstones occurs only in patients in whom the lithogenic disturbance that led to gallstone formation is transient. For most patients, gallstone formation represents an imbalance in biliary lipid excretion, gallbladder stasis, or infection of the bile (see Chapter 65). In these patients, successful dissolution is usually followed by recurrence of gallstones in 30% to 50% of patients within five years.4-7 Therefore, the proper choice of treatment must take into account the type and severity of symptoms, the physical characteristics of the stones, gallbladder function, and the characteristics and preference of the patient.

DISSOLUTION THERAPY

The rationale for oral dissolution therapy is the reversal of the condition that led to formation of cholesterol gallstones, namely, the supersaturation of bile with cholesterol (see Chapter 65). Cholesterol stones dissolve if the surrounding medium is capable of solubilizing the cholesterol in the stones. Both chenodeoxycholic acid and ursodeoxycholic acid dissolve gallstones by decreasing biliary cholesterol secretion and desaturating bile. These agents encourage the removal of cholesterol from stones via micellar solubilization, formation of a liquid crystalline phase, or both. Chenodeoxycholic acid was the first bile acid used for gallstone dissolution but has been abandoned because of side effects, including diarrhea and increased serum aminotransferase

1121

1122

Section VIII  Biliary Tract and cholesterol levels. Ursodeoxycholic acid is well tolerated and is currently used in oral dissolution regimens. In randomized comparisons, ursodeoxycholic acid was just as effective as chenodeoxycholic acid alone or in combination with ursodeoxycholic acid.8-10 The rate of stone dissolution is a function of (1) thermodynamic forces, including the degree of bile desaturation and the concentration of ursodeoxycholic acid in bile; (2) kinetic forces, including stirring of bile; and (3) the surfaceto-volume ratio of the stones. Oral dissolution targets the thermodynamic forces.11 Because small stones have a smaller surface-to-volume ratio, they respond more quickly and reliably to oral dissolution therapy. The use of oral dissolution therapy does not address the problem of gallbladder stasis.12 Although prokinetic agents, including alpha-adrenergic antagonists and clarithromycin, have been shown to increase gallbladder motility, their use in preventing and treating gallstones has not been studied.13,14

Patient Selection

Selection of patients for oral dissolution therapy is a function of the stage of gallstone disease, gallbladder function, and the characteristics of the stones. Selection criteria are summarized in Table 66-1. Oral dissolution therapy should be considered for patients with uncomplicated gallstone disease, including those with mild, infrequent biliary pain. Patients with severe or frequent biliary pain and patients with complications of gallstones, including cholecystitis, pancreatitis, and cholangitis, should not be treated with oral dissolution therapy; these patients should be referred for surgery. In addition, the gallbladder must function and the cystic duct must be patent to allow unsaturated bile and stones to clear from the gallbladder. The patency of the cystic duct is usually evaluated by oral cholecystography. More recently, stimulated cholescintigraphy and functional ultrasonography have been used. These latter modalities assess cystic duct patency as well as gallbladder function. The characteristics of the stones play an important role in determining the efficacy of dissolution treatment. Oral dissolution therapy works only on cholesterol stones. Although verifying the composition of gallstones can be difficult, the appearance of stones on plain films or computed tomography (CT) images can be useful. Cholesterol

Table 66-1  Selection Criteria for Oral Bile Acid Dissolution Therapy Stage of gallstone disease Gallbladder function*

Stone characteristics

Symptomatic (biliary pain) without complications Opacification of gallbladder on oral cholecystography (patent cystic duct) Normal result of stimulated cholescintigraphy (normal gallbladder emptying) Normal result of functional ultrasonography (normal gallbladder emptying after a test meal) Radiolucent on radiography Isodense or hypodense to bile and absence of calcification on computed tomography Diameter 4 mm) in the absence of ascites or hypoalbuminemia, (2) sonographic Murphy’s sign (defined as maximum tenderness over the ultrasonographically localized gallbladder), and (3) pericholecystic fluid collection. A thickened gallbladder wall (Fig. 67-1) is not specific for cholecystitis but in the proper clinical setting is suggestive of gallbladder involvement and should prompt further evaluation. A sonographic Murphy’s sign is operator dependent and requires a cooperative patient but, when present, is a reliable indicator of gallbladder inflammation.45 A pericholecystic fluid collection indicates advanced disease. Sensitivity rates of ultrasonography for detecting acute acalculous cholecystitis have been reported to range from 67% to 92%, with specificity rates of more than 90%.44 Investigators have proposed an ultrasonographic scoring system to improve the diagnostic accuracy of ultrasonography in critically

Chapter 67  Acalculous Biliary Pain, Acalculous Cholecystitis, Cholesterolosis, Adenomyomatosis

ill patients.46 Two points are given for distention of the gallbladder or thickening of the gallbladder wall, and 1 point each is given for “striated” thickening (alternating hypoechoic and hyperechoic layers) of the gallbladder wall, sludge, or pericholecystic fluid. Scores of 6 or higher accurately predict acalculous cholecystitis.

pathogenesis of which involves obstruction of the cystic duct by a stone, filling of the gallbladder on scintigraphy virtually excludes cholecystitis as the cause of the patient’s symptoms.48 Hepatobiliary scintigraphy is less precise in acute acalculous cholecystitis. Gallbladder and cystic wall edema can cause an obstructive picture similar to calculous cholecystitis on scintigraphy. Patients with acute acalculous cholecystitis have often fasted for prolonged periods, a state that can result in concentrated, viscous bile that flows poorly through the cystic duct and causes a false-positive hepatobiliary scan result. Most patients with acute acalculous cholecystitis (in contrast to those with calculi) do not have an obstructed cystic duct; hence, hepatobiliary scans can be falsely negative as well.49 The sensitivity of the test may exceed 90%, but the lack of specificity in fasted, critically ill patients limits the usefulness of the test primarily to excluding acute acalculous cholecystitis rather than confirming the diagnosis. A study in which ultrasonography and cholescintigraphy were performed in critically ill patients found cholescintigraphy to be useful for the early diagnosis of acute acalculous cholecystitis, whereas ultrasonography alone did not permit an early decision regarding the need for surgery.50 In an effort to improve the accuracy of biliary scintigraphy, investigators have proposed the use of morphineaugmented cholescintigraphy, in which morphine sulfate is administered intravenously (0.05 to 0.1 mg/kg) to patients in whom the gallbladder has not been visualized on standard cholescintigraphy.51 The rationale for this procedure is that morphine increases resistance to the flow of bile through the sphincter of Oddi and thus forces filling of the gall­ bladder if the cystic duct is patent, thereby reducing the likelihood of a false-positive result. In approximately 60% of critically ill patients with possible biliary tract sepsis and a nonvisualized gallbladder on standard cholescintigraphy, the gallbladder is visualized after morphine augmentation, and, therefore, acute cholecystitis can be excluded as the source of sepsis.

Computed Tomography

TREATMENT

Figure 67-1.  Ultrasonogram demonstrating thickening of the gallbladder wall to 17 mm (denoted by asterisks) characteristic of acute acalculous cholecystitis. Point tenderness was noted when the transducer was pressed onto the abdomen over the gallbladder (sonographic Murphy’s sign). The diagnosis was confirmed at laparotomy. (Courtesy of David Hurst, MD, Dallas, Tex.)

Computed tomography (CT) findings suggestive of cholecystitis include gallbladder wall thickening (>4 mm), pericholecystic fluid, subserosal edema (in the absence of ascites), intramural gas, and sloughed gallbladder mucosa. Sensitivity and specificity rates of these findings for predicting acute acalculous cholecystitis at surgery have been reported to exceed 95%. CT is also superior to ultrasonography in detecting disease elsewhere in the abdomen that could be the cause of a patient’s fever or abdominal pain.47 An obvious disadvantage of CT is that it cannot be performed at the bedside, which is necessary in many critically ill patients. Several investigators have emphasized that CT is complementary to ultrasonography and often detects gallbladder disease in high-risk patients with normal ultrasonographic findings.

Hepatobiliary Scintigraphy

Hepatobiliary scintigraphy may be useful for excluding cystic duct obstruction in patients with clinical features suggestive of acute cholecystitis. Under normal conditions, intravenously administered radionuclide is taken up by the liver, secreted into bile, concentrated in the gallbladder (where it produces a “hot spot” on a scan), and emptied into the duodenum. A positive scan result for cystic duct obstruction is defined as failure of filling of the gallbladder despite the normal passage of radionuclide into the duodenum. In suspected calculous cholecystitis, the

In light of the rapid progression of acute acalculous cholecystitis to gangrene and perforation, early recognition and intervention are required. Supportive medical care should include restoration of hemodynamic stability as well as antibiotic coverage for gram-negative enteric organisms and anaerobes if biliary tract infection is suspected.

Surgical Cholecystectomy and Cholecystostomy

Traditionally, the definitive therapeutic approach for acute acalculous cholecystitis has been urgent laparotomy and cholecystectomy (see Chapter 66). Nowadays, laparoscopic cholecystectomy is the standard approach.52 In patients too unstable to tolerate anesthesia, radiographically guided percutaneous cholecystostomy can be performed53; definitive cholecystectomy can be undertaken when the patient is stable, if necessary.

Percutaneous Cholecystostomy

Several investigators have reported favorable results with the ultrasonographically guided percutaneous transhepatic placement of a cholecystostomy drainage tube, coupled with intravenous administration of antibiotics, as definitive therapy in patients in whom surgery poses a high risk.43,54-55 Studies suggest that most patients with acute acalculous cholecystitis can be treated with percutaneous drainage; if the postdrainage cholangiogram is normal, the catheter can be removed, and cholecystectomy is not necessary.54,56

1143

1144

Section VIII  Biliary Tract Transpapillary Endoscopic Cholecystostomy

Some critically ill patients with suspected acute acalculous cholecystitis are poor candidates for ultrasonographically guided percutaneous cholecystostomy, let alone surgery, because of massive ascites or uncorrectable coagulopathy. Such patients may benefit from an endoscopic approach in which the cystic duct is selectively cannulated during endoscopic retrograde cholangiopancreatography with an obliquely angled guidewire that tracks along the lateral wall of the bile duct and facilitates cannulation of the cystic duct.57 If the wire can negotiate the spiral valves within the cystic duct successfully, a nasobiliary catheter is introduced over the guidewire into the gallbladder, the contents are aspirated, and the gallbladder is lavaged with 1% Nacetylcysteine in saline to dissolve mucus and sludge. The nasocholecystostomy catheter is allowed to drain by gravity for several days and can be easily removed when the patient has recovered and is stable. Studies have shown that successful intubation of the gallbladder can be achieved in 90% of attempts and that drainage and lavage of the viscous black bile and sludge from the gallbladder result in clinical resolution in most of these critically ill patients. The technique is more cumbersome and expensive than ultrasonographic placement of a cholecystostomy tube and should be reserved for patients who would not tolerate a percutaneous approach.58

PREVENTION

Daily stimulation of gallbladder contraction with intravenously administered CCK, 50 ng/kg over 10 minutes, has been shown to prevent the formation of gallbladder sludge in fasting patients receiving total parenteral nutrition (see Chapter 62).59 The efficacy and cost-effectiveness of such prophylaxis remain to be established.

CHOLESTEROLOSIS DEFINITION

Cholesterolosis is an acquired histologic abnormality of the gallbladder epithelium characterized by excessive accumulation of cholesterol esters and triglyceride within epithelial macrophages (Fig. 67-2).60 Clinicians generally encounter the lesion only as an incidental pathologic finding after surgical resection of the gallbladder, although the diagnosis may be suspected in certain patients before surgery. Cholesterolosis, as well as adenomyomatosis of the gallbladder (see later), has been classified as one of the hyperplastic cholecystoses, a term introduced in 1960 to describe several diseases of the gallbladder thought to share the common features of mucosal hyperplasia, hyperconcentration and hyperexcretion of dye on cholecystography, and absence of inflammation.61 The proponents of this concept believed that biliary pain, in the absence of gallstones, could be explained by the presence of one of the hyperplastic cholecystoses. Other investigators, citing the lack of a common etiology and the nonspecificity of the clinical features, have recommended that the term hyperplastic cholecystoses be abandoned.

EPIDEMIOLOGY

Although cholesterolosis has been recognized as a distinct pathologic entity for more than a century, its actual prevalence remains a matter of some dispute. Depending on whether gross or microscopic criteria are used for diagnosis, the frequency of cholesterolosis in autopsy specimens has ranged from 5% to 40%. A large autopsy series involving more than 1300 cases in which each gallbladder was examined microscopically found the prevalence of cholesterol­

Lipid-laden foamy macrophages

Epithelium

Lamina propria

Muscle layer

Adventitia

Normal gallbladder

Diffuse cholesterolosis

Cholesterol polyp

Figure 67-2.  Schematic representation of a normal gallbladder, diffuse cholesterolosis, and a cholesterol polyp. Note the distribution of lipid-laden foamy macrophages in cholesterolosis and the cholesterol polyp. The diffuse form of cholesterolosis (center; see also Fig. 67-3) accounts for 80% of cases and generally causes no symptoms. Cholesterol polyps (right), present in 20% of cases, are typically small, fragile excrescences that have a tendency to ulcerate or detach spontaneously from the mucosa. Although usually asymptomatic, these polyps have been associated with biliary pain and even acute pancreatitis.

Chapter 67  Acalculous Biliary Pain, Acalculous Cholecystitis, Cholesterolosis, Adenomyomatosis osis to be 12%.62 When surgically resected gallbladders were examined, the frequency was, not surprisingly, about 50% higher (18%) than that found in autopsy material.63 The incidence of cholesterolosis has not been calculated because its onset is rarely known. The epidemiology of cholesterolosis is analogous to that of cholesterol gallstone disease,64 in that similar groups of persons are predisposed; however, the two lesions occur independently and do not usually coexist in the same person. Like gallstone disease, cholesterolosis is uncommon in children (the youngest reported patient was a 13-year-old girl) and shows a marked predilection for women up to 60 years of age. After that, the gender differences are less pronounced. No racial, ethnic, or geographic differences in prevalence have been described, although if the analogy with cholesterol gallstone disease is extended, the prevalence would be expected to be higher in Western than nonWestern societies. Obesity also appears to be a risk factor for cholesterolosis; a frequency of 38% has been observed in gallbladders resected during weight loss surgery.65

PATHOLOGY

Cholesterolosis is defined pathologically by the accumulation of lipid (cholesteryl esters and triglyceride) within the gallbladder mucosa. The four patterns of lipid deposition are as follows60: Diffuse: The lipid is distributed throughout the epithelial lining of the gallbladder and ends abruptly at the cystic duct. This pattern accounts for 80% of all cases. Cholesterol polyps: The excess lipid is confined to one or more areas of the epithelium that eventually form excrescences into the lumen of the gallbladder. Isolated cholesterol polyps in the absence of diffuse cholesterolosis account for about 10% of the total cases. Combined diffuse cholesterolosis and cholesterol polyps: Cholesterol polyps occur on a background of diffuse cholesterolosis. This pattern accounts for about 10% of cases. Focal cholesterolosis: Excess lipid deposition is limited to a small area of the mucosa.

Figure 67-3.  Photomicrograph of diffuse cholesterolosis. Note the hyperplastic, elongated villi and the foamy macrophages (arrows). (Hematoxylin and eosin.) (Courtesy of Pamela Jensen, MD, Dallas, Tex.)

cosa. Although extracellular deposits of lipid are rare, small yellow particles (lipoidic corpuscles) representing detached masses of foam cells are occasionally seen floating in the bile.

PATHOGENESIS

When the gallbladder is inspected visually at the time of laparotomy or laparoscopy, a diagnosis of cholesterolosis can be made in 20% of the cases on the basis of the gross appearance of the gallbladder mucosa as seen through the translucent serosal surface. When the gallbladder is opened, the mucosa characteristically has pale, yellow linear streaks running longitudinally, giving rise to the term strawberry gallbladder (although the mucosa is usually bile stained rather than red). When cholesterolosis is diagnosed at the time of surgical resection of the gallbladder, gallstones are also present in 50% of cases. If the diagnosis of cholester­ olosis is made at autopsy, stones are present in only 10%,62 demonstrating that the two disease processes are independent of each other.

The cause of the accumulation of cholesteryl esters and triglyceride in cholesterolosis remains obscure.66 Postulated mechanisms are that the cholesterol is derived from the blood67 or that mechanical factors that impede emptying of the gallbladder lead to local deposition of lipid.68 Data have shown unequivocally that the gallbladder epithelium is capable of absorbing cholesterol from the bile, as might be expected in epithelium that is embryologically and histologically similar to intestinal absorptive cells.69,70 Moreover, the cholesterol in gallbladder bile is already in the ideal physical state for absorption (i.e., a mixed micelle). The question remains as to why, in some patients, resorbed biliary cholesterol is esterified and then stored in foamy macrophages as cholesterolosis.71 Like cholesterol stones, cholesterolosis is frequently, but not always, found in gallbladders exposed to bile that is supersaturated with cholesterol.72 The two disorders (cholesterolosis and stone disease), both of which lead to the ectopic accumulation of cholesterol, probably share common pathogenic mechanisms (such as the secretion of abnormal bile) but progress independently in a given patient, depending on other factors such as the presence of nucleating proteins in bile and the rate of mucosal esterification of cholesterol.73 Cholesterolosis is not associated with high serum cholesterol levels.64

Microscopic Appearance

CLINICAL FEATURES

Gross Appearance

Hyperplasia of the mucosa is invariably present and is described as marked in 50% of cases. Usually, the hyperplasia is of the villous type. The most prominent feature is an abundance of macrophages within the elongated villi. Each macrophage is stuffed with lipid droplets and has a characteristic appearance of a foam cell (Fig. 67-3). In milder cases, the foam cells are limited to the tips of the villi (accounting for the linear streaks seen on gross examination); with more severe involvement, the foam cells may fill the entire villi and spill over into the underlying submu-

Cholesterolosis usually does not cause symptoms, as is evident by how frequently autopsy specimens show the lesion in patients who never had biliary symptoms. On occasion, individual patients have dull, vague, right upper quadrant or epigastric pain that resembles biliary pain and are found subsequently to have cholesterolosis without stones or gallbladder inflammation after cholecystectomy. Of the patients who undergo cholecystectomy for the syndrome of acalculous biliary pain, pain is more likely to resolve in those in whom incidental cholesterolosis is found

1145

1146

Section VIII  Biliary Tract on pathologic examination of the gallbladder than in those in whom cholesterolosis is not found.74 In retrospective surgical series of nearly 4000 gallbladders removed by cholecystectomy, 55 patients with acalculous cholesterolosis were identified.75 The investigators found that nearly half of these patients had presented with recurrent pancreatitis of unknown etiology and speculated that small cholesterol polyps had detached from the gallbladder wall and transiently obstructed the sphincter of Oddi, thereby provoking the acute pancreatitis. In 5 years of postoperative follow-up, pancreatitis did not recur. These investigators and others76,77 have suggested that cholester­ olosis (or more specifically, cholesterol polyps) should be considered in the differential diagnosis of idiopathic pancreatitis.

DIAGNOSIS

Diffuse cholesterolosis (which, as noted earlier, constitutes 80% of cases) is only rarely detectable by either ultrasonography or oral cholecystography. In the polypoid form, however, polyps of sufficient size have a characteristic appearance on ultrasonography as single or multiple, nonshadowing, fixed echoes that project into the lumen of the gallbladder.78 Most of the polyps are small (2 to 10 mm). The polyps can be identified accurately as cholesterolosis polyps by endoscopic ultrasonography, which demonstrates a characteristic aggregation of hyperechoic spots.79 On oral cholecystography, the polyps appear as small, round radiolucencies in the lumen of the opacified gallbladder and are best demonstrated after the gallbladder has emptied partially and abdominal compression has been applied.

TREATMENT

Because cholesterolosis is only rarely diagnosed before resection of the gallbladder, the issue of treatment is usually irrelevant. In the rare case of polypoid cholesterolosis diagnosed on ultrasonography or cholecystography, the absence of biliary tract symptoms argues against any intervention. If the patient has symptoms consistent with biliary pain or pancreatitis, a cholecystectomy is indicated.75 There is no medical therapy for cholesterolosis.

ADENOMYOMATOSIS DEFINITION

Adenomyomatosis (an unwieldy term that obscures its meaning) of the gallbladder is an acquired, hyperplastic lesion characterized by excessive proliferation of surface epithelium with invaginations into the thickened muscularis or even more deeply.80 Despite the prefix adeno-, the lesion is generally benign and unrelated to adenomatous epithelia elsewhere in the gastrointestinal tract. Simple adenomyomatosis is not thought to have the potential for malignant transformation. The literature on this obscure condition is complicated by the use of a number of different terms to describe the same lesion. One researcher noted that adenomyomatosis has been described by at least 18 distinct names, the more common of which are adenomyoma (used when the lesion is localized to the gallbladder fundus), diverticulosis of the gallbladder (ignores the hyperplasia), cholecystitis glandularis proliferans (overemphasizes the role of inflammation), Rokitansky-Aschoff sinuses (familiar but anatomically incorrect), adenomyosis, and adenomyomatous hyperplasia.81 Some terms are used in the radiologic literature,

whereas others are used exclusively by pathologists. None is familiar to most gastroenterologists.

EPIDEMIOLOGY

The prevalence of adenomyomatosis of the gallbladder varies greatly according to the criteria used for diagnosis and whether resected gallbladders or autopsy specimens are examined. In a large series of more than 10,000 cholecystectomy specimens, Shepard and associates82 found only 103 cases of adenomyomatosis, for a prevalence of about 1%. The lesion is more common in women than men by a 3 : 1 ratio, and the prevalence rises with age. Neither ethnic nor geographic differences in prevalence have been described.

PATHOLOGY

A review of the normal histologic architecture of the gallbladder and Rokitansky-Aschoff sinuses is useful for understanding the pathology of adenomyomatosis (Fig. 67-4). Unlike the small intestine, the gallbladder has no muscularis mucosa, and the lamina propria abuts directly on the muscular layer. In childhood, the epithelial layer is cast up into folds and supported by the lamina propria. As the gallbladder ages, the valleys of the epithelial layer may deepen so that they penetrate into the muscular layer and form Rokitansky-Aschoff sinuses. These sinuses are acquired lesions present in about 90% of resected gallbladders. If the Rokitansky-Aschoff sinuses are deep and branching and are accompanied by thickening (hyperplasia) of the muscular layer, a diagnosis of adenomyomatosis can be made.80 Rupture of Rokitansky-Aschoff sinuses is thought to underlie the rare entity xanthogranulomatous cholecystitis, in which the gallbladder is involved in an inflammatory process with lipid-laden macrophages.

Gross Appearance

Adenomyomatosis may involve the entire gallbladder (diffuse or generalized adenomyomatosis) or, more commonly, may be localized to the gallbladder fundus, in which case the lesion is often termed adenomyoma. On rare occasions, the process may be limited to an annular segment of the gallbladder wall (segmental adenomyomatosis) and may give rise to luminal narrowing and a “dumbbell-shaped” gallbladder (Fig. 67-5). In any case, the involved portion of the gallbladder wall is thickened to 10 mm or more, and the muscle layer is three to five times its normal thickness. On cut sections, cystic dilatations of the Rokitansky-Aschoff sinuses are evident and may be filled with pigmented debris or calculi.

Microscopic Appearance

Hyperplasia of the muscle layer is invariably present, and the epithelial lining occasionally undergoes intestinal metaplasia. Mild chronic inflammation is often present.

PATHOGENESIS

The pathogenesis of adenomyomatosis is unknown. Increased intraluminal pressure in the gallbladder from mechanical obstruction (e.g., from an obstructing calculus, kink in the cystic duct, or congenital septum) has been postulated to result in cystic dilatation of the RokitanskyAschoff sinuses, subsequent hyperplasia of the muscle layer, and adenomyomatosis.80 Like pressure-related colonic diverticula, Rokitansky-Aschoff sinuses are most likely to be found where the muscle layer is weakest (at the site of a penetrating blood vessel). Nevertheless, evidence of outflow obstruction of the gallbladder is not always found; for example, calculi are present in only about 60% of cases of

Chapter 67  Acalculous Biliary Pain, Acalculous Cholecystitis, Cholesterolosis, Adenomyomatosis Lumen

Epithelium

Lumen

Lumen

Lamina propria

Adventitia

Muscle layer

Normal gallbladder

Rokitansky-Aschoff sinus

Adenomyomatosis

Figure 67-4.  Schematic representation of a normal gallbladder, Rokitansky-Aschoff sinus, and adenomyomatosis. Rokitansky-Aschoff sinuses, which are present in about 90% of resected gallbladders, consist of invaginations of the epithelium into the muscle layer to produce tiny intramural diverticula. By themselves, they have no clinical significance. A histologic diagnosis of adenomyomatosis requires that the Rokitansky-Aschoff sinuses be deep, branching, and accompanied by hyperplasia of the muscle layer.

Lamina propria

Muscle layer

Epithelium

Normal gallbladder

Fundic adenomyomatosis (adenomyoma)

adenomyomatosis.82 Some investigators have proposed that adenomyomatosis is a consequence of chronic inflammation, but inflammation is not always present, particularly when the lesion is localized to the fundus.83 Finally, several investigators have noted an association between adenomyomatosis and anomalous pancreaticobiliary ductal union (see Chapter 55). In one study, half of the patients with adenomyomatosis had anomalous pancreaticobiliary ductal union,84 and in another study, one third of patients with anomalous pancreaticobiliary ductal union had adenomyo-

Generalized adenomyomatosis

Segmental adenomyomatosis

Figure 67-5.  Schematic representation showing the different patterns of adenomyomatosis. Most of the cases are localized to the fundus of the gallbladder (in which case the lesion is termed an adenomyoma); generalized and segmental patterns are much less common. An adenomyoma is usually 10 to 20 mm in diameter and may be largely confined to the wall or may project into the lumen to produce a polypoid lesion.

matosis.85 The pathogenic link between these two peculiar entities is unclear.

CLINICAL FEATURES

Adenomyomatosis, like cholesterolosis, usually causes no symptoms and is typically an incidental finding at autopsy or surgical resection. As noted earlier, gallstones are present in more than half of the resected gallbladders that are found to have adenomyomatosis; in these cases the symptoms can be ascribed to the stones.82 Uncommonly,

1147

1148

Section VIII  Biliary Tract A

B

Figure 67-6.  A, Gross pathologic appearance of a gallbladder adenomyoma involved by adenocarcinoma. B, Histologic examination shows a moderately differentiated adenocarcinoma of the gallbladder undermining the mucosa of the adenomyoma. (Hematoxylin and eosin.) (Courtesy of Aviva Hopkowitz, MD, Dallas, Tex.)

acalculous adenomyomatosis appears to cause symptoms indistinguishable from the biliary pain of cholelithiasis. On rare occasions, adenocarcinoma of the gallbladder has been found in association with adenomyomatosis (Fig. 67-6)86; however, the malignancy is often far removed from the localized area of adenomyomatosis, and the association has been thought to be coincidental rather than causal. Nevertheless, several reports of adenocarcinoma occurring in an area of gallbladder wall involved with adenomyomatosis have created diagnostic uncertainty on ultrasonography or cholecystography.87 A retrospective review of more than 3000 resected gallbladders revealed a significantly higher frequency (6.4%) of gallbladder cancer in gallbladders with the segmental form of adenomyomatosis than would have been expected by chance alone. The investigators proposed that segmental adenomyomatosis should be considered a potentially premalignant lesion.88 A second review of gallbladder cancers associated with segmental adenomyomatosis revealed a spectrum of cytologic atypia in the specimens ranging from hyperplastic to malignant epithelium, suggestive of neoplastic progression.89 When simple adenomyomatosis of the gallbladder is discovered incidentally, the lesion is likely to be benign. If there is any suspicion of an associated mass lesion, particularly one larger than 10 mm, or if segmental adenomyomatosis is found, however, a thorough radiologic evaluation of the gallbladder is warranted, and cholecystectomy should be considered.

Figure 67-7.  Oral cholecystogram showing segmental adenomyomatosis in a 28-year-old man with postprandial epigastric pain radiating through to the back. The film demonstrates an annular segment of the gallbladder wall (arrowhead) involved with adenomyomatosis, which has produced a constriction of the lumen. Although no gallstones were present, a cholecystectomy was performed, and the patient’s symptoms were relieved. (Courtesy of W. J. Kilman, MD, Dallas, Tex.)

DIAGNOSIS

As noted previously, adenomyomatosis is frequently diagnosed only after resection and direct examination of the gallbladder; however, several specific radiologic and ultrasonographic findings may, if present, allow the diagnosis to be made preoperatively. On oral cholecystography (see Chapter 65), the mural diverticula that constitute Rokitansky-Aschoff sinuses may fill with contrast material and produce characteristic radiopaque dots that parallel the margin of the gallbladder lumen.90 Any portion of the gallbladder wall may be involved (Fig. 67-7). Localized, fundal adenomyomatosis (adenomyoma) may manifest as a filling defect in the fundus, whereas segmental adenomyomatosis may appear as a circumferential narrowing of the gallbladder lumen. As is the

case with cholesterolosis, the radiologic findings in adenomyomatosis are best appreciated when the gallbladder has partially emptied of contrast material and external pressure has been applied during the examination.90 Although ultrasonography has largely replaced oral cholecystography in the evaluation of the gallbladder, the ultrasonographic findings in adenomyomatosis are less specific. A thickened gallbladder wall (>4 mm) is not specific for adenomyomatosis and can also be seen in many other conditions such as liver disease with ascites.91 Carefully performed studies in which radiologic and ultrasonographic findings of adenomyomatosis were correlated with pathologic findings have shown that diffuse or segmental thicken-

Chapter 67  Acalculous Biliary Pain, Acalculous Cholecystitis, Cholesterolosis, Adenomyomatosis ing of the gallbladder wall in association with intramural diverticula (seen as round anechoic foci) accurately predicts adenomyomatosis.92 If the intramural diverticula (dilated Rokitansky-Aschoff sinuses) are filled with sludge or small calculi, the lesions may appear echogenic with acoustic shadowing or a reverberation artifact.93 Endoscopic ultrasonography may demonstrate the characteristic finding of multiple microcysts, corresponding to the proliferated Rokitansky-Aschoff sinuses.79 CT and magnetic resonance imaging findings in adenomyomatosis include94 differential enhancement of gallbladder wall layers, detection of Rokitansky-Aschoff sinuses within a thickened gallbladder wall,95 and subserosal fatty proliferation.96 In a study of 20 patients with surgically proved adenomyomatosis who had preoperative ultrasound, helical CT, and MRI evaluation, the diagnostic accuracies of the three modalities were 66%, 75%, and 93%, respectively.97 In one case report, an adenomyoma without histologic evidence of cancer was the cause of a false-positive finding on 18-fluorodeoxyglucose positron emission tomography (PET), likely because of the associated inflammatory activity.98

variants of cholesterolosis that result from infiltration of the lamina propria with lipid-laden foamy macrophages. The pathogenesis of cholesterol polyps is discussed in the section on cholesterolosis (see earlier). Cholesterol polyps are typically small (>10 mm in diameter), pedunculated polyps that are attached to the mucosa by a thin, fragile stalk.102 Frequently, detached tiny cholesterol polyps are found floating in the bile when the gallbladder is opened in the operating room.103 Although they may be solitary in 20% of cases, the mean number of cholesterol polyps present in one series was eight.104

TREATMENT

Inflammatory Polyps

In the absence of biliary tract symptoms, adenomyomatosis requires no treatment. If the patient has biliary pain and radiographic or ultrasonographic evidence of adenomyomatosis with calculi, a cholecystectomy is indicated. A more difficult clinical problem arises when a patient is symptomatic and has suspected adenomyomatosis but no stones.87 In such cases, the more extensive or severe the adenomyomatosis appears to be, the more likely that the symptoms are related to the lesion and that the patient will benefit from cholecystectomy. Fear of malignant transformation is not a reason to operate, unless an ultrasonographic or radiologic image suggests a mass or perhaps shows the segmental form of adenomyomatosis.99

POLYPS OF THE GALLBLADDER DEFINITION

The term polyp of the gallbladder is used to describe any mucosal projection into the lumen of the gallbladder.100 The vast majority of gallbladder polyps are the result of lipid deposits or inflammation, rather than neoplasms. Because the nature of a polyp cannot be defined without histologic evaluation, however, clinicians must decide whether the concern of malignancy is sufficient to perform cholecys­ tectomy based on indirect information such as the radiographic appearance of the polyp, patient demographics, and symptoms.

EPIDEMIOLOGY

The frequency of gallbladder polyps, defined either pathologically or radiologically,101 ranges from 1% to 4%. Often, gallbladder polyps are an incidental finding at the time of cholecystectomy.

PATHOLOGY

Polyps of the gallbladder may be classified as shown in Table 67-3 as either non-neoplastic (95% of all gallbladder polyps) or neoplastic.102

Cholesterol Polyps

Cholesterol polyps (also known as papillomas of the gall­ bladder, although the term should be discarded) are the most common type of gallbladder polyp. They are benign

Adenomyomas

Adenomyomatosis of the gallbladder localized to the fundus may produce a hemispheric projection into the lumen that resembles a polyp. Such a lesion has come to be known as an adenomyoma, although it is not neoplastic in origin. The pathogenesis of an adenomyoma is discussed in the section on adenomyomatosis (see earlier). The lesion is usually approximately 15 mm in size, and its bulk is confined to the muscular wall of the gallbladder.102 Inflammatory polyps are small sessile lesions that consist of granulation and fibrous tissue infiltrated with lymphocytes and plasma cells. The average size is 5 to 10 mm. A solitary polyp is found in 50% of cases, and two to five polyps are found in the remainder.102 When discovered at the time of cholecystectomy, an inflammatory polyp is almost always an incidental finding.

Adenomas

In light of the high frequency of adenomatous polyps in the gastrointestinal tract, gallbladder adenomas are surprisingly uncommon. Their frequency in resected gallbladder specimens is only about 0.15%.105 Adenomas are typically solitary, pedunculated masses from 5 to 20 mm in diameter. They may occur anywhere in the gallbladder. When multiple, as they are in approximately one third of cases, two to five polyps are usually present. Histologically, they are classified as either papillary or nonpapillary. The former type consists of a branching, tree-like skeleton of connective tissue covered with tall columnar cells, whereas the latter consists of a proliferation of glands encased by a fibrous stroma. On rare occasions, the entire gallbladder mucosa may undergo adenomatous transformation that results in innumerable tiny mucosal polyps termed multicentric papillomatosis. Notably, gallstones are present in half of cases of adenomatous polyps.102 Unlike the colon, in which adenomas are much more common than adenocarcinomas, the gallbladder is affected less commonly by adenomas than by carcinomas (by a 1 : 4 ratio). The frequency of progression from adenoma to adenocarcinoma is not well defined. In a series of more than 1600 consecutive cholecystectomies from Japan, 18 of the operated patients were found to have gallbladder adenomas.106 Seven of the adenomas contained foci of carcinoma. In the same series, 79 cases of invasive carcinoma were found; 15 (19%) of the lesions were thought to have residual adenomatous tissue within the cancer, suggesting that the initial lesion may have been an adenoma. Notably, all the adenomas that contained foci of carcinoma were larger than 12 mm, a finding that suggests that large adenomas may represent premalignant lesions.

Miscellaneous Polyps

Although a wide variety of benign lesions may manifest as polyps in the gallbladder, these lesions are rare. Fibromas,

1149

1150

Section VIII  Biliary Tract Table 67-3  Types of Gallbladder Polyps HISTOLOGIC TYPE

RELATIVE FREQUENCY (%)

NEOPLASTIC

SIZE RANGE (mm)

NUMBER OF POLYPS

Cholesterol polyp (a polypoid form of cholesterolosis)

60

No

2-10

Multiple (average of 8)

Adenomyoma (a localized form of adenomyomatosis)

25

No

10-20

One

Inflammatory polyp

10

No

5-10

One in half of cases (2-5 in remainder)

Adenoma

4

Yes

5-20

One in two thirds of cases (2-5 in remainder)

Miscellaneous neoplasms

12 mm in diameter; lesions 10-18 mm in size, laparoscopic cholecystectomy should be considered in good surgical candidates For lesions >18 mm in size, open rather than laparoscopic cholecystectomy should be considered, because invasive cancer is more likely and extended resection may be required Extremely rare lesions (see text), with frequencies of 10 mm), patient age greater than 60 years is the strongest predictor of neoplastic disease. The presence of concurrent gallstones is also associated with a higher risk of malignancy.115 Single polyps and symptomatic polyps may be more likely to be malignant than multiple polyps and asymptomatic polyps, respectively.

NATURAL HISTORY

The few studies that have attempted to define the natural history of untreated gallbladder polyps highlight the benign nature of most polyps and support a “watch and wait” approach in most cases.116 On the basis of records at the Mayo Clinic, one study identified approximately 200 patients in whom cholecystograms demonstrated gallbladder polyps and immediate cholecystectomy was not performed.117 After 15 years of follow-up, symptoms sufficient to warrant surgery developed in fewer than 10% of the patients, and none of the patients available for follow-up had evidence of gallbladder cancer. One group of investigators performed annual or semiannual ultrasound for a 5-year period on 109 patients with polyps smaller than 10 mm. During this time, gallbladder cancer developed in no patient, and the polyp exhibited no growth in more than 88% of patients.118 Another study identified 224 patients with gallbladder polyps, 95% of which were predicted to be cholesterol polyps on the basis of the ultrasonographic appearance and the remainder of which were classified as “polypoid lesions of uncertain benignity.”119 After an average follow-up of 9 months, all the polyps thought initially to be benign remained the same size or were proved to be benign at resection. Two thirds of the polypoid lesions in which a benign nature was uncertain were found to be adenomas or carcinomas when resected. These findings suggest that although most gallbladder polyps are benign, high-risk polyps often have an identifiable characteristic such as larger size.

Patients who are symptomatic with biliary pain and have ultrasonographic evidence of both polyps and stones in the gallbladder should undergo elective cholecystectomy. The decision is more complicated for patients in whom gallbladder polyps without concurrent gallstones are discovered. For these patients, the decision to operate depends on the severity of symptoms, the confidence of the clinician that the symptoms are biliary in origin, and ultrasonographic features (particularly the size) of the polyp. The likelihood of malignancy is correlated with polyp size. Polyps less than 10 mm in diameter are unlikely to be cancerous and generally do not require intervention in the absence of symptoms. Because polyps larger than 10 mm have a greater likelihood of being cancerous, elective laparoscopic cholecystectomy should be considered in acceptable surgical candidates.120-122 In a patient who is a poor surgical risk with a polyp larger than 10 mm, periodic monitoring for polyp growth (perhaps every 6 to 12 months) with ultrasound or endoscopic ultrasound may be reasonable.120,122 Polyps larger than 18 mm in diameter pose a significant risk of malignancy and should be resected if possible. One study found that lesions of this size often contain advanced, invasive cancer that involves the serosal surface of the gallbladder and requires a more extensive dissection than can be accomplished by laparoscopy.123 As a result, the investigators advocate open cholecystectomy for these large polypoid lesions of the gallbladder. Because of the uncertainty regarding the neoplastic potential of polyps, common practice is to offer cholecystectomy to any good surgical candidate with a polyp greater than 10 mm in diameter. Patients should be warned, however, that false-positive results for polypoid lesions on ultrasonography are common and that polyps may not be found in all resected gallbladders. The 10-mm cut-off rule for following gallbladder polyps expectantly may not apply to patients with primary sclerosing cholangitis, in whom the risk of malignancy in polypoid lesions of the gallbladder may be as high as 60%.124 In this high-risk population, cholecystectomy for polyps smaller than 10 mm may be justifiable. In low-risk populations, such as asymptomatic patients with small, presumably benign polyps, periodic surveillance for polyp growth may be prudent. One group of investigators recommends conventional ultrasonographic evaluation every 3 to 6 months in the immediate post-diagnostic period to exclude a rapidly growing tumor, but less frequent or no investigation after 1 to 2 years of stability in polyp size.125

KEY REFERENCES

Akatsu T, Aiura K, Shimazu M, et al. Can endoscopic ultrasonography differentiate nonneoplastic from neoplastic gallbladder polyps? Dig Dis Sci 2006; 51:416-21. (Ref 79.) Albores-Saavedra J, Shukla D, Carrick K, Henson D. In situ and invasive adenocarcinomas of the gallbladder extending into or arising from Rokitansky-Aschoff sinuses: A clinicopathologic study of 49 cases. Am J Surg Pathol 2004; 28:621-8. (Ref 89.) Boulton R, Adams D. Gallbladder polyps: When to wait and when to act. Lancet 1997; 349:817-18. (Ref 116.) Hansel S, DiBaise J. Gallbladder dyskinesia. Curr Treat Options Gastroenterol 2008; 11:78-84. (Ref 6.) Kalliafas S, Ziegler DW, Flancbaum L, Choban PS. Acute acalculous cholecystitis: Incidence, risk factors, diagnosis, and outcome. Am Surg 1998; 64:471-5. (Ref 31.) Kmiot WA, Perry EP, Donovan IA, et al. Cholesterolosis in patients with chronic acalculous biliary pain. Br J Surg 1994; 81:112-15. (Ref 74.) Laurila J, Syrajala H, Laurila P. Acute acalculous cholecystitis in critically ill patients. Acta Anaesth Scand 2004; 48:986-91. (Ref 52.) Lee K, Wong J, Li J, Lai P. Polypoid lesions of the gallbladder. Am J Surg 2004; 188:186-90. (Ref 125.)

1151

1152

Section VIII  Biliary Tract Mirvis SE, Vainright JR, Nelson AW, et al. The diagnosis of acute acalculous cholecystitis: A comparison of sonography, scintigraphy, and CT. AJR Am J Roentgenol 1986; 147:1171-5. (Ref 44.) Rastogi A, Slivka A, Moser A, Wald A. Controversies concerning pathophysiology and management of acalculous biliary-type abdominal pain. Dig Dis Sci 2005; 50:1391-401. (Ref 16.) Terzi C, Sökmen, Seckin S, et al. Polypoid lesions of the gallbladder: report of 100 cases with special reference to operative indications. Surgery 2000; 127:622-7. (Ref 115.)

Weedon D. Adenomyomatosis. In: Sternberg SS (ed): Pathology of the Gallbladder. New York: Masson; 1984. pp 185-94. (Ref 80.) Weedon D. Cholesterolosis. In: Sternberg SS (ed): Pathology of the Gallbladder. New York: Masson; 1984. pp 161-5. (Ref 60.) Yap L, Wycherley AG, Morphett AD, Toouli J. Acalculous biliary pain: Cholecystectomy alleviates symptoms in patients with abnormal cholescintigraphy. Gastroenterology 1991; 3:786-93. (Ref 5.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

68 Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis Andrew S. Ross and Kris V. Kowdley

CHAPTER OUTLINE Primary Sclerosing Cholangitis  1153 Diagnosis  1153 Epidemiology  1155 Primary Sclerosing Cholangitis and Inflammatory Bowel Disease  1156 Etiology and Pathogenesis  1156 Natural History and Prognostic Models  1157 Clinical Features  1159 Pathology  1160 Complications  1161 Treatment  1164

Sclerosing cholangitis encompasses a spectrum of cholestatic conditions that are characterized by patchy inflammation, fibrosis, and destruction of the intrahepatic and extrahepatic bile ducts. These conditions are typically chronic, progressive disorders in which persistent biliary damage may lead to biliary obstruction, biliary cirrhosis, and hepatic failure, with associated complications. The first description of sclerosing cholangitis is credited to Delbet in 1924.1 Although considered for many years to be an extremely rare disorder, the advent of endoscopic retrograde cholangiopancreatography (ERCP) in the 1970s has allowed an improved understanding of the true prevalence of this disorder and facilitated careful study of its natural history. Nevertheless, many aspects of sclerosing cholangitis remain poorly understood; most notably lacking are a detailed knowledge of its etiology and proven effective medical therapy. A cholangiographic appearance of diffuse stricturing and segmental dilatation of the biliary system, designated sclerosing cholangitis, may be observed in many distinct conditions. The most frequent is primary sclerosing cholangitis (PSC), an idiopathic disorder that usually occurs in association with inflammatory bowel disease (IBD) but may develop independently. PSC may also be associated with a wide variety of fibrotic, autoimmune, and infiltrative disorders, although whether such associations imply a common pathogenesis or epiphenomena is unclear (Table 68-1). PSC is also associated with various immunodeficiency states; in such cases biliary abnormalities may be caused by infection with Dr. Bruce Y. Tung contributed to this chapter in the previous edition of this book.

Recurrent Pyogenic Cholangitis  1167 Epidemiology  1167 Etiology and Pathogenesis  1167 Clinical Features  1168 Pathology  1169 Treatment  1169 Prognosis and Complications  1169

an opportunistic pathogen. The term secondary sclerosing cholangitis refers to a clinical and radiologic syndrome that is similar to PSC but develops as a consequence of a known pathogenesis or injury. Obstructive, toxic, ischemic, and neoplastic causes of secondary sclerosing cholangitis have been described (see Table 68-1). This chapter focuses on PSC and recurrent pyogenic cholangitis.

PRIMARY SCLEROSING CHOLANGITIS DIAGNOSIS

No standardized criteria for the diagnosis of PSC have been universally adopted. Early diagnostic criteria included diffuse intra- and extrahepatic bile duct strictures occurring in the absence of prior biliary surgery or cholelithiasis and after exclusion of cholangiocarcinoma.2 These criteria were later modified because of the recognition that the clinical spectrum of PSC is broader than initially appreciated, and strict adherence to the original criteria underestimates the prevalence of the disease. It is now apparent that a form of PSC, termed small-duct PSC, involves only the intra­ hepatic biliary tree, without obvious extrahepatic duct abnormalities.3 In addition, both cholelithiasis and chole­ docholithiasis may develop as a consequence of PSC, and their presence does not exclude a diagnosis of underlying PSC.4,5 Furthermore, cholangiocarcinoma is a relatively common complication of PSC, and both conditions frequently coexist.6 The diagnosis of PSC is based on typical cholangiographic findings in the setting of consistent clinical, biochemical, serologic, and histologic findings as well as exclusion of

1153

1154

Section VIII  Biliary Tract Table 68-1  Classification and Diseases Associated with Sclerosing Cholangitis Primary Sclerosing Cholangitis (PSC) Principal Disease Associations Inflammatory bowel disease Crohn’s colitis or ileocolitis Ulcerative colitis Other Disease Associations Systemic disease with fibrosis Inflammatory pseudotumor Mediastinal fibrosis Peyronie’s disease Pseudotumor of the orbit Retroperitoneal fibrosis Riedel’s thyroiditis Autoimmune or Collagen Vascular Disorders Autoimmune hemolytic anemia Celiac disease Chronic sclerosing sialadenitis Membranous nephropathy Rapidly progressive glomerulonephritis Rheumatoid arthritis Sjögren’s syndrome Systemic lupus erythematosus Systemic sclerosis Type I diabetes mellitus Alloimmune Diseases Hepatic allograft rejection Hepatic graft-versus-host disease after bone marrow transplantation Infiltrative Disease Hypereosinophilic syndrome Histiocytosis X Sarcoidosis Systemic mastocytosis Immunodeficiency Congenital immunodeficiency Combined immunodeficiency Dysgammaglobulinemia X-lined agammaglobulinemia Acquired immunodeficiency Acquired immunodeficiency syndrome Angioimmunoblastic lymphadenopathy Selective IgA deficiency Secondary Sclerosing Cholangitis Obstructive Autoimmune pancreatitis Biliary parasites Caroli’s disease Choledocholithiasis Chronic pancreatitis Congenital abnormalities Cystic fibrosis Choledochal cyst Fungal infection Recurrent pyogenic cholangitis Surgical stricture Toxic Intra-arterial floxuridine (FUDR) Intraductal formaldehyde or hypertonic saline (echinococcal cyst removal) Ischemic Hepatic allograft arterial occlusion Paroxysmal nocturnal hemoglobinuria Toxic vasculitis (FUDR) Vascular trauma Neoplastic Cholangiocarcinoma Hepatocellular carcinoma Lymphoma Metastatic cancer

secondary causes of sclerosing cholangitis. The characteri­ stic cholangiographic findings are multifocal stricturing and ectasia of the biliary tree. Areas of narrowing are interspersed with areas of normal or near-normal caliber and of post-stenotic dilatation. Although the majority of patients with PSC have coexisting abnormalities of the intra- and extrahepatic bile ducts, a small percentage have an isolated lesion. Patients with small-duct PSC may have a normal cholangiogram. Gallbladder abnormalities, including tumors, may exist in up to 41% of patients with PSC.7 ERCP is considered the standard for establishing a diagnosis of PSC but carries a risk for complications of up to 10% in patients with PSC.8,9 Magnetic resonance cholangiopancreatography (MRCP) has largely replaced ERCP for diagnostic cholangiography as a result of improvements in image quality and the noninvasive nature of MRCP (Fig. 68-1). In the few studies that have compared MRCP and ERCP in patients with PSC, MRCP has demonstrated comparable sensitivity for the detection of biliary structuring,10-13 although performance and interpretation of magnetic resonance cholangiograms vary with the technique and institution. ERCP has the advantage of combining highresolution cholangiography with the potential for advanced diagnostic and therapeutic interventions, including brush cytology or intraductal biopsy for the diagnosis of cholangiocarcinoma, balloon or catheter dilation of strictures, biliary stent placement, sphincterotomy, and stone removal. Percutaneous transhepatic cholangiography (THC) may also yield diagnostic images and allow therapeutic intervention but requires percutaneous puncture and may be technically difficult if the intrahepatic bile ducts are not sufficiently dilated (see Chapter 70). Patients with IBD and a cholestatic pattern of liver biochemical test elevations should undergo imaging of the hepatobiliary system because of the relatively high pretest probability of PSC. Ultrasonography or computed tomography (CT) may be useful for planning further diagnostic and therapeutic strategies in selected patients, but they are usually insufficient for a diagnosis of PSC because normal findings do not exclude the diagnosis. The decision as to which method of cholangiography to perform must be individualized. In most cases, ERCP is the initial test of choice for patients in whom a therapeutic intervention or the need for brush cytology is anticipated. In an asymptomatic patient with mild liver biochemical abnormalities who is unlikely to require therapeutic intervention, MRCP is the preferred initial test if the images are reliable.13 When MRCP is nondiagnostic and clinical suspicion for PSC remains, diagnostic ERCP is indicated.

Differential Diagnosis

In a patient with a cholangiographic appearance characteristic of sclerosing cholangitis, secondary causes of sclerosing cholangitis must be excluded (see Table 68-1). Patients with the acquired immunodeficiency syndrome (AIDS) and a CD4+ T-lymphocyte count below 100/mm3 can exhibit a cholangiographic appearance identical to that of PSC; this entity is termed AIDS cholangiopathy. Cryptosporidium, Microsporidium, cytomegalovirus, and other organisms have been isolated from the bile of affected patients.14,15 Exposure of the bile ducts to toxins such as intra-arterial floxuridine (FUDR)16 and formaldehyde administered to treat a hydatid cyst, when the cyst communicates with the biliary tract,17 can produce a similar cholangiographic appearance. After these secondary causes of sclerosing cholangitis are excluded, distinguishing PSC from other disorders of

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis

A

B

Figure 68-1.  A, Reconstructed images from a magnetic resonance cholangiopancreatogram (MRCP) performed in a patient with primary sclerosing cholangitis. The image suggests atrophy of the right-sided intrahepatic bile ducts with slight dilatation of the ducts on the left side. B, A film from an endoscopic retrograde cholangiogram from the same patient is shown, revealing findings similar to those on MRCP.

the bile ducts may still be challenging. Choledocholithiasis and cholangiocarcinoma may both develop in conjunction with, or independent of, PSC. In the presence of extensive choledocholithiasis or diffuse cholangiocarcinoma, identifying underlying PSC may be difficult. Cholangiographic findings in patients with cirrhosis from causes other than PSC may at times be mistaken for PSC; however, cholangiography in cirrhotic patients without PSC typically shows diffuse intrahepatic attenuation of bile ducts without the ductal irregularity or stricturing seen in patients with PSC. Primary biliary cirrhosis (PBC) is another chronic cholestatic condition that shares some clinical features with PSC (see Chapter 89); however, PBC predominantly affects middle-aged women, has no association with IBD, and is associated strongly with high titers of antimitochondrial antibodies. Whereas liver histologic findings in the two disorders overlap substantially,18 the distinction between the two is readily apparent on cholangiography. Patients with advanced PBC may demonstrate smooth tapering and narrowing of the intrahepatic bile ducts, but ductal irregularity or strictures are not seen and extrahepatic lesions do not occur. Antimitochondrial antibody-negative PBC (autoimmune cholangitis) may be difficult to distinguish from small-duct PSC because serologic profiles and cholangiographic findings may overlap, but the demographic and histologic features of the two disorders are distinct (see Chapter 89). Autoimmune hepatitis may also be difficult to distinguish from PSC (see Chapter 88). In the pediatric population, PSC typically manifests with features of autoimmune hepatitis, and cholangiography is necessary to distinguish the two disorders (see Chapter 62).19 With use of a standardized scoring system for the diagnosis of autoimmune hepatitis, 7.5% of patients with PSC are characterized as “definite” or “probable” for the diagnosis of autoimmune hepatitis, thereby underscoring the need for cholangiography when PSC is suspected.20 Features suggestive of autoimmune hepatitis include female predominance, a hepatocellular rather than cholestatic pattern of liver biochemical test abnormalities, hypergammaglobulinemia, high titers of antinuclear and anti-smooth muscle antibodies, histologic evidence of

periportal necroinflammation, and clinical response to glucocorticoid therapy. An overlap syndrome between PSC and autoimmune hepatitis has been described; it consists of a mixed cholestatic and hepatocellular pattern of liver biochemical test abnormalities, the presence in serum of autoantibodies including antineutrophil cytoplasmic antibodies (ANCA), cholangiography consistent with PSC, and histologic evidence of periductular fibrosis as well as periportal necroinflammation.21,22 A disorder of the pancreaticobiliary tree termed autoimmune pancreatitis, sclerosing pancreatocholangitis, or immunoglobulin (Ig) G4–associated cholangitis has been described.23 This disorder shares cholangiographic and clinical features with PSC but differs in its responsiveness to glucocorticoid therapy. Serum levels of IgG4 are often elevated in this disorder, and high numbers of IgG4 positive lymphocytes (>20 per high-powered field) are identified in pinch biopsies obtained from the major papilla or bile duct and may be diagnostic. Although specific diagnostic criteria for this disorder are still emerging, persons without IBD who present with symptoms and cholangiographic findings consistent with PSC should undergo measurement of serum IgG4 levels as well as endoscopy and biopsy of the major papilla to exclude IgG4-associated cholangitis (see Chapter 59).23-25

EPIDEMIOLOGY

Determination of the true incidence and prevalence of PSC is complicated by the variable presentation of the disease, inconsistent diagnostic criteria, and referral bias inherent in many published studies. Two population-based studies have provided the most accurate epidemiologic estimates of PSC in Western populations. On the basis of these studies performed in the United States and Norway, the incidence of PSC is estimated to be 0.9 to 1.3 per 100,000, and the point prevalence is estimated to be 8.5 to 13.6 per 100,000.26,27 Although PSC has been diagnosed in neonates and as late as the eighth decade of life, most patients present between the ages of 25 and 45 years, with a mean age of approximately 39 years.19,28-33 Approximately 70% of patients with PSC are men,27-31 but in the subset of patients without IBD, the male-to-female ratio is lower (0.72:1).34 Women with

1155

1156

Section VIII  Biliary Tract PSC are generally older at diagnosis.27,35 PSC is also associated with nonsmoking, an effect that cannot be explained entirely by the association between ulcerative colitis (UC) and nonsmoking.36,37

PRIMARY SCLEROSING CHOLANGITIS AND INFLAMMATORY BOWEL DISEASE

The relationship between PSC and IBD is striking and incompletely understood. Approximately 80% of all patients with PSC have concomitant IBD.27-29,31,38,39 Conversely, PSC is present in 2.4% to 4.0% of all patients with chronic UC and 1.4% to 3.4% of patients with Crohn’s disease.35,38,40,41 Of patients with both PSC and IBD, approximately 85% to 90% have UC and the remainder have Crohn’s colitis or ileocolitis. The association with IBD is stronger with more extensive colonic involvement; the prevalence of PSC is approximately 5.5% in those with pancolitis, in contrast to 0.5% in those with only distal colitis.35 PSC is not thought to occur in association with Crohn’s disease isolated to the small intestine. Racial differences in the association between PSC and IBD may exist; concomitant IBD is seen in only 21% of Japanese patients with PSC.42 Despite the strong association between PSC and UC, the two diseases often progress independently of each other.43 Although IBD is typically diagnosed before PSC, UC may be newly diagnosed years after liver transplantation for endstage liver disease caused by PSC. Conversely, PSC may be diagnosed years after total proctocolectomy for UC.44,45 Whether PSC differs clinically in patients with and without concomitant IBD is unclear. Older reports demonstrated no histologic46 or cholangiographic47 differences between patients with or without IBD. One study,34 however, suggested that patients without IBD are more likely to be female, have disease isolated to the extrahepatic ducts, and be symptomatic at the time of diagnosis. Of the multiple multivariate analyses performed to identify risk factors for progression of PSC (see later), only one found that the presence of IBD has a significant independent effect on progression of PSC.29 Some patients without overt IBD may have subclinical histologic changes detected in the colon or may develop overt colitis at a later date.43 Therefore, a high index of suspicion for the emergence of IBD is warranted, and colonoscopy with random biopsies of the colonic mucosa is recommended for all patients with a new diagnosis of PSC.

ETIOLOGY AND PATHOGENESIS

The etiology and pathogenesis of PSC remain poorly understood. Genetic and immunologic factors appear to play key roles in disease susceptibility and progression. The importance of nonimmunogenetic (infectious, vascular, toxic) factors remains controversial. Currently, the most attractive model of disease pathogenesis postulates that PSC represents an immunologic reaction that develops in immuno­ genetically susceptible persons who are exposed to an environmental or toxic trigger, such as bacterial cell wall products. Any theory of the pathogenesis of PSC must explain the strong association with IBD.

Genetic Factors

The importance of genetic factors in the pathogenesis of PSC is demonstrated by familial occurrence of the disease and its associations with specific human leukocyte antigen (HLA) haplotypes. Although uncommon, familial clustering of cases of PSC have been reported.48,49 Furthermore, PSC is strongly associated with specific HLA haplotypes. Early

studies described an overrepresentation of HLA B8 and DR3 in patients with PSC; these haplotypes are also associated with other autoimmune disorders such as myasthenia gravis and autoimmune hepatitis.50,51 These findings are not explained simply by the association between PSC and IBD because HLA B8 and DR3 are not overrepresented in patients with IBD but without PSC. The subsequent development of molecular genotyping demonstrated that the most common allele in patients with PSC is DRB3*0101, which encodes the DRw52a antigen. One study found this allele in 100% of 29 patients with PSC who underwent liver transplantation,52 but subsequent studies have demonstrated this allele in only 50% to 55% of patients with PSC.53-55 Currently, the extended HLA haplotypes that are most strongly associated with PSC are as follows:53,55,56 B8-TNF*2-DRB3*0101-DRB1*0301-DQA1*0501DQB1*0201; DRB3*0101-DRB1*1301-DQA1*0103-DQB1*0603; and DRB5*0101-DRB1*1501-DQA1*0102-DQB1*0602. Haplotypes associated with protection from PSC include the following: DRB4*0103-DRB1*0401-DQA1*03-DQB1*0302 and MICA*002. The strongest association maps to the HLA class I/III boundary on chromosome 6p21. Strong disease associations have been identified with the MICA*008 allele57 and the tumor necrosis factor a-2 allele.58,59 Despite the multiple HLA associations described, however, a single HLA-encoded gene that determines susceptibility to PSC appears unlikely. More likely are multiple HLA susceptibility loci, which may explain in part why PSC is a relatively rare disease even though the HLA haplotypes associated with PSC are relatively common in populations of Northern European descent. Also controversial is whether specific haplotypes are associated with disease outcomes. One study suggested a poor prognosis in patients with PSC and HLA DR4,60 but this finding was not confirmed.55 A more recent multicenter study involving 256 patients with PSC showed that the heterozygous haplotype DR3,DQ2 was associated with a greater risk of liver transplantation or death and the DQ6 haplotype was associated with a decreased risk of disease progression.61 The relationship between several non-major histocompatibility complex (MHC) genes and susceptibility to PSC has also been investigated. An initial study reported an association with polymorphisms in the gene encoding matrix metalloproteinase 3 (MMP-3) and postulated a role for MMP-3 in progression of PSC because of its ability to regulate fibrosis and immune activation.62 A subsequent report, however, did not confirm an association between either MMP-1 or MMP-3 polymorphisms and PSC.63 Similarly, no associations between PSC and polymorphisms in the interleukin (IL)-1 or IL-10 genes have been noted.64

Immunologic Factors

Evidence suggests that the immune system plays a key role in the etiology and pathogenesis of PSC, including the multiple associations between PSC and other autoimmune disorders. The most frequently associated autoimmune disorders include type I diabetes mellitus and Graves’ disease, which are more common in patients with PSC and IBD than in patients with IBD alone.65 In addition, as described earlier, an overlap syndrome that includes features of both PSC and autoimmune hepatitis has been

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis described.21,22,66 In rare cases, well-characterized autoimmune hepatitis may evolve into sclerosing cholangitis, suggesting that both diseases may be part of the same clinical spectrum.67 Unlike most other autoimmune disorders, however, PSC has an approximately 2:1 male predominance, is not associated with disease-specific autoanti­ bodies, and does not exhibit a consistent clinical response to immunosuppressive therapy. A wide range of serum autoantibodies are found in patients with PSC, although none is specific for the disease. Whether any of these associated antibodies plays a key role in the pathogenesis of the disease process or whether they represent simple epiphenomena is unclear. Antinuclear antibodies may be present in 24% to 53%, anti-smooth muscle antibodies in 13% to 20%, and an atypical perinuclear ANCA (pANCA) in 65% to 88% of patients with PSC.68-75 Antibodies directed against cardiolipin, bactericidal/permeability-increasing protein, cathepsin G, and lactoferrin have also been detected.74-75 Antibodies directed against an epitope shared by colonic and biliary epithelial cells have been demonstrated and may suggest a mechanism for the association between IBD and PSC.76 Autoantibodies that bind to human biliary epithelial cells (anti-BEC) have been shown to induce expression of IL-6 and the cell adhesion molecule CD44; this finding could represent a potential mechanism for the inflammatory bile duct destruction seen in patients with PSC.77 Abnormalities of both humoral and cellular immunity have been described in patients with PSC. They include an increase in circulating immune complexes, deficient clearance of immune complexes, and activation of the classical pathway of the complement system.78-80 Serum elevations of IL-8 and IL-10 also suggest exaggerated humoral immunity.81 Some of the abnormalities in cellular-mediated immunity that have been described include a decrease in circulating CD8+ cytotoxic T cells,82 increased numbers of γδ T cells in peripheral blood as well as portal areas of the liver,83 and overrepresentation of Vβ3 T-cell receptor gene segments in hepatic (but not peripheral) T-cell populations.84

Biliary Epithelial Cells

The role of biliary epithelial cells in the pathogenesis of PSC remains unclear. Biliary epithelial cells could serve as a trigger and a target for immune-mediated injury. Biliary epithelial cells have been shown to express MHC class II antigens85 and adhesion molecules such as intracellular adhesion molecule-1 (ICAM-1)86 and could play a role as antigen-presenting cells to T lymphocytes. The expression of these molecules can be regulated on biliary epithelial cells by various cytokines, including IL-2 and interferon-γ.87 Biliary epithelial cells, however, may not express the costimulatory ligands necessary for activation of T lymphocytes.88 In addition, many of the same findings are seen in patients with PBC and extrahepatic bile duct obstruction as well, making it less likely that they play a primary pathogenic role in PSC.85

Infectious and Toxic Factors

The strong association between PSC and colitis has provoked the theory that penetration of infectious or toxic agents through an inflamed colon into the portal system may play an important role in the pathogenesis of PSC. Bile culture results have been positive in explanted livers in a majority of patients with PSC, although the number of bacterial strains has correlated inversely with the time since the last endoscopic intervention.89 In addition, bacterial endo-

toxin has been shown to accumulate in biliary epithelial cells in patients with PSC and PBC.90 In patients with AIDS cholangiopathy, a variety of organisms, including Cryptosporidium, Microsporidium, and cytomegalovirus, have been isolated from the bile.14,15 A study that evaluated serologic profiles in 41 patients with PSC found a higher percentage with Chlamydia lipopolysaccharide antibodies than in a large control population. No association was seen with any other microorganisms, including Mycoplasma and 22 viruses tested.91 Further study is necessary before a direct link between PSC and Chlamydia, or any other infectious agent, can be established. A loss of normal colonic mucosal barrier because of inflammation could allow portal inflow of noninfectious toxins. Toxic damage leading to sclerosing cholangitis has been demonstrated in humans as well as animal models. Biliary exposure to caustic agents17 or hepatic artery infusion of chemotherapeutic agents such as FUDR16 can produce a cholangiographic appearance identical to that of PSC. In a rat model, administration of the biliary toxin α-naphthylisothiocyanate led to the development of a chronic cholangitis similar to sclerosing cholangitis in humans.92 The toxic injury hypothesis, however, does not explain why PSC is not associated with the severity of colonic inflammation in patients with IBD and why PSC may develop years after a patient has undergone total proctocolectomy.

Vascular Factors

Ischemia has been postulated to play a role in the pathogenesis of PSC because a similar cholangiographic appearance may be found after surgical trauma to the biliary vascular supply93 and after hepatic artery thrombosis or arterial fibrointimal hyperplasia after liver transplantation.94,95 In addition, PSC is associated with the presence of autoantibodies such as pANCA and anti-cardiolipin antibodies. These autoantibodies, in turn, are strongly associated with vasculitides such as Wegener’s granulomatosis, polyarteritis nodosa, and thrombotic syndromes. These associations suggest that immune-mediated vascular injury plays a role in the pathogenesis of PSC.

NATURAL HISTORY AND PROGNOSTIC MODELS

PSC is typically a progressive disease, although the natural history is incompletely understood.29-31,96-98 The disease may be considered to progress through the following four clinical phases, although some phases may not develop or be apparent in an individual patient: 1. Asymptomatic phase: Patients may have cholangiographic evidence of PSC but normal serum liver biochemical values and no symptoms. These patients typically are identified as a result of incidental findings on imaging studies. 2. Biochemical phase: Patients remain asymptomatic but have biochemical abnormalities, typically elevations of serum alkaline phosphatase levels with variable elevations of serum bilirubin and aminotransferase levels. 3. Symptomatic phase: Symptoms of cholestasis or liver injury, or both, develop. Pruritus, fatigue, symptoms of cholangitis, and jaundice may occur, often in combination. 4. Decompensated cirrhosis: The final phase is characterized by worsening symptoms and complications of endstage liver disease, such as ascites, encephalopathy, and variceal bleeding.

1157

1158

Section VIII  Biliary Tract This model of disease progression provides a useful framework for understanding the variability of the natural history in studies of patients with PSC.

Asymptomatic Primary Sclerosing Cholangitis

Asymptomatic patients with PSC make up 15% to 44% of cohorts examined in published studies29,31,32,98,99 Some reports have suggested that asymptomatic patients typically have a benign course of disease. Helzberg and colleagues98 reported on 11 asymptomatic patients with PSC who were followed for a mean of 37 months, and all 11 remained asymptomatic without evidence of progressive disease. By contrast, Porayko and colleagues99 followed 45 asymptomatic patients with PSC for a median of 6.25 years, and during the surveillance period, liver failure, resulting in liver transplantation or death, developed in 13 (31%). Overall, symptoms developed in 24 (53%), and progressive liver disease, demonstrated by new symptoms or signs, worsening cholangiographic findings, or progressive liver histologic abnormalities, developed in 34 (76%) patients. The Kaplan-Meier estimate of median survival free of liver failure in this study was 71% at seven years for the asymptomatic patients, significantly lower than the 96% expected on the basis of an age-, sex-, and race-matched U.S. control population. Differences in the rates of progression between these studies98,99 may be the result of differences in patient populations, the definition of “asymptomatic,” and the duration of clinical follow-up.

Symptomatic Primary Sclerosing Cholangitis

Patients with symptoms at the time of diagnosis generally have a worse prognosis than asymptomatic patients.29,30 The clinical stage is likely more advanced at the time of diagnosis in symptomatic patients, who have more severe biochemical derangements, more abnormalities on cholangiography, and a higher histologic stage on liver biopsy specimens than asymptomatic patients. Wiesner and colleagues29 compared the natural history of PSC in 37 asymptomatic patients with that in 137 patients who were symptomatic at the time of diagnosis. After a mean followup of six years, 55 (40%) of the symptomatic patients had died, compared with 4 (11%) in the asymptomatic group. The Kaplan-Meier estimate of median survival for the entire cohort was 11.9 years; for the symptomatic cohort, the estimated median survival was between 8 and 9 years. Farrant and colleagues30 described the natural history of PSC in 126 patients, of whom 84% were symptomatic. After a median follow-up of 5.8 years, the estimated median survival was 12 years. Similar findings were reported in a large study by Broome and colleagues.31 In 305 patients with PSC followed for a median of 5.25 years, of whom 44% were asymptomatic, the estimated median survival was 12 years. Patients who were symptomatic at the time of entry into the study had a significantly worse expected survival (9.3 years) than asymptomatic patients. A study of 174 patients with PSC by Ponsioen and colleagues97 suggested a better overall prognosis, with a median expected survival of 18 years. The reason for improved survival in this most recent study is not known, but patient data were predominantly from the 1990s, compared with data from the 1970s and 1980s in the other studies described. Although therapeutic advances were not dramatic in the interim, earlier diagnosis in the 1990s may have led to differences in patient selection that appeared to affect outcomes.

Small-Duct Primary Sclerosing Cholangitis

Patients who have histologic, biochemical, and clinical features of PSC but a normal cholangiogram are considered

to have small-duct PSC, which accounts for 5% to 20% of all patients with PSC.3,100 Three studies have performed extended clinical follow-up in patients with small-duct PSC.100-102 In these studies, 12% to 17% of patients progressed to classic large-duct PSC over long-term follow-up, although the true rate may be higher because cholangiograms were not obtained routinely in all patients. Cholangiocarcinoma did not develop in any patient over a median follow-up of 63 to 126 months, and survival in the smallduct PSC group was better than that of matched control groups with classic PSC.100-102 Therefore, small-duct appears to represent an early stage of PSC, may progress to largeduct PSC in a small percentage of patients, and is associated with a better prognosis than classic PSC.

Prognostic Models

Natural history studies have provided insight into specific clinical, biochemical, and histologic features of PSC that may influence prognosis. Whereas early studies described individual factors that were associated with poor survival in PSC, subsequent studies utilized multivariable regression analysis techniques, such as Cox proportional hazards analysis, to develop more sophisticated mathematical models to predict survival. Such models predict expected survival for a specific patient at a specific time. This information is essential for counseling patients about their prognosis and planning future care, such as liver transplantation. Multivariable prognostic models that have been developed to predict survival in patients with PSC are shown in Table 68-2. In an early multivariable analysis, hepatomegaly and a serum bilirubin level > 1.5 mg/dL were found to be independently associated with a poor prognosis in PSC. The patient’s age, histologic findings, presence of concomitant IBD, and pattern of cholangiographic involvement did not correlate independently with survival in this study.98 Wiesner and colleagues29 developed a prognostic model based on age, serum bilirubin level, hemoglobin value, presence or absence of IBD, and histologic stage. With this model, three risk groups (low, intermediate, high) were formed, and predicted survival curves were shown to be similar to observed survival curves. Farrant and colleagues30 developed a multivariable prognostic model in which hepatomegaly, splenomegaly, serum alkaline phosphatase level, histologic stage, and age at presentation were found to be important independent factors. Dickson and colleagues103 then presented a model developed from a multicenter collaboration in which data from 426 patients with PSC were pooled. In this analysis, the patient’s age, serum bilirubin level, histologic stage, and presence of splenomegaly were found to correlate independently with survival, and the model was validated against the observed survival data in a subgroup of the entire cohort. Broome and colleagues31 found the patient’s age, histologic stage, and serum bilirubin level to be independent predictors of survival in 305 patients with PSC, but this prognostic model was not validated independently. Most recently, Kim and colleagues,104 using easily obtainable clinical and biochemical factors, revised an earlier predictive model that did not require liver biopsy and did not rely on subjective physical findings such as splenomegaly or hepatomegaly. This revised natural history model (revised Mayo risk model) found the patient’s age, serum bilirubin level, serum aspartate aminotransferase (AST) level, and serum albumin level and a history of variceal bleeding to be independent predictors of survival. The model was generated from data on 529 patients from 5 centers and was validated using data from another center that had not been used in the development of the model.

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis

Table 68-2  Independent Predictors of Survival and Prognostic Index Formulas Used in Natural History Models of Primary Sclerosing Cholangitis* MAYO CLINIC MODEL29 Predictors of Survival† Age Bilirubin Histologic stage Hemoglobin IBD Prognostic Index Formula‡ R = 0.06 × age + 0.85 × loge(min[bilirubin or 10]) 4.39 × loge(min[hemoglobin or 12]) + 0.51 × histologic stage + 1.59 × indicator for IBD

KING’S COLLEGE MODEL30

MULTICENTER MODEL103

REVISED MAYO MODEL104

Age Hepatomegaly Histologic stage Splenomegaly Alkaline phosphatase

Age Bilirubin Histologic stage Splenomegaly

Age Bilirubin Albumin AST Variceal bleeding

R = 1.81 × hepatomegaly + 0.88 × splenomegaly + 2.66 × log(alk phos) + 0.58 × histologic stage + 0.04 × age

R = 0.535 × loge(bilirubin) + 0.486 × histologic stage + 0.041 × age + 0.705 × splenomegaly

R = 0.03 × age + 0.54 × loge(bilirubin) + 0.54 × loge(AST) + 1.24 × variceal bleeding − 0.84 × albumin

*Superscript numbers indicate references. † Age expressed in years, bilirubin in mg/dL; hemoglobin in gm/dL; alkaline phosphatase in U/L; AST in U/L; albumin in g/dL. ‡ Values for IBD, hepatomegaly, splenomegaly, and variceal bleeding are 1 if present, 0 if absent. Alk phos, alkaline phosphatase; AST, aspartate aminotransferase; IBD, inflammatory bowel disease; min, minimum of; R, risk score.

The Child (Child-Pugh) classification may also be used to predict survival in patients with PSC (see Chapter 90). Shetty and colleagues105 found that Kaplan-Meier sevenyear survival rates for patients with Child class A, B, and C cirrhosis caused by PSC were 89.8%, 68%, and 24.9%, respectively. Subsequent evaluation, however, suggested that the Child classification is less accurate than the revised Mayo risk model, especially for patients with early-stage PSC.106 Despite the cumbersome nature of the mathematical formulas used in the various natural history models, these models can be useful in the clinical care of patients with PSC. They may facilitate selection for and timing of liver transplantation by comparing predicted survival with readily available post-liver transplantation survival rates. The availability of multiple models with differing prognostic variables, however, may be confusing in clinical practice. The models also may not account for other clinical events, such as the development of cholangiocarcinoma or variceal bleeding, that may affect prognosis in patients with PSC. Further refinement of these prognostic models, including consensus on the use of specific prognostic variables, may ultimately clarify their role in clinical practice.

CLINICAL FEATURES Symptoms

The initial clinical presentation of PSC can be quite varied and may run the gamut from asymptomatic elevations of serum alkaline phosphatase levels to decompensated cirrhosis with jaundice, ascites, hepatic ence­ phalopathy, or variceal bleeding. The most common symptoms at the time of presentation include jaundice, fatigue, pruritus, and abdominal pain.19,28-33,107,108 Other associated symptoms may include fever, chills, night sweats, and weight loss (Table 68-3). The onset of these symptoms is typically insidious, although an acute hepatitis-like presentation has been described.40 Increasingly, PSC is diagnosed in an asymptomatic or minimally symptomatic stage. Large series have shown that 15% to 44% of patients with PSC are asymptomatic at the time of diagnosis,29,31,32,98,99 probably because of the routine liver biochemical screening in patients with IBD, as well as the

Table 68-3  Most Common Symptoms and Signs at the Time of Diagnosis of Primary Sclerosing Cholangitis Symptoms Fatigue Abdominal pain Pruritus Fever/night sweats Asymptomatic Weight loss Signs Jaundice Hepatomegaly Splenomegaly Hyperpigmentation Ascites

Rate (%) 65-75 24-72 15-69 13-45 15-44 10-34 30-73 34-62 32-34 14-25 4-7

Data from references 19, 29-33, 98, 99, 107, 108.

widespread availability of MRCP and ERCP for evaluating elevated serum alkaline phosphatase levels. Symptoms of PSC are often intermittent. Episodes of pruritus, jaundice, abdominal pain, and fever are typically interspersed with asymptomatic periods of varying duration.39,107,108 The intermittency of the symptoms is thought to reflect intermittent biliary obstruction caused by microlithiasis and sludge.5,109 This obstruction may predispose to cholestasis and induce an acute inflammatory reaction. Secondary bacterial infection may result in low-grade cholangitis and predispose to pigment stone formation.5

Physical Examination

Physical findings may be normal in patients with PSC, particularly those who are asymptomatic. When physical abnormalities are present, the most common include hepatomegaly, jaundice, and splenomegaly (see Table 68-3). Skin findings are common and include cutaneous hyperpigmentation, excoriations resulting from pruritus, and xanthomata. As liver disease progresses, spider angiomas, muscular atrophy, peripheral edema, ascites, and other signs of advanced liver disease may appear.28-30

1159

1160

Section VIII  Biliary Tract Laboratory Findings

Chronic elevation of serum alkaline phosphatase levels, typically three to five times normal, is the biochemical hallmark of PSC. A normal alkaline phosphatase level, however, may be found in up to 6% of patients with cholangiographically proved PSC.110,111 In some cases, an advanced histologic stage has been demonstrated on a liver biopsy specimen despite normal serum alkaline phosphatase levels.110 Serum aminotransferase levels are typically elevated, although rarely above four to five times normal except in the pediatric population.112 The serum bilirubin level may be normal or elevated and often fluctuates. When the serum bilirubin level is elevated, the bilirubin is predominantly conjugated. Reductions in the serum albumin level and prolongation of the prothrombin time may reflect hepatic synthetic dysfunction with advanced liver disease. In addition, malnutrition and underlying IBD may lower serum albumin levels. Vitamin K malabsorption related to cholestasis may play a role in prolonging the prothrombin time. Other nonspecific consequences of cholestasis are elevations in serum copper, serum ceruloplasmin, and hepatic copper levels, increased urinary copper excretion, and elevated serum cholesterol levels. Several immunologic markers and serum autoantibodies are found in the majority of patients with PSC, although none is specific for the disease. Hyperglobulinemia is frequent; serum IgM levels are elevated in up to 50% of patients, and IgG and IgA levels also may be elevated.28,29,28,107 Antinuclear antibodies, often in low titer, may be detected in 24% to 53% of patients. Anti-smooth muscle antibodies are found in 13% to 20% of patients, but antimitochondrial antibodies are found in less than 10%.26,28,70,73,75 Most commonly found in patients with PSC are pANCA,72 which are detected in 65% to 88% of patients and appear to react to a heterogeneous group of antigens.70,71,74,75 These antigens have been found to represent neutrophil nuclear envelope proteins predominantly, and the corresponding antibodies have been referred to as “antineutrophil nuclear antibodies” (ANNA).113 In contrast to Wegener’s granulomatosis, titers of pANCA do not appear to correlate with disease activity, severity, or response to medical therapy in patients with PSC.72 Furthermore, the presence of autoantibodies does not appear to differ in patients with and without IBD. Anticardiolipin antibodies are detected in 66% of patients with PSC, and the titer has been reported to correlate with disease severity.75 In general, despite the high frequency of autoantibodies in patients with PSC, a clear association between the presence of these antibodies, pathogenesis of the disease, and prognosis or response to treatment remains unproved. Measurement of autoantibodies is therefore of limited clinical value in patients with PSC.

Imaging Findings

Cholangiography by ERCP, MRCP, or percutaneous THC establishes a diagnosis of PSC and provides information regarding the distribution and extent of disease. The characteristic cholangiographic findings include multifocal stricturing and ectasia of the biliary tree. Areas of narrowing are interspersed with areas of normal or near-normal caliber and areas of post-stenotic dilatation. The result is a classic “beaded” appearance to the biliary tree. The strictured segments are usually short, annular, or band-like in appearance (Fig. 68-2), although longer confluent strictures may be seen in more advanced disease. Localized segments of dilated ducts may have a saccular or diverticular appearance. Major areas of focal, tight narrowing known as dominant strictures may be seen and often involve the bifurcation of the hepatic duct.114 At times, diffuse involvement of the intrahepatic

A

B Figure 68-2.  Endoscopic retrograde cholangiopancreatography (ERCP) in two patients with primary sclerosing cholangitis (PSC). A, ERCP with contrast injected through a balloon catheter (seen in distal bile duct). The intrahepatic ducts are mainly affected and show diminished arborization (pruning), with diffuse segmental strictures alternating with normal-caliber or mildly dilated duct segments (cholangiectasias), resulting in a beaded appearance. B, ERCP in a patient with PSC. Radiologic features include diffuse irregularity of the intrahepatic ducts, multiple short strictures and cholangiectasias, small diverticula in the wall of the common hepatic duct (arrow), and clips from a prior cholecystectomy.

biliary tree may give a pruned appearance that is difficult to distinguish from the diffuse intrahepatic duct attenuation seen in patients with cirrhosis of any cause; irregularity of the duct wall or concomitant involvement of the extrahepatic bile duct supports a diagnosis of PSC. Both the extrahepatic and intrahepatic bile ducts are abnormal in approximately 75% of cases. The intrahepatic ducts alone may be involved in 15% to 20% of cases.28,31,35,107 Abnormalities of the extrahepatic biliary tree in the absence of intrahepatic involvement are less common.98,99 The cystic duct and gallbladder may be involved in up to 15% of patients but may not be well visualized on routine cholangiography.115 Pancreatic duct irregularities similar to those seen in chronic pancreatitis may rarely be noted.116

PATHOLOGY

Gross and histologic specimens from the extrahepatic bile ducts demonstrate a diffusely thickened, fibrotic duct wall.

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis

A

B Figure 68-3.  Liver histopathology in primary sclerosing cholangitis. A, A segmental bile duct is completely obliterated by fibrosis (arrow), demonstrating “fibro-obliterative cholangitis.” (Hematoxylin and eosin, ×200.) B, A medium-sized bile duct is surrounded by concentric fibrosis with an onionskin appearance. (Hematoxylin and eosin, ×400.) (A and B, Courtesy of Matthew Yeh, MD, PhD, Seattle, Wash.)

The fibrosis is accompanied by a mixed inflammatory infiltrate that may involve the epithelium and biliary glands.117,118 Florid hyperplasia of the biliary glands with accompanying neural proliferation has been described.119 Examination of PSC explants removed at the time of liver transplantation has demonstrated areas of thin-walled saccular dilatation, termed “cholangiectasias,” that correspond to the beaded appearance on cholangiography.120 A wide range of liver biopsy findings may be seen in patients with PSC. For this reason, histologic findings are not typically diagnostic for PSC. The characteristic bile duct lesion is a fibro-obliterative process that may lead to an “onionskin” appearance of concentric fibrosis surrounding medium-sized bile ducts (Fig. 68-3); however, this finding is seen in less than one half of biopsy specimens.117,121,122 The smaller interlobular and septal bile duct branches may be entirely obliterated by this process, resulting in fibroobliterative cholangitis. This finding is present in only 5% to 10% of biopsy specimens but is thought to be virtually pathognomonic of PSC.121 In this process, the biliary epithelium may degenerate and atrophy and be replaced entirely by fibrous cords. Other characteristic histopathologic find-

ings may include bile duct proliferation, periductal inflammation, and ductopenia. The degree of inflammation can be quite variable but is typically a portal-based mixture of lymphocytes, plasma cells, and neutrophils with a periductal focus. Lymphoid follicles or aggregates may also be seen.121,122 The histologic progression of PSC can be divided into four stages, analogous to a similar staging system in PBC46 (See Chapter 89). In stage I (portal stage) changes are confined to the portal tracts and consist of portal inflammation, connective tissue expansion, and cholangitis. Stage II (periportal stage) is characterized by expansion of inflammatory and fibrotic processes beyond the confines of the limiting plate, resulting in “piecemeal necrosis” (interface hepatitis) and periportal fibrosis. Depending on the degree of biliary obstruction, ductular proliferation and cholangitis may be of varying severity. Stage III (septal stage) is characterized by fibrous septa that bridge from one portal tract to the next. Bridging necrosis may occasionally be seen but is uncommon. Stage IV (cirrhotic stage) implies progression to biliary cirrhosis. The degree of inflammatory activity may subside as the stage of disease progresses, and focal bile ductular proliferation may be striking. A study that examined the time course of progression through the histologic stages revealed that for patients with stage II disease, 42%, 66%, and 93% progressed to a higher histologic stage at one, two, and five years, respectively.123 For patients initially with stage II disease, progression to biliary cirrhosis (stage IV) occurred in 14%, 25%, and 52%, respectively. Regression of stage was observed in 15% of patients and probably reflected sampling variability in the histologic assessment. Many of the histologic findings of PSC are nonspecific and may be seen in other disorders. In particular, the histologic distinction between PSC and PBC may be difficult to discern. In one study, histologic examination could classify only 28% of patients who had one of the two diseases.18 When lymphocytic interface hepatitis is prominent, the distinction from autoimmune hepatitis may be challenging, especially because hypergammaglobulinemia and autoantibodies may be present in both conditions. In addition, an overlap syndrome with features of both PSC and autoimmune hepatitis has been described.21,22,66 When severe cholestasis develops, hepatic copper accumulation can be dramatic and may mimic that seen in Wilson disease.124

COMPLICATIONS Cholestasis

The complications associated with all causes of chronic cholestasis may develop in patients with PSC (see also Chapters 20 and 89). Pruritus is one of the most common symptoms of PSC and may adversely affect a patient’s quality of life. Severe excoriations and debilitating symptoms may develop. The pathogenesis of pruritus in chronic cholestasis is poorly understood, and response to therapy is inconsistent (see Chapter 89). The accumulation of bile acids in the plasma and tissue of cholestatic patients has been cited as a potential cause of pruritus, and the pruritus of cholestasis is typically treated with oral administration of bile-acid binding resins such as cholestyramine. Not all patients with elevated serum bile acid levels itch, however. In addition, pruritus is frequently intermittent, despite the relative stability of serum bile acid levels. Several lines of evidence suggest that cholestasis is asso­ ciated with an increased level of endogenous opioids. In animal models, cholestasis is associated with an increase in plasma levels of endogenous opioids.125 In humans, cholestatic patients may experience opiate withdrawal-like

1161

1162

Section VIII  Biliary Tract symptoms after the administration of an opioid antagonist. In addition, administration of naloxone and naltrexone, which have opioid antagonist properties, has been reported to relieve pruritus in cholestatic patients in small clinical trials.126,127 Nutritional deficiencies may complicate chronic chole­ stasis in patients with PSC. Intestinal absorption of the fatsoluble vitamins A, D, E, and K is particularly affected and is thought to be related to decreased intestinal concentrations of conjugated bile acids.128 Concomitant disease such as IBD, chronic pancreatitis, and celiac disease may also contribute to intestinal malabsorption. Clinical consequences include night blindness (vitamin A deficiency) and coagulopathy (vitamin K deficiency). The importance of metabolic bone disease, also referred to as hepatic osteodystrophy, is often underrecognized in patients with PSC. Two forms of metabolic bone disease may develop: osteomalacia and osteoporosis. With improvements in nutritional management, osteomalacia (decreased bone mineralization) is now relatively rare, and most bone disease in cholestatic patients is osteoporosis. Bone mineral density is significantly lower in patients with PSC than in age- and sex-matched controls.129 The pathogenesis of bone density loss in PSC and other chronic cholestatic liver diseases is unknown. Intestinal malabsorption of vitamin D is probably not the primary abnormality because serum vitamin D levels are often normal and vitamin D repletion does not usually have a major impact on the severity of osteoporosis. In patients with concomitant IBD, the use of glucocorticoids may play a role in exacerbating bone loss in patients with PSC. Overall, severe osteoporosis is less common in patients with PSC than in those with PBC because a majority of patients with PSC are young men who have a higher baseline bone mineral density and a slower rate of bone loss than middle-aged women, who account for most cases of PBC.

Biliary Stones

Cholelithiasis and choledocholithiasis are more common in patients with PSC than in the general population. Gallstones are found in approximately 25% of patients with PSC and are often pigmented calcium bilirubinate stones.5 Biliary strictures may predispose to bile stasis and intraductal sludge and stone formation. Ultrasonography has only an intermediate sensitivity for detecting intraductal stones. Therefore, patients with PSC and worsening cholestasis or jaundice should undergo ERCP to distinguish biliary stone disease from the development of a dominant stricture or cholangiocarcinoma.

Cholangiocarcinoma

Cholangiocarcinoma is a feared complication of PSC and can arise from bile duct epithelium anywhere in the biliary tract (see Chapter 69). PSC should be considered a premalignant condition of the biliary tree, analogous to the relationship between UC and carcinoma of the colon. The reported frequency of cholangiocarcinoma in patients with PSC has ranged from 6% to 11% in natural history studies and from 7% to 36% in patients with PSC who undergo liver transplantation.130,131 Tumors are most commonly found in the common hepatic duct and perihilar region but may involve only the bile duct, intrahepatic ducts, cystic duct, or gallbladder. Cholangiocarcinoma remains the leading cause of death in patients with PSC. The pathogenesis of cholangiocarcinoma in PSC is poorly understood. Although cholangiocarcinoma may complicate any stage of the disease, chronic inflammation is thought to predispose to epithelial dysplasia and an increased risk of

malignant transformation. The role of chronic inflammation is supported by the observation that patients with chronic Clonorchis sinensis and other liver fluke infections are also at increased risk of cholangiocarcinoma (see Chapter 82).132 A role for proinflammatory cytokines in stimulating oxidative DNA damage and inactivation of DNA repair processes has been postulated. Biliary malignancy should be suspected when a patient with PSC exhibits rapid clinical deterioration with worsening jaundice, weight loss, and abdominal pain. Advanced PSC without cholangiocarcinoma may present with the identical clinical presentation. The diagnosis of cholan­ giocarcinoma presents a particular challenge in patients with PSC. A malignant biliary stricture may be indistinguishable cholangiographically from the underlying PSC (Fig. 68-4). Because of the tendency of cholangiocarcinoma to grow in sheets as opposed to a discrete mass, crosssectional imaging with CT or magnetic resonance imaging (MRI) may be insensitive for detection of cholangiocarcinoma. Several serum tumor markers of cholangiocarcinoma have been evaluated. Serum CA 19-9 has been the most commonly utilized tumor marker, with one small study reporting a sensitivity of 89% and specificity of 86% for a serum CA 19-9 level greater than 100 U/mL in diagnosing cholangiocarcinoma.133 A later study from the same group found a lower sensitivity for serum CA 19-9 but a correlation between the CA 19-9 level and tumor stage; no tumor was resectable in patients with a CA 19-9 level greater than 1000 U/mL.134 Another study described a biochemical index using CA 19-9 and carcinoembryonic antigen (CEA) levels (CA 19-9 + (CEA × 40) > 400), with a reported sensitivity of 86% for cholangiocarcinoma.135 More recent studies, however, suggested a poor sensitivity (33%) for this combined biochemical index despite a relatively high specificity (85%).136,137 Obtaining an adequate tissue sample presents a particular challenge in the diagnosis of cholangiocarcinoma. Dominant strictures resulting from PSC may be indistinguishable cholangiographically from those harboring cholangiocarcinoma. Brush cytology can be obtained at the time of ERCP, but the sensitivity of this approach is only 50% to 60% at best.138,139 False-positive results are also possible because chronically inflamed cells may take on a malignant cytologic appearance. The addition of a sampling technique to brush cytology, such as endobiliary biopsy or fine-needle aspiration (FNA), improves sensitivity,138-140 and when clinical or cholangiographic suspicion for cholangiocarcinoma is high, two tissue sampling techniques should be used. The sensitivity of cytologic examination is increased by use of specialized techniques such as fluorescent in situ hybridization and digital image analysis. Endoscopic ultrasound (EUS) with FNA has an emerging role in the evaluation of suspected cholangiocarcinoma when brush cytology and other methods have failed to yield a diagnosis.141,142 The sensitivity of EUS-FNA for diagnosing cholangiocarcinoma in patients with PSC has been reported to be as high as 89%.141,142 High-frequency intraductal ultrasound (IDUS) is under study as a means of discriminating benign from malignant dominant bile duct strictures, and several studies143-145 have demonstrated that the addition of IDUS to ERCP improves the ability to distinguish benign from malignant dominant bile duct strictures in patients with PSC. Direct endoscopic visualization of bile duct strictures is also possible with the availability of choledoscopy (cholangioscopy) in clinical practice. A single small study has shown that the addition of choledoscopy to ERCP in patients with PSC and a dominant stricture

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis The development of cholangiocarcinoma is an ominous sign, with a median survival of five months after diagnosis.6 In addition, recurrence of cholangiocarcinoma after liver transplantation is nearly universal, and cholangiocarcinoma is generally considered a contraindication to liver transplantation. Early identification of patients with PSC who are at high risk of cholangiocarcinoma is crucial so that liver transplantation may be undertaken before the bile duct cancer develops. One study reported a strong association between current or former smoking and the development of cholangiocarcinoma in patients with PSC.147 A second study, however, could not confirm an association with smoking.130 In addition, duration of PSC, distribution of biliary strictures, and past medical or surgical intervention do not appear to be associated with an increased risk of cholangiocarcinoma. Longer duration of IBD has been shown to be a risk factor for cholangiocarcinoma in some,148 but not all,130 studies. Overall, therefore, studies to date have not defined clear risk factors for cholangiocarcinoma that are clinically useful for identifying patients with PSC at particularly high risk. Nor has the optimal surveillance strategy been delineated.

A

B Figure 68-4.  Cholangiographic progression of cholangiocarcinoma complicating primary sclerosing cholangitis (PSC). This 43-year-old man with a history of a mild ulcerative colitis for 3 years presented with pruritus and an elevated serum alkaline phosphatase level. A, Initial endoscopic retrograde cholangiopancreatography (ERCP) showed mild, diffuse changes compatible with uncomplicated PSC. Several short, annular strictures and cholangiectasias are present in the intrahepatic ducts (arrows), with a single, short, annular stricture of the bile duct. The cystic duct is not filled and may already be obstructed by tumor. B, Second ERCP performed seven months later after progressive jaundice and weight loss developed rapidly. Now a 2-cm mass is projecting into, and obstructing, the common hepatic duct (arrows). A catheter has been passed beyond the obstructing mass. There is marked dilatation of the left main duct proximal to the obstruction; the right main duct is completely occluded.

enhanced detection of malignancy, with an overall sensitivity of 92%.146 Therefore, in patients with PSC, suspected cholangiocarcinoma, and negative brush cytology or endobiliary biopsy results, repeat ERCP plus EUS-FNA, IDUS, or choledoscopy improves the sensitivity of diagnosing cholangiocarcinoma.

Colonic Neoplasia

Most patients with PSC also have IBD; the majority have ulcerative pancolitis. UC is known to be associated with an increased risk of colonic dysplasia and carcinoma (see Chapter 112); the risk of colon cancer increases with the duration, extent, and severity of colitis. For patients with pancolitis, the cumulative risk of colon cancer is approximately 5% to 10% after 20 years and 12% to 20% after 30 years.149-151 A growing body of evidence suggests that patients with concomitant PSC and UC are at significantly higher risk for developing colonic neoplasia (dysplasia or carcinoma) than patients with UC alone.152-157 In one report, 132 patients with UC and PSC were compared with a randomly selected historical cohort of patients with UC but without PSC.155 Colonic carcinoma or dysplasia developed in significantly more patients with both UC and PSC than those with UC alone (25% versus 6%), and significantly more deaths from colorectal cancer were observed in the patients with PSC (4.5% versus 0%). A similarly designed study included 20 patients with both UC and PSC and 25 matched controls with UC alone.153 Colonic dysplasia was observed significantly more often in patients with both UC and PSC (45% versus 16%), although the time to the development of dysplasia was similar in the two groups. Patients with PSC and UC were also more likely than patients with UC alone to have synchronous sites of dysplasia in the colon. Another study examined 40 patients with PSC and UC matched with 80 control subjects with UC alone.152 The cumulative risk of colorectal dysplasia or carcinoma in patients with both UC and PSC was 9%, 31%, and 50% after 10, 20, and 25 years, respectively. These rates were significantly higher than those for the control group (2%, 5%, and 10%, respectively). The mechanisms by which PSC confers an added risk of colonic neoplasia are not well understood. A high colonic concentration of secondary bile acids may play a role because patients with UC and colonic dysplasia or carcinoma have higher fecal bile acid concentrations than patients with UC who do not have dysplasia or carcinoma.158 This theory is supported by the higher incidence of rightsided colon cancer in patients with UC and PSC than in those with UC alone.155,156 In the study by Shetty and colleagues,155 76% of the colon cancers in patients with UC and PSC occurred proximal to the splenic flexure, compared with only 20% in patients with UC alone.155 Increased

1163

1164

Section VIII  Biliary Tract colonic secondary bile acid concentrations may also explain the apparent chemoprotective effect of ursodeoxycholic acid (UDCA) against the development of colonic neoplasia. Two studies have reported that UDCA use is associated with a lower risk of colonic dysplasia or cancer in patients with UC and PSC.159,160 UDCA may confer protection against colonic neoplasia by reducing colonic concentrations of secondary bile acids, as well as by affecting expression of protein kinase C isoforms, metabolism of arachidonic acid, and expression of cyclooxygenase-2.159-162 These data need to be confirmed in larger-scale, prospective trials before UDCA use can be recommended routinely for preventing colon cancer in patients with PSC. Whether UDCA has a chemoprotective effect in patients with UC but without PSC is unknown. Patients with PSC who have UC should undergo annual colonoscopic surveillance for the detection of colonic dysplasia or cancer. As for colonoscopic surveillance in patients with UC alone, multiple mucosal biopsy specimens should be obtained (see Chapter 112).163 Most experts agree that colonoscopic surveillance should start immediately after the diagnosis of PSC. Surveillance should continue even after liver transplantation because these patients remain at increased risk for colonic neoplasia.164-166

Peristomal Varices

Varices at the stoma may develop in patients with PSC and portal hypertension who have previously undergone proctocolectomy with ileostomy for IBD.167,168 These varices may bleed spontaneously, and the bleeding may be dramatic. Treatment modalities that may initially be effective in achieving hemostasis include injection sclerotherapy,169 percutaneous transhepatic coil embolization,170 surgical stomal revision,171 and transjugular intrahepatic portosystemic shunt placement (see Chapter 90).172 Nevertheless, recurrent bleeding is common, and liver transplantation should be considered to relieve portal hypertension and treat the underlying liver disease.

TREATMENT

Except for liver transplantation, no specific therapy has proved effective for treating PSC. The objectives of management should be to treat the complications of disease, such as flares of bacterial cholangitis, jaundice, and pruritus, and prevent complications, such as osteoporosis and nutritional deficiencies. Other complications such as cholangiocarcinoma and liver failure should be diagnosed as early as possible to allow the possibility of treatment.

Medical Treatment of Underlying Disease

A wide variety of medications have been studied in patients with PSC (Table 68-4). Many of the published studies have been small and uncontrolled with limited follow-up. Because of the varied course of PSC, with spontaneous remissions and unpredictable flares, adequate clinical trials in patients with PSC require long-term follow-up. In addition, the defined study endpoints, whether clinical, biochemical, histologic, or a mathematical risk score, have varied greatly among published studies. To date, no medical treatment has been shown clearly to alter the course of PSC. UDCA has been the most extensively studied drug in patients with PSC. At least five controlled clinical trials have been reported, with varying doses of UDCA.173-177 The mechanisms by which UDCA is thought to exert a beneficial effect in cholestatic conditions include protection of cholangiocytes against cytotoxic hydrophobic bile acids, stimulation of hepatobiliary secretion, protection of hepatocytes against bile-acid induced apoptosis, and induction of anti-

Table 68-4  Medical Therapy for Primary Sclerosing Cholangitis* NO PROVEN BENEFIT

POSSIBLE BENEFIT

Antibiotics Cholestyramine Glucocorticoids183,184 Azathioprine Methotrexate187 Cyclosporine Tacrolimus188 Pentoxifylline189 Colchicine191 d-penicillamine192 Nicotine193,194 Perfenidone195

Ursodeoxycholic acid (20-30 mg/kg/day)173-177,180

*Superscript numbers indicate references.

oxidants (see also Chapter 89).178,179 In a randomized, controlled trial by Beuers and colleagues,173 patients with PSC treated with UDCA, 13 to 15 mg/kg/day for one year, had significant improvements in biochemical and histologic endpoints in comparison with those given placebo, but no effect on symptoms was noted. In the largest controlled trial of UDCA for PSC to date, Lindor and colleagues176 randomized 105 patients to UDCA, 13 to 15 mg/kg/day, or placebo for a median of 2.2 years. Significant biochemical improvements were seen with UDCA therapy, but no difference was seen in the primary endpoint, which was a composite of death, liver transplantation, histologic progression of at least two stages, hepatic decompensation, or quadrupling of the serum bilirubin level. Because of the disappointing results with standard-dose UDCA, several groups have published results on the use of high-dose UDCA in the treatment of PSC. Mitchell and colleagues177 performed a two-year controlled trial of high-dose UDCA, 20 mg/kg/day, versus placebo in 26 patients with PSC. Compared with placebo, treatment with high-dose UDCA led to improvement in biochemical markers and a reduction in the progression of fibrosis and cholangiographic changes. In addition, high-dose UDCA was well tolerated, with no significant adverse events. Harnois and colleagues180 performed an open-label pilot study of UDCA, 25 to 30 mg/kg/day, for one year in 30 patients with PSC and compared changes in the revised Mayo risk score with those in patients from a previous randomized trial of standard-dose UDCA. After one year of treatment, improvement in the revised Mayo risk score in the high-dose UDCA group was significantly greater than that seen in placebotreated patients from the prior study but no better than that in the patients treated with standard-dose UDCA. With the encouraging results of these two small studies, a larger study was undertaken in which 219 patients with PSC were randomized to high-dose UDCA or placebo. After five years, no differences in symptoms, liver biochemical abnormalities, quality of life, or transplant-free survival were found between the two groups.181 A subsequent study in 31 patients randomized to low-dose, standard-dose, and high-dose UDCA found improvement in the revised Mayo risk score in all groups, with statistical significance only in the group receiving high-dose UDCA.182 Therefore, the precise benefit of UDCA in general, and high-dose UDCA in particular, remains uncertain. Given the immunologic alterations in patients with PSC, immunosuppressive therapy would appear to be a reasonable consideration. Glucocorticoids, administered both

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis orally and via nasobiliary lavage, have not shown a clear benefit in uncontrolled studies,183,184 although subgroups of patients may have a clinical response.185 Lack of long-term data demonstrating clear response and concerns about longterm adverse effects, including exacerbation of metabolic bone disease, have limited the use of glucocorticoids. Oral budesonide, a newer glucocorticoid with limited systemic toxicity, has been evaluated in an uncontrolled pilot study in 21 patients with PSC.186 After one year of therapy, treated subjects had minimal biochemical or histologic improvement and no change in revised Mayo risk score. In addition, significant loss of bone mass was seen with use of budesonide. Other immunomodulators also have been evaluated. In a small prospective, controlled trial of methotrexate, no biochemical, histologic, or cholangiographic differences from therapy with placebo were seen after two years of treatment.187 In a small pilot study of tacrolimus therapy, significant biochemical improvement was observed after one year, but no change in cholangiographic or histologic severity was documented.188 Pentoxifylline, which inhibits tumor necrosis factor-α (TNF-α), led to no biochemical or symptomatic improvements in a one-year pilot study in 20 patients with PSC.189 Etanercept, a recombinant inhibitor of TNF-α, also showed no benefit in a small number of patients with PSC.190 Colchicine has been evaluated as a potential therapy for PSC because of its antifibrogenic potential. In a randomized, controlled clinical trial comparing colchicine therapy with placebo for three years, no differences were seen between the two groups in rates of mortality or liver transplantation or in symptoms or biochemical and histologic findings.191 D-penicillamine has also been studied in a randomized, controlled clinical trial192 because of the increased hepatic copper concentrations seen in patients with PSC and other chronic cholestatic conditions. In addition to its cupruretic effects, penicillamine may have antifibrogenic and immunosuppressive properties. Therapy with penicillamine therapy for three years, however, led to no difference in mortality or in biochemical or histologic progression as compared with therapy with placebo. In addition, penicillamine was associated with substantial toxicity. Other studies have failed to demonstrate a significant response to nicotine193,194 or the antifibrotic drug pirfenidone.195 Finally, combinations of various agents such as azathioprine, glucocorticoids, UDCA, and antibiotics have been studied in a limited fashion.196-198 The results of these studies have been mixed, with some showing no benefit and others demonstrating histologic improvement in small numbers of patients. A problem with combining agents is an increased risk of adverse drug reactions.

Medical Treatment of Complications

An important component in the medical care of patients with PSC is the management of complications of the disease, such as pruritus, nutritional deficiencies, and bacterial cholangitis. As discussed earlier, therapy with UDCA does not have a consistent effect on pruritus, although some patients may notice symptomatic improvement. Although treatment with antihistamines may improve pruritus, anion-exchange resins such as cholestyramine, colestipol hydrochloride, or colesevelam are typically more effective, although compliance is a problem with the use of bile acid resins. These drugs are relatively unpalatable, frequently produce con­ stipation, and may interfere with the absorption of other medications. Rifampin appears to be an effective and safe alternative for patients who do not respond to the preceding measures.199,200 Opiate antagonists such as naloxone and

naltrexone have also been shown to be effective for cholestatic pruritus, although self-limited episodes of opioid withdrawal-like symptoms may occur.126,127,201 Patients who are unresponsive to these measures and who do not obtain relief from endoscopic therapy of a dominant stricture (see later) may need to be considered for plasmapheresis (which has shown benefit in anecdotal reports) or even liver transplantation (see also Chapter 89). Patients with PSC should be screened for nutritional deficiencies by measurement of fat-soluble vitamin levels and the prothrombin time. In most patients, vitamin supplements are given orally, but a parenteral route may be necessary in patients with severe intestinal fat malabsorption. Administration of vitamin A is usually effective for correcting subclinical vitamin A deficiency. Correction of vitamin D deficiency, with or without calcium supplements, is of unproven benefit in cholestatic liver disease but is generally recommended because of its safety.202 The use of bisphosphonates and other agents to promote bone formation requires further study in patients with PSC and osteoporosis. In patients with PSC, prolongation of the prothrombin time is more likely to be the result of advanced liver disease than of vitamin K deficiency, although a trial of oral vitamin K is warranted in patients with coagulopathy (see Chapters 89 and 92). Bacterial cholangitis may develop spontaneously or after manipulation of the biliary tree. In some patients, recurring bouts of bacterial cholangitis can be debilitating. Antibiotic prophylaxis is indicated in any patient with known or suspected PSC who undergoes manipulation of the biliary tree via ERCP, percutaneous THC, or surgery. Typically, a fiveto seven-day course of a broad-spectrum antibiotic such as a fluoroquinolone, cephalosporin, or beta-lactamase inhibitor is prescribed following biliary manipulation. For patients with recurring cholangitis, long-term antibiotic prophylaxis may be helpful. The standard approach involves rotating antibiotics (e.g., amoxicillin-clavulanate, ciprofloxacin, and trimethoprim-sulfamethoxazole), given in three- to fourweek cycles, in an attempt to reduce the risk of resistance. In patients with severe, recurrent bacterial cholangitis that does not respond to this approach, liver transplantation may be the only beneficial option.

Endoscopic Management

Endoscopic therapy for PSC carries the potential to relieve jaundice, pruritus, and abdominal pain; improve bio­ chemical cholestasis; decrease the frequency of episodes of bacterial cholangitis; and improve biliary flow. In theory, improved long-term biliary patency could slow the progression of the disease and prevent or delay biliary cirrhosis. Studies of endoscopic intervention in patients with PSC, however, have typically been small, retrospective series and uncontrolled trials, and firm conclusions are not available. Patients most likely to benefit from endoscopic intervention are those with one or more dominant stricture. These patients are more likely to present with specific symptoms such as worsening jaundice or pruritus, cholangitis, or abdominal pain. Multiple studies have reported significant improvements in clinical, biochemical, and cholangiographic endpoints in patients with a dominant stricture treated with endoscopic therapy,203-207 usually dilation with a balloon or graduated dilators, with or without temporary placement of a biliary stent. Sphincterotomy is often performed for improved access and to treat choledocholithiasis, if present. One retrospective study suggested that balloon dilation followed by stent placement offered no improvement and increased the risk of complications

1165

1166

Section VIII  Biliary Tract compared with balloon dilation alone.208 Because the study was not randomized, however, these findings may have been attributable to differences between the treatment groups. Three studies have suggested that progression of the underlying disease process may be slowed by endoscopic therapy of a dominant stricture. Baluyut and colleagues209 performed graduated and balloon dilation, with or without stent placement, in 63 patients with PSC with a median follow-up of 34 months, and demonstrated an observed 5-year survival that was significantly better than survival predicted from the revised Mayo model score (see Table 68-3). Stiehl and colleagues210 performed endoscopic balloon dilation and occasional stent placement in 52 patients with PSC in whom a dominant stricture developed while the patients were on therapy with UDCA. Actuarial survival free of liver transplantation at three, five, and seven years was significantly better than that predicted from the multicenter model score (see Table 68-3). Finally, a retrospective chart review by Gluck and colleagues8 reported a single-center 20-year experience with endoscopic therapy for PSC. Endoscopic therapy for a dominant stricture was performed in 84 of 106 patients who underwent a total of 317 procedures during the 20-year observation period. The patients in whom endoscopic therapy was performed had a significantly higher survival rate than that predicted by the revised Mayo model score at years three and four. These studies were not randomized trials, and in some cases were retrospective, but they provide some supporting evidence to suggest that endoscopic management of a dominant stricture may alter the course of PSC. Endoscopic therapy in PSC also has important limitations. Complications of ERCP, such as pancreatitis, cholangitis, worsening cholestasis, and perforation, occur at an overall rate of 7.3% to 10%.8,210,211 Patients with diffuse intrahepatic biliary stricturing and no dominant stricture are less likely to derive benefit from endoscopic intervention and may be at higher risk for post-ERCP cholangitis.208 If ERCP is performed in expert hands and only for specific indications such as worsening of jaundice, pruritus, or cholangitis—that is, in the subgroup of patients who are most likely to benefit from therapy—the risks in patients unlikely to benefit will be minimized (see Chapter 70).

Percutaneous Management

Percutaneous THC with balloon dilation, stenting, or both, can also be undertaken to treat biliary strictures in patients with PSC. This approach is typically recommended only when endoscopic intervention is contraindicated or unsuccessful because of the added risks of bleeding and bile peritonitis, as well as increased patient discomfort, associated with percutaneous intervention (see Chapter 70).

Surgical Management

Biliary Surgery The role of biliary surgery in PSC has diminished considerably with improvements in endoscopic therapy and liver transplantation. Resections of a dominant stricture of the bile duct or near the hepatic bifurcation followed by hepaticojejunostomy or choledochojejunostomy have been the most commonly performed operations.212,213 Postoperative mortality is increased significantly in patients with PSC and cirrhosis.214 In addition, biliary surgery may complicate future liver transplantation. Currently, biliary surgery in patients with PSC is rarely indicated and should be reserved for the small subset of patients who have early-stage PSC and biliary strictures that are not amenable to endoscopic or percutaneous intervention.

Liver Transplantation Liver transplantation (see also Chapter 95) is the only therapy that has been shown conclusively to improve the natural history of PSC. In addition, quality of life improves after liver transplantation.215,216 The procedure is recommended for patients with PSC in whom decompensated cirrhosis and complications of portal hypertension develop. Recurrent cholangitis or pruritus that is refractory to medical and endoscopic management rarely may also be indications for liver transplantation. Determination of the appropriate timing for liver transplantation in patients with PSC may be challenging, although use of available natural history models can be helpful. When a patient’s expected survival after liver transplantation exceeds survival predicted from the natural history models, liver transplantation should be undertaken, in the absence of contraindications. Intraoperatively, patients who undergo liver transplantation for PSC should have a Roux-en-Y choledochojejunostomy anastomosis, instead of a standard choledochocholedochostomy. This approach is recommended to allow removal of as much of the native biliary tree as possible, to reduce the risk of recurrent strictures and cholangiocarcinoma.217 Patient and graft survival after liver transplantation for PSC is excellent. A large single-center experience demonstrated 1-, 5-, and 10-year actuarial patient survival rates of 93.7%, 86.4%, and 69.8%, respectively. Corresponding graft survival rates were 83.4%, 79.0%, and 60.5%.218 Similar results have been reported in other series.219,220 Overall, survival rates after liver transplantation for PSC are significantly better than those for any other disease except PBC.221 Recipient factors that have been associated with a worse prognosis after liver transplantation for PSC are older age, decreased serum albumin level, renal failure, Child’s class C cirrhosis, and advanced United Network for Organ Sharing status.221,222 The presence of cholangiocarcinoma has a major impact on the outcome after liver transplantation for PSC. Early studies demonstrated that even in cases in which cholangiocarcinoma was discovered incidentally in the explant, recipient survival was poor, with a one-year survival rate of 30% in one series.223 On the basis of such studies, cholangiocarcinoma has generally been considered a contraindication to liver transplantation. Another report confirmed the poor post-liver transplantation outcome in patients with known cholangiocarcinoma but suggested a good survival rate for those who had a small cholangiocarcinoma found incidentally at the time of transplantation.220 Subsequent studies have demonstrated improved results of liver transplantation in patients with cholangiocarcinoma, with oneand five-year survival rates of 65% to 82% and 35% to 42%, respectively.224,225 Preoperative chemoradiation in highly selected patients with cholangiocarcinoma has shown promise in reducing the rate of tumor recurrence after liver transplantation (see Chapter 69).226,227 Biliary strictures commonly recur after liver transplantation for PSC and may represent recurrent PSC. In addition to recurrent PSC, potential causes of biliary strictures after liver transplantation include ABO blood group incompatibility, hepatic artery occlusion, chronic ductopenic graft rejection, Roux-en-Y–related cholangitis, and preservationrelated ischemia. The diagnosis of recurrent PSC has been proposed to be confined to those patients who have a consistent cholangiographic pattern and compatible liver histology showing fibrous cholangitis, fibro-obliterative lesions, biliary fibrosis, or biliary cirrhosis, and who lack other risk factors for biliary strictures, such as hepatic artery occlusion, ABO incompatibility, or ductopenic graft rejection, or who develop non-anastomotic strictures within 90 days of

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis transplantation.228 With these stringent criteria, the risk of recurrent PSC after liver transplantation ranges from 5.7% to 21.1%,219,220,229-230 and patient and graft survival do not appear to be adversely affected for the first five years of follow-up.220,230 No specific risk factors for the development of recurrent PSC have been clearly identified. In addition, no specific therapy has been shown to treat or prevent recurrent PSC effectively after liver transplantation. The effect of liver transplantation on the course of underlying IBD and risk for subsequent colonic neoplasia remains controversial. The clinical course of IBD after liver transplantation has varied greatly among studies.45,219,232,233 Immunosuppression was thought possibly to ameliorate the clinical course of IBD but does not in most instances. A large single-center study showed that liver transplantation was associated with an increased risk for colectomy.234 The risk of colorectal neoplasia is generally agreed to be increased after liver transplantation in patients with UC,164-166,219 although this finding has not been confirmed in all studies.234 Annual surveillance colonoscopy is recommended after liver transplantation for PSC in patients with UC.

RECURRENT PYOGENIC CHOLANGITIS Recurrent pyogenic cholangitis (RPC) was originally defined by Cook and colleagues as a syndrome characterized by recurrent bacterial cholangitis, intrahepatic pigment stones, and biliary strictures, possibly leading to chronic liver disease and cho­langiocarcinoma.235 RPC has also been called oriental cholangiohepatitis, “Hong Kong disease,” “biliary obstruction syndrome of the Chinese,” and hepatolithiasis.236

EPIDEMIOLOGY

Digby first described RPC in 1930 in Chinese patients in Hong Kong and was the first to recognize that RPC was clinically and pathologically distinct from biliary disease caused by gallstones in Western populations.237 Subsequently, most cases have occurred in patients from East Asia; men and women are affected equally. Patients in rural areas and those of lower socioeconomic status appear to be at increased risk, and the incidence is highest among persons between ages 30 and 40.238,239 In certain parts of Southeast Asia such as Taiwan, over one half of the cases of gallstone disease are estimated to be caused by RPC.240 In Singapore, Japan, and Hong Kong, 2% to 5% of biliary calculous disease is attributed to RPC.241 The incidence of RPC in East Asia appears to be decreasing. A study from the Queen Mary Hospital, a major referral center for RPC, reported that only 6.7% of a total of 490 patients who underwent surgery for hepatobiliary disease were classified as having RPC.241 By contrast, typical gallstone disease was the reason for surgery in 44%. These numbers likely reflect referral bias; a nationwide survey found that RPC or intrahepatic stones represented only about 20% of the total cases of biliary tract disease. The incidence of intrahepatic stones appears to be decreasing in Taiwan as well; one study reported that the incidence of hepatolithiasis decreased from 21% to 18% between 1981 and 1989.240 By contrast, the incidence and prevalence of RPC are increasing in Western countries, reflecting immigration patterns. One study from San Francisco found that the prevalence of RPC doubled between 1983 and 1995.242 Similarly, a review from a county hospital in Los Angeles reported that 57% of 18 patients with RPC seen over a seven-year period were of Asian descent, and 36% of the patients were of Hispanic descent.243

The reasons for the changing epidemiology of intrahepatic stones in East Asia are unclear. Several possible explanations have been proposed, including adoption of a Western-style diet with a higher protein content; improved hygiene, which may lead to reduced gastrointestinal infections and consequently decreased entry of bacteria into the portal circulation (an important cause of cholangitis); and, less likely, reduction in disease burden related to Clonorchis sinensis and Ascaris lumbricoides. A low-protein diet may lead to reduced biliary levels of glucaro-1:4-lactone, an inhibitor or bacterial b-glucuronidase, which helps promote the formation of calcium bilirubinate stones by decon­ jugating bilirubin into unconjugated bilirubin.244 Furthermore, a diet low in fat may be associated with reduced gallbladder contractility and thus promote stasis, which also is a factor in stone formation.245 The rising incidence of cholesterol gallstones in Asia suggests that environmental factors, such as adoption of a more Western-style diet, rather than genetic factors, are major factors in the pathogenesis of RPC.241

ETIOLOGY AND PATHOGENESIS

The etiology of RPC remains unknown. The most attractive hypothesis links biliary tract infection with the parasites Clonorchis sinensis, Opisthorchis species, and Ascaris lumbricoides (see Chapters 82 and 110). Infection with these parasites is endemic in the same geographic region where RPC is prevalent. C. sinensis, a trematode (liver fluke), is endemic in China, Japan, Taiwan, Korea, and Vietnam.246,247 O. felineus and O. viverrini are the two species most commonly implicated in opisthorchiasis. O. felineus has been described in Southeast Asia and in parts of the former Soviet Union, and A. lumbricoides, a roundworm, is a ubiquitous parasite and may infect over one billion people throughout the world.248,249 Both organisms colonize the biliary tree and lead to infection, biliary tract obstruction, and secondary bacterial cholangitis. The Clonorchis worm can survive for decades in the biliary tree and may lead to inflammatory changes in the bile ducts, as well as direct bile duct obstruction by the flukes, and shows a specific predilection for the left hepatic ducts.248 Analysis of biliary stones from patients with RPC has demonstrated evidence of ova.250 Patients with RPC, however, have not been shown to have an increased prevalence of infection with these parasites when compared with the general population, and approximately one half of patients with RPC demonstrate no evidence of infection.251-253 Furthermore, some parts of Asia with a high prevalence of RPC have low or undetectable rates of infection with C. sinensis.254 Nevertheless, it remains possible that parasitic infection may account for a significant proportion of cases of RPC. Moreover, currently available serologic tests lack precision to detect prior infection, and structural or functional changes in the bile ducts, gallbladder, or sphincter of Oddi caused by infection in the remote past may alter the biliary epithelia and thereby promote the later formation of intrahepatic stones. Bacterial infections have also been proposed as a cause of RPC because of the high incidence of bacterial cholangitis in patients with RPC. Portal bacteremia, possibly related to gastrointestinal infection and bacterial translocation, has been associated with low socioeconomic status and malnutrition.239 The hypothesis that bacterial infections lead to RPC, however, does not explain the unique epidemiologic features of the disease. Dietary factors and hygiene have also been implicated in the pathogenesis of RPC. Because cholecystokinin, a potent mediator of gallbladder contractility, is secreted in response

1167

1168

Section VIII  Biliary Tract to dietary fat, diets high in carbohydrate and low in saturated fat may be associated with reduced gallbladder contractility and promote stone formation. Furthermore, deconjugation of bilirubin by bacteria or endogenous enzymes, as noted earlier, may be related to dietary and environmental factors and may facilitate formation of pigment stones. Several investigators have suggested that patients with RPC may lack an inhibitor of bacterial b-glucuronidase.255,256 Another factor that may contribute to the pathogenesis of RPC is abnormal motor activity of the sphincter of Oddi. A study of biliary sphincter manometry in 15 patients with RPC and 15 normal control subjects257 demonstrated various abnormalities, including retrograde phasic waves, abnormal propagation of phasic contractions, elevated basal sphincter pressure, and increased frequency of phasic contractions in the sphincter among the patients with RPC (see Chapter 63). Almost one half of the patients had evidence of papillitis at endoscopy. The investigators suggested that papillitis may lead to altered function in the sphincter of Oddi, resulting in turn in delayed biliary drainage and recurrent cholangitis. Whether papillitis and sphincter of Oddi dysfunction are a cause or effect of RPC is difficult to ascertain from the study; recurrent bouts of cholangitis may, in fact, have led to changes in the biliary tract, and recurrent passage of small pigment stones and debris may have resulted in endoscopic and manometric changes.

Figure 68-5.  Computed tomographic scan with intravenous contrast in a patient with recurrent pyogenic cholangitis. The scan demonstrates marked dilatation of the biliary tree, most prominent in the left lobe of the liver. High-density foci in the dilated segments represent calcified stones.

CLINICAL FEATURES

Patients with RPC often present with symptoms of ascending cholangitis (45% of patients).258 Patients characteri­ stically present with fever, right upper quadrant pain, and jaundice, also referred to as Charcot’s triad. A prior history of such attacks is elicited in the majority of patients, whereas up to 30% of patients present with an initial episode.242,250 Patients may also present with abdominal pain or pancreatitis. On physical examination, abdominal tenderness is a common finding. Hepatomegaly is present in approximately 20% of patients. A palpable gallbladder may be present in approximately 10% of patients and may point to emphysematous cholecystitis.250,251,253 Laboratory findings are compatible with biliary obstruction, with elevation of serum total and direct bilirubin, aminotransferase, and alkaline phosphatase levels; leukocytosis may be present. Imaging findings in patients with RPC are characteristic (Figs. 68-5 and 68-6). The majority of patients (75% to 80%) have intrahepatic stones, with predominant involvement of the left hepatic duct. Up to 70% may also have associated stones in the gallbladder, and cholecystitis (calculous or acalculous) is a frequent associated finding.259-261 Dilatation of the bile ducts is found almost universally. The central bile ducts are dilated disproportionately, with abrupt tapering and attenuation of more peripheral bile ducts within the liver. The presence of bile duct calculi is usually associated with intrahepatic bile duct dilatation and downstream strictures. The left hepatic ducts are involved more often than are the right.250 Cross-sectional imaging studies of the liver, such as ultrasonography and CT, as well as direct cholangiography, are useful for evaluating a patient with known or suspected RPC. Ultrasonography is a reasonable screening test early in the evaluation of the patient and may identify intra- and extrahepatic bile duct dilatation in the majority of cases. CT is helpful in delineating the complete segmental anatomy of the liver and is invaluable in planning surgical therapy (see Figure 68-5). CT may also demonstrate associated atrophic segments of the liver, abscesses, and cholangiocarci-

Figure 68-6.  Endoscopic retrograde cholangiopancreatography in a patient with recurrent pyogenic cholangitis. Multiple filling defects are seen in the bile duct, consistent with stones and/or air bubbles. The poorly opacified left hepatic duct is strictured at its takeoff, markedly dilated, and filled with intrahepatic stones. The right intrahepatic ducts demonstrate characteristic reduction in arborization, widening of the branching angles, and abrupt peripheral tapering.

noma.250 A study of more than 1300 patients with RPC from Korea described characteristic CT findings in 82 patients in whom cholangiocarcinoma was associated with RPC; cholangiocarcinoma tended to be located in atrophic segments associated with biliary calculi and was often accompanied by portal vein occlusion or narrowing.262

Chapter 68  Sclerosing Cholangitis and Recurrent Pyogenic Cholangitis Direct cholangiography, whether performed by the percutaneous or endoscopic route, allows localization of intrahepatic stones and strictures and placement of drains or extraction of stones (see Figure 68-6). ERCP is the first choice for therapeutic intervention because it is less invasive than percutaneous THC.263 Removal of intrahepatic stones by this technique may be exceedingly difficult, however, because of the presence of a tight stricture and the sharp angulation of the intrahepatic bile ducts. MRCP permits selective visualization of the biliary tree and has become the preferred test for diagnostic evaluation of the bile ducts. In a study from Korea,264 18 of 24 patients who had MRCP before undergoing surgical treatment for RPC also underwent direct cholangiography. Two examiners evaluated MRCP images and direct cholangiograms and compared them with surgical findings. All dilated bile ducts and 98% of focal duct strictures and intraductal stones were identified by MRCP. By contrast, direct cholangiography identified only 44% to 47% of segmental bile duct abnormalities. The authors concluded that MRCP was more sensitive than direct cholangiography for complete evaluation of the biliary tree. MRCP is an important complementary technique for examination of the biliary tree, especially in patients being considered for surgery because of its ability to visualize the bile ducts upstream from obstructed areas completely and to correlate the observed abnormalities with the segmental anatomy of the liver.265 Patients who present with an initial episode of cholangitis associated with intrahepatic stones and strictures should undergo evaluation for infection with Clonorchis and Opisthorchis species, particularly if the patient comes from or has traveled to an endemic area. The diagnosis of a parasitic infection is made by identification of eggs in fecal specimens; concentrated stool may be required. Eggs are present in stool after four weeks of infection.250 Duo­ denal or biliary fluid also may demonstrate eggs or intact worms. Peripheral eosinophilia may be present in cases of parasitic infection and may be associated with elevated serum IgE levels.246

PATHOLOGY

The characteristic findings in RPC include strictures and dilatation of the intra- and extrahepatic bile ducts. In classic cases, the left hepatic duct is more commonly and more severely affected than the right.250,251 With chronic disease, atrophy of the left lobe or the left lateral segment of the liver may occur and may be the site at which cholangiocarcinoma develops.250,251 The bile ducts are often obstructed with pigment stones, in addition to sludge and inspissated bile, and are frequently contaminated by bacteria and purulent material. The stones are composed predominantly of calcium bilirubinate, although cholesterol has been reported to be found in increasing proportions.239 Ova of parasites with a predilection for the biliary tree, such as C. senensis, have been described within the stones.250,252 An inflammatory infiltrate is frequently present in the walls of involved bile ducts and may be associated with periductal fibrosis and abscesses.250,251

TREATMENT

Antibiotic therapy should be initiated promptly, after cultures of blood and bile (if accessible) have been obtained. Abdominal ultrasonography is a reasonable initial imaging modality.259 In patients with evidence of cholangitis and dilated intra- and extrahepatic bile ducts, ERCP is the preferred interventional procedure. ERCP with sphincterotomy, with or without placement of a nasobiliary drain or an endobiliary drain, may adequately remove bile

duct stones and traverse strictures. After adequate biliary drainage has been achieved, additional cross-sectional imaging with CT or MRI, with or without MRCP, may be considered. Several studies have reviewed the success of various nonoperative interventions for initial management of patients with RPC presenting acutely. Sperling and colleagues258 compared outcomes in 41 patients with RPC based on whether they received immediate therapeutic ERCP, hepatobiliary surgery, or no intervention. Symptoms recurred in 62% of patients who underwent only diagnostic ERCP but one half as often in those treated with therapeutic ERCP or surgery. Therapeutic ERCP was particularly effective in patients with disease involving the extrahepatic bile ducts and was comparable in efficacy to surgery. Patients with disease involving both the right and left hepatic bile ducts tend to undergo more imaging studies, percutaneous cholangiograms, and endoscopic or surgical procedures.242 Hepaticojejunostomy has been a commonly used surgical procedure for the treatment of intrahepatic stones in patients with RPC. Kusano and colleagues,266 from Japan, reviewed the long-term outcomes of hepaticojejunostomy for intrahepatic hepatolithiasis in 159 patients over a 23-year period. Surgical approaches included hepatectomy (n = 94), biliary lithotripsy (n = 65), or hepaticojejunostomy (n = 72) (or combinations). Residual or recurrent stones were identified in approximately one third of patients after presumably complete removal. The rate of cholangitis was higher among patients who underwent a biliary-enteric anastomosis (22 of 72; 31%) than in those who did not (3 of 87; 3.4%). This study suggests that surgical therapy involving a biliaryenteric anastomosis may be associated with a higher rate of cholangitis than nonsurgical therapy. Laparoscopic biliary bypass surgery has been proposed as a technically feasible and effective option for patients with RPC.267 A large retrospective case series from Canada of 42 patients who underwent surgical intervention for RPC demonstrated excellent postoperative outcomes for both liver resection and bile duct exploration with biliary bypass; the overall operative mortality rate was 0%, and the complication rate was 35% for resection and 30% for bile duct exploration.268 Treatment with an antihelminthic agent is indicated in patients with evidence of active parasitic infection. Prazi­ quantel, 75 mg/kg in three divided doses for one day, is the treatment of choice (see also Chapter 82). The treatment is almost universally effective in both Clonorchis and Opisthorchis infections.269 Side effects of the medication include headache and gastrointestinal symptoms. Even patients without gastrointestinal symptoms but with Clonorchis or Opisthorchis in the stool should be treated to reduce the risk of cholangiocarcinoma.132

PROGNOSIS AND COMPLICATIONS

The natural history of RPC has been evaluated in a large series from Korea. The records of 193 patients with newly diagnosed RPC were reviewed to examine the rates and risk factors for recurrence.270 The mean follow-up period was 56 months (range, 1 to 242 months). Cumulative recurrence rates of cholangitis were 25% at three years and 37% at five years, with an overall rate of 45% during the follow-up period. Factors associated with cholangitis included recurrent stones (hazard ratio = 4.02; 95% CI: 1.3 to 12.4), residual stones (hazard ratio = 1.77; 95% CI: 1.1 to 3.0); prior hepatic resection was associated with a lower rate of recurrent cholangitis (hazard ratio = 0.28; 95% CI: 0.1 to 0.7). Strictures in the extrahepatic bile ducts were associated

1169

1170

Section VIII  Biliary Tract with a higher rate of cholangitis, whereas strictures in the intrahepatic ducts and disruption of the sphincter of Oddi (e.g., sphincterotomy) were not. Cholangiocarcinoma is another long-term complication in patients with RPC (see Chapter 69). The frequency of cholangiocarcinoma has not been clearly delineated. One large study found that the frequency of cholangiocarcinoma among patients with RPC was approximately 3%.271 Other studies have reported the rate to be as high as 9%.272,273 Secondary biliary cirrhosis may develop and require liver transplantation.

KEY REFERENCES

Angulo P, Pearce DH, Johnson CD, et al. Magnetic resonance cholangiography in patients with biliary disease: Its role in primary sclerosing cholangitis. J Hepatol 2000; 33:520-7. (Ref 10.) Bambha K, Kim WR, Talwalkar J, et al. Incidence, clinical spectrum, and outcomes of primary sclerosing cholangitis in a United States community. Gastroenterology 2003; 125:1364-9. (Ref 27.) Bjornsson E. Immunoglobulin G4-associated cholangitis. Curr Opin Gastroenterol 2008; 24:389-94. (Ref 25.) Bjornsson E, Chari ST, Smyrk TC, Lindor K. Immunoglobulin G4 associated cholangitis: Description of an emerging clinical entity based on review of the literature. Hepatology 2007; 45:1547-54. (Ref 23.) Feldstein AE, Perrault J, El-Youssif M, et al. Primary sclerosing cholangitis in children: A long-term follow-up study. Hepatology 2003; 38:210-17. (Ref 33.) Gluck M, Cantone N, Brandabur J, et al. A twenty-year experience with endoscopic therapy for symptomatic primary sclerosing cholangitis. J Clin Gastroenterol 2008; 42:1032-9. (Ref 8.) Gregorio GV, Portmann B, Karani J, et al. Autoimmune hepatitis/ sclerosing cholangitis overlap syndrome in childhood: A 16-year prospective study. Hepatology 2001; 33:544-53. (Ref 66.)

Kamisawa T, Okamoto A. IgG4-related sclerosing disease. World J Gastroenterol 2008; 14:3948-55. (Ref 24.) Kaya M, Angulo P, Lindor KD. Overlap of autoimmune hepatitis and primary sclerosing cholangitis: An evaluation of a modified scoring system. J Hepatol 2000; 33:537-42. (Ref 20.) Mitchell SA, Grove J, Spurkland A, et al. Association of the tumour necrosis factor alpha -308 but not the interleukin 10 -627 promoter polymorphism with genetic susceptibility to primary sclerosing cholangitis. Gut 2001; 49:288-94. (Ref 59.) Mitchell SA, Bansi DS, Hunt N, et al. A preliminary trial of high-dose ursodeoxycholic acid in primary sclerosing cholangitis. Gastroenterology 2001; 121:900-7. (Ref 177.) Moff SL, Kamel IR, Eustace J, et al. Diagnosis of primary sclerosing cholangitis: A blinded comparative study using magnetic resonance cholangiography and endoscopic retrograde cholangiography. Gastrointest Endosc 2006; 64:219-23. (Ref 13.) Olsson R, Boberg KM, de Muckadell OS, et al. High dose ursodeoxycholic acid for the treatment of primary sclerosing cholangitis: A 5-year multicenter, randomized, controlled study. Gastroenterology 2005; 129:1464-72. (Ref 181.) Said K, Glaumann H, Bergquist A. Gallbladder disease in patients with primary sclerosing cholangitis. J Hepatol 2008; 48:598-605. (Ref 7.) Stiehl A, Rudolph G, Kloters-Plachky P, et al. Development of dominant bile duct stenoses in patients with primary sclerosing cholangitis treated with ursodeoxycholic acid: Outcome after endoscopic treatment. J Hepatol 2002; 36: 151-6. (Ref 9.) Textor HJ, Flacke S, Pauleit D, et al. Three-dimensional magnetic resonance cholangiopancreatography with respiratory triggering in the diagnosis of primary sclerosing cholangitis: Comparison with endoscopic retrograde cholangiography. Endoscopy 2002; 34:984-90. (Ref 11.) Tung BY, Emond MJ, Haggitt RC, et al. Ursodiol use is associated with lower prevalence of colonic neoplasia in patients with ulcerative colitis and primary sclerosing cholangitis. Ann Intern Med 2001; 134:89-95. (Ref 159.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

69 Tumors of the Bile Ducts, Gallbladder, and Ampulla Boris Blechacz and Gregory J. Gores

CHAPTER OUTLINE Cholangiocarcinoma  1171 Epidemiology  1171 Etiology  1172 Pathology  1172 Pathogenesis  1172 Clinical Features and Diagnosis  1173 Staging  1175 Treatment  1176 Gallbladder Carcinoma  1177 Epidemiology  1177 Etiology  1177 Pathology  1178 Pathogenesis  1178

Biliary malignancies comprise the vast majority of biliary neoplasms and are divided into the following three categories: (1) carcinomas of the intra- and extrahepatic bile ducts; (2) carcinoma of the gallbladder; and (3) carcinoma of the ampulla of Vater.1 In the United States and other Western nations, biliary malignancies are rare. In certain parts of the world, however, their prevalence rates are high, making them leading causes of cancer death. Biliary cancers are highly aggressive with dismal prognoses. Often these cancers are diagnosed at an advanced stage. In general, they are resistant to chemotherapy. Advances in the understanding of the molecular pathogenesis of these tumors has allowed the development of new experimental models and targeted therapies, and advances in diagnosis and surgical treatment have resulted in improved outcomes for subsets of patients.

CHOLANGIOCARCINOMA Cholangiocarcinoma is an epithelial carcinoma with differentiated features of biliary epithelium that arises from the intra- and extrahepatic biliary tree.2 It is the most common bile duct tumor and second most common primary hepatic malignancy (after hepatocellular carcinoma, see Chapter 94). Since the 1970s, the incidence has increased substantially in Western societies.3 The classification of cholangiocarcinomas into intra- and extrahepatic cancers is based on differences in anatomy, etiology, pathogenesis, molecular signature, and treatment. The second-order bile ducts are the anatomic margin for the distinction between these two subsets (see Chapter 62). Extrahepatic cholangiocarcinomas account for 80% to 90%

Clinical Features and Diagnosis  1179 Staging  1180 Treatment  1180 Ampullary Carcinoma  1181 Epidemiology  1181 Etiology  1181 Pathology  1181 Pathogenesis  1182 Clinical Features and Diagnosis  1182 Staging  1182 Treatment  1182 Other Tumors of the Bile Ducts and Gallbladder  1183

of these cancers and can be further subclassified into hilar or distal bile duct cancers. Hilar cholangiocarcinoma, also referred to as Klatskin tumors, are described clinically according to the Bismuth-Corlette classification as types I to IV (Fig. 69-1). Type I cholangiocarcinomas involve the common hepatic duct distal to the union of the right and left hepatic ducts; type II tumors involve the union of the right and left hepatic ducts; type IIIa tumors involve the union of the right and left hepatic ducts and extend up the right hepatic duct; type IIIb tumors involve the union of the right and left hepatic ducts and extend up the left hepatic duct; and type IV tumors are multifocal or involve the biliary confluence and extend up the right and left hepatic ducts. The natural course of cholangiocarcinoma is aggressive, with a median survival of less than 24 months following diagnosis.4 The only potentially curative treatment is surgical. Unfortunately, the majority of patients are diagnosed at an advanced stage that precludes curative surgery.

EPIDEMIOLOGY

Cholangiocarcinoma is an uncommon cancer that accounts for less than 2% of all malignancies. It is the ninth most common gastroenterologic malignancy and the second most common primary hepatic and biliary malignancy. Hepatobiliary malignancies account for 13% and 3% of overall cancer-related mortality in the world and in the United States, respectively; 10% to 20% of these deaths are caused by cholangiocarcinoma. Global incidence rates for cholangiocarcinoma are heterogeneous. The highest incidence is observed in Southeast Asia, with rates up to 96 per 100,000 population, and the lowest incidence is observed in Australia, with rates as low as 0.1 per 100,000 population.5 In the United States,

1171

Section VIII  Biliary Tract Type I

Type II

Type IIIa

Type IIIb

Intrahepatic

Liver Common hepatic duct Gallbladder

Bile duct

Hilar Distal

Ampulla of Vater

Extrahepatic

1172

Type IV

Duodenum

A

B

Figure 69-1.  Classification of cholangiocarcinoma. A, Anatomic classification of intra- and extrahepatic cholangiocarcinoma. Extrahepatic cholangiocarcinoma is further subclassified as hilar and distal. B, Bismuth-Corlette classification of hilar cholangiocarcinoma as types I to IV. Tumor is depicted in yellow and normal bile ducts in green. (From Blechacz BR, Gores GJ. Cholangiocarcinoma. Clin Liver Dis 2008; 12:131-150.)

incidence rates are 0.95 per 100,000 population for intra­ hepatic cholangiocarcinoma and 0.82 per 100,000 for extrahepatic cholangiocarcinoma. Ethnic differences in the incidence of cholangiocarcinoma have been observed in the United States, with the highest rates observed in patients of Hispanic descent and the lowest in African Americans. Since the 1980s, age-adjusted incidence rates for intrahepatic cholangiocarcinoma have increased, whereas those for extrahepatic cholangiocarcinoma have remained stable; the cause of the overall increase in incidence is unknown.3 Fifty-two percent to 54% of patients with cholangiocarcinoma are male. Globally, the average age at diagnosis is older than 50 years. In Western industrialized nations, most cases are diagnosed around age 65, and cholangiocarcinoma is uncommon before age 40 except in patients with primary sclerosing cholangitis (see Chapter 68).

ETIOLOGY

In the majority of cases, the etiology of cholangiocarcinoma is unknown. Several risk factors have been identified (Table 69-1). These risk factors are characterized by their association with inflammation and cholestasis. Primary sclerosing cholangitis (PSC) is one of the most common risk factors (see Chapter 68). In patients with PSC, the annual frequency rate of cholangiocarcinoma is 0.6% to 1.5%, and the overall prevalence rate of cholangiocarcinoma is 5% to 15%. In the majority of patients with PSC in whom cholangiocarcinoma develops, the diagnosis of cholangiocarcinoma is made within two-and-a-half years of the diagnosis of PSC.6 Other risk factors for cholangiocarcinoma include biliary infections with Opisthorchis viverrini and Clonorchis sinensis, which are endemic in East Asia (see Chapter 82).7 Biliary malformations such as Caroli’s disease and choledochal cysts are associated with a 10% to 15% risk for the development of cholangiocarcinoma (see Chapter 62),8 whereas hepatolithiasis (“recurrent pyogenic cholangitis”) carries a 10% risk for the development of cholangiocarcinoma (see Chapter 68).9 Recurrent bacterial cholangitis in the setting of biliary-enteric drainage procedures has also been associated with the development of cholangiocarcinoma.10 Carcinogens such as thorotrast (used as a radiologic contrast agent in the past) and dioxins have been associated with an

Table 69-1  Risk Factors for Cholangiocarcinoma Biliary-enteric drainage procedures Caroli’s disease Choledochal cyst Cirrhosis Clonorchis sinensis infection Hepatitis C Hepatolithiasis Opisthorchis viverrini infection Primary sclerosing cholangitis Thorotrast Toxins (dioxins, polyvinyl chloride)

increased risk of cholangiocarcinoma.11 Hepatitis C and cirrhosis are considered possible risk factors for cholangiocarcinoma (see Chapter 79).5

PATHOLOGY

Cholangiocarcinoma is a paucicellular, highly desmoplastic tumor. Macroscopically, it can be described according to its growth characteristics as mass forming, periductalinfiltrating, or intraductal-papillary. Intrahepatic cholangiocarcinomas are typically mass forming, whereas ductal carcinomas can present in any of the three growth forms. Histologically, 90% of cholangiocarcinomas are adenocarcinomas. Other histologic types include intestinal-type adenocarcinoma, clear cell adenocarcinoma, signet-ring cell carcinoma, adenosquamous carcinoma, squamous cell carcinoma, and small cell carcinoma.12 Intraductal-papillary adenocarcinomas spread superficially along the biliary mucosa without deep invasion of the fibromuscular wall layers and have a better prognosis than nonpapillary cancers.13 Metastases to regional and peripancreatic lymph nodes are frequently observed with this growth type.

PATHOGENESIS

As noted earlier, etiologic associations as well as experimental data provide evidence for inflammation and bile acids as key factors in the molecular pathogenesis of cholangiocarcinoma. Inflammation provides a microenvi-

Chapter 69  Tumors of the Bile Ducts, Gallbladder, and Ampulla Table 69-2  Molecular Changes in Cholangiocarcinoma

Table 69-3  Diagnostic Criteria for Cholangiocarcinoma

MALIGNANT PHENOTYPE

DYSREGULATED GENE products AND PATHWAYS

Proliferation

Interleukin (IL)-6, IL-6 receptor (gp130) Hepatocyte growth factor (HGF)/c-Met ErbB2 K-ras BRAF Cyclooxygenase-2 (COX-2) Caspase-9 Mcl-1 bcl-2 Bcl-XL COX-2 Telomerase Cyclin D1 p21waf1/cip1 p27kip1 pRb p53 mdm-2 DCP4/Smad4 p16INK4a E-cadherin α/β-catenin Aspartyl (asparaginyl) β-hydroxylase WISP1v Matrix metalloproteinases Vascular endothelial growth factor (VEGF) Transforming growth factor-β (TGF-β)

Malignant appearing stricture AND persistent serum CA 19-9 level >129 U/mL in the absence of bacterial cholangitis* Mass lesion on cross-sectional imaging Positive conventional cytology result Positive (transluminal) biopsy specimen Stricture plus polysomy on fluorescence in situ hybridization (FISH)

Apoptosis evasion

Senescence evasion Cell cycle dysregulation

Invasion/Metastases

Angiogenesis

ronment that promotes malignant transformation of bile duct–associated cells such as cholangiocytes, biliary stem cells, or epithelial cells within peribiliary glands (see Chapter 62). In subsets of cholangiocarcinoma and mixed hepatocellular-cholangiocarcinoma, hepatic stem/progenitor cells have been suggested to be the cells of origin.14 On a molecular level, increased tissue concentrations of cytokines, bile acids, and growth factors can lead to functional inactivation of mismatch DNA repair genes and tumor suppressor genes and promote the expression of protooncogenes. Cytokines stimulate expression of inducible nitric oxide synthase (iNOS) in epithelial cells, thereby resulting in increased intracellular levels of nitric oxide (NO) and reactive nitrogen oxide species (RNOS). NO and RNOS interact with cellular DNA and proteins and thus result in DNA mutations and strand breaks, with inactivation of DNA repair proteins.15 A variety of different oncogenic mutations have been described in cholangiocarcinoma (Table 69-2). Their frequencies differ depending on the anatomic location, stage, type, and etiology of the tumor and the ethnicity of the patient. Growth factors, dysregulated signaling pathways, and tyrosine kinases promote further tumor cell proliferation, as well as inhibition of senescence and apoptosis. Key signaling pathways in cholangiocarcinoma involve interleukin (IL)-6, cyclooxygenase-2 (COX-2), epidermal growth factor receptor (EGFR), ErbB2 (also known as human epidermal growth factor receptor 2 [HER 2]), hepatocyte growth factor (HGF), and vascular endothelial cell growth factor receptor (VEGFR). The IL-6 signaling axis in particular has been shown to be critical to the pathogenesis of cholangiocarcinoma.16,17 Increased IL-6 signaling is sustained in cholangiocarcinoma cells by the following mechanisms: (1) autocrine secretion of high levels of IL-6; (2) COX-2-mediated up-

*Serum CA 19-9 >100 U/mL in patients with primary sclerosing cholangitis.

regulation of the IL-6 receptor subunit gp130; and (3) inactivation of the negative feedback loop through epigenetic silencing of suppressor of cytokine signaling 3 (SOCS3). Increased IL-6 signaling constitutively activates the PI3K pathway, Janus Kinase (JAK) signal transducers and activators of transcription (STAT) pathway, and p38- and p42/44mitogen-activated protein kinase (MAPK) pathways, which in turn mediate cell survival, cell proliferation, and evasion of cell senescence. The EGFR pathway is also constitutively activated, resulting in further activation of p38- and p42/44MAPK.18 The EGFR homolog ErbB2 is overexpressed in many cholangiocarcinomas, resulting in downstream induction of Raf/MAPK pathways. In addition to its oncogenic effects, MAPK induces overexpression of COX-2, which is further stimulated by increased concentrations of bile acids, oxysterols, and iNOS.19,20 COX-2 also mediates cell survival and proliferation. Sustained cell proliferation is further mediated by increased HGF secretion and overexpression of its receptor, c-Met.21 The multiple pathways form a complex interacting network.

CLINICAL FEATURES AND DIAGNOSIS

The diagnosis of cholangiocarcinoma is challenging because of the paucicellular character of the malignancy and requires a multidisciplinary approach that includes clinical evaluation and laboratory, endoscopic, and radiologic studies (Fig. 69-2 and Table 69-3). Intrahepatic cholangiocarcinoma manifests predominantly with abdominal pain and systemic symptoms such as cachexia, malaise, and fatigue. Extrahepatic cholangiocarcinoma manifests in most cases with painless jaundice secondary to malignant biliary obstruction. In 10% of patients, bacterial cholangitis is the initial presenting symptom. An “atrophy-hypertrophy” complex can be documented by physical examination as palpable hypertrophy of the contralateral, unaffected lobe of the liver, with atrophy of the affected lobe as a result of vascular encasement and bile ductal obstruction. Laboratory analysis may reveal evidence of obstructive cholestasis (e.g., elevation of serum alkaline phosphatase and bilirubin levels). Serum levels of several serum tumor markers (CA 19-9, carcinoembryonic antigen [CEA], and Ca-125) may be elevated in patients with cholangiocarcinoma; however, none of these serum markers is specific, and they each can be elevated in other gastroenterologic or gynecologic malignancies and in the setting of biliary inflammation or infection.22 The most commonly used marker is CA 19-9. In patients with PSC, the sensitivity and specificity of CA 19-9 for the diagnosis of cholangiocarcinoma is 79% and 98%, respectively, when the cutoff value is 129 U/mL. In patients without PSC, the sensitivity is 53% with a cutoff value of 100 U/mL.23,24 Importantly, patients with a negative Lewis antigen status do not express CA 19-9, and significant CA 19-9 elevations can occasionally be observed in patients with bacterial cholangitis and choledocholithiasis.

1173

1174

Section VIII  Biliary Tract Clinical suspicion of CCA

CA 19-9 level Endoscopic cholangiography (brushing, cytology, DIA, FISH)

Figure 69-2.  Diagnostic algorithm for hilar cholangiocarcinoma. In cases of clinically suspected hilar cholangiocarcinoma, a serum CA 19-9 level, endoscopic retrograde cholangiopancreatography, and conventional as well as molecular cytologic analysis of endoscopically obtained biliary brushings of malignant-appearing areas should be performed. If results of these tests are normal or negative, observation is recommended. Management of cholangiocarcinoma should be prompted by identification of a dominant stricture, serum CA 19-9 level >129 U/mL, or a biopsy or cytology result that is positive for carcinoma or polysomy. In indeterminate cases, gadolinium-enhanced magnetic resonance imaging (MRI) of the liver with ferumoxide contrast is recommended. If a mass lesion or vascular encasement is identified, management of cho­ langiocarcinoma should be initiated. If the MRI study is negative but clinical concern about cholangiocarcinoma persists, positron emission tomography (PET) can be performed. If “hot spots” are identified on PET (positive result), further management should be directed toward cholangiocarcinoma. If the result of the PET scan is negative, close follow-up is recommended. If MRI is negative and cholangiocarcinoma is considered unlikely, the patient can be observed expectantly. CCA, cholangiocarcinoma; DIA, digital image analysis; FISH, fluorescence in situ hybridization; pos., positive. (Published with permission from Wiley InterScience.)

Dominant stricture CA 19-9 >129 U/mL Pos. biopsy, cytology, or FISH polysomy

Indeterminate

MRI

Mass Vascular encasement

No dominant stricture Neg. biopsy/cytology/ advanced cytology CA 19-9 1 cm) Chronic Salmonella typhi or paratyphi carrier status First-degree relative with gallbladder cancer Inflammatory bowel disease Intrahepatic biliary dysplasia Porcelain gallbladder Primary sclerosing cholangitis *Methylcholanthrene, O-aminoazotoluene, nitrosamines, possibly others.

logic studies in the United States have observed only a marginally significant three-fold increased risk of gallbladder carcinoma in men with cholelithiasis.43,44 Gallbladder carcinoma actually develops in only 1% to 3% of patients with cholelithiasis, and 20% of patients with gallbladder carcinoma do not have evidence of cholelithiasis. Therefore, a prophylactic cholecystectomy in an asymptomatic patient with gallstones to prevent gallbladder carcinoma cannot be recommended. A positive correlation between the risk of gallbladder carcinoma and the size and number of gallstones has been reported but reflects the duration of cholelithiasis.45 No differences in the risk of gallbladder carcinoma have been observed with different types of gallstones. Porcelain gallbladder (extensive calcification of the gallbladder wall) is a classic, although controversial, risk factor for gallbladder carcinoma.46 Although an increased risk of gallbladder carcinoma has been reported in patients with a porcelain gallbladder, the risk may be limited to patients with selective mucosal calcification (types II and III porcelain gallbladder) rather than those with diffuse mucosal calcification (type I).47 Adenomatous polyps of the gallbladder constitute another risk factor for gallbladder carcinoma (see Chapter 67). The risk correlates positively with the size, type, and growth rate of the polyps. Patients with polyps that are greater than 1 cm in size, sessile, and associated with gallstones, exhibit a rapid increase in size, demonstrate arterial flow on Doppler ultrasonography, or are symptomatic are at increased risk of malignant transformation, and patients with such polyps warrant prophylactic cholecystectomy.46,48 Anomalous union of the pancreaticobiliary ductal system (AUPBD) has been associated with the development of gallbladder carcinoma. In this congenital defect, the pancreatic and bile ducts unite outside the duodenal wall in a long common channel. The anomaly is found incidentally in 1.5% to 2% of patients who undergo ERCP and leads to cholestasis and reflux of pancreatic secretions into the gallbladder with resulting chronic inflammation of the mucosa. Approximately 10% of patients with gallbladder carcinoma have coexisting AUPBD, and gallbladder carcinoma develops in 15% to 40% of those with AUPBD. Patients with an associated choledochal cyst have a lower frequency of gallbladder carcinoma than those without a choledochal cyst.49 Patients with AUPBD are usually 10 years younger at the time of diagnosis of gallbladder carcinoma and have a lower frequency of cholelithiasis than those without AUPBD. On the basis of a significantly increased risk of gallbladder carcinoma, several Japanese hepatobiliary oncology associations have recommended an aggressive, prophylactic surgical approach to patients with AUPBD.46

PSC has been associated with gallbladder carcinoma, and studies have reported that adenocarcinoma of the gallbladder develops in up to 20% of patients with PSC and that 40% to 60% of gallbladder masses in patients with PSC are malignant.50,51 Therefore, patients with PSC and a gallbladder mass of any size should undergo prophylactic cholecystectomy or be monitored closely for gallbladder carcinoma. Adenomyomatosis of the gallbladder, which is characterized by microscopic invaginations (Rokitansky-Aschoff sinuses) of the mucosa with cyst formation in the muscularis propria, has been linked to gallbladder carcinoma (see Chapter 67). The magnitude of the malignant potential of adenomyomatosis has not been clearly defined and appears to depend on morphologic features and the patient’s age.52 Adenomyomatosis generally is viewed as a benign condition. Other conditions associated with gallbladder carcinoma include inflammatory bowel disease, intrahepatic biliary dysplasia, and cholangiocarcinoma.51 Chronic carriers of Salmonella typhi or paratyphi have been shown to be at increased risk for the development of gallbladder carcinoma.53 Other bacteria such as Escherichia coli and Helicobacter pylori also have been associated with gallbladder carcinoma, but the data are not conclusive. First-degree relatives of patients with gallbladder carcinoma have a relative risk of 13.9 for developing this malignancy.54 Carcinogens, including methylcholanthrene, O-aminoazotoluene, and nitrosamines, have been identified in animal models of gallbladder carcinoma. Other potential carcinogens include mustard oil, products of free radical oxidation, and secondary bile acids.55

PATHOLOGY

From 80% to 95% of gallbladder carcinomas are adenocarcinomas; the majority of these are moderately-to-well differentiated.42 Less common types, in order of frequency, include undifferentiated or anaplastic carcinoma, squamous cell carcinoma, and adenosquamous carcinoma. Rare types include carcinoids, small cell carcinomas, malignant melanomas, lymphomas, and sarcomas.55 Sixty percent of gallbladder carcinomas are located in the gallbladder fundus, 30% in the body, and 10% in the gallbladder neck.56 Analogous to cholangiocarcinoma, the papillary form of gall­bladder carcinoma has a lower potential for invasion and metastatic spread to lymph nodes.57 Gallbladder carcinoma spreads via direct invasion, lymphogenic or hematogenic metastasis, perineural invasion, and intraperitoneal or intraductal invasion. Lymphatic tumor cell spread is determined by the physiologic gallbladder lymphatic plexus, including the first-level lymph nodes along the biliary tract (cystic duct, bile duct, and hepatic duct) followed by pancreaticoduodenal lymph nodes as well as lymph nodes along the common hepatic artery and celiac axis. Lymph node metastases are described in 54% to 64% of patients and correlate with the depth of invasion. Gallbladder carcinoma has a predisposition to involve the liver bed because of venous drainage, predominantly into hepatic segments IVb and V (see Chapter 71), and the anatomic approximation that allows direct hepatic invasion. Perineural spread is observed in 24% and intraductal spread in 19% of cases.

PATHOGENESIS

Gallbladder carcinoma can develop from foci of mucosal dysplasia or carcinoma in situ (CIS) that progress to adenocarcinoma or from an adenoma-carcinoma sequence similar

Chapter 69  Tumors of the Bile Ducts, Gallbladder, and Ampulla to that seen with colon cancer (see Chapter 123).58 Foci of dysplasia and CIS are frequently found adjacent to gallbladder carcinoma in surgically resected gallbladder specimens and are thought to be precursors of invasive adenocarcinoma.42 The time of progression of dysplasia to carcinoma is estimated to be 10 to 15 years.59 Like cholangiocarcinoma, the major pathogenic factor is inflammation. Increased iNOS and COX-2 expression has been demonstrated immunohistochemically in gallbladder carcinoma samples as well as in hyperplastic gallbladder mucosa from patients with AUPBD and also has been associated with TP53 tumor suppressor gene mutations in patients with gallbladder carcinoma. High expression and mutation rates of the TP53 gene have been demonstrated in 35% to 92% of gallbladder carcinomas, 86% of carcinomas in situ, and 28% of dysplastic foci, supporting an early role for TP53 mutation in the dysplasia-carcinoma progression sequence.60,61 In up to 60% of patients with gallbladder carcinoma, mutations of the K-ras oncogene have been detected; the frequency is highest in patients with AUPBD.62 Single studies have reported upregulation of ErbB2 and the nm23 metastasis suppressor protein.63,64 Mutations and increased expression of the nuclear oncogene that encodes p16INK4 have been shown in several studies.65 Other studies have demonstrated loss of heterozygosity or microsatellite instabilities in chromosomal regions that harbor known or putative tumor suppressor genes.66 Therefore, available data indicate pathogenetic mechanisms similar to those of other biliary malignancies, but the molecular pathogenesis appears to depend on the specific etiology and mode of progression.

A

CLINICAL FEATURES AND DIAGNOSIS

In two thirds of cases, gallbladder carcinoma is diagnosed incidentally during or after cholecystectomy for presumed benign disease, reflecting the initial clinically silent nature of this malignancy. Common clinical presentations include biliary or abdominal pain and jaundice secondary to direct invasion of the biliary ducts or metastases to the hepatoduodenal ligament. Weight loss, abdominal distention, or other symptoms resulting from compression or invasion of adjacent organs indicate more advanced disease. CEA and CA 19-9 are the most commonly used tumor markers in gallbladder carcinoma. As indicated earlier, these tests aid in diagnosis but should not be relied on, because levels can be elevated in inflammatory conditions and gastroenterologic and gynecologic malignancies; moreover, a subset of the population does not produce CEA. Abdominal ultrasound is often one of the first imaging studies performed in a patient who presents with the aforementioned symptoms. The sensitivity and accuracy of ultrasound for gallbladder carcinoma are 85% and 80%, respectively; early cancers, especially sessile polyps, can be missed. Typical imaging presentations of gallbladder carcinoma include focal or diffuse mural thickening of the gallbladder, an intraluminal mass greater than 2 cm in size that originates in the gallbladder wall, and a subhepatic mass that replaces or obscures the gallbladder and often invades adjacent organs (Fig. 69-6). Findings indicative of the malignant nature of a gallbladder lesion include irregular, asymmetrical mural thickening greater than 1 cm in depth and a nodular or smooth intraluminal mass greater than 1 cm in size, with fixation to the gallbladder wall, that is not displaced by the patient’s movements and has no acoustic shadow. In indeterminate cases, Doppler ultrasound can be attempted to differentiate a malignant from a benign gallbladder lesion on the basis of the pattern of the color signal, blood flow velocity, and resistive index (a measure of

B

C Figure 69-6.  Imaging of gallbladder carcinoma by computed tomography (CT) and ultrasonography. A, Axial CT view of the abdomen. Cholelithiasis is seen inferior to the gallbladder mass (arrow). B, Coronal view of the same patient. C, Ultrasonogram in the same patient showing a large mass (arrow) originating from the gallbladder wall and protruding into the lumen.

resistance to arterial blood flow).67 MRI and CT can be helpful in the diagnosis if the ultrasound findings are indeterminate. The role of PET in gallbladder carcinoma is evolving, and currently PET is not part of the routine evaluation. The sensitivity of PET for detecting gallbladder carcinoma is only 75% to 78%.68,69 Its main impact is in the

1179

1180

Section VIII  Biliary Tract Table 69-7  TNM and American Joint Committee on Cancer (AJCC)/International Union Against Cancer (UICC) Staging Systems for Gallbladder Carcinoma TNM STAGE

CRITERIA

Tx T0 Tis T1a T1b T2 T3

Primary tumor cannot be assessed No evidence of primary tumor Carcinoma in situ Tumor invades lamina propria Tumor invades muscularis propria Tumor invades perimuscular connective tissue without extension beyond serosa or into liver Tumor invades serosa OR Tumor invades one adjacent organ OR Both (extension ≤2 cm into liver) Tumor extends >2 cm into liver AND/OR Tumor extends into ≥2 adjacent organs (stomach, duodenum, colon, pancreas, omentum, extrahepatic bile ducts, liver)

T4

Nx N0 N1 N2

Regional lymph nodes cannot be assessed No regional lymph node metastases Metastases in cystic duct, pericholedochal, and/or hilar lymph nodes Metastases into peripancreatic (head only), periduodenal, periportal, celiac, and/or superior mesenteric lymph nodes

Mx M0 M1

Distant metastases cannot be assessed No distant metastases Distant metastases

AJCC/UICC Stage

Tumor

Node

Metastasis

0 IA IB II A II B III IV

Tis T1 T2 T3 T1-3 T4 Any T

N0 N0 N0 N0 N1 Any N Any N

M0 M0 M0 M0 M0 M0 M1

TNM, tumor, node, metastasis.

detection of distant metastases that result in a change in management.26

STAGING

The most commonly used staging system is the TNM system described by the AJCC and UICC. The TNM-based staging system correlates with survival. Reported five-year survival rates for patients with stages 0, I, II, III, and IV gallbladder carcinoma are 60%, 39%, 15%, 5%, and 1%, respectively (Table 69-7).

TREATMENT

Surgery is the only potential curative therapeutic option for gallbladder carcinoma. Only 15% to 47% of patients are candidates for surgery at the time of diagnosis because of the advanced stage of their disease. Contraindications to resection include multiple hepatic or peritoneal metastases, malignant ascites, distant metastases, extensive involvement of the hepatoduodenal ligament, encasement or occlusion of major vessels, and poor performance status. Direct involvement of the colon, duodenum, or liver is not considered an absolute contraindication to surgical resection, however. The goal of surgical treatment is an R0 resection, defined as negative margins and nodal dissection one level past microscopically involved lymph nodes. R0 resection in gallbladder carcinoma has been shown to correlate with survival and with significantly increased five-year survival rates.70 Surgical procedures with curative extent include (1) simple cholecystectomy; (2) extended or radical cholecystectomy with additional resection of greater than 2 cm

of the gallbladder bed plus lymphadenectomy of the hepatoduodenal ligament behind the second part of the duodenum, head of the pancreas, and celiac axis; (3) extended cholecystectomy with hepatic, segmental, or lobar resection; (4) extended cholecystectomy with extensive para-aortic lymph node resection; and (5) extended cholecystectomy with bile duct resection or pancreaticoduodenectomy. The surgical approach is dictated by the extent of tumor. Stage Tis and T1a (see Table 69-7) gallbladder cancer can be treated with simple cholecystectomy, with five-year survival rates of 85% to 100%. A few reports favor simple cholecystectomy also for stage 1b gallbladder carcinoma and report similar survival rates after either simple or radical cholecystectomy.71,72 Up to 15% of patients with stage 1b gallbladder carcinoma, however, are positive for lymph node metastases, compared with 2.5% of patients with stage 1a gallbladder carcinoma. Also, higher recurrence rates have been observed after simple (versus radical) cholecystectomy; therefore, radical cholecystectomy is preferred for stage 1b gallbladder carcinoma.73,74 Invasion of the muscularis propria, as in stage T2 or higher tumors, requires radical cholecystectomy. The extent and benefit of partial hepatectomy in stage T2 tumors are controversial, and partial hepatectomy has not been shown to prolong survival. Also, the surgical approach to stage T3 and T4 tumors is controversial. Some studies show no five-year survival benefit after radical cholecystectomy for stage T3 and T4 tumors, but other studies report five-year survival rates of 15% to 63% and 7% to 25%, respectively. Because of the poor prognosis of gallbladder carcinoma and the possibility of a survival benefit, as well as prolongation of

Chapter 69  Tumors of the Bile Ducts, Gallbladder, and Ampulla Postoperative diagnosis of gallbladder carcinoma

Staging

T2, T3, T4

T1

T1a

Re-exploration

T1b

Resectable

No further treatment if margins are negative

M1

Radical cholecystectomy

Unresectable

Palliative treatment

survival until recurrence, a radical surgical approach to these advanced-stage gallbladder carcinomas is recommended by many centers. When gallbladder carcinoma is diagnosed during laparoscopy, the procedure should be converted to an open procedure, and the laparoscopic port sites should be resected, because tumor may recur at these sites secondary to iatrogenic dissemination.75 Further surgical management then depends on the tumor stage, as outlined above and in Figure 69-7. When gallbladder carcinoma is diagnosed postoperatively, further management depends on the tumor stage and the presence or absence of tumor at the margins of the surgical specimen. Currently, chemotherapy and radiation therapy are not considered standard treatment for gallbladder carcinoma. A variety of different chemotherapeutic drugs, either alone or in combination, have been evaluated, including gemcitabine, fluoropyrimidines, mitomycin, and platinumbased compounds. Partial response rates of up to 36% and stabilization of disease in up to 48% of patients have been reported. The number of trials is limited, the number of patients in these trials has been small, and none of the studies has been randomized and controlled. Similarly, the role of radiation therapy is not defined in gallbladder carcinoma. In general, gallbladder carcinoma is considered radioresistant. Use of radiation therapy is further limited by the radiosensitivity of the surrounding tissue.

AMPULLARY CARCINOMA Carcinomas of the ampulla of Vater belong to the family of periampullary carcinomas. This family includes carcinomas of the duodenum, ampulla of Vater, distal bile duct, and pancreas. Ampullary carcinomas are the second most common form of periampullary carcinoma (after pancreatic head cancer). The distinction between the different forms is important because ampullary carcinomas are often diagnosed earlier than the others and therefore at a resectable stage, thus resulting in a better prognosis.

Figure 69-7.  Diagnostic algorithm for gallbladder carcinoma diagnosed intra- or postoperatively at laparoscopic cholecystectomy. In cases in which pathologic examination of the surgical cholecystectomy specimen identifies a stage T1a tumor with negative surgical margins, no further treatment is indicated. If the tumor is found to be a stage T1b tumor or the margins of resection are positive for malignant tissue, re-exploration for further resection is indicated. Similarly, patients with gallbladder carcinoma found to be stage T2, T3, or T4 should undergo surgical re-exploration. If re-exploration reveals resectable gallbladder carcinoma, radical cholecystectomy should be performed. If the tumor is deemed unresectable, palliative management is indicated. When postoperative staging reveals metastatic spread, palliative management is indicated. M, metastasis stage; T, tumor stage. (Modified from Misra S, Chaturvedi A, Misra NC, Sharma ID. Carcinoma of the gallbladder. Lancet Oncol 2003; 4:167-76.)

EPIDEMIOLOGY

Ampullary carcinomas are rare, accounting for fewer than 1% of all gastrointestinal cancers and 4% to 8% of periampullary carcinomas. The annual incidence has been estimated to be 0.6 per 100,000 population.76,77 Peak incidence is in the seventh decade of life. There is a slight male predominance, with a male-to-female ratio of 1.48 : 1.77 Racial heterogeneity has been observed; the vast majority of patients are white, followed by patients of Hispanic and Asian descent. African Americans have the lowest incidence rates in the United States.76

ETIOLOGY

Although the etiology of ampullary carcinomas is unknown in the majority of cases, several conditions have been associated with this malignancy, mostly in case reports or small series. Familial adenomatous polyposis (FAP) is an important risk factor for the development of ampullary carcinomas, with a relative risk of 124 (see Chapter 122).78 Periampullary carcinoma is the second most common cause of death (after colon cancer) in patients with FAP. Usually, periampullary carcinoma arises later than colorectal carcinoma in this patient group but earlier in comparison with sporadic ampullary carcinomas.77 Screening for upper gastrointestinal lesions (polyps or carcinoma) at regular intervals of six months to four years, depending on the degree of duodenal polyposis, is therefore recommended in patients with FAP. Similarly, increased rates of ampullary carcinoma have been described in patients with Gardner’s syndrome, a variant of FAP (see Chapter 122).79 Hereditary nonpolyposis colorectal cancer does not appear to pre­ dispose to ampullary carcinoma.80 Other genetic diseases reported to predispose to the development of ampullary carcinoma include neurofibromatosis type I and Muir-Torre syndrome.81,82 As for cholangiocarcinoma, chronic liver fluke infection has been reported to be a risk factor for ampullary carcinoma (see Chapter 82).77

PATHOLOGY

The ampulla of Vater is an anatomically complex area that consists of the papilla, common pancreaticobiliary channel,

1181

1182

Section VIII  Biliary Tract type typically develop from adenomas, whereas pancreaticobiliary and ulcerating carcinomas often lack a precursor lesion.77 On the molecular level, an increased frequency of K-ras mutations has been observed in 24% to 47% of tumors and is more common in the intestinal than pancreaticobiliary type.88 Also, p53 overexpression has been observed in 46% of tumors and is thought to be associated with ulcerating ampullary carcinomas. In an immunohistochemical study, aberrant expression of cell cycle regulators (i.e., p21WAF1/CIP1, p27Kip1, p16INK4, cyclin D1, type 8, and cyclin E, and the retinoblastoma protein [pRb]), was observed.89 These changes are similar to those seen in colorectal and pancreatic carcinomas, yet are distinctive. More research is necessary to understand the development of these tumors on a molecular level.

CLINICAL FEATURES AND DIAGNOSIS

Figure 69-8.  Endoscopic appearance of ampullary carcinoma. A catheter has been placed in the ampulla of Vater for biliary drainage after a sphincterotomy was performed.

distal bile duct, and distal main pancreatic duct. Macroscopically, ampullary carcinomas are classified into the following three subtypes: (1) intramural protruding (intraampullary), (2) extramural protruding (periampullary), and (3) ulcerating ampullary77 (Fig. 69-8). The ulcerating type is usually diagnosed at an advanced stage and has the highest rates of lymph node metastases. Consistent with its anatomic heterogeneity, the ampulla includes several different histologic cell types, such as epithelia of the common pancreaticobiliary channel, bile duct, pancreatic duct, or duodenal mucosa, Brunner’s glands, and aberrant pancreatic acini in the wall of the bile duct. The most common site of cellular atypia is found in the area of the common pancreaticobiliary channel, followed by the pancreatic duct, duodenal epithelium, and Brunner’s glands.83 Seventy-five percent of ampullary neoplasias are adenocarcinomas, 20% are benign adenomas, and 5% are neuroendocrine tumors.84 Adenocarcinomas account for 90% of ampullary malignancies; the rest include unusual types, such as mucinous, signet-ring cell, and undifferentiated carcinomas.84 Histopathologically, 90% of ampullary adenocarcinomas can be classified into pancreaticobiliary or intestinal types.85,86 Immunohistochemically, the two types can be distinguished by high cytokeratin 7 expression and a lack of intestinal apomucin (MUC2) in the pancreaticobiliary type and cytokeratin 20 as well as MUC2 expression in the intestinal type.87 The frequencies of the two different histologic types, as well as their correlation with invasion, lymph node metastases, and prognosis, differ in various studies, with some studies indicating that the frequency is higher and the prognosis is worse for the pancreaticobiliary type and other studies finding the opposite results.85-87

PATHOGENESIS

The majority of ampullary carcinomas follow an adenomacarcinoma sequence. In 30% to 91% of ampullary carcinomas, residual adenomatous tissue is found.77 Although precursor lesions can develop from intestinal as well as pancreaticobiliary-type tissue, carcinomas of the intestinal

Like the other periampullary and biliary malignancies, ampullary carcinomas present initially with obstructive jaundice in 70% to 82% of cases. Pancreaticobiliary ampullary carcinomas in particular have been reported to present initially with obstructive jaundice.85 Because of their anatomic location, cholestasis develops at an earlier stage than in other periampullary and biliary malignancies, and the resectability rate is therefore higher at the time of diagnosis. Nonicteric patients may present with bacterial cholangitis. Rare patients have “silver stools” as a result of the combination of acholic stools that result from bile duct obstruction and bleeding of the tumor. When obstructive cholangitis is suspected, further diagnostic evaluation is similar to that for other biliary malignancies. Immunohistochemical analysis has shown high expression of CEA and CA 19-9 in the tumor,90 but no studies have evaluated the serum concentrations of these markers in patients with ampullary carcinoma. Usually, ampullary carcinomas are diagnosed by endoscopy on the basis of their macroscopic appearance and findings on biopsy specimens (see Fig. 69-8). Subsequent diagnostic tests are directed toward an assessment of resectability and detection of metastases. As for other biliary and periampullary carcinomas, radiologic techniques such as CT and MRI are used commonly in this setting. On MRI/ MRCP, ampullary carcinoma usually is seen as a discrete, hypodense mass on T2-weighted images. Occasionally, the tumor can present as irregular thickening around the bile duct or bulging into the duodenum. Frequently, dilatation of both the bile and pancreatic ducts (“double-duct sign”) or only the bile duct is seen; dilatation of the pancreatic duct alone is seen rarely.91 Often, EUS is used in the preoperative evaluation. Its accuracy for detecting invasion of adjacent organs is 80% to 90%, and its sensitivity and specificity for detecting vascular invasion are 73% and 90%, respectively.92,93

STAGING

Ampullary carcinomas are classified according to the AJCC/ UICC TNM classification.28 The T stage was shown to be predictive of survival in a univariate analysis but not in a multivariate analysis.85,87 Independent predictors of survival by multivariate analysis are lymphovascular invasion, perineural invasion, stage greater than or equal to III, and pancreaticobiliary subtype.85 Several studies have confirmed the significance of nodal involvement in predicting survival.87,94

TREATMENT

As for other biliary malignancies, surgical resection is the only curative treatment for ampullary carcinomas. In

Chapter 69  Tumors of the Bile Ducts, Gallbladder, and Ampulla contrast to the other biliary malignancies, however, 77% to 88% of ampullary carcinomas are resectable at the time of diagnosis.95 The standard surgical approach is pancreaticoduodenectomy. Outcomes are excellent in the absence of lymph node metastases, with five-year survival rates of 68% to 78%.96,97 In the presence of lymph node–positive disease, the prognosis worsens significantly, with five-year survival rates of 16% to 25%.96,97 Extracapsular lymph node involvement results in further worsening of the prognosis, with a five-year survival rate of only 9%.90 Nodal microinvolvement has been reported to be an adverse prognostic factor, and immunohistochemical analysis of resected nodes has been recommended.98 Limited surgical or endoscopic papillectomy has been reported but is not recommended, because recurrence rates are higher than with pancreaticoduodenectomy.1 Chemotherapy and radiation therapy have not been evaluated in randomized controlled trials. A few studies have used 5-fluorouracil and radiation as adjuvant therapy in patients with node-positive disease and reported improved survival.99,100 In the absence of large, randomized controlled trials, however, such treatment is not standard. Palliative treatment should be directed at alleviating tumor-associated complications with the goal of optimizing the patient’s quality of life. Obstructive cholestasis is a major cause of morbidity and can usually be treated palliatively either by endoscopic or percutaneous placement of a biliary stent or by a surgical bypass similar to that carried out for other biliary or periampullary malignancies.

OTHER TUMORS OF THE BILE DUCTS AND GALLBLADDER Other neoplastic diseases may involve the biliary tract (Table 69-8). Their inclusion in the differential diagnosis of biliary tumors is essential, because management differs depending on the tumor type. Tumors of neuroectodermal origin such as carcinoids (see Chapter 31) and paragangliomas are rare and typically nonfunctioning.101 They are located most commonly in the ampulla of Vater. Occasionally, carcinoids develop in the extrahepatic biliary tree, predominantly in the bile duct. Patients are usually female and young. Primary carcinoids of the biliary tract constitute less than 1% of all gastrointestinal carcinoids and usually are not associated with the carcinoid syndrome.102,103 Approximately one third of patients have metastases at diagnosis. The treatment of choice is surgical resection, and the prognosis is generally good.104-106 Patients with paragangliomas often present with gastrointestinal bleeding; only 25% present with jaundice. Their malignant potential has been estimated to be 33%, and some investigators recommend pancreaticoduodenectomy as the treatment of choice.107 Granular cell tumors, which are of neuronal derivation, are extremely rare; only a few cases have been described. Usually, they are located in the extrahepatic biliary tree, particularly at the junction of the cystic duct and the bile duct.108 Occasionally, they can cause biliary obstruction, as occurs when they are located in the hepatic hilum.108 Because of their benign character, resection is usually curative.109 Rarely, neuromas of the extrahepatic biliary tree develop after cholecystectomy.110 Mesenchymal tumors, such as lipomas, leiomyomas, hemangiomas, and lymphangiomas, have been described in the gallbladder. In general, mesenchymal tumors are extremely rare and restricted to case reports. Lymphangiomas are often asymptomatic and only incidentally detected;

Table 69-8  Other Tumors of the Gallbladder and Bile Ducts GALLBLADDER Benign

Malignant

Tumor-like lesions

Adenoma Granular cell tumor Mesenchymal tumor (lipoma, leiomyoma, hemangioma, lymphangioma) Paraganglioma Adenosquamous carcinoma Carcinoid Small cell carcinoma Spindle cell sarcomatoid carcinoma Others (angiosarcoma, carcinosarcoma, Kaposi’s sarcoma, leiomyosarcoma, malignant fibrous histiocytoma, melanoma, metastatic tumors, non-Hodgkin’s lymphoma, rhabdomyosarcoma) Adenomyoma/adenomyomatosis Cholesterol polyp Heterotopia (gastric, pancreatic, liver, adrenal, thyroid) Inflammatory polyp

BILE DUCTS Benign

Malignant

Precursor lesions

Adenomyoma Bile duct adenoma Biliary adenofibroma Biliary cystadenoma and cystadenocarcinoma Biliary hamartoma Ciliated hepatic foregut cyst Granular cell tumor Neuroma Serous cystadenoma Solitary or multiple cysts Carcinoid Embryonal (botryoid) rhabdosarcoma Leukemia Lymphoma Melanoma Metastatic tumor Paraganglioma Biliary dysplasia (intraepithelial neoplasia/ atypical hyperplasia) Intraductal papillary neoplasia

however, they can increase in size and result in abdominal pain or jaundice. Ultrasound, CT, and MRI/MRCP aid in the preoperative diagnosis. Usually, they manifest as a multilocular, fluid-filled, cystic mass with thin walls and septa and show enhanced signal density with contrast administration.111 Most of the reported cases have been treated successfully with surgical resection, including cholecystectomy if the tumor is located within the gallbladder, or endoscopic resection if the tumor is in the area of the ampulla of Vater.112-115 Hamartomas also have been reported in the area of the ampulla of Vater and have been resected successfully by endoscopy.116 Heterotopia of the gallbladder may be caused by gastric, pancreatic, hepatic, adrenal, or thyroid tissue. Clinical complications such as hemorrhage are extremely rare.117,118 Benign bile duct lesions include adenomas, cystadenomas, adenofibromas, cysts, and granular cell tumors. Adenomyomas are found more commonly in the ampulla of Vater. Cystadenomas are more common in women and manifest primarily with abdominal pain. They are found predominantly in the intrahepatic biliary tree and are characterized on ultrasound by papillary extrusions of the wall

1183

1184

Section VIII  Biliary Tract and septa. They are considered premalignant because of their potential to transform into cystadenocarcinomas; hence, the treatment of choice is complete resection (see Chapter 94).119-121 Malignant tumors of the biliary tree other than cholangiocarcinoma include cystadenocarcinomas, lymphomas, and malignant melanomas. These malignancies arise primarily in the extrahepatic bile ducts. Cystadenocarcinomas can be distinguished morphologically from cholangiocarcinomas by their cystic character.122 They are rarely located in the gallbladder.123 Symptoms are nonspecific, and CT and MRI can be helpful in making the diagnosis. The treatment of choice is surgical resection.123 Malignant melanoma of the biliary tree is uncommon and should prompt investigation for a cutaneous melanoma, because cases of metastatic spread to the bile ducts have been described.124,125 Lymphomas can occasionally involve the extrahepatic biliary tree and often are mistaken for cholangiocarcinoma.126,127 In general, biliary lymphomas are very rare and account for less than 1% of lymphomas.127,128 Embryonal rhabdomyosarcoma of the biliary tree is extremely rare in adults but is the most common malignant tumor at this anatomic location in children.129 Frequently, it is misdiagnosed preoperatively as a choledochal cyst.130 Complete surgical resection is rarely possible, and a multidisciplinary approach to treatment is recommended. The prognosis of biliary rhabdosarcomas is good, with reported five-year survival rates of up to 78%.131 Few reports exist of follicular lymphomas originating in the extrahepatic biliary tree and gallbladder. Often, these tumors are diagnosed after resection.

ACKNOWLEDGMENTS

This work was supported by a grant from the NIH DK59427, the Mayo Clinic Clinical Investigator Program, and the Mayo Foundation.

KEY REFERENCES

Blechacz B, Gores GJ. Cholangiocarcinoma: Advances in pathogenesis, diagnosis, and treatment. Hepatology 2008; 48:308-21. (Ref 2.)

Buckles DC, Lindor KD, Larusso NF, et al. In primary sclerosing cholangitis, gallbladder polyps are frequently malignant. Am J Gastroenterol 2002; 97:1138-42. (Ref 50.) Csendes A, Becerra M, Rojas J, Medina E. Number and size of stones in patients with asymptomatic and symptomatic gallstones and gallbladder carcinoma: A prospective study of 592 cases. J Gastrointest Surg 2000; 4:481-5. (Ref 45.) Gores GJ, Nagorney DM, Rosen CB. Cholangiocarcinoma: Is transplantation an option? For whom? J Hepatol 2007; 47:455-9. (Ref 35.) Isomoto H, Kobayashi S, Werneburg NW, et al. Interleukin 6 upregulates myeloid cell leukemia-1 expression through a STAT3 pathway in cholangiocarcinoma cells. Hepatology 2005; 42:1329-38. (Ref 16.) Ito H, Matros E, Brooks DC, et al. Treatment outcomes associated with surgery for gallbladder cancer: A 20-year experience. J Gastrointest Surg 2004; 8:183-90. (Ref 70.) Jaiswal M, LaRusso NF, Burgart LJ, Gores GJ. Inflammatory cytokines induce DNA damage and inhibit DNA repair in cholangiocarcinoma cells by a nitric oxide-dependent mechanism. Cancer Res 2000; 60:184-90. (Ref 15.) Kimura W, Futakawa N, Yamagata S, et al. Different clinicopathologic findings in two histologic types of carcinoma of papilla of Vater. Jpn J Cancer Res 1994; 85:161-6. (Ref 86.) Levy C, Lymp J, Angulo P, et al. The value of serum CA 19-9 in predicting cholangiocarcinomas in patients with primary sclerosing cholangitis. Dig Dis Sci 2005; 50:1734-40. (Ref 23.) Moreno Luna LE, Kipp B, Halling KC, et al. Advanced cytologic techniques for the detection of malignant pancreatobiliary strictures. Gastroenterology 2006; 131:1064-72. (Ref 27.) O’Connell JB, Maggard MA, Manunga J, et al. Survival after resection of ampullary carcinoma: A national population-based study. Ann Surg Oncol 2008; 15:1820-7. (Ref 76.) Petrowsky H, Wildbrett P, Husarik DB, et al. Impact of integrated positron emission tomography and computed tomography on staging and management of gallbladder cancer and cholangiocarcinoma. J Hepatol 2006; 45:43-50. (Ref 26.) van der Gaag NA, ten Kate FJ, Lagarde SM, et al. Prognostic significance of extracapsular lymph node involvement in patients with adeno­ carcinoma of the ampulla of Vater. Br J Surg 2008; 95:735-43. (Ref 94.) Yoon JH, Canbay AE, Werneburg NW, et al. Oxysterols induce cyclooxygenase-2 expression in cholangiocytes: Implications for biliary tract carcinogenesis. Hepatology 2004; 39:732-8. (Ref 19.) Yoon JH, Gwak GY, Lee HS, et al. Enhanced epidermal growth factor receptor activation in human cholangiocarcinoma cells. J Hepatol 2004; 41:808-14. (Ref 18.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

70 Endoscopic and Radiologic Treatment of Biliary Disease Andrew H. Stockland and Todd H. Baron

CHAPTER OUTLINE Imaging of the Biliary Tract  1185 Ultrasonography  1185 Magnetic Resonance Cholangiopancreatography and Multidetector Computed Tomography Cholangiography  1185 Percutaneous Transhepatic Cholangiography  1186 Technique  1186 Postoperative Biliary Strictures  1188 Primary Sclerosing Cholangitis  1188 Bile Leaks  1188 Bile Duct Injury  1188 Bile Duct Stones  1189 Malignant Biliary Obstruction  1189 Percutaneous Cholecystostomy Tube Placement  1191

Endoscopic therapy and radiologic treatment of biliary disease have evolved in separate but parallel manners. Endoscopic therapy is performed using endoscopic retrograde cholangiopancreatography (ERCP) and, more recently, using endoscopic ultrasound (EUS)-guided techniques. ERCP is performed primarily by endoscopists trained in a gastroenterology training program, but in some centers it is performed by surgeons. ERCP is one of the most technically demanding endoscopic procedures, and for the successful management of complex cases, the learning curve is steep. Radiologic therapy of the biliary tree is performed via a percutaneous approach by interventional radiologists. The two approaches should be seen as complementary rather than competitive. The decision to proceed with an endoscopic or radiologic approach is often based on local expertise; other considerations include physician referral patterns, location of a lesion within the biliary tree, failure of one method, and altered anatomy as a result of prior surgery.

IMAGING OF THE BILIARY TRACT Imaging of the biliary tree is of utmost importance in planning the management approach to patients with biliary disorders and is discussed briefly in this context.

ULTRASONOGRAPHY

Noninvasive imaging of the biliary tree frequently begins with transabdominal ultrasound, which provides a global

Endoscopic Retrograde Cholangiopancreatography  1191 Bile Duct Stones  1191 Bile Leaks  1191 Primary Sclerosing Cholangitis  1192 Benign Biliary Strictures  1193 Indeterminate Biliary Strictures  1193 Malignant Biliary Strictures  1193 Sphincter of Oddi Dysfunction  1196 Complications  1197 Combined Percutaneous and Endoscopic Approaches  1197 Endoscopic Ultrasonography  1197 Diagnostic Role  1197 Therapeutic Role  1197

picture of the liver and is nearly universally available. There is no radiation exposure, and contrast agents are not required. Intrahepatic ductal dilatation can be visualized easily and the size of the bile duct can be documented. Ultrasound also provides imaging of the gallbladder and detects gallstones. For detection of choledocholithiasis, ultrasound has a high specificity, but the sensitivity does not exceed 68% and is often lower than 50%.1,2 The sensitivity decreases if the stones are small and the bile ducts are not dilated. Ultrasound is highly accurate (78% to 98%) for detecting extrahepatic biliary obstruction.2 When used in conjunction with the clinical evaluation, ultrasound allows differentiation between liver parenchymal disease and extrahepatic biliary obstruction with a reasonable sensi­ tivity and high specificity.2 Ultrasound is less accurate, however, at defining the level and cause of obstruction, with accuracy rates ranging from 27% to 95% and 23% to 88%, respectively.2 In addition, ultrasound is limited in the ability to distinguish malignant from benign causes of obstruction.2

MAGNETIC RESONANCE CHOLANGIOPANCREATOGRAPHY AND MULTIDETECTOR COMPUTED TOMOGRAPHY CHOLANGIOGRAPHY

Magnetic resonance cholangiopancreatography (MRCP) is a magnetic resonance imaging (MRI) study (and thus noninvasive) that is dependent on the high T2-signal charac­ teristics of bile. It does not require administration of oral or

1185

1186

Section VIII  Biliary Tract

Figure 70-1.  Choledocholithiasis on magnetic reso­ nance cholangiopancreatography (MRCP) and endo­ scopic retrograde cholangiopancreatography (ERCP). A, MRCP showing filling defect in distal bile duct (arrow). B, Corresponding ERCP with same filling defect.

A

intravenous contrast material. For the detection of choledocholithiasis, MRCP has a sensitivity ranging from 81% to 100%, a specificity ranging from 96% to 100%, and high overall diagnostic accuracy (Fig. 70-1).3 In addition, MRCP is highly accurate in demonstrating the presence of benign and malignant strictures4 and allows a thorough evaluation of the intrahepatic bile ducts. In patients suspected of having post-liver transplant biliary complications, intravenous administration of mangafodipir trisodium (Teslascan, Amersham Health, Princeton, NJ) may be used. This agent is excreted primarily in the bile and may improve imaging sensitivity for post-liver transplant biliary leaks and strictures.5 An MRI can be performed as well with an intravenous contrast agent, such as gadodiamide (Omniscan, GE Healthcare, United Kingdom) or gadopentetate dimeglumine (Magnevist, Bayer Healthcare, Leverkusen, Germany or Multihance, Bracco, Princeton, NJ), to detect and characterize mass lesions in the liver, porta hepatis, or pancreas. Contraindications to MRI include a cardiac pacemaker, automatic implantable cardioverter defibrillator, and some types of cerebral aneurysm clips. A particular concern about gadolinium-based intravenous contrast agents is that they may precipitate nephrogenic systemic fibrosis, a rare scleroderma-like disease manifested by hardening of the skin and fibrotic changes that affect multiple organs. The cause remains unclear, but reports suggest that patients with preexisting kidney disease (renal failure) are at greatest risk.6,7 Multidetector computed tomography cholangiography (MDCT) with multiplanar reformation is a computed tomography (CT)-based imaging study. MDCT is a combination of rapid volume acquisition and thin-slice imaging. Water is used as an oral contrast agent for the biliary tree, and intravenous iodinated contrast is also administered. Images acquired in the axial plane can be reconstructed sagittally or coronally and reformatted three dimensionally. The intravenous contrast dye is not excreted in bile but enhances adjacent surrounding visceral structures such as the liver,

B

pancreas, and other soft tissues. Bile ducts thus appear as low attenuation structures that are best visualized if dilated. The sensitivity and specificity of MDCT for bile duct strictures have been reported to be 85.7% and 100%, respectively.8 MDCT also has a high sensitivity and specificity for the detection of bile duct stones.9

PERCUTANEOUS TRANSHEPATIC CHOLANGIOGRAPHY Percutaneous transhepatic cholangiography (THC) is an invasive diagnostic test and can be therapeutic if necessary. In light of the array of noninvasive imaging studies, percutaneous THC is rarely performed purely for diagnostic purposes. Subsequent decompression of biliary obstruction, removal of a stone, balloon dilation of a stricture, and placement of a stent for a stricture can be performed. The procedure is generally reserved for patients for whom ERCP is precluded because of difficult endoscopic access across a biliary-enteric (Roux-en-Y) anastomosis, gastric bypass, or extrahepatic biliary stricture that cannot be traversed endoscopically. Serious procedure-related complications such as bleeding, sepsis, or bile leakage occur in approximately 2% to 4% of cases.10 The procedure generally can be performed with monitored moderate (“conscious”) sedation.11 Broadspectrum intravenous antibiotics are usually administered prophylactically.

TECHNIQUE

Review of CT and MR imaging of the liver prior to percutaneous THC can help determine the best approach (i.e., from the right or left side) and the location of the dominant dilated ducts and help avoid traversing adjacent structures, such as the colon, unintentionally. Dilated bile ducts on the left side may be easily accessible with a minimal number of needle passes with use of ultrasound guidance from a

Chapter 70  Endoscopic and Radiologic Treatment of Biliary Disease subxiphoid approach.12 A standard right-sided approach is used most frequently, however, and is performed from an intracostal approach, usually via the mid-axillary region below the 10th intercostal space. Higher punctures increase the risk of pneumothorax or biliary pleural effusion. From either side, the procedure is initiated by advancing a 22-gauge needle under fluoroscopic guidance centrally toward the liver hilum and gently injecting contrast as the needle is withdrawn slowly. The initial use of such a small needle reduces hepatic trauma as well as the likelihood of bleeding despite the potential need for multiple needle passes to cannulate a bile duct, particularly when the bile ducts are not dilated. Ultrasound guidance and CT guidance can be used to access nondilated bile ducts.13,14 When a bile duct is cannulated, a diagnostic cholangiogram can be performed. Isolated ducts, because of strictures or stones that do not communicate with the rest of the biliary tree, may need to be opacified via additional needle passes. If the procedure is only diagnostic, and biliary obstruction is not evident, the needle is simply withdrawn. If the biliary system is obstructed, however, serious consideration should be given to traversing the obstruction and leaving a decompressive “external-internal” tube in place; abandoning an obstructed biliary system may lead to bile leakage from the puncture site. The risk of hepatic arterial injury is reduced by using a peripheral intrahepatic bile duct for final access. If the duct cannulated initially is too central (the larger branches of the hepatic artery tend to be more central), a more peripheral duct should be chosen for access into the biliary tree. Frequently, use of a second needle to puncture a more peripheral duct is required, and the initial needle is used to opacify and visualize this new and safer access duct. A 0.018-inch “micro” wire is then advanced via the needle into the biliary tree and the access system “upsized” by the passage of catheters of increasing diameter over the wire. When access is gained, the obstruction can be traversed and an external-internal biliary tube can be placed (Fig. 70-2). These tubes provide drainage

A

C

holes positioned above the level of obstruction; the distal pigtail is configured within the small intestine. The size of the tube usually ranges from 8 to 12 French. A larger tube may yield better decompression, but care must be taken not to place a tube in which the size may actually obstruct drainage of smaller ducts, particularly in the setting of primary sclerosing cholangitis (PSC), in which many of the obstructed ducts are not dilated. If the obstruction cannot be traversed during the initial attempt, a drainage catheter can be left proximal to the obstruction in the biliary tree (external drainage), and subsequent attempts can be made via this access after several days of drainage. This delay often allows inflammation to decrease and increases the likelihood of subsequent internalization of a catheter. Generally, the external-internal drainage tube is left to external drainage until fever or blood in the biliary tree resolves. Capping of the external end of the tube to permit internal drainage only decreases biliary fluid losses, which can be more than 1 liter per day, and prevents associated dehydration or electrolyte abnormalities. Bile samples obtained during the initial procedure can be sent for culture or cytology. Contraindications to percutaneous THC include coagulopathy. Generally the procedure is thought to be safe with an international normalized ratio (INR) of less than 1.8 and platelet count greater than 50,000/mm3. Any abnormalities should be corrected immediately before the procedure. Marked ascites between the liver and puncture site increases the risk of bile leakage, whereas a tortuous biliary catheter course may lead to malposition of the catheter or difficulty with future manipulations. In the presence of a substantial amount of perihepatic ascites, a pre-procedure paracentesis can be performed or ultrasound guidance can be used to place a small temporary peritoneal drainage catheter adjacent to the liver for the duration of the procedure. Biliary sepsis can be minimized by avoiding overdistention of the bile ducts and limiting the number of manipulations during the procedure. As soon as a tube is placed, it

B

D

Figure 70-2.  Schematic showing percutaneous trans­ hepatic cholangiography. A, A peripheral bile duct is identified and entered with a needle. B, A guidewire is passed through the needle across the obstructing lesion into the duodenum. C, The needle has been withdrawn. D, An internal-external catheter is inserted over the guidewire.

1187

1188

Section VIII  Biliary Tract can be used as an access for further manipulations or interventions. Following initial biliary decompression, further intervention should be avoided until fever and sepsis have resolved. Patients need to be monitored closely for the first 24 to 48 hours following the procedure. Brisk bleeding around the catheter site, through the catheter itself, or from the gastrointestinal tract suggests the possibility of hepatic arterial injury.15 Presentation of a hepatic artery pseudoaneurysm can be delayed, sometimes for a week or two after the initial procedure. If bleeding persists, the hemoglobin level drops substantially or the patient becomes hemodynamically unstable, hepatic angiography should be considered, and, if an injured arterial branch is demonstrated, embolization should be performed. A small amount of blood in the biliary tube or bile ducts following the original procedure, or during subsequent manipulations, is frequently self-limited and clears within one or two days.

POSTOPERATIVE BILIARY STRICTURES

Postoperative strictures may occur following laparoscopic cholecystectomy, major hepatic resection, and liver transplantation at a choledochocholedochal anastomosis or within an intrahepatic duct as a result of ischemia or recurrent PSC (Table 70-1) (see Chapters 66, 68, and 95). Dilation of a postoperative or other benign biliary stricture can be performed via percutaneous THC or through a mature, surgically-placed T-tube tract. Maturation of the T-tube tract usually requires six weeks. Percutaneous THC and biliary balloon dilation may be performed at the same session in the absence of clinical signs of cholangitis or sepsis. An 8-French or 10-French transhepatic tube is left in place, and the patient returns for repeat cholangiography six weeks later, at which time further stricture dilation is performed if bile duct narrowing of 30% or greater persists. The tube is then repeatedly upsized to a 12-French tube to facilitate healing of the stricture at a larger diameter. If the stricture resolves on follow-up, the biliary tube can be removed; otherwise, a similar procedure should be performed after six to eight weeks. In one of the largest series published with long-term follow-up, percutaneous biliary balloon dilation was performed in 85 patients with a benign biliary stricture.16 In the 75 patients with follow-up, 205 percutaneous procedures were performed during 112 treatments of 84 biliary strictures. Stricture balloon dilation from 8 to 12 mm was performed. Procedures were repeated at 2- to 14-day intervals until cholangiography demonstrated free drainage of contrast material to the small intestine and no residual stenosis. An internal-external biliary drain was left in place for a mean of 14 to 22 days and removed if the patient did well when the catheter was clamped and had a normal cholangiogram. All procedures were technically successful. A total of 52, 11, 10, and 2 patients underwent a total of one, two, three, and four dilations, respectively. Major complications occurred in 2% of procedures: two subphrenic abscesses,

Table 70-1  Principal Causes of Benign Biliary Strictures Chronic pancreatitis Postoperative Bile duct resection with choledochocholedochal anastomosis Laparoscopic cholecystectomy Liver transplantation Primary sclerosing cholangitis

one hepatic arterial pseudoaneurysm, and one case of hemobilia. The probability that clinically significant restenosis did not develop at 5, 10, 15, 20, and 25 years was 0.52, 0.49, 0.49, 0.41, and 0.41, respectively, after the first treatment, and 0.43, 0.30, 0.20, 0.20, and 0.20, respectively, after the second treatment. No significant difference was found in the rate of restenosis for strictures at anastomotic and nonanastomotic sites. Overall, 56 of 75 patients (75%) had successful management with percutaneous therapy. Following liver transplantation, percutaneous THC is used for treating complications in patients with a ductto-duct anastomosis and especially in patients in whom hepaticojejunostomy has been performed; these latter anastomoses frequently cannot be accessed via an endoscopic approach. Hepaticojejunal anastomosis is performed at the time of liver transplantation in children, persons with PSC, persons who undergo reoperation for a complication of a duct-to-duct (choledococholedochal) anastomosis, and living-related donors. For treatment of both duct-to-duct and hepaticojejunal anastomotic strictures, percutaneous therapy provides a high nonoperative success rate.17-19 In addition, in those patients in whom the bile duct is approachable via ERCP but who fail an endoscopic approach, a percutaneous approach is often successful.20

PRIMARY SCLEROSING CHOLANGITIS

Most nonsurgical therapeutic interventions for PSC are now performed via ERCP (see Chapter 68). In the past, percutaneous therapy for a dominant stricture using balloon dilation followed by biliary drain placement for two to three months was found to be highly effective for treating obstructive biliary symptoms in patients with PSC21 but less effective in patients with jaundice for more than six months because of liver parenchymal dysfunction. More recently, only case reports of percutaneous therapy for PSC have appeared in the literature.22 In our experience, percutaneous therapy is useful for patients with a dominant stricture that cannot be accessed endoscopically (Fig. 70-3). In these cases, a guidewire or catheter passed percutaneously can be left in the duodenum to facilitate future endoscopic access (see later).

BILE LEAKS

Bile leaks are almost always postsurgical in etiology and arise from anastomotic (e.g., post-liver transplant) and nonanastomotic sites. The latter include cut surfaces of the liver and bile ducts following hepatectomy and laparoscopic injury. Percutaneous management may include drainage of free bile from the peritoneal cavity and of localized bile collections (bilomas) as well as placement of a biliary catheter above or across the leaking site to allow successful closure in the majority of cases.23-25

BILE DUCT INJURY

Widespread performance of laparoscopic cholecystectomy has led to an increased frequency of major bile duct injuries (see Chapter 66). Other causes of bile duct injury include bile duct exploration or biliary injury resulting from abdominal surgery or trauma. Percutaneous transhepatic biliary drain placement can be used as primary treatment of the injury or to augment surgical repair. Misra and colleagues26 retrospectively evaluated 51 patients who underwent percutaneous biliary management following laparoscopic cholecystectomy-related bile duct injuries over a 10-year period; 45 had operative repair prior to referral. Overall, 46 of the 51 were initially managed percutaneously, and 5 were managed percutaneously following failed hepaticojejunostomy. Nonoperative percutaneous management with balloon

Chapter 70  Endoscopic and Radiologic Treatment of Biliary Disease

A

B

C Figure 70-3.  Percutaneous therapy for recurrent primary sclerosing cholangitis after liver transplantation. The patient has a hepaticojejunostomy. A, Initial puncture and passage of a guidewire into the jejunum. B, Balloon dilation of right and left hepatic ducts. C, Follow-up cholangiogram with indwelling internal-external biliary catheter.

dilation resulted in an overall success rate of 58.8% at a mean follow-up of 76 months.

BILE DUCT STONES

Bile duct stones can be managed percutaneously via cholecystostomy tubes, percutaneous placed drains, or surgical T-tubes. Gallbladder tube or T-tube tracts require approximately six weeks to mature prior to use. In many cases, bile duct stones can be cleared percutaneously by dilating the papilla from an antegrade approach.27,28 The stones, which may require mechanical fragmentation, are flushed into the duodenum. The percutaneous catheter is replaced for several days and then removed. With this approach, in one study27 stones were removed in 95 of 100 patients. In some cases, particularly in the setting of complex intrahepatic stones, a small-caliber choledochoscope (cholangioscope) can be passed through a mature percutaneous tract. Stones are then fragmented using a variety of techniques, with a high rate of success (see Chapter 65).29,30

MALIGNANT BILIARY OBSTRUCTION

Stents can be placed percutaneously for relief of malignant biliary obstruction, either preoperatively or for palliation. Stents are composed of either rigid plastic or self-expanding metal. Self-expanding metal stents (SEMS) were designed to avoid occlusion from bacterial biofilm, which invariably occurs in plastic stents and results in the need for re-intervention. Distal bile duct strictures (e.g., caused by pancreatic head cancer) are preferably managed via ERCP (see later) because endoscopic stent placement is less painful than percutaneous stent placement and is associated with fewer compli­ cations. This conclusion is based on a randomized trial of endoscopic and percutaneous approaches using plastic stents.31 Percutaneous stent placement can easily be achieved in these patients, however. Multiple interventions are often needed to place plastic stents prior to final internalization of the stent because large-bore (≥10-French) stents require dilation of a tract through the liver, which

1189

1190

Section VIII  Biliary Tract

A

B

C Figure 70-4.  Percutaneous management of calculous cholecystitis. A, Percutaneous catheter is placed into the gallbladder. Note multiple gallstones. B, A large-caliber tube allows extraction of stones. C, Complete clearance of gallstones is demonstrated.

often cannot be accomplished in one stage. Bleeding, which occurs with such aggressive dilation, often requires main­ tenance of an external catheter to drain blood within the biliary tree. More recently, SEMS have been used. In a classic study, percutaneous placement of SEMS was asso­ ciated with significantly longer stent patency, reduction in the need for re-intervention, and shorter hospital stays.32 In addition, advantages of SEMS when placed percutaneously are the availability of small-diameter pre-deployment delivery systems, so that the percutaneous tract does not require dilation, and the capability for stent insertion in one step.32,33 In a randomized trial34 of endoscopic versus percutaneous palliation of malignant bile duct obstruction in which metal biliary stents were placed percutaneously in one step and plastic stents were placed endoscopically,

percutaneous placement of SEMS was associated with a 34% lower rate of recurrent biliary obstruction. Relief of hilar biliary obstruction (e.g., caused by hilar cholangiocarcinomas, or Klatskin tumors) is more difficult to achieve endoscopically than relief of distal bile duct obstruction. Several studies have suggested that the percutaneous approach to these tumors is superior to the endoscopic approach, with a lower rate of post-procedure cholangitis.34,35 Covered SEMS were designed to improve stent patency by reducing the frequency of occlusion resulting from tumor ingrowth and tissue hyperplasia. Studies have shown promising results,36,37 although no randomized trials of covered versus uncovered stents placed via the percutaneous approach have been published.

Chapter 70  Endoscopic and Radiologic Treatment of Biliary Disease PERCUTANEOUS CHOLECYSTOSTOMY TUBE PLACEMENT The standard treatment of acute calculous cholecystitis is cholecystectomy (see Chapters 65 and 66). Even with the advent of laparoscopic cholecystectomy, some patients are still not surgical candidates. Percutaneous cholecystostomy tube placement is a minimally invasive way to treat these patients and can be performed with a local anesthetic or with moderate sedation. Tube placement enables immediate decompression of the gallbladder. Bile samples obtained during tube placement can be used to guide antimicrobial therapy, and the tube can be used for cholangiography to confirm cystic duct obstruction or, if the cystic duct becomes patent, bile duct obstruction. Percutaneous gallbladder therapy is useful for the management of severe acute calculous cholecystitis as a nonoperative approach in elderly patients or persons who are poor candidates for surgery and as a way to avoid emergency surgery.38 In the last situation, an elective cholecystectomy can be performed subsequently, often laparoscopically.39,40 If the patient remains a poor surgical candidate, the cholecystostomy tube can remain in place long term. Alternatively, stones in the gallbladder, cystic duct, or bile duct can be managed percutaneously. The mature percutaneous tract can be dilated, after which the stones can be extracted (Fig. 70-4). Patients with intrahepatic gallbladders and small, shrunken, thick-walled gallbladders are not candidates for this approach. Despite the high success rate of percutaneous stone removal, stones can recur.41 An additional challenge in patients in an intensive care unit is the management of suspected acute acalculous cholecystitis (see Chapter 67). A Murphy sign can be difficult, if not impossible, to demonstrate, particularly in intubated or unresponsive patients. Delayed diagnosis and treatment can lead to gallbladder gangrene and perforation and to mortality. In patients in whom clinical suspicion for acute acalculous cholecystitis is high, a gallbladder tube should be placed percutaneously. If the gallbladder is not the source of the patient’s clinical problem, the cholecystostomy tube remains in place for six weeks. In one study of 55 critically ill patients with suspected acute acalculous cholecystitis who underwent percutaneous gallbladder tube placement, clinical improvement was seen in 58.7% within 24 hours and 95.7% within 72 hours.42

ENDOSCOPIC RETROGRADE CHOLANGIOPANCREATOGRAPHY Endoscopic retrograde cholangiopancreatography has evolved from a purely diagnostic to an almost exclusively therapeutic procedure. ERCP is commonly performed using moderate sedation,43 although in severely ill patients and in cases anticipated to be complex, an anesthesiologist is often needed. ERCP is performed with a side-viewing duodenoscope that allows identification of the major papilla. The bile duct is cannulated under endoscopic and fluoroscopic guidance. A variety of catheters, guidewires, and stents are available to allow therapeutic interventions to be performed. Diagnostic ERCP is still used for facilitating manometry in patients with suspected sphincter of Oddi dysfunction (see Chapter 63) and for establishing the diagnosis of PSC when other imaging techniques have been nondiagnostic (see Chapter 68).44 A variety of biliary indications for ERCP45 will each be discussed.

BILE DUCT STONES

ERCP is usually performed in patients with known choledocholithiasis or in those patients with at least a moderate clinical suspicion of choledocholithiasis (see Chapter 65). In patients with gallbladder stones and a low clinical suspicion of choledocholithiasis, noninvasive imaging studies (MRCP, MDCT) or EUS are preferred to minimize the potential for complications of ERCP.46 In patients with a low clinical suspicion of choledocholithiasis in whom cholecystectomy is planned, intraoperative cholangiography can be performed, and, if stones are identified, laparoscopic exploration and stone removal can be undertaken. ERCP can then be reserved for patients in whom the stones are not extracted.46 The standard method for stone removal is endoscopic biliary sphincterotomy to allow enlargement of the papilla and subsequent extraction of stones with a balloon or basket (Fig. 70-5). With this approach, more than 80% of all stones can be removed successfully.47 Larger stones may require additional removal techniques (discussed later). An alternative to biliary sphincterotomy is balloon dilation of the papilla (balloon sphincteroplasty), which can be performed using small-diameter balloons (4 to 8 mm). The technique was introduced as a way to preserve sphincter of Oddi function, especially in young patients. The stones are removed using balloon or basket techniques. Most of the literature on balloon sphincteroplasty comes from outside the United States. Two meta-analyses of randomized trials of balloon sphincteroplasty versus sphincterotomy have shown that the rates of pancreatitis and need for mechanical lithotripsy are significantly higher, but the risk of bleeding is significantly lower, with balloon sphincteroplasty than with sphincterotomy.48,49 In the United States, the only randomized trial that compared balloon sphincteroplasty and sphincterotomy was closed prematurely because of two deaths in young patients from post-ERCP pancreatitis after sphincteroplasty.50 Sphincteroplasty still remains, however, an alternative approach in patients with coagulopathy,48 persons with underlying cirrhosis (particularly Child’s class C48 [see Chapter 90]), and those with altered anatomy (e.g., Billroth II gastrojejunostomy [see Chapter 53]), in which sphincterotomy is technically difficult.48 Removal of large bile duct stones (defined arbitrarily as ≥1.5 cm in diameter) may require additional techniques than those described earlier to be removed successfully. One such technique is lithotripsy. One form of lithotripsy is mechanical lithotripsy, in which the stone is captured in a specialized large basket and crushed51 (Fig. 70-6). The fragments are removed using standard extraction techniques. Another form of lithotripsy is intraductal lithotripsy, which is performed by passing laser or electrohydraulic catheters into the bile duct. The stones are fragmented under direct endoscopic visualization using a transpapillary choledochoscope.52 Direct visualization is necessary to ensure that the lithotripsy device is directed at the stone and not the bile duct wall. More recently, the combination of biliary sphincterotomy and large-diameter (>12 mm) balloon dilation has been used to remove large stones and decrease the need for mechanical lithotripsy.53 This large-diameter dilation method appears to be safe and not associated with an increased risk of post-ERCP pancreatitis.54 If large stones cannot be removed, a biliary stent is placed to relieve the obstruction.55 Additional procedures can then be undertaken electively to remove residual stones.

BILE LEAKS

As discussed previously, bile leaks arise as a result of postsurgical complications and trauma. Most commonly, post-

1191

1192

Section VIII  Biliary Tract

A

B

Figure 70-5.  Endoscopic images during removal of a bile duct stone. A, Bulging papilla consistent with an impacted stone is seen. B, After endoscopic sphincterotomy, the stone is extracted.

A

B

C

Figure 70-6.  Mechanical lithotripsy of a bile duct stone. A, Endoscopic retrograde cholangiogram showing a large stone (arrow). B, Lithotripsy basket crushing the stone. C, Follow-up cholangiogram shows clearance of the stone from the bile duct.

cholecystectomy leaks arise from either the cystic duct or duct of Luschka (see Chapter 62). These smaller leaks can usually be managed with biliary sphincterotomy alone or placement of a small-caliber (7-French) plastic biliary stent (or both).56 This approach diverts bile away from the leak into the duodenum and negates the effect of the otherwise high-pressure biliary sphincter. More complex leaks usually require placement of one or more large-caliber plastic

biliary stents in combination with biliary sphincterotomy (Fig. 70-7).57 The use of removable, covered SEMS for treatment of refractory leaks has also been described.58

PRIMARY SCLEROSING CHOLANGITIS

Patients with PSC may benefit from endoscopic intervention to treat a dominant stricture or biliary lithiasis.59 Patients with a dominant stricture usually present with progressive

Chapter 70  Endoscopic and Radiologic Treatment of Biliary Disease large diameter (10 mm) of the SEMS results in dilation of the stricture over time. The stent is removed after an interval of three to six months. Results using this approach have been encouraging, although these devices are not yet approved for use in benign diseases.

INDETERMINATE BILIARY STRICTURES

A

Some biliary strictures cannot be readily classified as benign or malignant on the basis of imaging studies and tissue sampling. Tissue sampling techniques at ERCP consist of wire-guided biliary brush cytology and intraductal forceps biopsy.67 Additional techniques that can be used to assess indeterminate strictures include intraductal ultrasonography68 and direct choledochoscopy with or without directed biopsy.69 In a small percentage of patients, the diagnosis still remains unclear. In some patients, the final diagnosis can only be established during long-term follow-up or at surgical exploration and resection.

MALIGNANT BILIARY STRICTURES

Endoscopic relief of malignant biliary obstruction is achieved by placement of large-bore plastic stents or SEMS across the malignant stricture. The approach to the patient depends on whether the stricture is distal to the bifurcation of the common hepatic duct or involves the bifurcation (hilar obstruction).

Distal Bile Duct Strictures

B Figure 70-7.  Intrahepatic bile leak treated endoscopically after a hepatec­ tomy. A, Contrast extravasation (long arrow) is seen near the percutaneous drain; an internal biliary stent (short arrows) is placed. B, Follow-up chol­ angiogram shows resolution of the leak.

biliary obstruction. Cholangiocarcinoma must be considered in these patients. Routine brush cytology has a low sensitivity in these patients, but fluorescence in situ hybri­ dization (FISH) has been shown to have a high sensitivity for the detection of cholangiocarcinoma (see Chapter 69).60 Choledochoscopy also may improve detection of malignancy in these patients.61 Endoscopic treatment of a dominant stricture involves balloon dilation, often in combination with short-term (50%, the diagnosis is either the uncommon Dubin-Johnson syndrome or the even rarer Rotor’s syndrome (see Fig. 73-1, Table 20-2, and Table 64-4). The defect in Dubin-Johnson syndrome is in the MRP2 gene. The defect in Rotor’s syndrome has yet to be defined, but in both syndromes excretion of conjugated bilirubin across the bile canalicular membrane is reduced, resulting in an increased conjugated serum bilirubin level. Neither syndrome is associated with adverse clinical outcomes. Additional genetic disorders of bile acid transport that may be associated with hyperbilirubinemia are discussed in Chapters 64 and 76.

AMINOTRANSFERASES The serum aminotransferases (also called transaminases), the most sensitive markers of acute hepatocellular injury, have been used to identify liver disease since the 1950s.9 ALT (formerly serum glutamic pyruvic transaminase, or SGPT) and AST (formerly serum glutamic oxaloacetic transaminase, or SGOT) catalyze the transfer of the α-amino groups of alanine and l-aspartic acid, respectively, to the α-keto group of ketoglutaric acid. AST, found in cytosol and mitochondria, is widely distributed throughout the body; it is found, in order of decreasing concentration, in liver, cardiac muscle, skeletal muscle, kidney, brain, pancreas, lung, leukocytes, and erythrocytes. ALT, a cytosolic enzyme also found in many organs, is present in greatest concentration by far in the liver and is, therefore, a

Indirect Hyperbilirubinemia Hemolytic Disorders   Inherited    Red cell enzyme defects (e.g., glucose-6-phosphate dehydrogenase deficiency)    Sickle cell disease    Spherocytosis and elliptocytosis   Acquired    Drugs and toxins    Hypersplenism    Immune mediated    Paroxysmal nocturnal hemoglobinuria    Traumatic: macro- or microvascular injury Ineffective Erythropoiesis   Cobalamin deficiency   Folate deficiency   Profound iron deficiency   Thalassemia Drugs: Rifampin, Probenecid Inherited Conditions   Crigler-Najjar syndrome types I and II   Gilbert’s syndrome Other   Hematoma Direct Hyperbilirubinemia Inherited Conditions   Dubin-Johnson syndrome   Rotor’s syndrome

MECHANISM Overproduction of bilirubin

Overproduction of bilirubin

Impaired hepatocellular uptake Impaired conjugation of bilirubin

Overproduction of bilirubin Impaired excretion of conjugated bilirubin

more specific indicator of liver injury. Increases in serum values of the aminotransferases reflect either damage to tissues rich in these enzymes or changes in cell membrane permeability that allow ALT and AST to leak into serum; hepatocyte necrosis is not required for the release of aminotransferases, and the degree of elevation of the aminotransferases does not correlate with the extent of liver injury.10 Aminotransferases have no function in serum and act like other serum proteins. They are distributed in plasma and interstitial fluid and have half-lives measured in days. The activity of ALT and AST at any moment reflects the relative rate at which they enter and leave the circulation. They are probably cleared by cells in the reticuloendothelial system, with AST cleared more rapidly than ALT. Normal values for aminotransferases in serum vary widely among laboratories, but values gaining general acceptance are AST ( AST (>1000 U/L or >20-25 × normal) Hepatic Causes Acute bile duct obstruction Acute Budd-Chiari syndrome Acute viral hepatitis Autoimmune hepatitis Hepatic artery ligation Ischemic hepatitis Medications/toxins Wilson disease Severe, Acute Elevations, AST > ALT (>1000 U/L or >20-25 × normal) Hepatic Cause Medications or toxins in a patient with underlying alcoholic liver injury Nonhepatic Cause Acute rhabdomyolysis Chronic, Mild Elevations, AST > ALT ( C in younger age group in Taiwan but B < C in older age group in Japan HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus.

is especially prevalent in the Mediterranean area, Middle East, and south Asia. Genotype E is indigenous to western sub-Saharan areas, and genotype F prevails in Central America. Cases of genotype G have been reported in the United States and France. Genotype H has been described in Mexico. Clinical associations appear to exist with the various genotypes (see Table 78-1). Currently, the strongest clinical associations appear to be that (1) HBeAg seroconversion occurs earlier in patients with genotype B than in those with genotype C and (2) response to therapy with interferon varies with genotype (see later).25 The viral genotype also has implications for the frequency of precore and core mutations (see later) and may have an effect on the frequency of HCC.

Chapter 78  Hepatitis B and D Mutations of the Hepatitis B Virus Genome

The vast majority of mutations in the HBV genome identified by comparing nucleotide sequences with those of wildtype HBV are silent or do not alter the amino acid sequence in a particular ORF. Some mutations have potentially important disease associations, however, and are described next. Hepatitis B Surface Antigen Mutants Mutations in the surface gene can result in changes in the antibody-binding domain. Accordingly, both virus neutralization by polyclonal antibody to HBsAg and testing for HBsAg by methods that depend on antibody binding can be affected. Large-scale vaccination programs in regions endemic for HBV have revealed a 2% to 3% frequency of vaccine escape mutants that result from alterations in the “a” determinant of the HBsAg protein, which is the major neutralizable epitope. The typical mutation results in the substitution of glycine for arginine at amino acid position 145; this substitution prevents binding of neutralizing antibodies (i.e., antibody to HBsAg [anti-HBs]). The clinical significance of these mutants for neonatal vaccination programs is highly controversial because the frequency of these variants among HBV-infected mothers whose infants respond to vaccination has been found to be similar to that of mothers whose infants do not respond. The “a” determinant mutants also are proposed to have clinical relevance after liver transplantation for hepatitis B. As many as 50% of patients in whom recurrent HBV infection develops despite the use of HBIG have been shown to have these escape mutants, and the rate at which the mutations are detected appears to correlate with the length of time over which HBIG is repeatedly administered.26 Mutations in the Precore, Basal Core Promoter, and Core Genes Mutations in the precore and basal core promoter regions of the HBV genome can influence the production of HBeAg. A precore mutation results in a stop codon at nucleotide 1896 that abolishes the synthesis of HBeAg,27 whereas mutations in the basal core promoter at nucleotides 1762 and 1764 decrease HBeAg synthesis by approximately 70% while maintaining pregenomic RNA levels.28 Both types of mutations have been observed in cases of severe or fulminant hepatitis, which has been attributed to the loss of the immune-tolerizing effects of HBeAg antigen (see later). The presence of core promoter mutations has been linked to a significantly increased risk of HCC.29 Precore and basal core promoter mutants have been described in the same patients and are particularly common in Asian and European patients with chronic hepatitis B.30 A large serosurvey of HBV carriers residing in the United States has found that precore and core promoter mutations are common (fre­ quencies of 27% and 44%, respectively), depending on the ethnicity and places of birth of the patients. Both mutant forms of HBV were observed to occur far more commonly in HBeAg-negative patients (precore mutation in 38% of HBeAg-negative versus 9% of HBeAg-positive patients; core promoter mutation in 51% versus 36%).31 In addition to these mutations, upstream mutations in the core gene can influence immunologic responses to HBV. Core gene mutations have been shown to block recognition of HBV by cytotoxic T lymphocytes (CTLs), a key mode of viral clearance. Therefore, the mutations contribute to HBV immune escape and possibly influence the response to interferon.32 Core gene mutations within the immunodominant epitopes of the HBV nucleocapsid also can affect CD4+ T-cell reactivity.33

In patients with perinatally acquired chronic hepatitis B, a prolonged immune tolerant phase with minimal to absent hepatic necroinflammatory activity is typically seen for the first 20 to 30 years of HBV infection. Sequencing studies have shown stable core gene sequences during this phase. Precore mutations are also uncommon during this phase. Core gene mutations become more common as patients pass from the immune tolerant phase, at which time a growing number of mutations are observed in the region of the core gene that includes many B- and T-cell epitopes. Both precore stop codon mutants and core gene mutants have been associated with a poor response to interferon therapy. Hepatitis B Virus DNA Polymerase Mutants The polymerase gene encodes a DNA polymerase enzyme needed for encapsidation of viral RNA into core particles, conversion of the pregenomic viral RNA into a negative strand of viral DNA (reverse transcription), and conversion of this first HBV DNA strand into a second DNA strand of positive polarity. In general, the HBV reverse transcriptase function of the polymerase gene is highly conserved because major mutations that impair the efficiency of viral replication lead to selection pressure against such variant forms. As indicated earlier, HBV has a high rate of replication (1011 virions per day) and low replication fidelity, meaning that it has a propensity to mispair nucleotide bases when it reverse transcribes viral RNA to DNA. HBV DNA polym­ erase also lacks any proofreading activity, so it cannot repair its mistakes. Therefore, when a nucleotide base is misplaced, it remains in the growing viral DNA strand as a base mutation, and the new HBV DNA molecule has a different sequence from the original (wild-type) genome. The overall error rate of HBV DNA polymerase is estimated to be 1 per 10,000 nucleotides copied, which translates to the potential for 10 million base-pair errors per day in an infected person. All possible single-base mutations can be produced in a 24-hour period, although many such mutations will yield nonviable viruses.34 Mutations in the sequence of HBV polymerase can lead to drug resistance to nucleoside analogs used to treat HBV infection because some HBV polymerase mutants have decreased susceptibility to these drugs and are selected during treatment. The mutations in the sequence of HBV DNA polymerase that confer drug resistance result in amino acid substitutions in the reverse transcriptase domain of the enzyme. The changes in the structure of the enzyme, in turn, are thought to sterically inhibit binding of the drugs to their active sites. The amino acids in the reverse transcriptase are numbered from 1 to 344, and an amino acid identity is given by the single letter amino acid code. By convention, substitutions in reverse transcriptase are designated by the wildtype amino acid, followed by the number of the amino acid, followed by the substituted amino acid. For example, for the nucleoside analog lamivudine (see later), two types of mutations occur at nucleotide position 204 of domain C (the catalytic site of the polymerase) that result in substitution of the amino acid methionine (M) for either isoleucine (I) or valine (V). These mutations are designated M204I and M204V, respectively, and are referred to collectively as YMDD mutants; the letters stand for the amino acids (Y = tyrosine, M = methionine, D = aspartate) in the C domain. The M204V mutation tends to occur in conjunction with a mutation in domain B that results in substitution of leucine (L) with methionine (L180M). The M204I mutation or the combined M204V-L180M mutations result in marked resistance to the effect of lamivudine (>10,000-fold reduction in susceptibility). After prolonged exposure to adefovir, drug resistant mutations in domains B and D are selected

1291

1292

Section IX  Liver L180M A181V/T T184G/S S202I M204V/I N236T M250V LAM LdT ADV

Table 78-2  Hepatitis Flares in Patients with Chronic Hepatitis B CAUSE OF FLARES

COMMENT

Spontaneous

Factors that precipitate antecedent viral replication are unclear Flares are often observed during withdrawal; requires preemptive antiviral therapy

TDF

Immunosuppressive therapy

ETV

Antiviral therapy for HBV   Interferon

Figure 78-3.  Common amino acid changes (top row) that result from nucleotide substitutions in the hepatitis B virus polymerase gene and confer drug resistance to oral antiviral agents (first column) used to treat hepatitis B (see text for description of nomenclature). Note that cross resistance exists among lamivudine, telbivudine, and entecavir, but entecavir requires the presence of one or more secondary substitutions at codons 184, 202, or 250 for clinically apparent drug resistance to appear. Note also that the potential for cross-resistance exists between adefovir and tenofovir, but the greater antiviral potency of tenofovir reduces the chances of resistance greatly. A, alanine; ADV, adefovir; ETV, entecavir; G, glycine; I, isoleucine; L, leucine; LAM, lamivudine; LdT, telbivudine; M, methionine; TDF, tenofovir; N, asparagine; S, serine; T, threonine; V, valine.

(A181V/T and N236T). Other nucleotide substitutions have been described that are instrumental for telbivudine and entecavir resistance (Fig. 78-3). The inherent mutability of HBV indicates that single and even double polymerase mutants preexist as minor “quasispecies” even before treatment of HBV infection is begun. Because of the limitations in the sensitivities of current genotype assays, these mutants would not be detectable until they are selected and expanded under the pressure of drug treatment. Resistance to lamivudine is found in approximately 20% of patients after one year of treatment but in nearly 70% after five years (see Fig. 78-3).35 Rates of resistance to adefovir, a nucleotide analog, are 0% at one year and 29% after five years.36 The efficacy of both of these drugs against HBV is impaired by a single nucleotide substitution. The more mutations necessary for drug resistance (indicating a higher genetic barrier to resistance), the slower the emergence of and lower overall rate of resistance. For example, resistance to entecavir, another nucleoside analog, occurs in less than 1% of patients at five years because the preexistence of lamivudine-resistant mutations and one or more additional mutations in the viral polymerase gene are required for resistance.37 Persistent infection with drugresistant HBV ultimately is associated with progression of disease and blunting of hepatic histologic improvement with antiviral therapy.38 Severe flares of hepatitis have also been reported after the emergence of drug-resistant mutants,39 and acquisition of these mutants may lead to rapidly progressive liver disease after liver transplantation (Table 78-2).40 Horizontal transmission of these mutants, which can complicate drug therapy in secondarily infected persons, is also possible.

PATHOGENESIS

HBV is generally not a cytopathic virus, and the severity of HBV-associated liver disease is considered to be related to the intensity of the host immunologic response to the virus. Whereas both humoral and cellular immune responses are needed for effective clearance of the virus, the cellular immune response appears to be the arm principally involved in the pathogenesis of disease. The immunologic response to HBV encompasses both an innate, or nonspecific, response

Lamivudine   During treatment   YMDD mutant   On withdrawal*

HIV treatment

Genotypic variation   Precore and core promoter mutants Superinfection with other hepatitis viruses

Flares are often observed during the second to third month; may herald virologic response Flares are no more common than with placebo Can have severe consequences in patients with advanced liver disease Flares are caused by rapid re-emergence of wild-type HBV; can have severe consequences in patients with advanced liver disease Flares can occur with HAART or with immune reconstitution; in addition, HBV increases the risk of antiretroviral drug hepatotoxicity Fluctuations in serum ALT levels are common with precore mutants May be associated with suppression of HBV replication

*Has also been reported with adefovir and entecavir. ALT, alanine aminotransferase; HIV, human immunodeficiency virus; HAART, highly active antiretroviral therapy; HBV, hepatitis B virus; YMDD, tyrosine-methionine-aspartate-aspartate.

(for example, natural killer cells and interferons) and an adaptive immune response, including antibodies to viral antigens, human leukocyte antigen (HLA) class II–restricted CD4+ T cells, and HLA class I–restricted CD8+ CTLs.41 Induction of the antigen-specific T-cell response is thought to occur in lymphoid organs, where the host T cells encounter viral peptide antigens (or epitopes) that are presented by antigen-presenting cells such as dendritic cells, B cells, and macrophages. This process results in the maturation and expansion of T cells that are specific for these viral epitopes and is followed by their migration to the liver, where they perform their effector function. During acute HBV infection, most HBV DNA molecules are cleared rapidly from the liver via noncytopathic mechanisms mediated by cytokines that are released initially by cells of the innate immune system42 and later by liverinfiltrating HBV-specific CD8+ cells. Cell-mediated immune responses are efficient in self-limited infection because the responses are vigorous, multispecific, and oriented toward type 1 helper T (Th1) cells. Persons with chronic HBV infection, by contrast, exhibit infrequent, narrowly focused, and weak HBV-specific T-cell responses.43 In chronic hepatitis B, the majority of mononuclear cells in liver infiltrates of patients with chronic hepatitis B at any given time are non–antigen-specific.44 CD8+ CTLs are thought to contribute to the disease process in the liver and result in apoptosis of infected hepatocytes. To be recognized by the CD8+ CTLs, targeted hepatocytes must present viral epitopes as short peptides that have been endogenously processed and fit within the peptide-binding groove of the class I major histocompatibility complex (MHC) molecules.45 The binding of the CTL

Chapter 78  Hepatitis B and D HBeAg

Anti-HBe

Normal range

HBV DNA

ALT

Phase

Immune tolerant

Immune clearance

Inactive carrier state

Optimal treatment time Liver

Minimal, if any, inflammation

Chronic inflammation/ fibrosis

Reactivation Optimal treatment time

Mild inflammation and minimal fibrosis

Active inflammation

Figure 78-4.  Natural evolution and phases of chronic hepatitis B in a person with perinatal or early-life acquisition of the infection. The four phases are designated the immune tolerant phase, immune clearance phase, inactive carrier state, and reactivation phase, as determined by biochemical, virologic, and histologic activity of the disease. In some patients, hepatitis B e antigen (HBeAg) seroconversion to antibody to HBeAg (anti-HBe) is followed by the selection of precore and/or core promoter mutant forms of HBV and continuing hepatitis. This evolution generally occurs after several decades of infection and accounts for the frequent occurrence of active HBeAg-negative hepatitis in middle age. At this point, some patients may have high-normal or slightly elevated serum ALT levels despite having significant underlying fibrosis and inflammation. These patients are not in the immune tolerant phase but instead should be considered to be in the immune clearance phase with mild disease. Ultimately, prolonged histologic activity (necroinflammation) leads to cirrhosis in at least 20% of patients; cirrhosis, in turn, as well as ongoing necroinflammatory activity in the liver, can result in hepatocellular carcinoma. The optimal times for antiviral therapy are during the immune clearance and reactivation phases.

T-cell receptor (TCR) to the peptide-MHC complex on the hepatocyte surface can then result in the direct killing of the infected cell and release of potent antiviral cytokines by the activated CTL. Recognition by MHC class II– restricted CD4+ helper T cells requires the appropriate presentation of viral peptides in the context of class II MHC molecules. The CD4+ cells produce antiviral cytokines and provide help in neutralizing antibody production. Antibody neutralization limits intrahepatic spread of virus during primary infection and serves an important role in preventing reinfection.

NATURAL HISTORY

Four phases of HBV infection have been described: immune tolerance, immune clearance, the inactive carrier state, and reactivation (Fig. 78-4). Patients who acquire the infection in the perinatal period often have high serum levels of HBV DNA without biochemical evidence of active hepatitis and are considered to be immune tolerant to HBV. When followed longitudinally, many of these patients ultimately exhibit elevated serum aminotransferase levels in association with histologic evidence of chronic hepatitis. The trigger mechanisms for this apparent change in tolerance are poorly understood but likely reflect changes in the immune reactivity of the host. Experiments in transgenic mice suggest that HBeAg induces a state of immunologic tolerance to HBV in neonates.46 Perinatal transmission of HBeAg has been considered to be a potential mechanism for the immune-tolerant state. As persons enter the immune clearance phase, HBV DNA concentrations diminish, serum alanine aminotransferase (ALT) levels rise, and hepatic histologic activity, reflecting

immune-mediated lysis of infected hepatocytes, increases. The duration of this second phase varies, often lasting many years. The third phase (inactive HBV carrier state) occurs after seroconversion from HBeAg to anti-HBe and is usually preceded by a marked reduction in serum HBV DNA to levels that are detectable only by PCR methodology, followed by normalization of serum ALT levels and resolution of liver necroinflammation. This phase may last a lifetime, but a proportion of patients ultimately undergo spontaneous or immunosuppression-mediated reactivation of HBV replication with reappearance of high levels of HBV DNA in serum, with or without HBeAg seroreversion and a rise in serum ALT levels. For unclear reasons, precore or core promoter mutants that prevent or down-regulate HBeAg production may be selected during or after HBeAg seroconversion (see earlier).47 A key event in the natural history of HBeAg-positive chronic hepatitis is seroconversion of HBeAg to anti-HBe, which is associated with marked reduction in HBV replication and biochemical and histologic remission in the majority of patients. Regression of liver fibrosis occurs gradually months to years after HBeAg seroconversion.48 Most studies have found that the mean annual rate of spontaneous HBeAg seroconversion ranges from 8% to 15% in HBV-infected children or adults with serum ALT elevations. Longitudinal studies of untreated patients with predominantly HBeAg-positive chronic hepatitis B have shown that the frequency of development of cirrhosis ranges from 2 to 5 per 100 person-years and the five-year cumulative frequency of progression to cirrhosis from 8% to 20%.49 The rate of cirrhosis has been suggested to be higher in HBeAg-

1293

1294

Section IX  Liver negative patients than in HBeAg-positive patients. Risk factors for the development of cirrhosis have been identified; of these, older age, the stage of fibrosis at presentation, and ongoing HBV replication with persistent or intermittent detection of HBV DNA by a non–PCR-based assay are perhaps the most important clinically. Combined infection with hepatitis D virus (HDV [see later]), HCV, or HIV and concomitant alcohol abuse have also been linked to a higher rate of development of cirrhosis. When cirrhosis develops, two major complications may occur: hepatic decompensation and HCC. In a large European cohort with HBV-related compensated cirrhosis, the five-year cumulative frequency of hepatic decompensation was 16%, and the incidence per 100 person-years was 3.3.50 Similar rates have been reported in Asians. The cumulative five-year frequency of HCC can be as high as 14%.50 Factors associated with an increased risk of HCC include male gender, age more than 45 years, having a first-degree relative with HCC, the presence of cirrhosis, HBeAg positivity, and reversion from anti-HBe to HBeAg positivity.51 HCC can still develop in HBsAg-positive persons with none of the identified risk factors, but less frequently. In addition, HCC has been described in persons who lose HBsAg. Recommendations about ultrasonography and alpha fetoprotein screening for HCC are controversial, but in general, screening is recommended in all patients with cirrhosis and in male HBV carriers older than age 40 years in whom the likely route of transmission has been perinatal or early childhood exposure; some authorities recommend screening after age 30 years in highly viremic patients when perinatal acquisition is suspected (see Chapter 94). ALT as a Surrogate Marker for Disease Activity The serum ALT level has been used conventionally as a measure of disease activity in patients with chronic hepatitis B. Use of the standard reference range (0 to 40  U/L), however, can be misleading for evaluating HBV-related disease activity. A serum ALT level within the normal range is an imperfect surrogate marker for the absence of disease activity because determination of standard reference ranges has not traditionally taken into account increased body mass index (BMI), diabetes mellitus, and other features, such as alcohol intake, that tend to inflate values in a “normal” reference population (see Chapter 73). An insurance record-based study in Korea that included more than 140,000 persons who were followed for eight years demonstrated that all-cause liver-related mortality is increased when the serum ALT level exceeds 20  U/L in women and 30  U/L in men.52 This finding is particularly relevant to the management of Asians with hepatitis B and viremia. Because these patients tend to have a small body mass, a normal serum ALT value according to the standard laboratory reference range may be misleading. Such patients often have been excluded from treatment and assumed to be immune tolerant (that is, without liver disease). Studies in Asia and the United States have shown that as many as 20% to 30% of HBV carriers with persistently normal serum ALT levels and serum HBV DNA levels >104 copies/ mL have stage 2 or greater (of 4) inflammation and stage 2 or greater (of 4) fibrosis on a liver biopsy specimen.53 Moreover, Asian HBV carriers with high-normal serum ALT levels (>0.5 × upper limit of normal [ULN]) have been shown to have more fibrosis on liver biopsy specimens than do those with low-normal serum ALT levels (80%) if not treated by liver transplantation. Patients older than age 40 years appear to be more susceptible than younger persons to “late-onset liver failure,” in which encephalopathy, renal dysfunction, and other extrahepatic complications of severe liver insufficiency become manifest over the course of several months. The pathogenic mechanisms of fulminant hepatitis are poorly understood but are presumed to involve massive immune-mediated lysis of infected hepatocytes. This proposed mechanism may explain why many patients with fulminant hepatitis B have no evidence of HBV replication in serum at presentation.

Chronic Hepatitis B

A history of acute or symptomatic hepatitis is often lacking in patients with chronic HBV infection. When symptoms are present, fatigue tends to predominate over other constitutional symptoms, such as poor appetite and malaise. Right upper quadrant pain also may occur but is generally low grade. Patients may remain asymptomatic even during

periods of reactivated hepatitis. In other instances, particularly when superimposed on cirrhosis, reactivation of HBV infection may be associated with frank jaundice and signs of liver failure (see later discussion of acute flares in chronic hepatitis B). Physical findings may be normal, or hepatosplenomegaly may be found. In decompensated cirrhosis, spider angiomata, jaundice, ascites, and peripheral edema are common. Liver biochemical test results are usually completely normal during the inactive HBV carrier state. In contrast with patients in the immune-tolerant phase of HBV infection, most patients in the immune-clearance phase of chronic HBV infection have mild to moderate elevations in serum AST and ALT levels. During exacerbations of disease, serum ALT levels may be as high as 1000 U/L or more, and the clinical and laboratory picture is indistinguishable from that of acute hepatitis B, including the presence in serum of IgM anti-HBc. Progression to cirrhosis should be suspected whenever hypersplenism, hypoalbuminemia (in the absence of nephropathy), or prolongation of the prothrombin time is found. The serum AST level is typically higher than the serum ALT level in patients with advanced cirrhosis (see Chapter 73). Extrahepatic Manifestations Extrahepatic syndromes seen in association with acute or chronic hepatitis B are important to recognize because they may occur without clinically apparent liver disease and can be mistaken for independent disease processes in other organ systems. The pathogenesis of these extra­ hepatic disorders has not been fully elucidated but likely involves an aberrant immunologic response to extrahepatic viral proteins.67 Many of the extrahepatic manifestations (e.g., arthritis, dermatitis, glomerulonephritis, polyarteritis nodosa, cryoglobulinemia, papular acrodermatitis, and polymyalgia rheumatica) are observed in association with circulating immune complexes that activate serum com­ plement. Antiviral therapy may be indicated for persistent symptoms. Arthritis-Dermatitis.  The constellation of fever, arthralgias, rash, angioneurotic edema, and, less commonly, hematuria and proteinuria is seen as a prodromal manifestation of acute hepatitis B and rarely in patients with chronic hepatitis B. The pro­ximal interphalangeal joints, knees, ankles, shoulders, and wrists are the joints most commonly affected. During the period of acute joint symptoms, HBsAg titers in the blood are high and complement levels are low. Correspondingly, HBsAg has been detected in synovial membranes, and complement levels in synovial fluid are low. Evidence of activation of the complement system by HBsAg–anti-HBs complexes exists. After the joint symptoms subside, complement levels return to normal, and HBsAg titers in serum begin to decline. This syndrome must be distinguished from inflammatory forms of arthritis because glucocorticoid therapy mistakenly given to patients with such HBV manifestations enhances HBV replication, and abrupt withdrawal of these agents may be associated with a flare in disease activity. Polyarteritis Nodosa.  One of the most serious extrahepatic syndromes associated with chronic HBV infection is polyarteritis nodosa. As many as 30% of patients with polyarteritis nodosa are infected with HBV, but the disorder develops in less than 1% of patients with HBV infection, either after acute or recent hepatitis B or, more commonly, in association with chronic HBV infection. Typical features include arthralgias, mononeuritis, fever, abdominal pain, renal disease, hypertension, central nervous system abnormalities, and rash. Medium to small arteries and arterioles are involved by fibrinoid necrosis and perivascular infiltra-

1295

1296

Section IX  Liver tion. The disease is thought to result from deposition of circulating immune complexes that contain HBsAg; for this reason, therapy with plasmapheresis may be indicated. Good therapeutic responses also have been observed with antiviral agents, given alone or in combination with plasmapheresis. No apparent relationship exists between the severity of the vasculitis and the severity of the hepatic disease, and the hepatic disease often is relatively mild despite high levels of viral replication. The course of polyarteritis nodosa is variable, but the prognosis is gravest for patients with substantial proteinuria (>1 g/day), renal insufficiency (serum creatinine > 1.6 mg/dL), gastrointe­ stinal involvement, cardiomyopathy, and involvement of the central nervous system. Glomerulonephritis.  Several types of glomerular lesions have been described in patients with chronic HBV infection; membranous glomerulonephritis and membranoproliferative glomerulonephritis are the most common.68 Renal biopsy specimens have demonstrated immune complex deposition and cytoplasmic inclusions in the glomerular basement membrane. The immune complexes activate complement and production of cytokines with a subsequent inflammatory response. Nephrotic syndrome is the most common presentation of HBV-associated glomerulonephritis. In affected children, renal failure at presentation is almost always mild, and a history of clinical liver disease is uncommon. Nevertheless, liver biopsy specimens almost always demonstrate varying degrees of chronic hepatitis. The diagnosis of HBV-associated glomerulonephropathy is usually established by serologic evidence of HBV antigens or antibodies, the presence of immune-complex glomerulonephritis in a renal biopsy specimen, and the demonstration of glomerular deposits of one or more HBV antigens, such as HBsAg, HBcAg, or HBeAg, by immunohistochemistry. Most patients have detectable HBeAg in serum and, in addition, demonstrate low serum C3 and occasionally low C4 levels. The renal disease typically resolves in months to several years, especially in children. Often, resolution occurs in conjunction with HBeAg seroconversion. Rarely, however, renal failure may ensue. The natural history of HBV-related glomerulonephritis in adults has not been well defined, but several reports suggest that glomerular disease is often slowly and relentlessly progressive.69 Successful treatment has been accomplished with interferon alpha and

has been linked to long-term control of HBV replication.70 Therapy with nucleoside analogs has resulted in improved renal function and diminished proteinuria. Cryoglobulinemia.  Type II cryoglobulins consist of a polyclonal IgG and monoclonal IgM, whereas type III cryoglobulins contain polyclonal IgG and rheumatoid factor. Type II and type III cryoglobulinemia have been associated with hepatitis B, but the association is uncommon. In a large patient cohort, the frequency of cryoglobulinemia was significantly higher in patients with chronic HCV infection (54%) than in patients with chronic HBV infection (15%) (see Chapter 79). Cryoglobulinemia may be associated with systemic vasculitis (purpura, arthralgias, peripheral neuropathy, and glomerulonephritis) but is often pauci­ symptomatic or asymptomatic. Interferon has been used successfully to treat symptomatic cryoglobulinemia in association with chronic hepatitis B. Experience with nucleoside analog therapy has not been reported. Histopathologic Features Chronic HBV infection is characterized by mononuclear cell infiltration in the portal tracts. Periportal inflammation often leads to the disruption of the limiting plate of hepatocytes (interface hepatitis), and inflammatory cells often can be seen at the interface between collagenous extensions from the portal tracts and liver parenchyma (referred to as active septa). During reactivated hepatitis B, lobular inflammation is more intense and reminiscent of that seen in acute viral hepatitis. Steatosis is not a feature of chronic hepatitis B, as it is in chronic hepatitis C. The only histologic feature noted on routine light microscopy that is specific for chronic hepatitis B is the presence of ground-glass hepatocytes (Fig. 78-5). This morphologic finding results from accumulation of HBsAg particles (20 to 30 nm in diameter) in the dilated endoplasmic reticulum. Because of high levels of cysteine in HBsAg, ground-glass cells have a high affinity for certain dyes, such as orcein, Victoria blue, and aldehyde fuchsin. Ground-glass hepatocytes also may be seen in HBV carriers, in whom they may be detected in up to 5% of cells. When present in abundance, ground-glass hepatocytes often indicate active viral replication.71 Immunofluorescence and electron microscopic studies have shown HBcAg inside the hepatocyte nuclei of affected cells. During periods of intense hepatitis

Hepatitis B surface antigen

Ground-glass inclusions

A

B

Figure 78-5.  A, Liver biopsy specimen showing ground-glass inclusions in hepatocytes. These inclusions represent large amounts of hepatitis B surface antigen (HBsAg) in the endoplasmic reticulum of infected hepatocytes (Hematoxylin and eosin, ×630.) B, Immunohistochemical stain for HBsAg. Note that the brownish inclusions correspond to the ground-glass inclusions seen in A (×630). (Courtesy of Dr. Gist Farr, New Orleans, La.)

Chapter 78  Hepatitis B and D activity, cytoplasmic core antigen staining is generally observed. After successful treatment of HBV infection with a nucleoside analog, the cytoplasmic core antigen staining often disappears, but nuclear core antigen staining may remain, indicating persistence of the HBV cccDNA template.

deterioration occurs, often without the subsequent loss of HBeAg. Multiple episodes of reactivation and remission have been shown to accelerate the progression of chronic hepatitis B and are particularly likely to occur in patients infected with the precore mutant form of chronic hepatitis B (see earlier).47

Acute Flares in Chronic Hepatitis B

Immunosuppressive Therapy-Induced Flares Reactivation of HBV replication is a well-recognized complication in patients with chronic HBV infection who receive cytotoxic or immunosuppressive therapy.76 Suppression of the normal immunologic responses to HBV leads to enhanced viral replication and is thought to result in widespread infection of hepatocytes by HBV. On dis­ continuation of immunosuppressive medications, such as cancer chemotherapy, antirejection drugs, and glucocorticoid therapy, immune competence is restored and infected hepatocytes are rapidly destroyed. The more potent the immunosuppression, the higher the level of viral replication and, thus, the greater the potential for serious clinical consequences of sudden withdrawal of the therapy and restoration of immunologic competence. Postmortem studies of liver tissue from patients with severe liver injury have documented sparse staining of viral antigens, suggesting that the patients were in an active state of immune clearance.77 The vast majority of patients who experience immunosuppressive therapy–induced flares have been positive for HBsAg in serum before treatment, but some studies have described the reappearance of HBsAg in patients who were initially positive for anti-HBs, anti-HBc, or both.78 Reactivated hepatitis in patients who are negative for HBsAg and positive for either anti-HBc or anti-HBs is explainable by the possible latency of HBV in liver and mononuclear cells and the large extrahepatic reservoir of HBV. Chemotherapy given to patients with cancer who are HBV carriers is associated with an increased risk of liver-related morbidity and mortality.79 Reactivated hepatitis B also occurs in patients who are given immunosuppressive medications to prevent organ transplant rejection. The frequency of reactivated hepatitis appears to be particularly high in patients who undergo bone marrow transplantation because of extensive immunologic conditioning before transplantation and treatment of graft-versus-host disease.80 Rarely, fibrosing cholestatic hepatitis, a rapidly progressive form of liver injury associated with inordinately high levels of HBsAg and HBcAg in liver tissue, may develop in such patients.81 Acute flares of hepatitis B resulting from cancer chemotherapy and other immunosuppressive drugs are often detected after substantial increases in serum aminotransferase levels have been noted. Initiation of antiviral treatment after detection of such biochemical abnormalities has little effect on reducing liver injury because much of the immunologic response to HBV and viral elimination has already occurred. Instead, the key to management lies in anticipating the occurrence of a flare, initiating antiviral treatment preemptively (e.g., 4 to 6 weeks before the start of chemotherapy), and continuing the treatment for 6 to 12 months after completion of chemotherapy.82

Chronic hepatitis B is often punctuated by sudden flares of disease activity that are reflected by a rise in serum aminotransferase levels. Although a uniform definition is lacking, a flare has frequently been described as a rise in serum ALT levels to at least two times the baseline value. Spontaneous flares are an important part of the natural history of hepatitis B because when they occur repeatedly, they lead to histologic progression. Acute flares in chronic hepatitis B occur in association with a number of circumstances and clinical situations (see Table 78-2). Most flares result from a change in the balance between immunologic responses to HBV and the level of viral proliferation. Acute flares in chronic hepatitis B that are not explainable by infection with other hepatotropic viruses often occur as a secondary response to increased levels of replicating wild-type or mutant HBV or as a result of therapeutic intervention with immunologic modifiers such as interferon, glucocorticoids, and cancer chemotherapy. In some instances, the event that initiates an acute exacerbation of chronic hepatitis B may not be readily identifiable, and the flare is considered spontaneous. Spontaneous Flares Spontaneous exacerbations of chronic hepatitis B often result from reactivated infection, and an increase in serum HBV DNA levels often precedes an increase in serum aminotransferase levels. Histologic evidence of acute lobular hepatitis superimposed on the changes of chronic viral hepatitis is frequently observed during these flares.72 IgM anti-HBc, a marker that is often diagnostic of acute viral hepatitis, may also appear in serum at this time.73 The reasons for reactivated infection are unknown but likely relate to subtle changes in the immunologic control of viral replication. Reactivation seems to occur more commonly in persons who are infected with HIV.74 In persons who acquire HBV infection early in life, flares become more common during adulthood, presumably because of a breakdown in immune tolerance to HBV.75 Fatigue may be reported during flares of chronic hepatitis B, but in many instances, patients remain asymptomatic. Occasionally symptoms and signs of frank liver failure become apparent, particularly when the flare is superimposed on advanced chronic hepatitis B. Most clinically recognizable flares occur in patients who are in the nonreplicative phase of HBV infection (i.e., initially testing positive for anti-HBe and negative for serum HBV DNA on a molecular hybridization assay). During such flares, serum HBV DNA levels increase, and HBeAg often reappears in serum (seroreversion). HBV DNA and HBeAg are often detectable in serum when the patient is first seen, but if the flare has been ongoing for several weeks or longer, the accompanying enhancement of the immune response may make it difficult to detect a rise in serum HBV DNA levels. Frequently, subsidence of these flares of hepatitis is accompanied by loss of HBV DNA and HBeAg in serum. Flares also can occur in patients who are in the replicative phase of infection (i.e., already positive for HBV DNA and HBeAg in serum). In these instances, HBV replication intensifies, serum HBV DNA levels rise, and liver biochemical

Antiviral Therapy–Induced Flares Antiviral treatment of chronic hepatitis B can be associated with flares of hepatitis in several circumstances. Flares may occur during interferon therapy, after withdrawal of nucleoside analogs or glucocorticoid therapy, and in association with lamivudine-resistant mutants (see also later discussion of treatment).

1297

1298

Section IX  Liver Interferon.  Interferon-induced flares of chronic hepatitis B occur in approximately one third of treated patients and result from the immunostimulatory properties of the drug. Flares generally occur during the second or third month of treatment with conventional preparations of interferon. Flares also occur in patients treated with pegylated interferon and have been reported to occur more frequently in patients infected with HBV genotype A than with other genotypes. This finding may explain the higher rate of sustained virologic response and HBsAg clearance in patients infected with HBV genotype A.83 Serum ALT flares have been shown to be a predictor of sustained virologic response, particularly in patients with high levels of viremia.84 Flares tend to be particularly common in patients who have decompensated liver disease, with rates as high as 50% reported in one series.85 Such flares are frequently associated with clinical deterioration in the patient. Nucleoside Analogs.  Serum ALT flares occur in approximately 20% to 25% of patients after withdrawal of nucleoside analogs such as lamivudine and adefovir. These flares probably are caused by rapid resurgence of wild-type HBV, and although generally well tolerated, they have been associated with serious clinical exacerbations in patients with advanced liver disease.86 Reinstitution of therapy is often associated with a decline in serum HBV DNA and aminotransferase levels. Flares have been seen to follow the emergence of YMDD mutant HBV during therapy with lamivudine (see earlier).87 Initial reports emphasized the temporal occurrence of these flares at the time of or shortly after detection of lamivudine resistance. Further follow-up of patients with lamivudine-resistant HBV mutants, however, has indicated that the frequency of moderate or severe serum ALT flares (defined as >5 and >10 × UNL, respectively) increases with time after detection of lamivudine resistance. In one long-term study, the cumulative frequencies of such ALT flares were as follows: 24% at less than one year, 29% at one to two years, 30% at two to three years, 37% at three to four years, and 61% at more than four years after detection of lamivudine resistance.88 Glucocorticoid Withdrawal.  Serum ALT levels increase, often with an inverse decline in HBsAg concentration and HBV DNA level, after withdrawal of glucocorticoids in patients with chronic HBV infection.89 In clinical trials of HBV therapy, a short course of glucocorticoid therapy given before conventional antiviral therapy was reported to enhance virologic response rates.90,91 The immune rebound following withdrawal of glucocorticoid therapy after a fourto eight-week course may result from increased activation of lymphocytes that promote Th1 cytokine responses at a time when viral antigen expression is increased. Serious hepatic decompensation has been reported in patients with advanced disease, however, and this therapeutic approach is no longer used. Antiretroviral Therapy.  Serum ALT flares occur in patients coinfected with HIV and HBV who receive highly active antiretroviral therapy (HAART).92 A number of potential causes have been identified. Lamivudine resistance and withdrawal may be associated with ALT flares. HBV infection raises the risk of toxicity from antiretroviral therapy, usually within six months after the initiation of treatment. Immune reconstitution resulting from HAART may also be associated with ALT flares. Affected patients may also be particularly susceptible to flares because of infection with other hepatitis viruses. Flares Associated with Genotypic Variation Chronic infection with precore mutant HBV (referred to as HBeAg-negative chronic hepatitis B) often is associated

with multiple flares of liver cell necrosis interspersed with periods of normal serum ALT and low serum HBV DNA levels.47 Approximately 45% of patients have episodic serum ALT flares with normal levels between episodes, and 20% have flares superimposed on persistent ALT elevations.93 These flares have been attributed to rises in the concentration of precore mutants in the liver and changes in the ratio of concentrations of precore to wild-type HBV. Mutations at the basal core promoter region of the HBV genome are associated with greater HBeAg synthesis, histologic evidence of liver inflammation, and increased viral replication.94 Multiple exacerbations of hepatitis resulting from reactivated HBV infection have been described in patients with basal core promoter mutations, either alone or in association with precore mutations. HBeAg-negative patients who have both precore and core promoter mutants may be particularly predisposed to episodes of severe reactivation after cancer chemotherapy.95 Flares Caused by Infection with Other Viruses Patients with chronic HBV infection may exhibit severe flares in serum aminotransferase levels and even frank liver failure when superinfected with other hepatotropic viruses, such as hepatitis A virus (HAV), HCV, and HDV. Increased mortality has been reported when HDV superinfection is superimposed on chronic hepatitis B, and chronic HDV infection is often associated with multiple fluctuations in serum aminotransferase levels (see later discussion of HDV).96 Acute hepatitis C superimposed on chronic hepatitis B has been reported to be as severe as HDV superinfection and has been associated with a high rate of liver failure (34%) and death (10%).97 A cumulative frequency of cirrhosis and HCC that is higher than that attributable to chronic HDV infection or chronic HBV infection alone has been demonstrated. Acute hepatitis C often leads to chronic HCV infection, and the subsequent course also may be characterized by frequent fluctuations in serum aminotransferase levels. Patients with chronic hepatitis B who become infected with other hepatotropic viruses (HAV, HCV, HDV, or both HCV and HDV) may test negative for both HBeAg and HBV DNA on non–PCR-based assays because of viral interference.

DIAGNOSIS

HBsAg appears in serum 2 to 10 weeks after exposure to HBV and before the onset of symptoms or elevation of serum aminotransferase levels. In self-limited acute hepa­ titis, HBsAg usually becomes undetectable after four to six months. Persistence of HBsAg for more than six months implies progression to chronic HBV infection. The disappearance of HBsAg is followed several weeks later by the appearance of anti-HBs. In most patients, antiHBs persists for life and provides long-term immunity. In some patients, anti-HBs may not become detectable after disappearance of HBsAg, but these patients do not appear to be susceptible to recurrent infection.98 Anti-HBs may not be detectable during a window period of several weeks to months after the disappearance of HBsAg. During this period, the diagnosis of acute HBV infection is made by the detection of IgM anti-HBc in serum.99 Coexistence of HBsAg and anti-HBs in serum has been reported in approximately 25% of HBsAg-positive persons and occurs more commonly in persons with chronic hepatitis B than in those with acute hepatitis B.100 In most instances, the anti-HBs is present in a low level, nonneutralizing, and heterotypic—that is, directed against a

Chapter 78  Hepatitis B and D subtype of HBsAg different from the subtype present in the infected patient. The mechanisms behind this finding are not clear but relate to antibody formed against minor variants of the HBsAg protein. The presence of these heterotypic antibodies is not associated with specific risk factors or changes in clinical course and may occur in patients with or without active liver disease and viral replication. Anti-HBc is detectable in acute and chronic HBV infection. During acute infection, anti-HBc is predominantly of the IgM class and is usually detectable for four to six months after an acute episode of hepatitis and rarely for up to two years. As noted earlier, IgM anti-HBc may become detectable during exacerbations of chronic hepatitis B. Anti-HBc persists in persons who recover from acute hepatitis B and also in association with HBsAg in those who progress to chronic infection. In areas where HBV is not endemic, isolated anti-HBc in serum has been detected in 1% to 4% of the general population. Isolated reactivity for anti-HBc may occur in the following situations: (1) during the window period of acute hepatitis B, when anti-HBc is predominantly of the IgM class; (2) many years after recovery from acute hepatitis B, when anti-HBs has fallen to undetectable levels; (3) as a false-positive serologic test result; (4) after many years of chronic HBV infection, when the HBsAg titer has fallen below the level of detection; (5) in HBV-infected persons who are coinfected with HCV; and (6) rarely, as a result of varying sensitivity of HBsAg assays.101 Evidence for coinfection with HCV has been demonstrated in as many as 60% of persons in whom anti-HBc is the only marker of HBV.102 Results of PCR testing of sera have shown that 0% to 30% of patients with isolated anti-HBc have HBV DNA in serum. Usually, the HBV DNA is detectable at a low level and not by standard hybridization assays, which are less sensitive than PCR assays.103 The presence of low-level viremia in these HBsAg-negative subjects has clinical implications with regard to potential infectivity. For example, in the past, anti-HBc testing of blood donors prevented some cases of post-transfusion hepatitis B.104 Also, the risk of transmission of HBV infection from a liver donor with isolated anti-HBc has been found to be as high as 50% to 70% in some series; lower rates of transmission have been observed in other forms of solid organ transplantation.105 Low-level viral replication also has implications with regard to the possibility of underlying liver disease. HBV DNA in serum and liver tissue has been confirmed by PCR methodology in some HBsAg-negative patients with cirrhosis and HCC and in some patients with fulminant non-A, non-B, non-C hepatitis as defined by conventional serologic testing.106 HBeAg is a soluble viral protein that is found in serum early during acute HBV infection. HBeAg reactivity usually disappears at or soon after the peak in serum aminotransferase levels, and persistence of HBeAg three or more months after the onset of illness indicates a high likelihood of transition to chronic HBV infection. The finding of HBeAg in the serum of an HBV carrier indicates greater infectivity, a high level of viral replication, and the need for antiviral therapy. With a commercially available PCR assay, nearly 90% of patients with HBeAg-positive chronic hepatitis B were found to have serum HBV DNA levels persistently above 105 copies/mL, with a mean value of 8.37 log10 (>108) copies/mL.107 By contrast, anti-HBe–positive patients had much lower serum HBV DNA levels; higher values were found in those with persistently or intermittently elevated serum ALT levels (mean of 5.1 log10 copies/mL) than in those with persistently normal ALT levels (3.10 log10 copies/mL).

Most HBeAg-positive patients have active liver disease; the exceptions are HBeAg-positive children and young adults with perinatally acquired HBV infection, who usually have normal serum ALT levels and minimal inflammation of the liver.14 In general, seroconversion from HBeAg to anti-HBe is associated with a reduction in serum HBV DNA levels of 3 log10 copies/mL or greater and remission of liver disease. Some patients, however, continue to have active liver disease and detectable HBV DNA in serum because of low levels of wild-type virus or the selection of precore or core promoter mutations that impair HBeAg secretion. HBV DNA can be measured in serum with qualitative or quantitative assays. The clinical utility of testing for serum HBV DNA has been hampered by the absence of a licensed test in the United States as well as an accepted international reference standard. A number of non–PCR-based assays are available with levels of sensitivity ranging from 103 to 105 genomic copies/mL of serum. Although these assays are less sensitive than those that are PCR-based, their results correlate with clinical response to antiviral therapy, and several of the currently available antiviral therapies were licensed on the basis of clinical trials in which these assays were used (see later). The use of these less sensitive non–PCRbased assays has several shortcomings, however, so most clinical laboratories use one of several commercially available PCR assays with enhanced sensitivity (102 genomic copies/mL or less). The measurement of serum HBV DNA is commonly used to evaluate a patient’s candidacy for antiviral therapy and to monitor response during treatment. Patients with a high serum HBV DNA level at baseline less commonly respond to therapy with conventional interferon than patients with a low level.108 With the use of solution hybridization testing, a baseline HBV DNA level of 200 pg/mL (roughly equivalent to 56 million copies/mL on a PCR assay) or greater has been found to be associated with a low rate of response to standard interferon. By contrast, baseline serum HBV DNA levels have not been shown to correlate with response to nucleoside analog therapy because of the more potent inhibition of viral replication by these agents. Monitoring of HBV DNA levels at key intervals during therapy allows one to predict the likelihood of HBeAg clearance. Several studies have found that the level of serum HBV DNA at 12 weeks of nucleoside analog treatment may help predict the likelihood of HBeAg seroconversion.109,110 Other studies have suggested that measuring the HBV DNA level at baseline or during treatment can be used to evaluate the likelihood of relapse after treatment is discontinued and development of resistance to lamivudine.111,112 Reappearance of HBV DNA in serum during treatment suggests that drug resistance has occurred,113 and high pretreatment levels of serum HBV DNA have been shown to correlate with a higher rate of recurrent HBV infection in liver transplant recipients who are treated with lamivudine.114 Qualitative PCR is an even more sensitive method of detecting HBV DNA than quantitative PCR. Use of qualitative PCR has altered traditional concepts about the clearance of HBV DNA from serum in acute and chronic HBV infection. Small amounts of HBV DNA can be detected in serum and peripheral mononuclear cells years after recovery from acute hepatitis B.66 Even after disappearance of HBsAg and apparent loss of HBV DNA from serum in patients with chronic hepatitis B, small amounts of HBV DNA persist in liver tissue and peripheral mononuclear cells years later.9 Detection of HBV DNA in serum by a qualitative PCR assay before liver transplantation may identify patients who are at increased risk of apparent de novo

1299

1300

Section IX  Liver hepatitis after transplantation and may pinpoint HBV as the cause of liver disease in HBsAg-negative patients.115,116 Finally, detection of minute amounts of HBV DNA may be particularly important in diagnosing patients with ful­ minant hepatitis B—who frequently have cleared HBsAg by the time they seek medical attention.117

TREATMENT Virologic Endpoints and Definitions of Response

The primary goal of treatment for chronic hepatitis B is durable suppression of serum HBV DNA to levels below those associated with liver disease. This goal can be accomplished with either interferon alpha or nucleoside analogs. The level at which serum HBV DNA is suppressed adequately is generally considered to be less than 105 copies/ mL (2000 IU/mL

1-2 × ULN

− + or − (cirrhosis)

≤2,000 IU/mL Detectable

≤ULN —

+ or − (cirrhosis)

Undetectable

—

Observe; consider treatment if serum ALT level rises Low efficacy of currently available treatments Pegylated interferon alpha, lamivudine, adefovir, entecavir, telbivudine, or tenofovir may be used as initial therapy but lamivudine and telbivudine are not preferred because of high rates of resistance Duration of therapy: Pegylated interferon alpha: 48 weeks Lamivudine: minimum of 1 year§ Adefovir: minimum of 1 year§ Entecavir: minimum of 1 year§ Telbivudine: minimum of 1 year§ Tenofovir: minimum of 1 year§ For interferon nonresponders or those with contraindications, use tenofovir or entecavir For lamivudine resistance, add adefovir or tenofovir; or switch to emtricitabine/tenofovir combination; or switch to high-dose entecavir For adefovir resistance, add lamivudine; or switch to emtricitabine/ tenofovir combination; or switch to or add entecavir For entecavir resistance, switch to or add adefovir or tenofovir For telbivudine resistance, add adefovir or tenofovir; or switch to emtricitabine/tenofovir combination; or switch to high-dose entecavir Pegylated interferon alpha, lamivudine, adefovir, entecavir, telbivudine or tenofovir may be used as initial therapy (see above) Endpoints of treatment: Sustained normalization of serum ALT level and undetectable HBV DNA on PCR assay Duration of therapy: Interferon alpha: 1 year Lamivudine: >1 year Adefovir: >1 year Entecavir: >1 year Telbivudine: >1 year Tenofovir: >1 year For pegylated interferon alpha nonresponders or those with contraindications to interferon alpha, use tenofovir or entecavir For drug resistance, considerations are same as above Consider liver biopsy and treat if moderate or severe inflammation or significant fibrosis Observe, treat if serum HBV DNA or ALT level increases Compensated cirrhosis: Treat if HBV DNA >2000 IU/mL (lamivudine and telbivudine not preferred) If HBV DNA 105 copies/mL. ‡ Some authorities treat these patients indefinitely to try to reduce the rate of future complications if the patient is older than age 40 and infection was acquired early in life. In this paradigm, patients with high-normal serum ALT levels are considered appropriate for treatment and liver biopsy is optional. Other authorities prefer to treat patients with a serum HBV DNA level >20,000 IU/mL and any degree of serum ALT elevation. Liver biopsy is preferred for patients older than 35-40 years of age (see text). § Treatment should be continued for 6-12 months after HBeAg seroconversion and may vary in duration with the severity of liver disease before treatment. ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; IU, international units; PCR, polymerase chain reaction; ULN, upper limit of normal.

1301

1302

Section IX  Liver recommendation is based on the observation that rates of sustained virologic response in patients with minimal pretreatment serum ALT elevations are low with either interferon or nucleoside analogs.108,124 All the guidelines indicate that treatment decisions ideally should be made in the context of liver histologic findings and that treatment should be directed preferentially to patients with histologically moderate to severe hepatitis, although currently treatment decisions are not based on specific grading or staging of liver histology.

Antiviral Agents

Seven antiviral agents have been licensed for the treatment of hepatitis B as of 2009. One of these drugs, standard or conventional interferon alpha-2b, largely has been replaced in clinical usage by pegylated interferon alpha-2a. The choice of drug therapy should be individualized based on patient and viral specific features whenever possible. Interferons Interferon alpha.  Interferon alpha was licensed for the treatment of chronic hepatitis B in 1992. Interferon is effective after a relatively short course of treatment (four months to one year) and, unlike the nucleoside analogs, has not been associated with drug resistance. Also, unlike nucleoside analogs, interferon has direct immunomodulatory properties. Interferon enhances HLA class I antigen expression on the surface of infected hepatocytes and augments CD8+ CTL activity. These effects could be important mechanistically in reducing the amount of the HBV cccDNA (the genomic template for viral transcription) and may thereby explain the loss of HBsAg that occurs in approximately 5% to 8% of interferon-treated patients. The major disadvantages of interferon relate to its poorer acceptance because of side effects (see Chapter 79) and lower level of HBV DNA suppression. Flares of serum ALT have been described during therapy with interferon alpha, and although these flares are potentially important in achieving a virologic response, they occur unpredictably and are inconsistently associated with antiviral efficacy. The magnitude of an ALT flare has been shown to predict the likelihood of a sustained virologic response in patients with high levels of viremia, suggesting that vigorous cell-mediated immune responses often are required to overcome high levels of viral replication.84 Pegylated Interferon Alpha.  Pegylated interferon has been found to be more potent than conventional or standard interferon and is currently licensed in more than 75 countries.125 Doses of 1.0 µg per kg of body weight of peginterferon alfa-2b and 180 µg of peginterferon alfa-2a given once weekly have been studied in clinical trials.126-128 No data are available for judging whether the increased effectiveness of pegylated interferon is primarily a function of a more pronounced effect on viral replication or of greater immunomodulatory action. Viral genotype appears to affect the response to interferon. In a report from Taiwan, patients with genotype-B, HBeAg-positive chronic hepatitis B were found to have a response to conventional interferon more frequently than patients with genotype-C chronic hepatitis B.129 A relationship between virologic response and genotype was reaffirmed in a large multicenter study of peginterferon alfa-2b. In this study, HBeAg-positive patients infected with HBV genotype A demonstrated HBeAg loss more frequently than those with genotypes B, C, and D (47% vs. 44%, 28%, and 25%, respectively).126 A subsequent follow-up study of this cohort after a mean interval of three years demonstrated that patients infected with HBV genotypes A or B had the highest

rates of durable virologic response (96% and 86%, respectively) and the highest rate of HBsAg clearance (58% and 14%, respectively).130 By contrast, rates for patients infected with HBV genotype D were 76% and 6%, respectively. These results confirm and extend those of earlier studies in HBeAg-positive patients that suggested that patients infected with HBV genotype A respond more frequently than those infected with genotype D. The relationship between sustained virologic response and HBV genotype is less clear in patients with HBeAg-negative hepatitis B, but a systematic analysis of more than 500 patients treated with either conventional or pegylated interferon alpha found that patients infected with HBV genotype C had the highest rates of sustained virologic and biochemical response and those infected with HBV genotype D had the lowest rates (50% and 21%, respectively).131 The effect that genotype exerts on the response to interferon could be particularly relevant to the treatment of North American patients with chronic HBV infection, in light of the influx of Asian HBV carriers infected with HBV genotypes B and C beginning in the last decades of the 20th century. Nucleoside and Nucleotide Analogs Nucleoside analogs have excellent oral bioavailability, a good safety record, and antiviral efficacy comparable to that observed with interferon alfa-2b. These drugs have proved to be particularly useful in the management of patients with decompensated cirrhosis, in whom even small doses of interferon can lead to worsening liver failure and severe infections. Nucleoside and nucleotide analogs replace natural nucleosides during the synthesis of the first or second strand (or both) of HBV DNA. They thus serve as competitive inhibitors of the viral reverse transcriptase and DNA polymerase. Because nucleos(t)ide analogs partially suppress viral replication, they have to be given for more than one year in most cases to achieve maximal efficacy. Unfortunately, drug resistance occurs with prolonged monotherapy. Figure 78-3 illustrates the common HBV nucleotide substitutions associated with drug resistance and the potential for cross resistance. Nucleos(t)ide analogs have several other limitations as well. With these agents, demonstrating the clearance of HBV cccDNA has been difficult, and in contrast to treatment with interferon, HBsAg clearance rarely occurs after one year of treatment with nucleoside analogs. These problems may result, in part, from the fact that these agents, unlike interferon, do not have a direct, enhancing effect on the immunologic response to HBV.78 Also, as indicated earlier, postwithdrawal serum ALT flares have been seen in approximately 25% of cases after discontinuation of nucleoside analog therapy. Lamivudine.  The approval of the nucleoside analog lamivudine in 1998 was a major breakthrough in the treatment of hepatitis B. The drug has been shown to be a relatively potent inhibitor of viral replication, convenient to administer, and free of severe adverse effects. Clinical trials demonstrated that a one-year course of lamivudine resulted in suppression of viral replication and improvement in histologic findings in the liver.132,133 In one study, HBeAg loss and HBeAg seroconversion occurred in 32% and 17% of patients, respectively.133 A two-year course of lamivudine proved to be more effective, resulting in an increase in the rate of HBeAg seroconversion from 17% at one year to 27% at two years.134 Prolongation of treatment beyond one year, however, has been associated with incremental changes in viral resistance (38% at two years), and the longer treatment is continued, the more frequently resistance is seen (65% at year five).88 Resistance is even more commonly encountered

Chapter 78  Hepatitis B and D (90% at four years) in patients coinfected with HIV because of the early use of lamivudine in HAART regimens.92 Lamivudine resistance for more than two years has been associated with a blunted histologic response, and patients in whom lamivudine resistance has developed experience more hepatitis flares.38,88 For these reasons, the drug is no longer recommended as first-line therapy. Fortunately, a number of alternative nucleos(t)ide analogs have considerably lower resistance profiles. Adefovir Dipivoxil.  Adefovir dipivoxil is the acyclic phosphonate nucleotide analog of adenosine monophosphate. The drug was approved in 2002 for the treatment of HBeAg-positive and HBeAg-negative chronic hepatitis B on the basis of the findings of randomized, controlled trials in the United States, Europe, and Asia.135,136 In these pivotal studies, treatment with adefovir for 48 weeks resulted in median serum HBV DNA reductions of 3.52 log10 and 3.91 log10 copies/mL in HBeAg-positive and HBeAg-negative patients, respectively. The rates of HBeAg loss and HBeAg seroconversion were slightly lower than those achieved with lamivudine for 52 weeks. A rise in the frequency of HBeAg seroconversion and nondetectability of HBV DNA by PCR methodology has been observed during the second year of adefovir treatment.137 The level of HBV DNA suppression has been the same irrespective of viral genotype.138 Although the extent of viral suppression is only 0.5 to 1.0 log value less with adefovir than with lamivudine, the two drugs differ greatly in their resistance profiles. Point mutations (A181V, N236T) in the B and D domains, respectively, of the HBV polymerase gene that affect HBV susceptibility to adefovir occur in only 3% of patients after two years of treatment but increase thereafter, with rates of 6% to 8% at three years, 15% to 18% at four years, and 20% to 29% at five years.139-143 HBV isolates with the N236T mutation have remained susceptible to lamivudine and appear to be sensitive to entecavir and telbivudine in vitro (see later).144 Adefovir has been shown to be clinically and virologically effective in patients with lamivudine-resistant HBV, whether they have clinically stable disease, decompensated cirrhosis, or recurrent hepatitis B after liver transplantation.145,146 Adefovir-resistant mutants, on the other hand, remain susceptible to lamivudine, and adefovir resistance has been shown to occur more frequently when lamivudine is discontinued in lamivudine-resistant patients than when lamivudine is continued.147,148 Therefore, most experts choose to continue lamivudine when starting adefovir therapy in patients who have become resistant to lamivudine. The sooner resistance to lamivudine is appreciated, the more rapid and complete the virologic response to the addition of adefovir.149 Adefovir has the disadvantage of potential nephrotoxicity, and dose reductions may be necessary in patients with or at risk of compromised renal function.146 Adefovir is still used frequently, but incomplete viral suppression occurs in 30% of patients, particularly those with high viral levels at the initiation of treatment. Therefore, adefovir is being replaced as first-line therapy with more potent drugs such as entecavir and tenofovir (see later). Entecavir.  Entecavir is a deoxyguanine nucleoside analog that inhibits HBV replication selectively. The drug blocks HBV replication by inhibiting the priming of HBV DNA polymerase and the synthesis of the first and second strands of HBV DNA. Entecavir was approved by the U.S. Food and Drug Administration (FDA) in 2005 on the basis of registration trials that demonstrated no resistance after one year of treatment and improved virologic efficacy when compared with lamivudine.150-151 Entecavir is effective against both

wild-type and lamivudine-resistant HBV. In phase 3 clinical trials, an entecavir dose of 0.5 mg was used to treat HBeAgpositive and HBeAg-negative patients who were previously untreated with a nucleoside analog, whereas 1 mg was used in patients who were resistant to lamivudine. Entecavir is more potent than either lamivudine or adefovir and results in nondetectable HBV DNA in 67% and 90% of previously untreated (“treatment-naïve”) HBeAg-positive and HBeAgnegative patients, respectively, after one year. In HBeAgpositive patients treated for an additional year, the cumulative rate of HBV DNA negativity by a PCR assay was 80% for entecavir-treated patients compared with 37% for lamivudine-treated patients.152 Entecavir resistance is rare (approximately 1% at five years) in treatment-naïve patients but is common in patients with prior lamivudine resistance. Virologic “rebound” as a result of resistance to entecavir has been demonstrated to occur in 1% of lamivudine-refractory patients after one year of treatment and in an additional 9% after two years, with further increases as treatment is continued.153 Telbivudine.  Telbivudine is an l-nucleoside analog of thymidine that has been shown to be more potent than lamivudine in randomized, controlled trials in patients with HBeAg-positive and HBeAg-negative hepatitis B.154,155 In the registration trial, 1367 patients with HBeAg-positive or HBeAg-negative hepatitis B were randomized to receive telbivudine 600 mg or lamivudine 100 mg, each once daily, for 104 weeks.155 Virologic and biochemical responses associated with telbivudine were superior to those with lamivudine. Several randomized studies also have reported rapid and marked reductions in serum HBV DNA levels in patients who had been switched from adefovir to telbivudine. Unfortunately, after one and two years of treatment, genotypic resistance was found in 5% and 25% of HBeAgpositive patients, respectively, and 2% and 11% of HBeAgnegative patients, respectively.155,156 The highest rates of virologic response and, conversely, the lowest rates of resistance, were found in patients who had less than 3 log10 copies of HBV DNA at week 24 of treatment, with the best responses found in those with a negative HBV DNA result by a PCR assay. Resistance to telbivudine is conferred by the M204I mutation either alone or in conjunction with the L180M mutation. These mutations also confer resistance to lamivudine, and for this reason switching to telbivudine is not preferred in cases of lamivudine resistance. Despite the greater antiviral potency and lower resistance profile of telbivudine compared with lamivudine, telbivudine has a higher rate of resistance than entecavir and tenofovir (see later) and has fallen out of favor as a treatment for HBV infection. In addition, a multicenter study of telbivudine in combination with pegylated interferon was stopped prematurely because of the development of myopathy and elevations in serum creatinine kinase levels in a higher percentage of patients than reported with other nucleoside analogs. Tenofovir Disoproxil Fumarate.  Tenofovir, an acyclic nucleotide inhibitor of HBV polymerase and HIV reverse transcriptase, is similar chemically to adefovir dipivoxil but is produced in a 300-mg rather than a 10-mg formulation. Tenofovir was originally licensed for the treatment of HIV infection and approved in 2008 for the treatment of HBV infection. Its antiviral activity against HBV has been reported to be significantly greater than that of the 10-mg dose of adefovir in lamivudine-resistant and treatment-naïve patients.157,158 Resistance to tenofovir has not occurred after two years of treatment in either HBeAg-positive or HBeAgnegative patients.158 Preliminary studies have suggested that adefovir-resistant HBV may be less sensitive than wild-type

1303

1304

Section IX  Liver HBV to tenofovir.159 Whether unique nucleotide substitutions in the HBV polymerase will ultimately be found that confer resistance to tenofovir in either previously untreated or adefovir-resistant cases is unknown. Prolonged tenofovir therapy has been associated with bone loss and decreased bone density in HIV-infected patients.160 As with adefovir, renal tubular damage and Fanconi syndrome have been associated with prolonged use of tenofovir in a small number of cases and may be a particular risk in the elderly or persons with existing mild renal disease.161 Serum creatinine levels should be monitored regularly during prolonged use of tenofovir, and if the serum creatinine level rises, the dose may need to be modified or the drug discontinued to reduce the risk of further nephrotoxicity. Concomitant use of ritonavir or didanosine in HIV-infected patients has been found to be associated with a greater risk of Fanconi syndrome.162 Other Nucleoside Analogs Emtricitabine.  Emtricitabine is a fluorinated cytosine analog that inhibits HBV DNA polymerase and HIV reverse transcriptase. This drug is currently licensed in the United States and other countries for the treatment of HIV type 1 (HIV-1) infection. Preliminary results from a placebocontrolled, phase 3 study in previously untreated patients with chronic HBV infection have demonstrated that emtri­ citabine, 200 mg daily for 48 weeks, reduces serum HBV DNA levels by a median of 3 log10 copies/mL and improves liver histology significantly.163 Emtricitabine treatment of patients coinfected with HBV and HIV has led to levels of suppression of HBV DNA similar to those that occur in treated patients infected with HBV alone.163 The drug is structurally related to lamivudine, however, and has similar mutational sites and rates of resistance. In HBV monoinfected patients, the frequencies of YMDD mutations in persons who receive emtricitabine, 200 mg daily, have been shown to be between 9% and 13% at week 48 and 19% at week 96.164 On the basis of these findings, this agent is unlikely to play an important role in the management of HBV monoinfection. Clevudine.  Clevudine, a pyrimidine analog, is a potent inhibitor of HBV replication both in vitro and in vivo. This drug has been studied in woodchucks and in humans and is licensed for treatment of hepatitis B in Korea on the basis of 24-week clinical trials. In one study, therapy with clevudine given daily for 12 weeks resulted in a reduction in serum HBV DNA levels of greater than 4 log10 copies/mL.165 After therapy was discontinued, rebound to pretreatment HBV DNA levels occurred slowly in this and other studies.166 In one study, a 3-log10 reduction in HBV DNA levels was still evident 6 months after discontinuation of clevudine after a 24-week course of treatment.167 This pattern is different from that reported with other nucleoside analogs and might be explainable by a suppressive effect of clevudine on HBV cccDNA, the genomic template that has been shown to be relatively resistant to treatment with other oral agents. To date, little is known about the safety and efficacy of longterm treatment with clevudine. In 2009, clinical trials were halted because of the occurrence of myopathy in patients treated with clevudine for more than 24 weeks. Combination of a Nucleoside and Nucleotide Agent.  The FDA has approved a combination formulation of tenofovir (300 mg) and emtricitabine (200 mg) for use in patients with HIV infection. This formulation is not specifically licensed for treatment of hepatitis B, but the drug could be useful in cases of resistance to lamivudine, telbivudine, entecavir, adefovir, or tenofovir. HBV mutants resistant to the first four drugs remain susceptible to tenofovir, and tenofovirresistant HBV remains susceptible to emtricitabine. The

drug is occasionally used as first-line therapy of HBV monoinfection in patients with advanced cirrhosis because drug resistance has greater consequences in this clinical setting. As a dual formulation tablet, the drug may improve patient adherence and reduce the cost of care when compared with combination therapy with two nucleoside analogs. Viral Resistance to Nucleoside and Nucleotide Analogs.  With the exception of tenofovir, all currently available nucleos(t)ide analogs have been associated with drug resistance when used as monotherapy. Genotypic resistance is the term that describes the finding of a nucleotide sub­stitution in the HBV DNA polymerase gene that has been associated with clinical evidence of drug resistance. Such a mutation can be detected by a commercially available reverse hybridization assay (InnoLipa, Innogenetics, Belgium). The detection method can detect only known mutants and is limited by the ability to detect only the most common mutations. Furthermore, to be detectable, the drugresistant mutant HBV has to constitute at least 10% of the viral population in an infected patient. Testing for drugresistant HBV should be done prior to discontinuation of the drug because of the rapid re-emergence of wild-type HBV on discontinuation of the drug, at which point the mutant HBV can no longer be detected. Phenotypic resistance refers to the findings of an in vitro assay that uses a human liver cell line transfected with the relevant HBV DNA polymerase mutant. Various concentrations of test drug are added to the culture to assay for loss of susceptibi­ lity of the virus to the drug. This test is a research tool and is helpful in determining whether new genotypic changes cause clinically meaningful drug resistance. Registration trials for nucleos(t)ide analogs have incor­ porated serial genotypic testing for drug-resistant HBV mutants. These studies have shown quite clearly that genotypic resistance to a nucleos(t)ide analog often occurs weeks to months before a virologic breakthrough (defined as greater than a 1-log10 [10-fold] increase in serum HBV DNA levels above the previous nadir) (Fig. 78-6). Ultimately, if the patient is continued on the drug, virologic rebound (defined as an increase in serum HBV DNA levels to greater than 100,000 copies or 20,000 IU/mL) will occur and will be followed by elevation of the serum ALT level (see Fig. 78-6). Rescue therapy can modify this sequence of events if a second agent that lacks cross resistance to the original drug is used alone to replace the first drug or added to the first drug. many experts prefer to add another nucleos(t)ide analog that does not share the same resistance pattern rather than switch to an alternative monotherapy in this clinical situation. A patient treated with a nucleos(t)ide analog should be tested at periodic intervals to assess for virologic breakthrough.168 Such testing enables assessment of suppression of viremia and permits detection of viral breakthrough as early during the course of treatment as possible. To do this, the serum HBV DNA level should be monitored at threemonth intervals. Patients who do not have at least a 1-log10 decline in the serum HBV DNA level after three months of treatment are considered to have primary treatment failure and should be given an alternative agent. The next key interval is at 24 weeks of treatment. In the GLOBE trial, the relative efficacies of lamivudine and telbivudine were compared, and persons who remained positive in serum for HBV DNA at 24 weeks of treatment with either drug were more likely to have a failed response and eventually developed resistance; the risk of a negative outcome to treatment was proportional to the level of detectable HBV DNA in serum at the 24-week point.155 In clinical practice, patients

Chapter 78  Hepatitis B and D Virologic rebound

HBV DNA

ALT Virologic breakthrough

Resistance mutation

Biochemical breakthrough

Time Figure 78-6.  Time course of events leading to clinically apparent drug resistance in patients treated for hepatitis B virus (HBV) infection. In this case, HBV DNA levels decline in serum during antiviral therapy. The levels may or may not become undetectable, and with continued selection pressure by the drug, compensatory mutations develop and improve the ability of the mutant to replicate. The drug-resistant HBV mutant expands within the viral quasispecies. Detection of genotypic resistance occurs weeks to months before virologic breakthrough (a rise in HBV DNA levels in serum), which in turn may precede biochemical breakthrough elevation of serum alanine aminotransferase (ALT) levels by weeks to months. Ultimately, if the patient continues to take the drug, virologic rebound (a marked rise in HBV DNA levels) and a further rise in serum ALT levels occur.

often have a serum HBV DNA level of greater than 10,000 copies/mL at week 24 if their baseline level of viremia is high, even with very potent drugs. If such a patient is taking a nucleos(t)ide analog with a high genetic barrier to resistance (e.g., entecavir or tenofovir), continuing the drug is reasonable, with the expectation of an eventual response. By contrast, if such a patient is taking a nucleos(t)ide analog that is associated with either a high rate of resistance (e.g., lamivudine) or limited antiviral potency (e.g., adefovir), switching to an alternative drug is probably best.169 Unless the choice of a particular nucleos(t)ide analog is restricted or the baseline serum HBV DNA level is relatively low (less than 106 copies/mL), first-line therapy with a highly potent nucleos(t)ide analog such as entecavir or tenofovir is preferred because the rapidity of HBV DNA suppression associated with these agents makes it far less likely that drug-resistant mutants will emerge. Whenever a virologic breakthrough occurs, the patient should be questioned about adherence to therapy. If poor adherence is not a factor, genotypic testing for resistance should be done, as described earlier. If drug resistance is confirmed, the treating clinician may prescribe an alternative nucleos(t)ide analog that lacks cross resistance to the first drug or add the new drug while continuing the first. Clinical experience has indicated, however, that adding a second drug may be preferred to switching to another single agent because sequential monotherapy can result in multidrug-resistant HBV.170 When multidrug resistance occurs, combination therapy is not likely to be effective. Combination Nucleoside or Nucleotide Analog Treatment The combination of two or more nucleos(t)ide analogs may be more effective in the treatment of HBV infection than a single agent. In vitro data and studies in the woodchuck

model of hepatitis B support a role for combination therapy of hepatitis B. Combination treatment is hoped to prevent or delay the emergence of drug resistance and lead to more rapid clinical stabilization. This outcome could be particularly important for patients with decompensated cirrhosis or those in urgent need of liver transplantation. Disadvantages of combination therapy are the added cost and the potential for greater toxicity. In addition, certain combinations could theoretically lead to multidrug resistance. Somewhat surprisingly, the results of early clinical trials of combination therapy in previously untreated patients with HBeAg-positive chronic hepatitis B have shown that the combination of two nucleoside analogs (telbivudine and lamivudine) or the combination of a nucleoside analog and a nucleotide analog (lamivudine and adefovir) does not lead to greater viral inhibition during the first year of treatment than that seen with monotherapy.154,171 The reasons for the lack of apparent additive effect in these studies remain unexplained. Possibly, nucleoside analogs such as telbivudine and lamivudine compete sterically for binding to the HBV DNA polymerase or compete for phosphorylation enzymes (kinases) required for drug activation. Another possible explanation is that a measurable increase in viral suppression may be difficult to demonstrate whenever a drug with substantially less antiviral activity is added to a more potent drug (for example, when lamivudine is added to telbivudine or when adefovir is added to lamivudine). A study in which the combination of lamivudine and adefovir was compared with lamivudine alone, each given for two years, clearly demonstrated a lower rate of lamivudine resistance in the combination therapy arm (15% vs. 43%, respectively).171 On the basis of these and similar observations and the clinical experience with combination reverse transcriptase inhibitors in HIV infection, some authorities have recommended combination therapy as a first-line approach to prevent drug resistance. With the newer nucleos(t)ide agents, however, resistance occurs so infrequently that combination therapy might be better reserved for patients with clinical and laboratory features that have been associated with drug resistance, such as a high level of viremia and high body mass index.168 Combination Interferon and Nucleoside Analog Treatment From a conceptual standpoint, treatment with the combination of interferon and a nucleoside analog might prove to be more effective than either drug alone because these agents have different mechanisms of action and might also permit a shorter course of nucleoside analog therapy, thereby reducing the risk of viral resistance. Three large multicenter studies evaluated these effects in patients given a combination of pegylated interferon and lamivudine. In one study, HBeAg-positive patients received peginterferon alfa-2b with either lamivudine or placebo for one year.126 At the end of treatment, 44% of the patients who received combination therapy had lost HBeAg, whereas only 29% who received peginterferon alone had done so; however, response rates in the two groups were no longer significantly different 26 weeks after the end of treatment (35% and 36%, respectively). It is possible that the low dose of peginterferon used in this study (100 µg weekly for eight months followed by 50 µg weekly until the end of treatment) may have contributed to the relatively high relapse rates after remission. In the second study,127 HBeAg-positive patients were treated with peginterferon alfa-2a, 180 µg weekly for 48 weeks, or combined therapy with lamivudine or with lamivudine monotherapy. Although the greatest degree of HBV DNA suppression was observed in the group

1305

1306

Section IX  Liver that received combination therapy, the proportion of patients who underwent HBeAg seroconversion did not differ significantly between the two interferon-containing arms six months after completion of therapy. In the third study,128 patients with HBeAg-negative chronic hepatitis B were treated with peginterferon alfa-2a in an identical design strategy to that of the HBeAg-positive trial. Again, viral suppression was greater in the combination therapy arm, but this advantage did not translate into a higher rate of sustained virologic response. An interesting finding from the three-arm studies that evaluated peginterferon alfa-2a was that the rate of lamivudine resistance was significantly lower with combination therapy than with lamivudine monotherapy. Taken together, these three studies provide proof of concept that pegylated interferon alpha and lamivudine have additive antiviral effects during treatment. Trials of combinations of pegylated interferon and entecavir or tenofovir given for more than one year are in progress.

Antiviral Therapy in Special Populations

Pregnant Women Nucleos(t)ide analog therapy may be considered during pregnancy for two reasons: to protect the health of the mother and to prevent breakthrough HBV infection in HBV-vaccinated newborns. Several studies have shown that treatment of mothers who have high serum levels of HBV DNA during the last 4 to 12 weeks of pregnancy decreases the rate of HBV infection in newborns vaccinated at birth against HBV.172,173 These studies have been small and generally have had problematic study designs or incomplete follow-up. Furthermore, breakthrough HBV infections develop in only 5% of newborns when the newborns are given a three-dose regimen of HBV vaccine and single dose of HBIG (see later). Therefore, antiviral treatment of the mother to prevent newborn HBV infection remains highly controversial and cannot be recommended at this time. None of the current antiviral agents is licensed for use in pregnancy, and prominent warnings exist for the potential risk of harmful effects on the fetus. Telbivudine and tenofovir are considered category B drugs by the FDA (defined as a lack of animal embryologic toxicity without studies in humans). Extensive experience with tenofovir exists in HIVHBV–coinfected mothers. Lamivudine is a category B drug in HIV-infected pregnant women but a category C drug in HBV-infected women (defined as embryotoxic or teratogenic in animals without study in humans). A large amount of safety data on lamivudine in HIV-infected mothers also exists. Women who are of child-bearing age should be warned not to become pregnant while undergoing treatment with a nucleos(t)ide analog. If pregnancy occurs, the risk of drug withdrawal, including ALT flares, must be balanced against the uncertainity of harmful effects to the fetus. Because lamivudine has a long record of safety and has had the most extensive use during pregnancy in HIV-infected women, many hepatologists prefer to prescribe this agent whenever they feel compelled to treat the hepatitis B in a pregnant woman. Because defects in bone mineral density, including osteomalacia, have been described with tenofovir in HIVinfected patients, this drug seems to be a poor choice during pregnancy because of uncertainty about the effects on fetal bone maturation. The degree of risk to the fetus from the use of nucleo­ s(t)ide analogs that are licensed only for hepatitis B (adefovir, telbivudine, entecavir) is likely to be small. The Antiretroviral Pregnancy Registry in the United States has been tracking spontaneously reported maternal and fetal outcomes in women receiving oral nucleos(t)ide

drugs since 1989. Of the nearly 10,000 reported pregnancies during which the mother had received an oral nucleos(t)ide analog, 95% had HIV infection, and less than 1% had HBV infection alone. Nonetheless, the overall frequency of birth defects in the infants of these women has not been shown to be significantly different from that reported in the general U.S. population. Interferon is contraindicated during pregnancy largely because of its antiproliferative effects. In the event of pregnancy, interferon should be discontinued. Breast feeding is not recommended during the first year of the infant’s life for mothers who are undergoing antiviral therapy. Persons with Acute Hepatitis Because of the high rate (>95%) of complete immunologic recovery from acute hepatitis B, definitive recommendations about the treatment of acute hepatitis B cannot be made. Some experts recommend nucleos(t)ide analog therapy when HBeAg remains detectable in serum for more than 10 to 12 weeks because of the high likelihood of evolution to the chronic HBV carrier state without treatment. A National Institutes of Health–sponsored clinical workshop on hepatitis B proposed that persons with acute viral hepatitis complicated by an increase in INR above 1.5 and deep jaundice persisting for more than four weeks should receive antiviral therapy.174 Antiviral treatment of patients with fulminant hepatitis B is recommended in the AASLD guidelines because of the safety of nucleos(t)ide analog therapy and the need for liver transplantation in many of these patients.120 Persons with Cirrhosis Nucleos(t)ide analog therapy has been shown to be safe in patients with cirrhosis and has made a major difference in the care of patients with advanced liver disease. Interferon is contraindicated in patients with even mildly decompensated cirrhosis because immune-mediated flares of serum ALT levels may occur and may be associated with further clinical deterioration. Also, serious infections have been reported in treated patients.175 Practice guidelines of the AASLD suggest that nucleos(t)ide analog therapy is preferred in all cases of HBV-related cirrhosis because of a greater chance of dose-limiting side effects and clinical worsening in the event of an ALT flare with interferon. Nevertheless, some studies have shown that patients with stable, well compensated cirrhosis can be treated safely and actually may have a higher rate of virologic response when compared with patients without cirrhosis.176,177 Treatment of patients with HBV cirrhosis needs to be individualized, and all of those treated with interferon should be carefully monitored. Persons with Human Immunodeficiency Virus–Hepatitis B Virus Coinfection With improved control of HIV disease with HAART, liver disease has emerged as one of the leading causes of death in patients with HIV.178 Antiviral therapy for hepatitis B should be considered for all HIV-HBV–coinfected patients with evidence of liver disease, irrespective of the CD4 count. In coinfected patients not requiring HAART, therapy for HBV should be based on agents with no HIV activity such as adefovir or pegylated interferon.179 Entecavir treatment is associated with a decline in HIV RNA levels; thus, it also is not recommended for use in patients who are not receiving concomitant HIV treatment.180 In patients with CD4 counts less than 350/mm3, the use of agents with dual anti-HIV and anti-HBV activity should be considered. Combination therapy with either emtricitabine and tenofovir or

Chapter 78  Hepatitis B and D lamivudine and tenofovir should ideally be used to avoid or delay the development of antiviral resistance. Persons with Hepatitis B Virus–Hepatitis C Virus Coinfection When compared with monoinfected patients, HBV-HCV– coinfected patients tend to have more severe liver injury and a higher probability of cirrhosis.181 Limited data are available, however, to define the best approach to treatment in this group of patients. In most instances, one virus, often HCV, is dominant through the process of viral interference. The typical patient is positive for HCV RNA but negative for HBV DNA in serum. Close monitoring has been recommended before treatment is initiated, however, because some patients exhibit alternating viremia. In a prospective clinical trial of 19 patients with combined HBV-HCV infection, all were positive for HCV RNA, and only 5 were positive for HBV DNA prior to treatment. A high rate of sustained virologic response (74%) was observed after a 48-week course of pegylated interferon alpha and ribavirin (see Chapter 79), and two of the five HBV DNA-positive patients also had a virologic response. Unfortunately, HBV DNA became detectable again in four patients who were initially negative, suggesting that a risk exists that HBV may reactivate with eradication of HCV.182 The optimal therapy for patients who are positive in serum for both HBV DNA and HCV RNA is also unclear. In such instances, the author has had success in treating both infections simultaneously with a combination of a nucleos(t)ide analog, pegylated interferon alpha, and ribavirin.

Unresolved Issues

Major advances in antiviral therapy have occurred since 2000, but many unresolved issues remain. Although clearance of HBsAg from serum occurs in some patients after a relatively short course of interferon therapy, interferon is used infrequently because of its adverse effects. Nevertheless, data based on serial monitoring of HBsAg and HBeAg concentrations during therapy with pegylated interferon strongly suggest that failure to achieve a decline in serum concentrations of these viral parameters at key treatment intervals (12 and 24 weeks) has a high negative predictive value for a sustained response and could provide stopping rules similar to those used for the treatment of HCV infection (see Chapter 79). Clinical trials also have suggested that serial monitoring of the HBsAg concentration allows determination of the appropriateness of extending the duration of interferon therapy in patients with HBeAg-negative hepatitis, a condition in which HBsAg clearance is the most reliable endpoint. Multinational trials to evaluate the clinical utility of these markers to monitor virologic response during therapy with interferon are in progress. Commercial assays for HBsAg or HBeAg concentrations are not yet available in the United States but are widely available in Europe and elsewhere in the world and are likely to become available in the United States. Whether 24 weeks of pegylated interferon therapy is preferable to 48 weeks of therapy in patients with a favorable HBV genotype (A or B) or in those who demonstrate an early loss of HBeAg in serum is unknown and is under study. Because combination therapy with pegylated interferon alpha and lamivudine has been associated with enhanced viral suppression, studies of pegylated interferon and newer, more potent nucleos(t)ide analogs are also is progress. The excellent tolerability of oral antiviral therapy makes long-term treatment of hepatitis B feasible. Long-term treatment could be especially important for middle-aged or older patients with high levels of viremia who acquired infection

early in life. Many such patients have normal or only slightly elevated serum ALT levels. For such patients, the intent of treatment would be to keep the serum HBV DNA at the lowest possible level to prevent disease complications and to prolong life, rather than to achieve potentially shortterm endpoints such as HBeAg seroconversion and normalization of serum ALT levels. Additional studies on survival and complication rates after prolonged treatment will determine whether these goals can be achieved, but a placebocontrolled design in patients with active disease will be difficult because data already exist to suggest that the progression of disease is slowed over a few years when HBVinfected patients with advanced fibrosis are treated. The lessons of HIV treatment with regard to the need for combination therapy to prevent drug resistance should not be ignored, but the availability of potent nucleos(t)ide analogs with a high genetic barrier to resistance makes the need for combination therapy a less pressing issue at this time. Because of the low resistance rates associated with these agents, study of the relative benefits of combination nucleos(t)ide analog therapy and monotherapy will require large numbers of patients followed for a long period of time. Retrospective analysis of the data from drug registration trials and prospective evaluation in large clinical trials are needed to define further the common predictors of drug resistance for each drug. Once this is done, combination therapy could be selected for those patients who are more likely to develop resistance with a given nucleos(t)ide analog, thereby limiting the use of combination therapy to selected patients. Finally, the availability of increasingly potent nucleo­ s(t)ide analogs has inadvertently led to diminished emphasis on the importance of the host immune response in achieving therapeutic endpoints. Pegylated interferon is currently the only approved therapeutic agent with immunomo­dulatory as well as viral inhibitory properties. The development and validation of practical tools to measure the immune response before, during, and after interferon therapy would provide a major step forward in understanding the relationship between the host immune response and the viral response, including HBsAg clearance.

PREVENTION

Immunoprophylaxis against HBV is of two types: passive immunization using HBIG and active immunization using inactive HBsAg. Active immunization gives long-term immunity, whereas passive immunization confers only immediate and short-lived protection.

Hepatitis B Immunoglobulin

HBIG is prepared from plasma that is known to contain high titers of anti-HBs. Numerous clinical trials have established the efficacy of HBIG in preventing HBV infection in highrisk persons, such as hemodialysis patients, sexual partners of patients with hepatitis B, and newborn infants of HBsAgpositive mothers. HBIG licensed in the United States has an anti-HBs titer of 1 : 100,000. In Europe, several preparations of HBIG with different concentrations and pharmacokinetic properties are available. HBIG is safe, although rare anaphylactic reactions can occur. Myalgias, rash, and arthralgias have also been reported and are believed to result from formation of antigen-antibody complexes.

Hepatitis B Vaccine

Currently marketed HBV vaccines make use of DNA recombinant technology by introducing the gene for HBsAg (S

1307

1308

Section IX  Liver gene) into the genome of the yeast Saccharomyces cerevisiae. The two vaccines available in the United States are Recombivax HB (Merck, licensed in 1986) and Engerix-B (GlaxoSmithKline, licensed by SmithKline Beecham in 1989). Aluminum hydroxide is used as an adjuvant in both vaccines. Because thimerosal, a preservative used in the vaccines, contains mercury, thimerosal-free vaccines have become available, especially for use in infants. The HBV vaccine is administered intramuscularly in the deltoid muscle of adults and the anterolateral thigh of infants and neonates. The vaccines induce HBsAg-specific helper T cells and T cell–dependent B cells to produce neutralizing antibody against the “a” epitope (amino acid sequence 124 to 148) of HBsAg as early as two weeks after the first injection.183 HBV vaccines are highly efficacious in preventing HBV infection.184 Because the vaccines contain HBsAg only, anti-HBs is the sole antibody produced. Consequently, a vaccinee who tests positive for anti-HBc after vaccination should be considered to have had a subclinical HBV infection. The vaccines typically achieve an anti-HBs titer greater than 100 mIU/mL. Antibody titers greater than 100 mIU/mL confer 100% protection against HBV infection, and a lower antibody titer (up to 10 mIU/mL) is seroprotective in most instances. Peak antibody titers and persistence of antibody levels vary among different persons. The titers drop steadily over the first two years after vaccination, sometimes to levels less than 10 mIU/mL. Two studies in different populations have demonstrated that anti-HBs titers decrease to nonprotective levels in at least 25% to 50% of recipients over a period of 5 to 10 years.185 Although protective anti-HBs response rates after HBV vaccination typically exceed 90%, a number of factors can impede an adequate antibody response. Smoking, obesity, injection into the buttock, chronic liver disease, presence of HLA-DR3, DR7, and DQ2 alleles, absence of the HLA-A2 allele, and extremes of age may be associated with reduced immunogenicity. Such “hyporesponders” may benefit from a higher dose of vaccine. Response rates also are lower in immunocompromised patients, such as transplant recipients, patients receiving chemotherapy, and those with endstage liver disease. Only 50% to 60% of patients on hemodialysis respond adequately to vaccination. Therefore, patients with chronic kidney disease should be vaccinated early in the course of their disease, before renal disease progresses, to ensure optimal response to vaccination.186 Five percent to 8% of HBV vaccine recipients do not achieve detectable anti-HBs levels (“nonresponders”). Studies conducted mostly in animals indicate that intradermal injection of the vaccine may produce a stronger humoral and cellular immune response than conventional intramuscular administration.187,188 Intradermal injection, by recruiting “professional” dendritic cells, stimulates primary MHC class I- and class II–restricted T cell responses. In one study, intradermal vaccination resulted in protective anti-HBs responses in nonresponders to intramuscular adminis­ tration.188 Repeated dosing with intradermal vaccination (5 µg every two weeks to provide an anti-HBs titer of 1000 mIU/mL or greater or a total of 52 doses) has resulted in a protective antibody response rate of nearly 100% in patients undergoing long-term hemodialysis. At present, intradermal HBV vaccination has not been recommended officially—in part because of concerns about standardization of the technique for intradermal delivery. Because HBV vaccination results in strong immunologic memory capable of preventing infection even in patients with low or undetectable antibody titers, no role exists for

Table 78-6  High-Risk Groups for Whom Hepatitis B Virus (HBV) Vaccination Should Be Considered Heath care workers Hemodialysis patients Household contacts and sexual partners of HBV carriers or patients with acute hepatitis B Injection drug users Inmates of correctional facilities International travelers to areas endemic for HBV who may have intimate contact with the local population or take part in medical activities Men who have sex with men Patients who are likely to require multiple transfusions with blood or blood products Patients with chronic liver disease (other than chronic hepatitis B) Potential organ transplant recipients Public safety workers with likelihood of exposure to blood Sexually active heterosexual men and women, if they have more than one partner Staff and clients of institutions for developmentally disabled

a booster vaccine dose in immunocompetent adults and children. Current recommendations include booster doses only for patients undergoing hemodialysis, in whom antiHBs titers should be tested annually and a booster dose given if the titer is lower than 10 mIU/mL.189 No serious side effects of the HBV vaccine have been reported. The frequency of neurologic diseases such as aseptic meningitis and Guillain-Barré syndrome is not increased in vaccine recipients. Targeted High-Risk Groups Table 78-6 lists the high-risk groups for whom HBV vaccination is recommended. Targeted vaccination has not achieved its objective in certain high-risk groups, such as injection drug users, but has achieved great success among health care workers and newborns. The CDC has extended its original recommendations for routine HBsAg screening to include persons born in countries in which the prevalence of hepatitis B exceeds 2%.190 This new recommendation will facilitate identification of susceptible persons who are in need of vaccination and those in need of antiviral therapy. Vaccination Schedule The doses of currently available HBV vaccines and recommendations for the schedules of administration are shown in Table 78-7. The typical vaccination schedule is zero, one, and six months. The first two doses have no effect on the final anti-HBs titer. The third dose acts as a booster to achieve a high anti-HBs titer. In immunocompromised patients and patients undergoing hemodialysis, four vaccine doses are recommended, with the fourth dose given to ensure the highest possible anti-HBs titer. If vaccination is interrupted, the second dose should be administered as soon as possible after the first.190 If the third dose is not given on schedule, it should be given at least two months after the second dose. The HBV vaccine is currently administered to all children and infants as a part of the universal immunization program. Combination HBV vaccines with diphtheria-pertussistetanus (DPT) and Haemophilus influenzae type B (Hib) (DTPw-HB/Hib), the vaccines in current use for immunization of infants, do not reduce the immunogenicity of any of the components of HBV infection.191 Adolescents who have not been vaccinated in infancy or childhood should also be vaccinated.

Chapter 78  Hepatitis B and D Postexposure and Perinatal Prophylaxis Table 78-8 summarizes recommendations for prevention of perinatal transmission of HBV. Table 78-9 lists recommendations for prophylaxis after exposure to a known HBsAg-positive source. Postexposure vaccination should be considered for any percutaneous, ocular, or mucous membrane exposure. The type of immunoprophylaxis is determined by the HBsAg status of the source and the vaccination-response status of the exposed person. Bivalent Vaccine A combined HAV and HBV vaccine has been licensed commercially (TWINRIX, GlaxoSmithKline, Research Triangle Park, NC) and has been shown to be highly immunogenic and protective against both infections (see Chapter 77). This vaccine offers ease of administration for persons at increased risk of both HAV and HBV infection (e.g., world travelers or men who have sex with men) and in patients with underlying chronic liver disease.192 HBV Escape Mutants and Implications for Immunization As described earlier, mutations in the HBV genome that encodes HBsAg can result in mutant HBV virus strains that can escape neutralization by anti-HBs. The mutation involves the “a” determinant and has shown decreased binding to monoclonal anti-a antibodies. Such mutants have been reported worldwide and are particularly common in areas endemic for HBV. These mutant viruses account for some instances in which the HBV vaccine has failed to prevent perinatal transmission. The frequency of this mutation is currently low, but in the future this HBV mutant could emerge as an important threat, in which case HBV vaccines may have to incorporate the mutant antigen to remain effective.

†

Infants and children age 400,000 IU/mL

Genotype 1 Liver biopsy

Consider

PEG-IFN and weightbased RBV 12–16 weeks

PEG-IFN and 800 mg/d RBV 24 weeks

Fibrosis Stage 3–4

Fibrosis Stage 0–2

No treatment required at this time PEG-IFN and weight-based RBV Check HCV RNA at 4 and 12 weeks

RVR

cEVR

pEVR

< 2-log drop

Treat 48 weeks 80–90% SVR

Treat 48 weeks 70% SVR

Treat 72 weeks 38% SVR

Nonresponder Discontinue drugs

Figure 79-9.  Algorithm for the treatment of hepatitis C virus (HCV) infection. Patients infected with HCV genotype 2 or 3 and with a low viral load (level of viremia 4 seconds. ALT, alanine aminotransferase; AST, aspartate aminotransferase; MCV, mean corpuscular volume; WBC, white blood cell. Data from Mendenhall CL. Alcoholic hepatitis. Clin Gastroenterol 1981; 10:417-41.

Alcoholic cirrhosis typically is micronodular or mixed micro- and macronodular. In patients with coexisting alcoholic hepatitis, alcoholic hyaline is almost universal, and sclerosing hyaline necrosis and moderate-to-severe fatty infiltration are common.7,8 In patients with alcoholic cirrhosis who abstain from alcohol for long periods, a frequent finding is a gradual transformation to macronodular cirrhosis, which is indistinguishable from cirrhosis caused by other forms of liver disease.

CONDITIONS THAT MAY RESEMBLE ALCOHOLIC LIVER DISEASE

Although the clinical diagnosis of alcoholic liver disease usually is quite straightforward, the similarity of clinical and histologic features of other disorders to those of alcoholic liver disease sometimes causes diagnostic confusion. The most commonly encountered conditions that have clinical or histologic features in common with alcoholic liver disease are nonalcoholic fatty liver disease (NAFLD), hereditary hemochromatosis, amiodarone hepatotoxicity, and Budd-Chiari syndrome.

Nonalcoholic Fatty Liver Disease

The condition that is most challenging to differentiate from alcoholic liver disease is NAFLD. The two conditions are histologically indistinguishable. As a consequence, the differentiation between alcoholic liver disease and NAFLD has to be made on clinical grounds. The strongest evidence in support of a diagnosis of NAFLD rather than alcoholic liver disease is a history of daily alcohol intake less than 20 g/ day. When a patient’s alcohol intake is questionable, differentiating the two conditions can be difficult, if not impossible. Patients with NAFLD are more likely than patients with alcoholic liver disease to be asymptomatic and often have peripheral insulin resistance, obesity, hypertension, and dyslipidemia.87,88 A model that incorporates the MCV, AST/ALT ratio, body mass index, and gender shows promise in more clearly differentiating patients with alcoholic liver disease from those with NAFLD.89 The serum CDT level can be useful for distinguishing heavy drinkers from abstinent patients with NAFLD; however, the accuracy of this test for detecting moderate but clinically significant levels of alcohol intake is less clear (see earlier and Chapter 85).80

Hereditary Hemochromatosis

On occasion, distinguishing patients with alcoholic liver disease and secondary iron overload from those with liver

disease caused by hereditary hemochromatosis, particularly those with decompensated cirrhosis, can be difficult. Patients with end-stage liver disease from alcoholic cirrhosis may have elevated serum iron and ferritin levels and increased hepatic iron levels suggestive of hereditary hemochromatosis.90 In fact, more than 20% of patients with endstage alcoholic cirrhosis have clinically important hepatic siderosis.91 To complicate matters further, 15% to 40% of patients with hereditary hemochromatosis consume more than 80 g of alcohol daily.92 The overlapping clinical features of hereditary hemochromatosis and alcoholic liver disease include hepatomegaly, testicular atrophy, cardiomyopathy, and glucose intolerance. Testing for mutations in the gene for hereditary hemochromatosis, HFE, is the best method for differentiating the two conditions among whites. Few patients with alcoholic cirrhosis and iron overload are homozygous for C282Y or heterozygous for the C282Y and H63D HFE genes, whereas some have a hepatic iron index value greater than 1.9 that might otherwise suggest hereditary hemochromatosis (see Chapter 74).90,91

Amiodarone Hepatotoxicity

Much less common and less difficult than NAFLD to distinguish from alcoholic liver disease is amiodarone hepatotoxicity. Although the hepatic histologic features of this condition may be similar to those of alcoholic hepatitis with or without cirrhosis, the clinical setting usually distinguishes amiodarone hepatotoxicity from alcoholic liver disease (see Chapter 86).8,93

Budd-Chiari Syndrome

Occasional patients with severe alcoholic liver disease can be misdiagnosed as having acute Budd-Chiari syndrome (hepatic vein thrombosis) on the basis of rapid clinical deterioration, marked hepatomegaly, caudate lobe hypertrophy, and failure to visualize the hepatic veins by Doppler ultrasonography.94 Careful evaluation of these patients usually reveals clinical and biochemical features typical of severe alcoholic hepatitis. Patent hepatic veins usually can be demonstrated by venography. Liver biopsy is particularly useful in distinguishing the characteristic histologic features of alcoholic liver disease from those of Budd-Chiari syndrome. Failure to recognize alcoholic hepatitis as the underlying cause of the liver disease before initiating anticoagulation or performing portacaval shunt surgery can result in high mortality rates (see Chapter 83).94

1391

Section IX  Liver DIFFERENTIAL DIAGNOSIS OF CLINICAL DETERIORATION

Occasionally, determining why a patient with stable, wellcompensated alcoholic liver disease has suddenly and inexplicably deteriorated may be difficult.

Acetaminophen Hepatotoxicity

The most common cause of severe drug-induced liver injury encountered in the United States is acetaminophen hepatotoxicity (see Chapter 86). Two clinical patterns of liver injury have been identified: (1) suicidal or accidental ingestion of large quantities of acetaminophen sufficient to cause hepatic injury and (2) ingestion of lesser quantities of acetaminophen by patients predisposed to injury because of up-regulation of the hepatic enzymes that convert acetaminophen to a hepatotoxic metabolite. The latter type of toxicity is seen most commonly in chronic alcoholics who take excessive acetaminophen over a period of days to weeks for relief of a headache, toothache, or other minor pain.95,96 The clinical features in these patients are indistinguishable from those of alcoholic liver disease, with one obvious exception: AST values are typically more than 1000 U/L, much higher than those in patients with alcoholic liver disease. Because liver injury typically has occurred by the time of hospita­ lization, acetaminophen levels are not helpful for diagnosis or management. Recognition of the cause of the unusually elevated serum aminotransferase levels comes from careful questioning of the patient and family about acetaminophen ingestion in the days to weeks before hospitalization. The morbidity and mortality associated with this condition are considerable.95,96 Because many of these patients have a history of recent heavy alcohol use, few are candidates for liver transplantation.

CHRONIC HEPATITIS C

The cofactor that influences progression of alcoholic liver disease most profoundly is HCV infection. Between one fourth and one third of patients with alcoholic liver disease have serologic or virologic evidence (or both) of HCV infection.100 The prevalence of HCV infection is highest in patients who have used injection drugs; however, the risk is high even among those who deny drug use. Histologic features of focal lymphoid aggregates, portal inflammation, and periportal or bridging fibrosis are common in liver biopsy specimens from alcoholics with HCV infection.101 Of greater importance, liver disease is more severe, advanced disease develops at a younger age, and survival is shorter in patients with both alcoholic liver disease and HCV infection than in patients with alcoholic liver disease and no evidence of HCV infection.100 In one of the more striking examples of the interaction between alcohol abuse and hepatitis C, Corrao and colleagues found that the .relative risk of cirrhosis was 10-fold higher among heavy drinkers with chronic hepatitis C than among those who had no evidence of HCV infection (Fig. 84-5).102 In addition, alcohol and HCV act synergistically in the development of hepatocellular carcinoma (see Chapter 79).103-105

OBESITY AND SMOKING

The risk of liver disease is two to three times higher in drinkers who are obese than in drinkers who have a normal body mass index.106 Although an increased risk of fatty liver is not surprising in obese persons (see Chapter 85), obesity also appears to be an independent risk factor for both alcoholic hepatitis and cirrhosis.106,107 Cigarette smoking also has been shown to accelerate the progression of fibrosis in patients with alcoholic liver disease,108,109 and smoking

Acute Viral Illness

Patients with alcoholic cirrhosis are vulnerable to decompensation from a variety of viral illnesses. Acute viral hepatitis can result in the sudden onset of liver failure, with extremely high mortality rates (see Chapters 77 through 81).97,98 Sudden decompensation and liver failure also have been reported during infection with influenza A virus.99 These dramatic cases illustrate the importance of routine immunization against hepatitis A and B and influenza in patients with alcoholic cirrhosis.

Hepatocellular Carcinoma

Occasional patients with alcoholic cirrhosis who have been abstinent for many years decompensate suddenly, with the abrupt onset of hepatic encephalopathy, variceal bleeding, or ascites. Not infrequently, the underlying cause is hepatocellular carcinoma. Unfortunately, the sudden onset of symptoms frequently results from tumor invasion of the portal or hepatic veins; as a result, the prognosis for these patients is dismal.3 The risk of hepatocellular carcinoma in patients with alcoholic cirrhosis underscores the need for surveillance for this neoplasm in these patients, especially those who abstain from alcohol and in whom the long-term prognosis is otherwise good (see Chapter 94).

160 140

HCV negative HCV positive

120 Odds ratio for cirrhosis

1392

100 80 60 40 20 0 None

COFACTORS THAT MAY INFLUENCE PROGRESSION OF ALCOHOLIC LIVER DISEASE A number of factors have been reported to have an adverse effect on the progression of liver disease in chronic alcoholics. The most important of these factors are chronic hepatitis C virus (HCV) infection, obesity, and smoking.

25 or 50

75 or 100 125 or 150

>175

Average lifetime daily alcohol intake (g) Figure 84-5.  Odds ratio for developing cirrhosis in patients who chronically drink varying amounts of alcohol based on the presence or absence of hepatitis C virus (HCV) infection. (Data from Corrao G, Lepore AR, Torchio P, et al. The effect of drinking coffee and smoking cigarettes on the risk of cirrhosis associated with alcohol consumption. A case-control study. Provincial Group for the Study of Chronic Liver Disease. Eur J Epidemiol 1994; 10:657-64.)

Chapter 84  Alcoholic Liver Disease appears to accelerate disease progression in patients with HCV infection who drink heavily.110

PROGNOSIS The prognosis for an individual patient with alcoholic liver disease depends on the degree of pathologic injury, patient’s nutritional status, presence of complications of advanced liver disease, presence of other comorbid conditions such as HCV infection, and patient’s ability to discontinue destructive patterns of drinking. In studies that have examined the natural history of alcoholic liver disease on the basis of histologic characteristics at diagnosis, patients with fatty liver have had the best outcome (70% to 80% survival rate at four to five years); those with alcoholic hepatitis or cirrhosis have had an intermediate outcome (50% to 75% survival rate at four to five years); and those with cirrhosis combined with alcoholic hepatitis have had the worst outcome (30% to 50% survival rate at four to five years) (Fig. 84-6).111 Among all patients with alcoholic liver disease, the average one-year and five-year survival rates are approximately 80% and 50%, respectively.80 Alcoholic cirrhosis also appears to be an independent risk factor for hepatocellular carcinoma.2,3,100-102 Among alcoholics, men older than 50 years of age appear to be most vulnerable to the development of hepatocellular carcinoma (see Chapter 94).112 Estimating the prognosis of patients with alcoholic liver disease is particularly important for determining the need for (1) specific therapy in a patient with severe alcoholic hepatitis and (2) liver transplantation in a patient with alcoholic cirrhosis.

ALCOHOLIC HEPATITIS

The prognosis among patients with alcoholic hepatitis can vary dramatically. In patients with severe disease, the mortality rate is high, approaching that for patients with fulminant hepatic failure. Clinical features associated with severe disease include hepatic encephalopathy, marked prolongation of the prothrombin time, elevation of the serum bilirubin level above 25 mg/dL, depression of the serum albumin level, an elevated serum creatinine level, and older age. Other important prognostic variables in patients with severe alcoholic hepatitis are spontaneous hepatic encephalopathy and hepatorenal syndrome (Fig. 84-7).79,81,82 The one-month mortality rate in patients with spontaneous hepatic encephalopathy is approximately 50%, and the rate in those with hepatorenal syndrome is 75%.79,81,82,113-115 Three models have been shown to predict short-term prognosis in these often critically ill patients. Maddrey and Boitnott discovered a simple formula they called the discriminant function (DF), calculated as [4.6 × prothrombin time − control value (seconds)] + serum bilirubin (mg/dL). The DF has proved useful for identifying patients with poor short-term survival rates.113 Three prospective studies have demonstrated that patients with a DF value of 32 or more have a poor prognosis, with one-month mortality rates of 35% to 45%.82,114,115 By contrast, patients with a DF value less than 32 have short-term survival rates of 90% to 100%.113,116 The Model for End-stage Liver Disease (MELD) score (which includes the serum bilirubin level, INR, and serum creatinine level) and the Glasgow alcoholic hepatitis score (which includes age, WBC count, blood urea nitrogen level, prothrombin time ratio [ratio of the patient’s prothrombin time to the control value], and serum bilirubin level) also have been shown to predict survival in patients with severe alcoholic hepatitis.117,118 Because both the MELD and Glasgow scores include measures of renal function, they appear to be more accurate than the DF in determining the prognosis of patients with concomitant kidney injury.

1.0

80

Mortality rate at day 28 (%)

Probability of survival

0.9

0.8

0.7

0.6

0.5 40

80

120

160

200

240

60

40

20

280

Weeks No cirrhosis, no hepatitis n = 58 Cirrhosis, no hepatitis n = 42 No cirrhosis, hepatitis n = 19 Cirrhosis, hepatitis n = 98 Figure 84-6.  Survival of patients with alcoholic liver disease stratified by histologic severity of disease. (From Orrego H, Black JE, Blendis LM, Medline A. Prognosis of alcoholic cirrhosis in the presence or absence of alcoholic hepatitis. Gastroenterology 1987; 92:208-14, with permission.)

0 DF > 32

Encephalopathy

Hepatorenal syndrome

Clinical features of severe alcoholic hepatitis Figure 84-7.  Mortality rate at day 28 in patients with severe alcoholic hepatitis who received no treatment and who had a discriminant function (DF) score greater than 32, spontaneous hepatic encephalopathy, or hepatorenal syndrome. See text for calculation of DF. (Data from references 84, 85, and 114.)

1393

Section IX  Liver 1.0

70 60 Probability of survival

0.8 5-year survival rate (%)

1394

50 40 30 20

Abstinence 0.6 Relapsing 0.4 Excessive 0.2

10

Unknown

0 0

0 5–7

8–10

11–15

Child-Turcotte-Pugh score Figure 84-8.  Five-year survival rates in patients with alcoholic cirrhosis according to their Child-Turcotte-Pugh scores. (Data from Poynard T, Naveau S, Doffoel M, et al. Evaluation of efficacy of liver transplantation in alcoholic cirrhosis using matched and simulated controls: Five-year survival. Multi-centre group. J Hepatol 1999; 30:1130-7.)

120 240 360 480 600 720 840 960 1080 Survival duration in days

Figure 84-9.  Survival curves during the three years following hospital discharge according to alcohol consumption: abstinence: patients who were abstinent; relapsing: patients with one or more periods of abstinence alternating with one or more periods of excessive consumption; excessive: patients with excessive consumption of alcohol at the first follow-up point. Survival differed significantly between abstinent and excessively drinking patients (P < .001). (Modified with permission of Veldt BJ, Laine F, Guillygomarc’h A, et al. Indication of liver transplantation in severe alcoholic liver cirrhosis: Qualitative evaluation and optimal timing. J Hepatol 2002; 36:93-8.)

ALCOHOLIC CIRRHOSIS

The clinical tool used most widely to determine prognosis in patients with alcoholic cirrhosis is the Child-TurcottePugh (CTP) score and Child (or Child-Pugh) classification (see Chapter 90). This simple classification system, which was designed specifically to assess the risk of mortality following portacaval shunt surgery in cirrhotic patients with variceal bleeding, has gained favor as a rapid method for determining the prognosis of patients with various chronic liver diseases. The Child classification is as effective as quantitative liver function tests (see Chapter 73) and disease-specific prognostic models for determining shortterm prognosis in groups of patients awaiting liver transplantation (see Chapter 95).119 Despite its limitations, the Child classification has been adopted widely for riskstratifying patients with cirrhosis because of its simplicity and ease of use. Five-year survival rates for patients with alcoholic cirrhosis decrease dramatically as the CTP score and Child’s class become higher at the time of clinical presentation (Fig. 84-8).117 The development of ascites, variceal bleeding, hepatic encephalopathy, spontaneous bacterial peritonitis, or hepatorenal syndrome also has a significant impact on the prognosis of patients with alcoholic cirrhosis. The five-year survival rate for persons in whom any of these complications develop is only 20% to 50% of that for patients with compensated cirrhosis.120 The most ominous complications are spontaneous bacterial peritonitis and rapid-onset hepatorenal syndrome (see Chapters 91 and 92). Fewer than half of the patients in whom spontaneous bacterial peritonitis develops can be expected to survive one year; the median survival of patients with hepatorenal syndrome is less than two weeks.121,122 Other models that have been used to predict prognosis in patients with alcoholic cirrhosis are the Beclere model, a proportional hazards model developed by Poynard and colleagues, and the MELD score.119,123 The Beclere model,

which was developed from a database of 818 patients with alcoholic cirrhosis who were followed prospectively for four years, includes the serum bilirubin level, serum albumin level, patient’s age, and presence or absence of hepatic encephalopathy.119 The MELD model was developed at the Mayo Clinic to assess short-term prognosis in patients undergoing transjugular intrahepatic portosystemic shunt placement.123 This model has been shown to be useful for predicting short-term survival in groups of patients with various liver diseases (see Chapter 90). Abstinence from continued excessive drinking is the most important predictor of survival in patients who survive an initial hospitalization for alcoholic cirrhosis.124 The rate of survival over the ensuing two years is 70% to 80% among patients who abstain or dramatically reduce their excessive drinking, compared with only 20% to 30% in those who continue to drink heavily (Fig. 84-9).124

TREATMENT ABSTINENCE AND LIFESTYLE MODIFICATION

Virtually every study of abstinence in alcoholic liver disease shows beneficial effects on patient survival, even in patients with decompensated cirrhosis (see Fig. 84-9). Reducing but not completely stopping alcohol consumption also has been shown to improve projected survival in patients with alcoholic liver disease.125,126 Heavy drinkers who receive socalled brief interventions, which are less than 1 hour in duration and incorporate simple motivational counseling techniques, are twice as likely as control patients to moderate or stop their drinking 6 to 12 months later.126 Regular meetings with a nurse or other health care professional to

Chapter 84  Alcoholic Liver Disease

NUTRITIONAL SUPPORT

Nutritional abnormalities are pervasive in alcoholics. Two divergent patterns are common: obesity and malnutrition. Persons who combine alcohol abuse with a high calorie diet frequently develop truncal obesity, which can accelerate the progression of the underlying alcoholic liver disease.41,42,125 The high frequency of obesity is not surprising given the high caloric content (7.1 kcal/g) of alcohol. By contrast, malnutrition is a widespread clinical problem among patients with alcoholic liver disease when a substantial proportion of nutrient-rich dietary calories are replaced with alcohol.131 Every patient with moderate to severe alcoholic hepatitis or cirrhosis shows some signs of malnutrition, with up to 50% of total energy intake derived from alcohol.132 The frequency of malnutrition increases dramatically with the severity of liver disease.128 For example, the frequency of profound malnutrition increases from 20% in patients with Child’s class A cirrhosis to 60% in those with Child’s C cirrhosis.127 Furthermore, a strong association exists between protein-calorie malnutrition and complications of alcoholic disease such as infections, encephalopathy, ascites, and variceal bleeding.131,132 Levels of folate, vitamin B6, thiamine, and vitamin A, as well as those of trace elements such as selenium, zinc, copper, and magnesium, are often severely reduced in patients with alcoholic liver disease.125 The mechanisms underlying the profound malnutrition frequently observed in patients with moderate to severe alcoholic liver disease are complex and multifactorial and include (1) anorexia induced by increased proinflammatory cytokines such as TNF and leptin; (2) intestinal fat and protein malabsorption; and (3) a catabolic state that promotes gluconeogenesis from skeletal and visceral proteins.131-133 Evaluating malnutrition in patients with liver disease can be difficult because the tests most commonly used to assess nutritional status (e.g., serum albumin concentration, anthropometry, immune status) often are affected by the liver disease. The creatinineheight index appears to be the most reliable indicator of loss of muscle mass in patients with moderate-to-severe liver disease.133,134

100 90 80 Percent surviving

emphasize abstinence and adherence to medication appear to be more effective than intense counseling by alcohol treatment specialists.124,127 Abstinence invariably causes resolution of hepatic steatosis. An additional goal of abstinence is to prevent ongoing injury, fibrosis, and the possible development of hepatocellular carcinoma, but few studies have addressed the effects of abstinence on disease progression. Short-term treatment with naltrexone, an opioid antagonist, decreases the chance of relapse in one third of heavy drinkers; however, this agent is contraindicated in patients with liver disease because of its extensive hepatic metabolism and potential hepatotoxicity.127,129 Baclofen, a gamma aminobutyric acid (GABA) B-receptor agonist, shows promise as the first safe and effective agent to improve abstinence and decrease the likelihood of relapse in patients with alcoholic cirrhosis.130 As discussed earlier, obesity, which is increasing in frequency among alcoholics as well as the general population, is associated with the development of fatty liver, steatohepatitis, and cirrhosis and appears to be a major risk factor for progression of alcoholic liver disease (see Chapter 85).41,42 The majority of alcoholics smoke cigarettes, another risk factor for more severe alcoholic liver disease. Therefore, lifestyle modifications including significant reduction or cessation of alcohol consumption, weight control, and elimination of cigarette smoking, are important initial approaches to the treatment of alcoholic liver disease.

70

TEN: 62%

60 50 40 P = 0.028

30

GC: 39%

20 10 0 0

30 60 90 120 150 180 210 240 270 300 330 360 Days

Figure 84-10.  Probability of survival for one year after randomization of 72 patients to total enteral nutrition (TEN) or to glucocorticoid therapy (GC). (From Cabre E, Rodriguez-Iglesias P, Caballeria J, et al. Short- and longterm outcome of severe alcohol-induced hepatitis treated with steroids or enteral nutrition: A multicenter randomized trial. Hepatology 2000; 32:3642, with permission.)

Treating nutritional abnormalities in alcoholics can be quite challenging. The management of obesity combines abstinence with dietary restriction and regular exercise— goals that are difficult to achieve in any group of patients, especially alcoholics. Malnutrition can be equally difficult to overcome. Two thirds of hospitalized patients provided with and encouraged to consume a balanced 2500-kcal diet with careful monitoring by a dietitian are unable to meet this objective. These difficulties have stimulated research into various forms of nutritional support for patients with severe alcoholic hepatitis and cirrhosis.135-138 In the first large study of intensive nutritional support in patients with moderate to severe alcoholic hepatitis, Mendenhall and colleagues demonstrated increased survival in patients provided aggressive nutritional support.132 In a pivotal multicenter study by Cabre and coworkers, patients were randomized to receive prednisone 40 mg daily (see later) or a liverspecific formula containing 2000 calories per day through a feeding tube.136 The one-month mortality rates were similar in both groups, but the one-year mortality rate was significantly lower in the patients who received the enteral nutrition, in great part because of reduced infectious complications, in comparison with patients who received glucocorticoids (Fig. 84-10). This study clearly demonstrates the important role of enteral nutrition in hospitalized patients with severe alcoholic liver disease. By contrast, no study has, to date, demonstrated a survival benefit for treatment with parenteral nutrition.131 Tube feeding in patients with alcoholic liver disease probably is underused in most hospitals because of concerns about precipitating hepatic encephalopathy or stimulating bleeding from esophageal varices, neither of which has been documented. Most patients probably can tolerate standard enteral products, and only selected patients with overt hepatic encephalopathy require liver-specific products rich in branched-chain amino acids. Studies of nutritional support in outpatients are limited. Hirsch and colleagues demonstrated that patients attending

1395

Section IX  Liver an outpatient liver clinic who took an enteral nutritional support product that contained 1000 kcal and 34 g of protein had significantly improved protein intake and fewer hospitalizations in comparison with those not receiving the supplement.137 Cirrhotics have decreased hepatic glycogen stores and an accelerated metabolic reaction to short-term fasting. Studies have shown that nighttime snacks (710 kcal = 2 cans of Ensure Plus) attenuate this catabolic response and significantly improve lean tissue over a one-year period of supplementation.138 Therefore, nutritional support can improve nutritional status and, in some patients, may enhance liver function and decrease the risk of death. Assessment of nutritional status and nutritional supplementation should be pursued aggressively in both inpatients and outpatients with alcoholic liver disease, especially those with moderate-to-severe alcoholic hepatitis and cirrhosis.

ANTI-INFLAMMATORY AND ANTICYTOKINE DRUGS Glucocorticoids

Glucocorticoid therapy has been the most extensively studied and the most controversial treatment for patients with alcoholic hepatitis. A total of 10 small, single-center, placebo-controlled randomized trials of glucocorticoid therapy were published from 1971 to 1984, and only two showed a benefit.114 Helman and colleagues demonstrated improved survival only in patients who had hepatic encephalopathy within the first 10 days after hospital admission.17 Maddrey and colleagues confirmed the prognostic importance of encephalopathy and found that a DF value greater than 32 (see earlier) was as effective as detecting encephalopathy in selecting patients at high risk for early mortality and that these patients appeared to benefit from gluco­ corticoid therapy.113 These two prognostic tools—hepatic encephalopathy and an elevated DF—were used to select patients for entry into a subsequent multicenter study that demonstrated a dramatic improvement in short-term survival with glucocorticoid therapy.114 The cumulative 28-day mortality rate for this severely ill group of patients was 35% in the placebo recipients, compared with only 6% in patients who received methylprednisolone. Using the same selection criteria for study entry, Ramond and colleagues confirmed the improvement in short-term survival and also demonstrated a continued survival benefit for up to six months after treatment with glucocorticoids (Fig. 84-11).115 Additional follow-up of these patients revealed that the survival benefit of glucocorticoid therapy persisted for one but not two years after treatment.133 No major complications were associated with glucocorticoid therapy in these studies.114,115 These two clinical trials included only patients with severe disease; patients with gastrointestinal bleeding requiring transfusions and active infection were excluded. Furthermore, none of the patients had evidence of hepatorenal syndrome before entry into the studies. Glucocorticoids should not be used in patients with mild alcoholic hepatitis; however, a short course of glucocorticoids (e.g., prednisone, 40 mg daily for 28 days, followed by 20 mg daily for 7 days and 10 mg daily for 7 days) may be beneficial in patients with severe disease. The DF, MELD score, and Glasgow index can each be used to select patients for treatment. Glucocorticoids should not be used in patients with gastrointestinal bleeding requiring blood transfusions or with evidence of active infection and probably are not effective in patients with hepatorenal syndrome.19,139 The response to glucocorticoids can be determined within seven

100

Glucocorticoid

Percent surviving

1396

50

Placebo

0 90

180

Day Figure 84-11.  Survival in 61 patients with alcoholic hepatitis randomly assigned to receive glucocorticoid therapy or placebo. Survival rates at six months were 84% in the glucocorticoid treatment group and 45% in the placebo group (P = .002). (From Raymond MJ, Poynard T, Rueff B, et al. A randomized trial of prednisone in patients with severe alcoholic hepatitis. N Engl J Med 1992; 326:507, with permission.)

days (primarily by a reduction in serum bilirubin levels), and treatment can be discontinued at that time if a response has not been achieved.140 Figure 84-12 illustrates the factors that should be taken into account when glucocorticoid therapy is considered in patients with severe alcoholic hepatitis.

Pentoxifylline

Pentoxifylline is a nonselective phosphodiesterase inhibitor that increases intracellular concentrations of adenosine 3′,5′-cyclic monophosphate (cAMP) and guanosine 3′,5′cyclic monophosphate (cGMP) and may thereby inhibit TNF production. Pentoxifylline also has been shown to decrease gene transcription and to affect multiple steps in the cytokine/chemokine inflammatory pathway, either directly or indirectly by inhibiting TNF.141 Selected effects of pentoxifylline include inhibition of cytokine/chemokine synthesis (e.g., MCP-1, IL-8, macrophage inflammatory protein [MIP]-1α and MIP-1b), decreased expression of adhesion molecules on endothelial cells, decreased activation of neutrophils, decreased proliferation of lymphocytes and monocytes, and decreased binding and transmigration of leukocytes. Pentoxifylline also reduces fibroblast proliferation and secretion of collagen and other interstitial matrix proteins. Akriviadis and colleagues performed a prospective, randomized, double-blind clinical trial of pentoxifylline in patients with severe alcoholic hepatitis (DF greater than 32).20 Forty-nine patients received pentoxifylline, 400 mg orally three times daily, and 52 received placebo (vitamin B12) for four weeks. Only 12 patients (24.5%) who received pentoxifylline died, compared with 24 (46%) who received placebo (Fig. 84-13). Pentoxifylline therapy was associated with a significant decrease in the frequency of hepatorenal syndrome as a cause of death and was well tolerated with

Chapter 84  Alcoholic Liver Disease no major side effects. On the basis of this single trial, pen­ toxifylline appears to be a viable alternative to glucocorticoids, particularly in patients with clinically important renal dysfunction.

Alcoholic hepatitis

Hepatic encephalopathy or at least one of the following: DF* ≥ 32, MELD† score ≥ 18, or Glasgow score‡ ≥ 9 Yes No Poor prognosis: consider specific therapy

Glucocorticoids Followed by Pentoxifylline

Good prognosis: nutritional support and conservative management

Louvet and his colleagues in France explored the possibility of switching patients to pentoxifylline if they failed to demonstrate a response to the first seven days of corticosteroid therapy.142 Unfortunately, these patients did not obtain any benefit from the early switch. Therefore, the optimal approach to the management of nonresponders to glucocorticoid therapy remains unresolved.

Specific Anti-Tumor Necrosis Factor Therapy

Active gastrointestinal bleeding, systemic infection, or renal insufficiency Yes

No

Pentoxifylline 400 mg three times daily for 28 days

Prednisone 40 mg for 7 days. If serum bilirubin level decreases, continue prednisone 40 mg daily for an additional 21 days, followed by a two-week taper. If bilirubin level does not decrease, stop treatment after 7 days.

Figure 84-12.  Algorithm for the management of patients with alcoholic hepatitis. *The DF is calculated as follows: 4.6 (prothrombin time of patient − prothrombin time of control) + serum bilirubin level (in mg/dL). † The Model for End-stage Liver Disease (MELD) score is based on the serum bilirubin level, international normalized ratio, and serum creatinine level (see Chapter 90). ‡ The Glasgow alcoholic hepatitis score is based on the patient’s age, white blood cell count, blood urea nitrogen level, ratio of prothrombin time to a control value, and serum bilirubin level. DF, discriminant function.

1.0 0.9

Probability of survival

0.8 0.7

Dysregulated cytokine metabolism was described in alcoholic hepatitis long before it was recognized in inflammatory bowel disease and rheumatoid arthritis. An initial concern in alcoholic liver disease arose from early observations that low (“basal”) amounts of TNF were important for liver regeneration.143 Therefore, many investigators suggested that down-regulating, without totally blocking, TNF activity would be a preferred therapeutic intervention. Indeed, many therapies used in alcoholic liver disease (e.g., glucocorticoids, pentoxifylline, S-adenosylmethionine) decrease but do not abolish TNF activity. Because therapy with anti-TNF antibodies has been shown to block development of alcohol-induced liver injury in rats, it was initially studied in small clinical trials in patients with alcoholic hepatitis, with apparent success.144 A large, double-blind, randomized controlled trial in France in which patients with acute alcoholic hepatitis were treated with prednisolone or prednisolone plus high-dose infliximab was terminated, however, because of an increased rate of infectious complications in the patients who received combined therapy.145 Etanercept was postulated to be more appropriate than infliximab in patients with alcoholic liver disease because of its shorter duration of action, but a National Institutes of Health (NIH)-sponsored multicenter trial reported similar one-month mortality rates and significantly worse six-month mortality rates in patients with moderate-to-severe alcoholic hepatitis treated with etanercept compared with those treated with placebo.146 Therefore, at the present time, agents that inhibit, but do not totally block, inflammatory mediators such as TNF appear to be preferable to those that totally block these mediators in patients with alcoholic hepatitis.

0.6

ANTIOXIDANTS S-adenosylmethionine

0.5 0.4 0.3 0.2 0.1 0 0

20

40

60

80

100

120

140

160

Time (days) Figure 84-13.  Probability of survival in 101 patients with alcoholic hepatitis treated with pentoxifylline (red line) or placebo (blue line) (P = .037). (From Akriviadis E, Botla R, Briggs W, et al. Pentoxifylline improves short-term survival in severe acute alcoholic hepatitis: A double blind, placebocontrolled trial. Gastroenterology 2000; 119:1637-48, with permission.)

As discussed earlier, alcoholic liver disease is characterized by elevated plasma methionine concentrations and decreased clearance of intravenously or orally administered methionine. Decreased MAT activity results in decreased plasma (and presumably intrahepatic) SAM levels.147 Administration of SAM has been reported to protect against experimental liver injury caused by alcohol, acetaminophen, carbon tetrachloride, and galactosamine.69,75 Theoretical benefits of SAM in alcoholic liver disease include its roles as an antioxidant and a critical methyl donor and its actions in maintaining mitochondrial function, decreasing TNF levels, and producing glutathione. A multicenter clinical study reported that SAM in a daily dose of 1200 mg significantly reduced the mortality rate and decreased the need for liver transplantation in patients with Child’s A and B alcoholic cirrhosis.148

1397

1398

Section IX  Liver Silymarin

Silymarin, the active ingredient extracted from Silybum marianum (also known as milk thistle), has been shown in experimental animals to protect against various hepatotoxins, including carbon tetrachloride, acetaminophen, iron (in iron overload), and poisonous mushrooms (see Chapter 87).149 It has antioxidant properties, protects against lipid peroxidation, and exerts anti-inflammatory and antifibrotic effects. Despite these properties, insufficient data are available from well-conducted clinical trials to demonstrate improvement in mortality, complications of cirrhosis, or histology in patients with alcoholic liver disease.150,151 Nevertheless, silymarin has become the most popular form of complementary and alternative medicine therapy for patients with liver disease because of its good safety profile, and ongoing NIH studies should answer questions concerning its efficacy (see Chapter 127).

Vitamin E

Vitamin E deficiency has been well documented in patients with alcoholic liver disease.152 Vitamin E has hepatoprotective effects in experimental liver injury, with potentially beneficial effects that include membrane stabilization, reduced NF-κB activation and TNF production, and inhibition of hepatic stellate cell activation and collagen production.152-154 Unfortunately, the largest randomized study of vitamin E supplementation in patients with alcoholic liver disease did not show a significant benefit, possibly because a relatively low dose was used.154

Combination Antioxidant Therapy

Oxidative stress plays an etiologic role in the development of alcoholic liver disease (see earlier), and antioxidants block the development of alcoholic liver disease in experimental animals. An initial study of a combination of antioxidants in patients with alcoholic hepatitis reported beneficial effects,155 whereas a subsequent trial in which glucocorticoids were compared with an antioxidant cocktail was stopped after an interim analysis found a significant benefit in the glucocorticoid-treated group.156 These inconsistent results led to a third trial in which patients with acute alcoholic hepatitis were randomized to antioxidant therapy alone or with glucocorticoids; neither treatment improved six-month survival.157 None of these studies evaluated whether or not antioxidant therapy actually decreases oxidative stress. At this time, antioxidants should not be used as sole therapy for alcoholic hepatitis; whether they are of benefit as adjunctive therapy in patients with alcoholic hepatitis or alcoholic cirrhosis and which antioxidant should be used remain unclear.

DRUGS OF UNLIKELY BENEFIT Colchicine

Colchicine has many potential therapeutic mechanisms of action in alcoholic liver disease, including inhibition of collagen production, enhancement of collagenase activity, and anti-inflammatory activity. Initial positive studies158 led to a large Veterans Administration (VA) Cooperative Study of colchicine therapy in patients with alcoholic cirrhosis that showed no beneficial effects on overall or liver-related mortality.159 A smaller study from Europe also showed no beneficial effects of colchicine therapy in patients with alcoholic liver disease.160

Propylthiouracil

Chronic alcohol feeding in animal models can induce a hypermetabolic state with increased oxygen consumption

similar to the hypermetabolic state associated with hyperthyroidism. This hypermetabolic state may lead to relative hypoxia in the centrilobular area of hepatic lobules. Propylthiouracil has been postulated to attenuate the hypermetabolic state, function as an antioxidant, and improve portal blood flow. Nevertheless, a Cochrane review of six randomized trials involving more than 700 patients found no beneficial effect of propylthiouracil therapy in patients with alcoholic liver disease.161

Anabolic Steroids

Anabolic steroids have been shown to decrease fatty infiltration in the liver and are hepatoprotective. As noted earlier, patients with end-stage liver disease frequently are malnourished and often have low circulating levels of the anabolic hormone insulin-like growth factor-1. These observations provide a rationale for using anabolic steroids to treat alcoholic liver disease.132 Nevertheless, a Cochrane review was not able to demonstrate efficacy for anabolic steroids (specifically oxandrolone) in patients with alcoholic liver disease, although such therapy did appear to be safe.162

Ursodeoxycholic Acid

Urosodeoxycholic acid, an agent used for a variety of cholestatic liver disorders (see Chapter 89), was evaluated in one large multicenter controlled clinical trial in patients with severe alcoholic liver disease and found to have no beneficial effect on six-month survival.163

Polyenylphosphatidylcholine

Polyenylphosphatidylcholine, or lecithin, a lipid extract obtained from soybeans, has been shown to prevent septal fibrosis and cirrhosis in alcohol-fed baboons and to stimulate the release of collagenase activity by cultured hepatic stellate cells. It also has antioxidant effects and decreases TNF production.164 Multiple positive studies of poly­ enylphosphatidylcholine in animal models of liver disease led to a VA Cooperative Study that evaluated the effects of this drug in humans with early alcoholic liver disease.165 Results of this study were negative; however, patients decreased their alcohol use markedly during the trial, thus decreasing the likelihood that a beneficial effect of poly­ enylphosphatidylcholine could be demonstrated.

LIVER TRANSPLANTATION

Liver transplantation for alcoholic liver disease remains one of the most contentious and controversial areas in transplantation medicine. Short-term outcomes following liver transplantation in patients with alcoholic liver disease are comparable to those for patients who receive transplants for most other conditions, with seven-year survival rates of 60% (see also Chapter 95).166 Profound confusion in the early postoperative period is more likely to develop in patients with alcoholic cirrhosis, however, than in those undergoing transplantation for other liver diseases.167 The result can be a prolonged hospitalization and an increase in the cost of transplantation. A number of transplant centers have reported distinctly lower long-term survival rates among patients with alcoholic liver disease, particularly among those who return to heavy drinking.168 In addition, patients with alcoholic liver disease have an increased risk of pharyngeal, esophageal, and gastric malignancies after transplantation.169 This risk is particularly high among persons who begin smoking heavily again after the operation.170 Many patients with apparently advanced alcoholic liver disease can recover to the degree that transplantation is not

Chapter 84  Alcoholic Liver Disease 1.00

0.8

0.75 Probability of survival

Probability of improving liver function

1.0

0.6

0.4

Transplanted: 58% 5-year survival

0.50

Simulated: 35% 5-year survival

Matched controls: 31% 5-year survival

0.25

0.2

0

0.00 0

60

120

180

240

300

360

0

500

required if they reduce their alcohol intake significantly or abstain completely (Fig. 84-14).124 Because the benefits of abstinence can be so dramatic, requiring a period of abstinence before proceeding with transplantation is reasonable for patients with alcoholic liver disease. Patients who have a CTP score of 11 or greater despite at least six months of abstinence have improved survival with liver transplantation compared with predicted survival based on the Beclere model (Fig. 84-15).119 Similar, although less impressive, results have been shown using other prognostic models.171 The optimal length of pretransplant abstinence remains controversial.40 Some experts have argued that patients with severe alcoholic hepatitis should be abstinent for one year before being considered for transplantation, whereas others have argued that patients should be considered for transplantation if they continue to have CTP scores of 11 or greater after only three months of abstinence.124,139 Evidence of a survival benefit following transplantation is less clear for patients with milder alcoholic liver disease, unless they have hepatocellular carcinoma. Patients with CTP scores of 5 to 7 do not benefit from liver transplantation.119 The survival benefit from transplantation for patients with a CTP score of 8 to 10 after 6 months of abstinence is minimal compared with predicted survival using the Beclere and MELD models.119,171 Furthermore, a trial in which patients with a CTP score of 8 to 10 were randomized to receive immediate transplantation or to be observed expectantly showed a lower two-year survival rate among the patients randomized to undergo immediate transplantation (73% versus 80%), primarily because of a high risk of postoperative malignancy.172 Efforts to refine risk scores in patients with severe alcoholic liver disease and incorporate them into standard transplantation selection criteria as well as efforts to select

1500

2000

Days

Follow-up duration in days Figure 84-14.  Probability that liver function will improve with time in patients with cirrhosis caused by alcoholic liver disease who reduce alcohol intake or abstain from alcohol. Improvement was defined as a change from Child-Turcotte-Pugh class C to class B or A. Red line, patients who reduced intake; blue line, abstinent patients; circles, deaths; triangle, patient undergoing transplantation. (From Veldt BJ, Laine F, Guillygomarc’h A, et al. Indication of liver transplantation in severe alcoholic liver cirrhosis: Qualitative evaluation and optimal timing. J Hepatol 2002; 36:93-8, with permission.)

1000

Figure 84-15.  Probability of survival over five years in patients with ChildTurcotte-Pugh scores of 11 to 15 after six months of abstinence from alcohol who underwent liver transplantation (top line), compared with matched control subjects (P = .008) and simulated control subjects (i.e., predicted from a model) (P = .001). (Modified from Poynard T, Naveau S, Doffoel M, et al. Evaluation of efficacy of liver transplantation in alcoholic cirrhosis using matched and simulated controls: 5 year survival. Multicentre group. J Hepatol 1999; 30:1130-7.)

patients who can maintain long-term abstinence for alcohol and tobacco abuse are ongoing.

OPTIMAL MANAGEMENT

Alcoholic liver disease accounts for 50,000 deaths annually in the United States and Europe. The optimal management of patients with alcoholic liver disease begins with a dramatic reduction in or elimination of alcohol intake, which often can be accomplished successfully using “brief interventions” by a nurse, primary care physician, or gastroenterologist. Abstinence can have a profound impact on survival even in patients with decompensated cirrhosis. The next important step is to eliminate other factors, such as cigarette smoking and obesity, which can enhance disease progression. Treatment of concomitant HCV infection may be an important aspect of management in some patients (see Chapter 79). Patients with severe alcoholic hepatitis should receive enteral feedings to ensure adequate calorie and protein intake. In patients with severe alcoholic hepatitis who do not have a systemic infection or gastrointestinal bleeding, a short course of glucocorticoid therapy should be considered. An alternative strategy is the use of pentoxifylline, especially in patients with marginal renal function or hepatorenal syndrome. For patients with alcoholic cirrhosis, no drugs available in the United States have documented efficacy. Many patients already are taking agents such as milk thistle or SAM that may be of benefit and appear to be safe. These agents are not covered by insurance, however, and access to such agents often depends on the socioeconomic status of the patient. Because of the risk of decompensation with superimposed infections, all patients with alcoholic cirrhosis should receive vaccinations for hepatitis A and B and annual vaccinations for influenza. In addition, they should

1399

1400

Section IX  Liver undergo regular screening for hepatocellular carcinoma (see Chapter 94). Screening is particularly important in older patients with cirrhosis who have been abstinent for sustained periods. Finally, liver transplantation has been shown to be effective in carefully selected patients who have discontinued drinking (see Chapter 95).

KEY REFERENCES

Akriviadis E, Botla R, Briggs W, et al. Pentoxifylline improves shortterm survival in severe acute alcoholic hepatitis: A double-blind, placebo-controlled trial. Gastroenterology 2000; 119:1637-48. (Ref 85.) Arteel GE. Oxidants and antioxidants in alcohol-induced liver disease. Gastroenterology 2003; 124:778-90. (Ref 13.) Boetticher NC, Peine CJ, Kwo P, et al. A randomized, double-blinded, placebo-controlled multicenter trial Etanercept in the treatment of alcoholic hepatitis. Gastroenterology 2008; 135;1953-60. (Ref 146.) Cabre E, Rodriguez-Iglesias P, Caballeria J, et al. Short- and long-term outcome of severe alcohol-induced hepatitis treated with steroids or enteral nutrition: A multicenter randomized trial. Hepatology 2000; 32:36-42. (Ref 136.) Ji C. Dissection of endoplasmic reticulum stress signaling in alcoholic and non-alcoholic liver injury. J Gastroenterol Hepatol 2008; 23(Suppl 1):S16-24. (Ref 70.) Kunos G, Osei-Hyiaman D, Batkai S, Gao B. Cannabanoids hurt, heal in cirrhosis. Nat Med 2006; 12:608-10. (Ref 71.) Mathurin P, Mendenhall CL, Carithers RL, et al. Glucocorticoids improve short-term survival in patients with severe alcoholic hepa-

titis (AH): Individual data analysis of the last three randomized placebo controlled double blind trials of glucocorticoids in severe AH. J Hepatol 2002; 36:480-7. (Ref 116.) Niemelä O. Biomarkers in alcoholism. Clin Chimica Acta 2007; 377:3949. (Ref 80.) Pfitzmann R, Schwenzer J, Rayes N, et al. Long-term survival and predictors of relapse after orthotopic liver transplantation for alcoholic liver disease. Liver Transplant 2007; 13:197-205. (Ref 168.) Plank LD, Gane EJ, Peng S, et al. Nocturnal nutritional supplementation improves total body protein status of patients with liver cirrhosis: A randomized 12-month trial. Hepatology 2008; 48:557-66. (Ref 138.) Purohit V, Abdelmalek MF, Barve S, et al. Role of S-adenosylmethionine, folate, and betaine in the treatment of alcoholic liver disease: Summary of a symposium. Am J Clin Nutr 2007; 86:14-24. (Ref 35.) Shukla SD, Velazquez J, French SW, et al. Emerging role of epigenetics in the actions of alcohol. Alcohol Clin Exp Res 2008; 32:1525-34. (Ref 69.) Stewart S, Prince M, Bassendine M, et al. A randomized trial of antioxidant therapy alone or with corticosteroids in acute alcoholic hepatitis. J Hepatol 2007; 47:277-83. (Ref 157.) Veldt BJ, Laine F, Guillygomarc’h A, et al. Indication of liver transplantation in severe alcoholic liver cirrhosis: Quantitative evaluation and optimal timing. J Hepatol 2002; 36:93-8. (Ref 124.) Yip WW, Burt AD. Alcoholic liver disease. Semin Diagn Pathol 2006; 23:149-60. (Ref 8.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

85 Nonalcoholic Fatty Liver Disease Andrea E. Reid

CHAPTER OUTLINE Nonalcoholic Fatty Liver and Steatohepatitis  1401 Epidemiology  1401 Etiology  1401 Pathogenesis  1402 Clinical, Laboratory, and Imaging Features  1405

In 1980, Ludwig and colleagues from the Mayo Clinic coined the term nonalcoholic steatohepatitis (NASH) to describe a form of liver disease observed in middle-aged patients with abnormal liver biochemical test results and histologic evidence of alcoholic hepatitis but no history of alcohol abuse.1 Much has been learned about NASH since this initial description. NASH is part of the spectrum of nonalcoholic fatty liver disease (NAFLD), which encompasses simple fatty liver, NASH, and NAFLD-associated cirrhosis. NAFLD has emerged as a burgeoning clinical entity, now recognized as an important component of the metabolic syndrome, as well as an exciting area of basic and clinical investigation in the field of hepatology.

NONALCOHOLIC FATTY LIVER AND STEATOHEPATITIS EPIDEMIOLOGY

The prevalence of NAFLD in the general population is undefined. Several studies have estimated the scope of this disorder in the United States. The Dallas Heart Study of more than 2200 adults documented hepatic triglyceride content by proton magnetic resonance spectroscopy (MRS) and found fatty liver in 31% of participants; the highest prevalence (45%) was among Hispanics.2 Most of the patients with fatty liver by MRS had normal liver biochemical test levels, although the normal range for serum aminotransferase levels in this study was wider than generally accepted. The National Health and Nutrition Examination Survey (NHANES) III, which included more than 15,700 adults, documented unexplained elevations of serum aminotransferase levels, presumably caused by NAFLD, in 2.8% to 5.5% of participants.3,4 Population-based estimates of NAFLD have been reported for other countries as well. These studies have documented NAFLD in 10% to 24% of the population, with the highest prevalence (up to 76%) among obese nondrinkers.5-7 Prevalence estimates vary widely depending on the information available in a given population and the diagnostic criteria that are used to estab-

Histopathologic Features  1405 Diagnosis  1407 Natural History  1408 Treatment  1409 Focal Fatty Liver  1410

lish the diagnosis (i.e., liver biochemical test levels, radiologic study results, or liver biopsy findings). Most cases of NAFLD are discovered in the fourth to sixth decades of life, although NAFLD is also described, with increasing frequency, in obese children and adolescents, as well as in older adults. NAFLD may be present long before a diagnosis is established. In early clinical studies, the majority of patients with NAFLD were female; however, subsequent data have suggested that men may be affected as often as women and may be at greater risk for advanced forms of NAFLD, including NASH. The prevalence of NAFLD appears to vary by ethnicity. In the Dallas Heart Study, Hispanics demonstrated the highest prevalence (45%) of NAFLD, compared with 33% for whites and 24% for African Americans. The reasons for racial and ethnic disparities in the prevalence of NAFLD is not known but may be related, at least in part, to racial differences in body fat distribution8 and the prevalence of the metabolic syndrome, which is greatest in people of Hispanic descent.9 Other studies have also shown that African Americans and Mexican Americans have higher frequencies of unexplained serum aminotransferase elevations than do whites.3,4,9,10 Familial clustering of NAFLD may occur,11,12 which likely reflects both genetic and environmental predisposition to the metabolic conditions associated with NAFLD (see later).13

ETIOLOGY

Many agents and conditions have been associated with NAFLD. The causes may be divided into two broad categories (1) drugs and toxins and (2) metabolic abnormalities, either acquired or congenital. Potential causes of NAFLD are listed in Table 85-1. Obesity is the condition most often reported in association with NAFLD. Since 1980, the proportion of Americans who are overweight (defined as a body mass index [BMI] > 25 kg/m2) or obese (BMI > 30 kg/m2) has increased markedly. In 2004, 66.2% of Americans adults were classified as overweight or obese, as were 17.4% of children ages 12 to 19 years.14 The health implications of the unremitting obesity epidemic are staggering, and NAFLD is

1401

1402

Section IX  Liver Table 85-1  Causes of Nonalcoholic Fatty Liver Disease Acquired Metabolic Disorders Diabetes mellitus Dyslipidemia Kwashiorkor and marasmus Obesity Starvation Cytotoxic and Cytostatic Drugs l-Asparaginase Azacitidine Azaserine Bleomycin Methotrexate Puromycin Tetracycline* Other Drugs and Toxins Amiodarone 4,4´-diethylaminoethoxyhexestrol Dichlorethylene Ethionine Ethyl bromide Estrogens Glucocorticoids Highly active antiretroviral therapy Hydrazine Hypoglycin Orotate Perhexilene maleate Safrole Tamoxifen Metals Antimony Barium salts Chromates Phosphorus Rare earths of low atomic number Thallium compounds Uranium compounds Inborn Errors of Metabolism Abetalipoproteinemia Familial hepatosteatosis Galactosemia Glycogen storage disease Hereditary fructose intolerance Homocystinuria Systemic carnitine deficiency Tyrosinemia Weber-Christian syndrome Wilson disease Surgical Procedures Biliopancreatic diversion Extensive small bowel resection Gastric bypass Jejunoileal bypass Miscellaneous Conditions Industrial exposure to petrochemicals Inflammatory bowel disease Partial lipodystrophy Jejunal diverticulosis with bacterial overgrowth Severe anemia Total parenteral nutrition *Tetracycline is cytotoxic by virtue of inhibiting mitochondrial b-oxidation.

a common byproduct in both adults and children. As noted earlier, most patients with NAFLD are obese. In morbidly obese patients (BMI > 35 kg/m2), including those referred for bariatric surgery, the frequency of NAFLD is as high as 90%, with advanced disease (i.e., NASH) seen in 9% to 40%.15-19 A correlation among BMI, degree of steatosis, and severity of liver injury has been demonstrated in several studies20-22; however, the distribution of body fat may be

more important than the total adipose mass for the development of hepatic steatosis. Studies have shown a significant correlation between the risk of the metabolic syndrome, degree of hepatic steatosis, and waist-to-hip ratio, thus highlighting the importance of intra-abdominal or visceral fat as a pre­dictor of NAFLD.23-25 NAFLD also is strongly associated with type 2 diabetes mellitus and glucose intolerance, with or without superimposed obesity.26 Type 2 diabetes mellitus, hyperglycemia, or glucose intolerance has been described in 20% to 75% of adult patients with NASH and may increase the risk of NASH more than twofold compared with that for nondiabetic persons. The presence of NAFLD in diabetic patients may also increase the risk of cardiovascular disease significantly.27 The association between type 2 diabetes mellitus and NAFLD appears strongest in morbidly obese patients.15 NAFLD has been associated with insulin resistance and hyperinsulinemia even in lean subjects with normal glucose tolerance.28 Diabetes mellitus may be an independent predictor of advanced NAFLD, including cirrhosis and hepatocellular carcinoma.29-32 Hyperlipidemia is found in a substantial proportion of patients with NAFLD. Data from the Dallas Heart Study revealed NAFLD in 60% of patients with mixed hyperlipidemia,33 and a study from Korea of potential living liver donors showed that hyperlipidemia was associated with a greater than twofold risk of significant (>30%) steatosis.34 Most patients with NAFLD have multiple risk factors, including central obesity, type 2 diabetes mellitus, and hyperlipidemia, although some affected persons lack all recognized risk factors. NAFLD has been associated with many drugs and toxins and metabolic, surgical, and genetic conditions (see Table 85-1) that have abnormal fat metabolism and mitochondrial injury or dysfunction in common. NAFLD is now recognized as the hepatic component of the metabolic syndrome, which includes hyperlipidemia, glucose intolerance, obesity, and systemic hypertension. The risk and severity of NAFLD increase with the number of components of the metabolic syndrome.28,35

PATHOGENESIS

The pathogenesis of NAFLD is poorly understood, in part because of a lack of suitable animal models that mimic human NAFLD. In light of the variety of conditions that have been associated with NAFLD, it is not surprising that no single pathogenic mechanism has been identified. The prevailing theory is the “two-hit hypothesis,” first proposed by Day and James in 1998.36 This hypothesis states that dysregulation of fatty acid metabolism leads to steatosis, which is the first hepatic insult in NAFLD. Steatosis is associated with several cellular adaptations and altered signaling pathways, which render hepatocytes vulnerable to a “second hit.” The second insult may be one or more environmental or genetic perturbations, which cause hepatocyte necrosis and inflammation and activate the fibrogenic cascade, thereby leading to fibrosis and cirrhosis in a minority of patients with NAFLD. Hepatic steatosis is the hallmark histologic feature of NAFLD. Normally, free fatty acid (FFA) is supplied to the liver through intestinal absorption (in the form of chylomicron remnants) or from lipolysis of adipose tissue, where FFA is stored as triglycerides. In the liver, FFA is oxidized by mitochondria, esterified into triglycerides, synthesized into phospholipids and cholesteryl esters, and secreted from the liver as very-low-density lipoprotein (VLDL). Under normal circumstances, fatty acid metabolism is under tight regulatory control by catecholamines, glucagon, growth hormone, and insulin. Hepatic triglyceride accumulation

Chapter 85  Nonalcoholic Fatty Liver Disease occurs when fatty acid metabolism shifts to favor net lipogenesis, rather than lipolysis. This shift occurs when the amount of FFA supplied to the liver from the intestine or adipose tissue exceeds the amount needed for mitochondrial oxidation, phospholipid synthesis, and synthesis of cholesteryl esters. Triglycerides also accumulate in the liver when synthesis of lipoprotein decreases or export of lipids from the liver is impeded (see also Chapter 72). Current evidence points to insulin resistance and hyperinsulinemia as the primary pathogenic factors in steatosis in most patients with NAFLD. Strong laboratory and clinical evidence supports the association of peripheral insulin resistance and hyperinsulinemia with NAFLD, even in lean patients without obvious glucose intolerance.37-39 The molecular mechanism leading to insulin resistance is complex and not understood completely. In the setting of obesity and hyperinsulinemia, alterations in several molecules, including FFA, tumor necrosis factor-α (TNF-α), membrane glycoprotein PC-1, and leptin, interfere with the insulin signaling pathway. Diabetes mellitus and obesity are associated with increased amounts of FFA in plasma, caused in part by abnormal release of FFA by insulin-resistant adipocytes. Excess FFA contributes to hepatic insulin resistance by down-regulating insulin receptor substrate-1 (IRS-1) signaling.40,41 Insulin resistance and hyperinsulinemia lead to steatosis by means of a number of aberrant mechanisms of FFA disposal. In the liver, insulin stimulates fatty acid synthesis, down-regulates mitochondrial boxidation of FFA, blocks the secretion of triglycerides from hepatocytes by increasing intracellular degradation of VLDL and apolipoprotein B-100 (apoB-100), and blocks exocytosis of VLDL-containing vesicles.40,42,43 Also, patients with NASH have impaired hepatic synthesis of apoB-100, which also may contribute to hepatic triglyceride accumulation.44 Insulin resistance in NAFLD may be potentiated by aberrant levels or function of several important peptide mediators secreted by adipocytes, including TNF-α, leptin, and adiponectin. In noninflammatory states, TNF-α is derived from adipose tissue (including adipose tissue macrophages), and plasma levels of TNF-α correlate with body fat mass.45 TNF-α interferes with insulin signaling by down-regulating IRS-1 signaling via serine phosphorylation, likely through activation of stress-related protein kinases including Jun N-terminal kinase (JNK), which plays a key role in obesityrelated insulin resistance. Activation of the inhibitor kappab kinase (IKK-b)/nuclear factor kappa b (NF-κb) pathway by FFA may also play a role in reduced hepatic insulin sensitivity,46,47 and may increase production of additional inflammatory cytokines such as TNF-α and interleukin (IL)-6.48 Elevated TNF-α levels have been demonstrated in several studies of NAFLD49-52; however, the independent contribution of TNF-α to the pathogenesis and risk of progression of NAFLD is still unclear. Adipocytokines are peptides produced by visceral adipose tissue. Adiponectin is secreted by adipocytes in inverse proportion to BMI and is a potent inhibitor of TNF-α. Serum adiponectin levels are reduced in obesity, insulin resistance, diabetes mellitus, and the metabolic syndrome.50 Delivery of recombinant adiponectin to mice fed a high-fat, alcohol-containing diet and to genetically obese (ob/ob) mice dramatically alleviates hepatomegaly, steatosis, inflammation, and elevated liver biochemical test levels in both murine populations.53 These therapeutic effects result in part from the ability of adiponectin to enhance hepatic fatty acid b-oxidation, decrease hepatic triglyceride content, and decrease hepatic insulin resistance. Furthermore, adiponectin suppresses hepatic and plasma concentrations of TNF-α. Studies have reported an inverse relationship

between serum adiponectin levels and the degree of steatosis and hepatocyte injury in humans with NAFLD, and this inverse association may be independent of insulin resistance.51,54 Further studies are needed to determine whether an increase in the TNF-α/adiponectin ratio has a primary pathogenic role in the development of steatosis or is more directly correlated with progression from steatosis to steatohepatitis. Leptin is a satiety hormone, derived from adipocytes, that controls food intake and energy regulation (see Chapter 1). Leptin is intimately involved with insulin signaling and regulation of glucose metabolism in peripheral tissues and may play an important role in regulating the partitioning of fat between mitochondrial b-oxidation and triglyceride synthesis in the liver.55 Severe steatosis and steatohepatitis develop in leptin-deficient (ob/ob) mice. Obesity in humans is associated with relative leptin resistance and high leptin levels, which may contribute to the genesis of steatosis by a negative impact on insulin signaling or may be a consequence of the chronic hyperinsulinemia associated with obesity. Several studies have examined the relationship between serum leptin levels and NAFLD, with conflicting results.56-59 One study has suggested that serum leptin levels in patients with NASH correlate with the severity of hepatic steatosis, independent of BMI, but not with the degree of hepatic inflammation or fibrosis.58 At present, the contri­bution of leptin to the pathogenesis of NAFLD is unclear. Although insulin resistance and hyperinsulinemia are clearly pivotal to the development of steatosis, consensus is lacking on the subsequent insults that cause steatosis to progress to steatohepatitis and fibrosis in some patients. Similarities in the histologic features and natural history of alcoholic liver disease and NAFLD suggest that common mechanisms may be involved in the pathogenesis of these disorders. Chronic oxidative stress is believed to be central to the pathogenesis of alcohol-related liver damage. Processes that increase the production of oxidants in the liver during chronic alcohol exposure include the metabolism of ethanol to its reactive intermediate acetaldehyde; induction of microsomal ethanol-oxidizing enzymes, such as cytochrome P450 2E1 (CYP2E1), which generates reactive oxygen species (ROS) that can peroxidize cellular membranes, thereby causing cellular injury60; inhibition of mitochondrial electron transport chain activity; and depletion of mitochondrial glutathione.61 Activation of microsomal enzymes, including CYP2E1, in patients with NAFLD62-63 and mitochondrial production of ROS in murine models of NAFLD64,65 suggest that chronic oxidative stress and lipid peroxidation may also be central to the pathogenesis of NAFLD. Increased levels of FFA can be directly toxic to hepatocytes through a number of mechanisms. An increased FFA concentration leads to lysosomal destabilization and stimulation of TNF-α.66 FFA also up-regulates cytochrome P450 isoenzymes, leading to enhanced generation of ROS and lipid peroxidation.67 An increased intracellular FFA concentration can lead to sustained up-regulation of peroxisomal proliferator-activated receptor-α (PPAR-α), which promotes fatty acid oxidation and disposal but also may increase oxidative stress through the production of dicarboxylic acid derivatives; PPAR-α also may predispose affected persons to carcinogenesis.44 FFA can be directly toxic to cellular membranes, lead to the formation of toxic fatty acid ethyl ethers, and cause overall disruption of mitochondrial function, thereby overwhelming the over­ lapping protective mechanisms designed to combat FFA hepatotoxicity.45

1403

1404

Section IX  Liver Endotoxin and endotoxin-mediated cytokine release are suspected in the pathogenesis of alcoholic steatohepatitis, in which increased serum levels of bacterial endotoxin and lipopolysaccharide (LPS) stimulate hepatic production of TNF-α, IL-6, and IL-8 and activate an inflammatory response that leads to hepatic necrosis (see Chapter 84).61 Bacterial endotoxin also may contribute to the development of NAFLD in some circumstances. Portal endotoxemia was believed to contribute to NASH and hepatic failure associated with surgical jejunoileal bypass (performed in the past to treat obesity), the risk of which was reduced with antibiotics. Yang and colleagues have demonstrated that ob/ob mice with steatosis are highly vulnerable to endotoxininduced hepatocyte damage, and NASH rapidly develops in these animals after exposure to low doses of bacterial LPS.68 In addition, Zucker diabetic (fa/fa) rats and ob/ob mice demonstrate decreased Kupffer cell function, which may increase the vulnerability of steatotic hepatocytes to TNFα–mediated liver damage.69 Small studies suggest a possible pathogenic role of bacterial endotoxins in human NAFLD as well,70,71 but these studies are far from conclusive. A growing body of evidence suggests that mitochondrial changes and altered hepatic energy homeostasis may play roles in the pathogenesis of NAFLD. Studies have shown a decrease in the activity of mitochondrial respiratory chain complexes in steatotic livers, with a concomitant increase in mitochondrial ROS formation; these changes correlate with serum TNF-α levels, insulin resistance, and BMI.72,73 Ob/ob mice have increased levels of uncoupling protein, UCP-2, an inner mitochondrial membrane protein that mediates proton leak, uncouples adenosine triphosphate (ATP) synthesis, regulates ROS production, and may render fatty hepatocytes vulnerable to metabolic stressors65; however, the role of UCP-2 in NAFLD in humans is unknown. Studies have shown that mice and humans with NAFLD have diminished capacity for replenishing ATP stores after ATP depletion. Mitochondrial structural defects may be one cause of reduced ATP stores. Megamitochondria and crystalline mitochondrial inclusions have been iden­ tified in patients with NAFLD and may represent an adaptive process to oxidative stress or secondary injury.39,74 Limited data in patients with NASH suggest differential expression of several genes important for proper mitochondrial functioning, including genes involved in ROS scavenging, glucose metabolism, and fatty acid metabolism.75 In addition, mitochondrial DNA damage similar to that found in alcoholic liver disease and Wilson disease also may contribute to the development of NASH. Further animal and human studies are needed to determine whether mitochondrial dysfunction and ATP depletion are causes or consequences of NAFLD. Fibrosis is a frequent histologic finding in advanced NAFLD but has not been well studied in this disease. Hepatic fibrosis results from activation and proliferation of hepatic stellate cells in the subendothelial space of Disse, with subsequent secretion of extracellular matrix components, including collagen types I and III (see Chapter 90). Factors proposed to initiate and perpetuate the fibrogenic process in stellate cells include inflammatory cytokines, angiotensin, alterations in the extracellular matrix, growth factors, and oxidative stress. In NAFLD, lipid peroxidation products may enhance hepatic production of transforming growth factor-b (TGF-b), which activates stellate cells. Endothelial cells, leukocytes, and Kupffer cells may stimulate the stellate cells to proliferate, possibly through the release of platelet-derived growth factor (PDGF), TGF-b, and other cytokines.76 In addition, hyperinsulinemia and hyperglycemia associated with NAFLD may stimulate release of

connective tissue growth factor, an intermediate molecule involved in fibrogenesis.77 Finally, animal data suggest that leptin may perpetuate fibrogenesis in NAFLD by stimulating Kupffer cells and sinusoidal endothelial cells to produce TGF-b.78 Research into the pathogenesis of NAFLD is proliferating at a rapid pace, but the picture is far from clear (Fig. 85-1). Any one of the putative mechanisms discussed here is unlikely to explain the pathogenesis of NAFLD in all affected patients. More likely, NAFLD develops as a con­ sequence of a “multi-hit” process. The first “hit” is steatosis

Obesity Insulin resistance Hyperinsulinemia Increased serum leptin levels Altered cytokine levels: Increased TNF-α Decreased adiponectin

Excessive dietary carbohydrates, dyslipidemia, genetic mutations, drugs, toxins, nutritional deficiencies, other factors

Fat-containing hepatocytes

Direct cytotoxic effects of increased FFA

Oxidative stress and lipid peroxidation

Impaired β-oxidation of FFA

Mitochondrial damage: Increased ROS Altered ATP homeostasis Structural abnormalities

Steatohepatitis ? Environmental, genetic, dietary factors

Increased TGF-β and other cytokines

Fibrosis/cirrhosis Figure 85-1.  Proposed pathogenesis of nonalcoholic fatty liver disease (NAFLD). NAFLD is believed to occur as a result of a multi-hit process. Insulin resistance and hyperinsulinemia are present in many patients with NAFLD and, possibly in association with other metabolic or genetic abnormalities and altered cytokine levels, may lead to hepatic steatosis. Hepatocellular injury leading to steatohepatitis occurs in a minority of patients with fatty liver, probably as a result of multiple overlapping insults. Increased levels of free fatty acid (FFA), uncompensated oxidative stress, lipid peroxidation, cytokine dysregulation, mitochondrial dysfunction, and other environmental and genetic factors may contribute to the development of hepatocellular injury and fibrosis in susceptible persons. ATP, adenosine triphosphate; ROS, reactive oxygen species; TNF-α, tumor necrosis factor-α; TGF-b, transforming growth factor-b.

Chapter 85  Nonalcoholic Fatty Liver Disease Table 85-2  Clinical and Laboratory Features of Nonalcoholic Fatty Liver Disease SYMPTOMS

SIGNS

LABORATORY FEATURES

Common None (48%-100% of patients)

Hepatomegaly

Two- to fourfold elevation of serum ALT and AST levels AST/ALT ratio less than 1 in most patients Serum alkaline phosphatase level is slightly elevated in one third of patients Normal serum bilirubin and serum albumin levels and prothrombin time Elevated serum ferritin level

Splenomegaly Spider angiomata Palmar erythema Ascites

Low-titer (less than 1 : 320) ANA Elevated transferrin saturation HFE gene mutation (C282Y)

Uncommon Vague right upper quadrant pain Fatigue Malaise

ANA, antinuclear antibodies; ALT, alanine aminotransferase; AST, aspartate aminotransferase.

induced primarily by insulin resistance and hyperin­ sulinemia. After steatosis develops, a number of factors, including lipid peroxidation, oxidative stress, cytokine alterations, mitochondrial dysfunction, and Kupffer cell activation, may incite an inflammatory response and fibrosis in some patients with genetic or environmental susceptibilities. The exact interplay among these and other factors remains to be elucidated, but understanding of the pathogenesis should be enhanced by the development and refinement of appropriate animal models, including the genetically obese, leptin-deficient ob/ob mouse and Zucker diabetic rats, in which steatosis develops; S-adenosylmethionine (SAM)-deficient mice, in which severe steatohepatitis deve­ lops79; the Otsuka-Long-Evans-Tokushima fatty (OLETF) rat model, in which a cholecystokinin-A receptor defect leads to hyperglycemia, obesity, insulin resistance, and hepatic steatosis80; and a high fat-fed rat model in which insulin resistance, elevated serum TNF-α levels, increased oxidative stress, mitochondrial lesions, and early fibrosis develop.81

CLINICAL, LABORATORY, AND IMAGING FEATURES

The clinical and laboratory features of NAFLD are summarized in Table 85-2. NAFLD usually is discovered incidentally because of elevated liver biochemical test levels or hepatomegaly noted during an evaluation for an unrelated medical condition. Most patients with NAFLD are asymptomatic, but some may describe vague right upper quadrant pain, fatigue, and malaise. Hepatomegaly has been described in up to 75% of patients with NAFLD but often is difficult to appreciate on physical examination because of obesity. Stigmata of chronic liver disease, such as splenomegaly, spider angiomata, and ascites, are rare, except in patients with NAFLD-associated cirrhosis. Elevated liver biochemical test levels may be found in up to 50% of patients with simple steatosis and are present in approximately 80% of patients with advanced NAFLD. A mild-to-moderate (1.5- to 4-fold) elevation of the serum aspartate aminotransferase (AST) or alanine aminotransferase (ALT) level, or both, is usual, and levels seldom exceed 10 times the upper limit of normal. The serum ALT level usually is greater than the AST level, in contrast with the pattern of alcoholic hepatitis, in which the AST level is at least twofold higher than the ALT level (see Chapters 73 and 84). The alkaline phosphatase and gamma glutamyl transpeptidase (GGTP) levels may be elevated, but the

serum bilirubin level, prothrombin time, and serum albumin level typically are normal, except in patients with NAFLDassociated cirrhosis. Up to one fourth of patients with NAFLD may have antinuclear antibodies (ANA) in low titers (less than 1 : 320).82 Antimitochondrial antibodies (AMA) and hepatitis B surface antigen are not detected. Antibody to hepatitis C virus (anti-HCV) must be absent to implicate NAFLD as the sole cause of abnormal liver biochemical test levels; however, steatosis, often in association with visceral obesity, frequently accompanies HCV infection and may be associated with a more aggressive course (see Chapter 79).83 Serum ceruloplasmin and α1-antitrypsin levels are within normal limits. Serum and hepatic iron levels may be elevated in patients with NAFLD. In particular, the serum ferritin level may be elevated in 20% to 50% of patients with NAFLD and may be a marker of more advanced disease.29,84 Nevertheless, the frequency of genetic hemochromatosis among patients with NAFLD is not increased. Clinical and laboratory findings do not correlate with the histologic severity of NAFLD. The entire histologic spectrum of NAFLD, including cirrhosis, can be seen in patients with normal or near-normal serum aminotransferase levels.31 Imaging studies often are obtained during the evaluation of unexplained liver biochemical abnormalities or suspected NAFLD. Hepatic ultrasonography, the imaging modality employed most commonly, may reveal a “bright” liver of increased echogenicity, consistent with hepatic steatosis (Fig. 85-2). Fatty liver also can be documented by abdominal computed tomography (CT) scan (a fatty liver is lower in density than the spleen), and by magnetic resonance imaging (MRI), with which fat appears bright on T1-weighted imaging. A study that assessed the sensitivities of MRI, abdominal CT, and ultrasonography for distinguishing advanced NASH from simple steatosis showed that ultrasonography and CT had sensitivity rates of 100% and 93% for detecting hepatic fat involving greater than 33% of the liver, with positive predictive values of 62% and 76%, respectively.85 No radiologic modality, however, was able to distinguish simple steatosis from more advanced forms of NAFLD. Imaging studies such as ultrasonography may support the diagnosis of NAFLD but cannot predict the severity of disease and cannot replace liver biopsy for establishing the diagnosis with certainty.

HISTOPATHOLOGIC FEATURES

The major histologic features of NAFLD resemble those of alcohol-induced liver disease and include steatosis (fatty

1405

1406

Section IX  Liver

A

B

Figure 85-2.  Imaging studies of fatty liver. A, Ultrasound demonstrating increased echogenicity. B, T1-weighted magnetic resonance image demonstrating a bright liver. (Courtesy of Dr. Mukesh Harisinghani, Boston, Mass.)

Figure 85-3.  Histologic features of simple steatosis (fatty liver). The characteristic feature is diffuse macrovesicular steatosis. Glycogenated nuclei are common. (Hematoxylin and eosin.) (Courtesy of Dr. Gregory Y. Lauwers, Boston, Mass.)

liver), steatohepatitis (fatty liver plus parenchymal inflammation with or without accompanying focal necrosis), and varying degrees of fibrosis, including cirrhosis. Steatosis is predominantly macrovesicular and usually is distributed diffusely throughout the liver lobule, although prominent microvesicular steatosis and zone 3 (perivenular) steatosis have been reported occasionally (Fig. 85-3). Mild lymphocytic, neutrophilic, or mixed inflammatory infiltrates also may be observed, and glycogenated nuclei are common. NASH, which is an advanced form of NAFLD, is indistinguishable histologically from alcoholic hepatitis (Fig. 85-4 and Table 85-3). Steatosis is present in all cases and can affect the hepatic lobules either diffusely or primarily in the central zones. The degree of steatosis may correlate with the patient’s BMI and generally is more severe in NASH than in alcoholic hepatitis.86 Lobular inflammation is a hallmark feature of NASH and is characterized by infiltration of lymphocytes, other mononuclear cells, and polymorphonuclear neutrophils. The intensity of the inflammation varies with

Figure 85-4.  Histologic features of nonalcoholic steatohepatitis. Diffuse or perivenular macrovesicular steatosis is present. Lobular inflammation consists of neutrophils, lymphocytes, and other mononuclear cells. Hepatocyte ballooning and necrosis of varied degrees are hallmark features. Glycogenated nuclei are present. Mallory bodies, which may be small, sparse, and inconspicuous, are seen. (Hematoxylin and eosin.) (Courtesy of Dr. Gregory Y. Lauwers, Boston, Mass.)

the severity of steatohepatitis and may be milder in NASH than in alcoholic hepatitis.87 Glycogenated nuclei may be present. Hepatocyte ballooning and hepatocyte necrosis of varying degrees often are present and may portend a worse prognosis.88,89 Mallory (or Mallory-Denk) bodies, which may be small, sparse, and inconspicuous, are seen frequently. Mild stainable iron may be present in up to 50% of the patients. Pericellular, perisinusoidal, and periportal fibrosis has been described in 37% to 84% of patients with NASH. The extent of fibrosis varies considerably, ranging from delicate strands surrounding small veins or groups of cells to densely fibrotic septa with distortion of the hepatic architecture. Perisinusoidal fibrosis is most common, especially in adults, is initially mild, and predominates in zone 3 around the terminal hepatic veins.41 Cirrhosis is found on initial biopsy in 7% to 16% of patients with NAFLD and

Chapter 85  Nonalcoholic Fatty Liver Disease Table 85-3  Histologic Features of Nonalcoholic Fatty Liver Disease Present in All or Most Cases Macrovesicular steatosis Diffuse or centrilobular steatosis; degree may correlate with BMI Parenchymal inflammation Polymorphonuclear neutrophils, lymphocytes, other mononuclear cells Hepatocyte necrosis Ballooning hepatocyte degeneration Observed with Varied Frequencies Perivenular, perisinusoidal, or periportal fibrosis (37%-84%), moderate to severe in 15%-50%; most prevalent in zone 3 (perivenular) Cirrhosis (7%-16% on index biopsy specimen) Mallory bodies Glycogenated nuclei Lipogranulomas Stainable hepatic iron

Elevated serum aminotransferase levels and/or hepatomegaly

Exclude excessive alcohol use and other forms of liver disease by history and laboratory tests

Image liver with US, CT, or MRI

Normal

Fatty liver present

Liver biopsy

Consider liver biopsy to stage disease and define risk of progression

BMI, body mass index.

abnormal liver biochemical test levels.15,29 The risk of cirrhosis in the setting of NAFLD may be greatest in morbidly obese patients. In NAFLD-associated cirrhosis, the typical histologic features of NAFLD may be minimal or absent, potentially leading to the misdiagnosis of cryptogenic cirrhosis. Strict histologic criteria for NASH have not yet been defined. At least two scoring systems have been proposed.90,91 To reach consensus on the pathologic classification of NASH, the Pathology Committee of the National Institutes of Health NASH Clinical Research Network has proposed a scoring system incorporating 14 histologic features.90 The unweighted sum of scores for steatosis, lobular inflammation, and hepatocellular ballooning is used to construct an activity score, which can be used to classify cases into categories designated “not NASH,” “borderline NASH,” or “NASH” with reasonable inter-rater agreement. Thus far, however, no scoring system for NAFLD or NASH has been widely adopted.

DIAGNOSIS

Establishing a definitive diagnosis of NAFLD requires both clinical and histologic data (Fig. 85-5). Most patients with NAFLD are evaluated because of chronically elevated liver biochemical test levels, with or without hepatomegaly. The combination of the patient’s history, physical examination, blood test results, and radiologic findings is useful for excluding other causes of liver disease. Laboratory testing should include liver biochemical tests, complete blood count, prothrombin time, anti-HCV, hepatitis B surface antigen, iron indices, ceruloplasmin in persons younger than 40 years of age, α1-antitrypsin, and AMA. Imaging studies may support the diagnosis (see earlier), but the absence of characteristic findings does not preclude a diagnosis of NAFLD. To establish a diagnosis of NAFLD, alcoholic liver disease must be excluded. Clinical and histologic data unreliably differentiate NAFLD from alcoholic liver disease in ambulatory patients. Therefore, the diagnosis of NAFLD should be entertained only in the absence of significant alcohol use (consumption of less than 20 to 40 g of alcohol per day in most clinical studies).

The Role of Liver Biopsy

The role of liver biopsy in establishing the diagnosis of NAFLD has been debated. Many practitioners consider NAFLD a diagnosis of exclusion, when clinical and labora-

Figure 85-5.  Diagnostic approach to patients with suspected nonalcoholic fatty liver disease (NAFLD). The diagnosis of NAFLD is based on clinical and histologic criteria. Most patients are evaluated because of elevated serum aminotransferase levels and/or hepatomegaly. The diagnosis of NAFLD should be considered when excessive alcohol use is absent and laboratory test results exclude other causes of liver disease. Radiologic studies may demonstrate fatty liver. Liver biopsy is the standard means of diagnosis and the only test that can reliably differentiate simple steatosis from advanced NAFLD (i.e., nonalcoholic steatohepatitis), although noninvasive methods for assessing fibrosis are under study. CT, computed tomography; MRI, magnetic resonance imaging; US, ultrasonography.

tory examinations fail to reveal another cause of chronic liver disease. The diagnosis of NAFLD is also suggested when an imaging study provides evidence of fatty liver. Most patients with NAFLD do not undergo a liver biopsy, largely because the results usually will not affect management (therapeutic options are limited, see later). The argument for liver biopsy in all patients with presumed NAFLD is centered on several lines of evidence: The correlation among clinical, laboratory, and histologic findings in NAFLD is poor, and patients with normal liver biochemical test results can have significant liver injury on biopsy specimens31; liver biopsy is the only diagnostic test that can reliably quantify hepatic steatosis, necrosis, and fibrosis; and the histologic stage of NAFLD is the best prognostic indicator.92 In addition, understanding of the natural history and treatment of NAFLD have been hampered by the lack of histologic data in most clinical studies. The most prudent approach may be to select patients for whom liver biopsy might influence management. Several clinical and biochemical risk factors for progressive disease have been identified and may facilitate selection of patients for liver biopsy. Patients with obesity, longstanding or persistent liver biochemical test abnormalities (even with good glycemic control and after weight loss), older age, multiple com­ ponents of the metabolic syndrome, an AST/ALT ratio greater than 1, markedly elevated liver biochemical test levels, symptoms and signs of portal hypertension, or evidence of fibrosis on an imaging study are more likely to have advanced disease (Table 85-4).29,89 The results of a liver biopsy in these patients might lead to a more aggressive treatment strategy, participation in clinical trials, or

1407

1408

Section IX  Liver Table 85-4  Risk Factors for Advanced* Nonalcoholic Fatty Liver Disease Clinical Older age (>50 years) Obesity Diabetes mellitus/insulin resistance Hypertension Laboratory AST/ALT ratio > 1 Serum ALT level > twice the upper limit of normal Serum triglyceride levels > 155 mg/dL Histologic Severe steatosis Necroinflammatory activity (hepatocyte ballooning, necrosis) Stainable iron ALT, alanine aminotransferase; AST, aspartate aminotransferase. *Nonalcoholic steatohepatitis and advanced fibrosis.

screening for hepatocellular carcinoma in the setting of cirrhosis.

Noninvasive Markers of Fibrosis in NAFLD

Although percutaneous liver biopsy remains the standard for the diagnosis of NAFLD, it is costly, invasive, and associated with a small risk of complications. Sampling variability is common, and the large number of persons with NAFLD far outstrips the manpower available to perform liver biopsies. Significant progress has been made in developing simple, noninvasive, and quantitative tests to estimate the degree of hepatic fibrosis in a number of liver diseases, including NAFLD. The FibroTest (called FibroSure in the United States) is the best studied of these noninvasive tests (see Chapters 73 and 79). The panel of blood tests used to estimate hepatic fibrosis includes serum α2-macroglobulin, apolipoprotein A-1, haptoglobin, total bilirubin, and GGTP levels, and the necroinflammatory activity index combines the same five markers plus the serum ALT level. In a study of 167 patients with NAFLD, FibroTest was highly sensitive for detecting bridging fibrosis and cirrhosis.93 FibroTest cutoff value of .70 had a positive predictive value of 73% and a specificity of 98% for advanced fibrosis. A cutoff value of 0.30 had a negative predictive value of 90% for advanced fibrosis. Unfortunately, 33% of patients had a FibroTest score between 0.30 and 0.70, and in this range, the test is inaccurate for assessing the stage of fibrosis. Therefore, patients with a score in this range would need a liver biopsy for accurate staging. Angulo and colleagues developed and validated another noninvasive fibrosis scoring system called the NAFLD Fibrosis Score, which is derived from clinical and laboratory information that is obtained easily in the context of any clinical encounter.94 Using this scoring algorithm, which incorporates age, BMI, hyperglycemia, AST/ALT ratio, platelet count, and serum albumin level, the authors defined a low cutoff value with a negative predictive value of 88% to 93% and a high cutoff value with a positive predictive value of 82% to 90%. Only 25% to 28% of cases were indeterminate and would therefore require liver biopsy for accurate staging. Additional noninvasive tests for fibrosis have been evaluated with variable success in small studies of NAFLD, including transient elastography (Fibroscan), which uses ultrasound to quantify liver stiffness and estimate fibrosis,95 serum dehydroepiandrosterone levels,96 and serum hyaluronic acid levels.97 One or more noninvasive indices of fibrosis is likely to be validated in the future and may supplant the need for liver biopsy in many, but not all, patients with NAFLD.

NATURAL HISTORY

The natural history of NAFLD is largely unknown because no prospective, longitudinal studies with histologic followup have been conducted in patients with NAFLD. In addition, the long-term complications of NAFLD are probably under-recognized and under-reported, in part because the characteristic feature of macrovesicular steatosis may diminish or disappear in late-stage NAFLD. The available data from retrospective studies suggest that NAFLD is a benign disease in most patients. The prognosis in patients with steatosis in the absence of hepatocyte necrosis and fibrosis clearly is favorable, with little potential for histologic or clinical progression.98 In some patients, however, NAFLD can lead to cirrhosis, liver failure, or hepatocellular carcinoma. Between 2003 and 2008, five retrospective studies were published that included a total of 215 patients with NAFLD who underwent paired liver biopsies.6,99-101 Follow-up periods ranged from 1 to 14 years, and histologic progression of fibrosis was documented in 32% to 41% of the patients. NAFLD was stable histologically in 34% to 50% and appeared to improve in 16% to 29%. Substantial variability in the rate of progression of fibrosis was observed among patients; histologic progression did not signal clinical deterioration in most cases, and no clinical or laboratory data reliably predicted the course of liver disease. These data must be interpreted with caution because variability in histologic staging may be the result of sampling error.102 A large population-based study from Olmsted County, Minnesota, included 420 patients with definite or presumed NAFLD based on imaging studies or, less frequently, liver biopsy findings.103 Cirrhosis developed in 3% of the patients over a mean follow-up of 7.6 years. Liver disease was the third leading cause of death in this cohort, behind malignancy and ischemic heart disease. These limited data suggest that NAFLD is an indolent condition, with few clinical sequelae in most patients, but can progress to irreversible, clinically important liver disease over a relatively short period of time in a minority of affected persons. This conclusion is supported by data from a classic study of 132 patients with NAFLD evaluated over an 18-year period. The study compared clinical outcomes based on the degree of injury on an index liver biopsy specimen.104 Each biopsy specimen was assigned to a histologic subgroup—types 1 to 4—that represented progressively severe disease as determined by the cumulative occurrence of steatosis, inflammation, hepatocyte necrosis, and fibrosis. Cirrhosis and liver-related deaths were more common (25% and 11%, respectively) in patients with NAFLD types 3 and 4 (steatosis, inflammation, and necrosis, with or without fibrosis) than in patients with NAFLD types 1 and 2 (steatosis without necrosis), although the mortality difference failed to reach statistical significance. In addition, the liver-related mortality rate of 11% among patients with NAFLD types 3 and 4 was higher than the age-adjusted death rate from chronic liver disease/cirrhosis in the general U.S. population (9.5 per 100,000 per year). These data suggest that the risk of liver-related complications in NAFLD correlates, at least to some extent, with the degree of hepatocellular injury and fibrosis found on an index liver biopsy specimen. NAFLD and alcoholic hepatitis are similar histologically, but differ substantially in clinical outcomes (see Chapter 84). The five-year survival rate of patients with alcoholic hepatitis is only 50% to 75% because of the large proportion of patients (greater than 50%) in whom cirrhosis and its complications develop. A study has shown that the longterm survival of patients with NASH is significantly better than the long-term survival of patients with alcoholic hepatitis.86 In the minority of patients in whom NAFLD-

Chapter 85  Nonalcoholic Fatty Liver Disease associated cirrhosis develops, however, the outcome may be similar to that for other causes of cirrhosis. Strong circumstantial evidence suggests that NAFLD is the likely cause of many cases of cryptogenic cirrhosis105-107 and may be associated with the development of hepatocellular carcinoma.30,106,108 One study showed that the 5- to 10-year outcome of NAFLD-associated cirrhosis was similar to that for HCV-associated cirrhosis, although hepatocellular carcinoma was significantly less common in the patients with NAFLD.109 Large prospective studies are needed to define the natural history of NAFLD, but emerging evidence confirms that NAFLD can be progressive and associated with significant morbidity and mortality in some patients. The risks of liverrelated morbidity and mortality are greatest in persons with evidence of advanced NAFLD (steatohepatitis with necrosis and fibrosis) on the initial liver biopsy specimen.104 If clinical and biochemical risk factors for progressive disease can be established, a subset of patients can be identified in whom a liver biopsy will have the greatest prognostic and therapeutic value (see earlier). Several potential risk factors have been identified in different populations (see Table 85-4). Women are over-represented in studies of NAFLD, but whether gender is an independent risk factor for advanced disease is unclear. Older age, obesity, diabetes mellitus, and an AST/ALT ratio greater than 1 were demonstrated in one study to be significant predictors of severe fibrosis (bridging/cirrhosis) in patients with NAFLD.29 In another study of overweight patients with abnormalities on liver biochemical tests, liver fibrosis was independently associated with hepatic necroinflammatory activity, BMI greater than 28 kg/m2, age older than 50 years, serum triglyceride level higher than 1.7 mmol/L, and serum ALT more than twice normal.43 In another study of morbidly obese patients referred for bariatric surgery, systemic hypertension, an elevated serum ALT level, and a high insulin resistance index were highly predictive of advanced NAFLD.15 As noted previously, however, the full spectrum of liver damage, including cirrhosis, has been documented in nonobese patients with near-normal liver biochemical test results.

TREATMENT

The optimal therapy for NAFLD has not been established. To date, no large, randomized treatment trials demonstrating resolution of steatosis, inflammation, and fibrosis have been conducted in patients with NAFLD. Small numbers of patients, varying inclusion criteria, and varying end points have limited the clinical impact of published studies. Historically, the treatment of NAFLD has consisted of weight loss, removal of offending drugs and toxins, and control of associated metabolic disorders, including diabetes mellitus and hyperlipidemia. Several case reports and small studies of diet and exercise have shown improvements in biochemical, ultrasonographic, and in some cases, histologic abnormalities in children and adults with NASH.52,110-112 Intensive nutritional counseling may lead to sustained weight loss and significant histologic improvement in some patients.113 Several small, largely uncontrolled studies also showed improvements in liver biochemical test results, steatosis, and fibrosis in a few patients who achieved modest weight loss with orlistat, a reversible inhibitor of gastric and pan­ creatic lipases.114,115 Bariatric surgery leads to massive weight loss and improves insulin sensitivity in most patients, normalizes some of the metabolic abnormalities involved in the pathogenesis of NAFLD, decreases the hepatic expression of mediators of liver inflammation and fibrosis, and improves hepatic histology in patients with NAFLD.116-120

Table 85-5  Potential Therapies for Nonalcoholic Fatty Liver Disease Avoidance of toxins Discontinue potentially offending medications/toxins Minimize alcohol intake Exercise and diet Moderate, sustained exercise and weight loss in overweight patients Effects of specific diets are not known Antidiabetic/insulin-sensitizing agents Metformin Thiazolidinediones Lipid-lowering agents Gemfibrozil Statins Antioxidants Betaine N-acetylcysteine Superoxide dismutase Vitamin E Iron reduction by phlebotomy Inflammatory mediators by: Agents that affect increasing mitochondrial ATP stores and/or activity Agents that affect modulating leptin activity Agents that affect modulating TNF-α activity Agents that affect raising adiponectin levels Bariatric surgery for morbid obesity ATP, adenosine triphosphate; TNF-α, tumor necrosis factor-α.

No scientific data exist to support a particular commercial or medicinal diet plan for NAFLD, and no one correct dietary approach is likely to be suitable for all patients with NAFLD.121 Unfortunately, the therapeutic benefit of weight loss achieved with diet, medicinal aids, exercise, or surgery has not been examined in randomized, prospective studies with firm histologic end points. Until such studies are performed, a recommendation for moderate weight loss is reasonable in overweight patients with NAFLD, although sustained weight loss is seldom achieved. Rapid weight loss can exacerbate steatohepatitis in morbidly obese patients, especially after bariatric surgery122; therefore, the rate of weight loss and serial liver biochemical test results should be monitored carefully in patients on a weight reduction regimen. New therapeutic methods should capitalize on today’s improved understanding of the pathogenesis of NAFLD (Table 85-5).

Antioxidants

Medications that minimize oxidative stress may prove useful. Vitamin E, an inexpensive yet potent antioxidant, has been examined as an agent for treatment of NAFLD in several small pediatric and adult studies, with varying results.52,123,124 In all studies, vitamin E was well tolerated, and most studies showed modest improvements in serum aminotransferase levels, ultrasonographic appearance of the liver, and, infrequently, histologic findings. Randomized controlled studies with histologic inclusion criteria and end points are needed, however, to determine if vitamin E, either alone or in combination with other medications, leads to histologic improvement in NAFLD. In light of the potentially negative effects of vitamin E on cardiovascular health,125 caution should be exercised in treating NAFLD with vitamin E until better studies are available. Betaine, a metabolite of choline that raises SAM levels and decreases cellular oxidative damage, has shown promise in a small pilot study as a therapeutic agent for NASH.126 N-acetylcysteine, superoxide dismutase, and PPAR-α

1409

1410

Section IX  Liver agonists such as ragaglitazar also may hold therapeutic promise,64,127,128 although clinical studies in humans are lacking.

Insulin-Sensitizing Agents

The association between hyperinsulinemic insulin resistance and NAFLD provides a logical target for treatment. Metformin, a biguanide that reduces hyperinsulinemia and improves hepatic insulin sensitivity, reduces hepatomegaly and hepatic steatosis in ob/ob mice,129 but results in human studies have been less impressive.130,131 Thiazolidinediones (TZDs), potent PPAR-α agonists, also are being investigated as possible agents for the treatment of NAFLD. PPAR-α is a nuclear receptor expressed in adipose tissue, muscle, and liver. In adipocytes, PPAR-α promotes cell differentiation and decreases lipolysis and FFA release. TZDs improve insulin sensitivity and hyperinsulinemia by increasing glucose disposal in muscle and decreasing hepatic glucose output. Treatment with troglitazone, a first-generation TZD, was associated with biochemical and histologic improvements in patients with NASH, but troglitazone subsequently was withdrawn from the market because of rare but serious hepatotoxicity. Rosiglitazone and pioglitazone, TZDs with low rates of hepatotoxicity, have been investigated in separate 48-week, single-arm treatment trials in patients with histologically proven NASH.132,133 In both studies, treatment was well tolerated and was associated with improved insulin sensitivity, normalization of liver biochemistries, and histologic improvement in most patients. A drawback of both TZDs, however, was substantial weight gain (4.0% to 7.3%) and increased total body adiposity. In addition, the durability of biochemical and histologic improvements after completion of therapy was not examined in either study. A follow-up study has suggested, however, that the beneficial effects of pioglitazone diminish when the drug is discontinued.134 A placebo-controlled European trial of rosiglitazone in 63 patients with NAFLD showed improvement in steatosis and serum aminotransferase levels but not in necroinflammation or fibrosis.135 Rosiglitazone was also associated with significant weight gain in the TZD-treated patients in this study. TZDs must be assessed in large, placebocontrolled trials before they can be recommended for routine use in patients with NAFLD.

untreated control subjects, but histologic features were not assessed.139 Treatment of NASH with atorvastatin, a 3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitor, showed promise in a small pilot study,140 but statins have not been assessed in a large clinical trial. Ursodeoxycholic acid, a cytoprotective agent, showed promise in a pilot study of NASH but was not effective in a randomized placebo-controlled trial.141 The combination of ursodeoxycholic acid and vitamin E has shown some efficacy. Future therapies for NAFLD might include agents that increase adiponectin levels, neutralize TNF-α, improve mitochondrial ATP homeostasis, or alter leptin levels. When the pathogenesis of NAFLD is elucidated further, new therapeutic agents will likely be developed.

Liver Transplantation

Patients with NAFLD in whom end-stage liver disease develops should be evaluated for liver transplantation. The outcome of liver transplantation in these patients is good, although NAFLD can recur after liver transplantation.142,143 The risk factors for recurrent or de novo NAFLD after liver transplantation probably are multifactorial and include hypertriglyceridemia, obesity, diabetes mellitus, and glucocorticoid therapy.

FOCAL FATTY LIVER In contrast with NAFLD, which is a diffuse parenchymal process, focal fatty liver is a localized or patchy process that simulates a space-occupying lesion in the liver on imaging studies. This condition has been recognized increasingly in adults and children as a result of the improved sensitivity of abdominal imaging. Focal fatty liver has characteristic patterns on CT: usually a nonspherical shape, absence of mass effect, and CT attenuation values consistent with those of soft tissue.144 The density of focal fatty liver is close to that of water, unlike that of liver metastases, which have a density that is closer to that of hepatocytes. Ultrasonography and MRI can help confirm a diagnosis of focal fatty liver (Fig. 85-6). A presumptive diagnosis of focal fatty liver

Iron Reduction

High serum iron and ferritin levels have been identified in some patients with NAFLD, most of whom do not have genetic hemochromatosis or hepatic iron overload. Most investigators believe that increased serum iron indices are a by-product of hepatic inflammation, rather than a contributor to the pathogenesis of NAFLD, but a few small studies have suggested that iron depletion may have a therapeutic role in NAFLD by decreasing plasma insulin, glucose, and serum aminotransferase levels.136,137 The relationship between iron and insulin is complex, but the insulinsparing effect of iron depletion may be the result of enhanced skeletal muscle glucose transport and metabolism and increased hepatic extraction and metabolism of insulin.138 The primary limitation of these studies is the lack of histologic inclusion criteria and end points, but the results are intriguing and merit further investigation.

Lipid-Lowering and Cytoprotective Agents

The usefulness of lipid-lowering and cytoprotective drugs for the treatment of NAFLD has been assessed in a few small trials, with varying results. Treatment with gemfibrozil was associated with biochemical improvement in 74% of patients in the treatment group, compared with 30% of

Figure 85-6.  Focal fatty liver. Focal fatty liver (arrow) on computed tomography (CT) scan. The characteristic features are the nonspherical shape, absence of a mass effect, and CT attenuation values consistent with those of soft tissue. Ultrasound studies and magnetic resonance imaging also may confirm the diagnosis of focal fatty liver. (Courtesy of Dr. Mukesh Harisinghani, Boston, Mass.)

Chapter 85  Nonalcoholic Fatty Liver Disease should not be made when a mass effect, areas of mixed hypo- and hyperechogenicity, an irregular shape, or a history of malignancy is present. In such cases, ultrasonographically guided fine-needle biopsy is recommended. No evidence exists to suggest that the pathogenesis of focal fatty liver is similar to that of NAFLD. In fact, the pathogenesis of focal fatty liver is uncertain and may involve altered venous blood flow to the liver, tissue hypoxia, and intestinal malabsorption of lipoproteins. Furthermore, in the absence of accompanying or background liver disease, the lesion often regresses. Therefore, no specific treatment is justified.

KEY REFERENCES

Adams LA, Lymp JF, St. Sauver J, et al. The natural history of nonalcoholic fatty liver disease: A population-based cohort study. Gastroenterology 2005; 129:113-21. (Ref 103.) Adams LA, Sanderson S, Lindor KD, Angulo P. The histological course of nonalcoholic fatty liver disease: A longitudinal study of 103 patients with sequential liver biopsies. J Hepatol 2005; 42:132-8. (Ref 99.) Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: A noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatology 2007; 45:846-54. (Ref 94.) Day C, James O. Steatohepatitis: A tale of two “hits”? [editorial]. Gastroenterology 1998; 114:842-5. (Ref 36.)

Dixon J, Bhathal P, O’Brien P. Nonalcoholic fatty liver disease: Predictors of nonalcoholic steatohepatitis and liver fibrosis in the severely obese. Gastroenterology 2001; 121:91-100. (Ref 15.) El-Serag H, Tran T, Everhart J. Diabetes increases the risk of chronic liver disease and hepatocellular carcinoma. Gastroenterology 2004; 126:460-8. (Ref 30.) Farrell GC, Larter CZ. Nonalcoholic fatty liver disease: From steatosis to cirrhosis. Hepatology 2006; 43:S99-S112. (Ref 47.) Hui J, Hodge A, Farrell G, et al. Beyond insulin resistance in NASH: TNF-alpha or adiponectin? Hepatology 2004; 40:46-54. (Ref 51.) Lin H, Yang S, Chuckaree C, et al. Metformin reverses fatty liver disease in obese, leptin-deficient mice. Nature Med 2000; 6:998-1003. (Ref 129.) Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology 2003; 37:917-23. (Ref 28.) Matteoni C, Younossi Z, Gramlich T, et al. Nonalcoholic fatty liver disease: A spectrum of clinical and pathological severity. Gastroenterology 1999; 116:1413-19. (Ref 104.) Promrat K, Lutchman G, Uwaifo G, et al. A pilot study of pioglitazone treatment for nonalcoholic steatohepatitis. Hepatology 2004; 39:18896. (Ref 133.) Ratziu V, Giral P, Charlotte F, et al. Liver fibrosis in overweight patients. Gastroenterology 2000; 118:1117-23. (Ref 89.) Ratziu V, Giral P, Jacqueminet S, et al. Rosiglitazone for nonalcoholic steatohepatitis: One-year results of the randomized placebocontrolled Fatty Liver Improvement with Rosiglitazone Therapy (FLIRT) Trial. Gastroenterology 2008; 135:100-10. (Ref 135.) Full references for this chapter can be found on www.expertconsult.com.

1411

CHAPTE R

86 Liver Disease Caused by Drugs Narci C. Teoh, Shivakumar Chitturi, and Geoffrey C. Farrell

CHAPTER OUTLINE Definitions and Importance  1413 Epidemiology  1414 Case Definition: Which Agent?  1414 Frequencies of Hepatic Drug Reactions  1414 Importance of Drugs as a Cause of Liver Disease  1415 Risk Factors  1415 Pathophysiology  1417 Role of the Liver in Drug Elimination  1417 Pathways of Drug Metabolism  1417 Toxic Mechanisms of Liver Injury  1419 Immunologic Mechanisms  1422 Clinicopathologic Features of Drug-Induced Liver Disease  1422 Classification  1422 Histopathologic Features  1424 Clinical Features  1424 A Practical Approach to Diagnosis  1424 Physician Awareness  1424 Exclusion of Other Disorders  1425 Extrahepatic Features  1425 Chronologic Relationships  1425 Which Drug?  1425 Indications for Liver Biopsy  1425 Considerations in Patients with Viral Hepatitis  1425 Prevention and Management  1426 Dose-Dependent Hepatotoxicity  1427

DEFINITIONS AND IMPORTANCE Drugs are a relatively common cause of liver injury, which usually is defined by abnormalities of liver biochemical test levels, particularly an increase in the serum alanine aminotransferase (ALT), alkaline phosphatase, or bilirubin level, to more than twice the upper limit of normal. Drug-induced liver injury can be difficult to define in clinical practice because the biochemical tests used to detect liver injury may also be elevated as part of an adaptive response to drugs. Further, the severity of drug-induced liver injury varies from minor nonspecific changes in hepatic structure and function to acute liver failure, cirrhosis, and liver cancer. The term drug-induced liver disease should be confined to cases in which the nature of liver injury has been characterized histologically. With the exception of acetaminophen, anticancer drugs, and some botanical or industrial hepatotoxins (see Chapter 87), most cases of drug-induced

Acetaminophen  1427 Other Types of Cytopathic Liver Injury  1429 Drug-Induced Acute Hepatitis  1431 Immunoallergic Reactions  1431 Metabolic Idiosyncrasy  1434 Drug-Induced Granulomatous Hepatitis  1437 Drug-Induced Chronic Hepatitis  1438 Diclofenac  1439 Minocycline  1439 Drug-Induced Acute Cholestasis  1439 Importance, Types of Reactions, and Diagnosis  1439 Cholestasis without Hepatitis  1439 Cholestasis with Hepatitis  1440 Cholestatic Hepatitis with Bile Duct Injury  1441 Drug-Induced Chronic Cholestasis  1441 Flucloxacillin  1441 Fibrotic Bile Duct Strictures  1441 Drug-Induced Steatohepatitis and Hepatic Fibrosis  1441 Amiodarone  1442 Tamoxifen and Other Causes of Drug-Induced Steatohepatitis  1442 Methotrexate  1443 Drug-Induced Vascular Toxicity  1444 Azathioprine  1445 Liver Tumors  1445

liver disease represent adverse drug reactions or hepatic drug reactions. These effects are noxious and unintentional and occur at doses recommended for prophylaxis or therapy. The latent period is longer (typically one week to three or six months) than that for direct hepatotoxins (hours to a few days), and extrahepatic features of drug hypersensitivity may be present. Although drug-induced liver disease is a relatively uncommon cause of jaundice or acute hepatitis in the community, it is an important cause of more severe types of acute liver disease, particularly among older people (see Epidemiology). The overall mortality rate among patients hospitalized for drug-induced liver injury is approximately 10%1 but varies greatly for individual drugs.2,3 The reported frequencies of individual hepatic drug reactions are often underestimated because of the inadequacy of spontaneous reporting by physicians and pharmacists.2,3 With more reliable prospective and epidemiologic techniques, the frequency (or risk) of most types of drug-induced liver disease is between one per 10,000 and one per 100,000 persons

1413

1414

Section IX  Liver exposed.4 Because these responses to drug exposure are clearly rare and unpredictable, they are often termed idiosyncratic drug reactions. Their rarity blunts diagnostic acumen because most clinicians will see few, if any, cases and therefore do not have an appropriate level of clinical suspicion. This concern applies especially to comple­ mentary and alternative medicine (CAM), as discussed in Chapter 87. Failure to withdraw the causative agent after the onset of symptoms of drug hepatitis or inadvertent reexposure to such a drug is a common and avoidable factor in acute liver failure attributable to drug-induced liver injury.5-8 Another challenge is that hepatic drug reactions produce an array of clinical syndromes and pathologic findings that mimic known hepatobiliary diseases. Furthermore, although individual agents (and some drug classes) typically produce a characteristic “signature syndrome,” they can also be associated with other and sometimes multiple clinicopathological syndromes. Drug-induced liver injury is the most common reason for withdrawal of an approved drug from the market. The subject therefore has medicoeconomic, legal, and regulatory ramifications. Because of the low frequency of most types of idiosyncratic drug reactions that involve the liver, serious hepatotoxicity is not usually detected until post-marketing surveillance is conducted. Historically, drugs that have developed a reputation for potential hepatotoxicity usually have been replaced by more acceptable alternatives. Examples include troglitazone, the prototypic thiazolidine­ dione and bromfenac, a nonsteroidal anti-inflammatory drug (NSAID), both of which were withdrawn from the market because of several cases of fatal acute liver failure.5-9 The burgeoning number of available conventional medications and CAM preparations now includes many hundreds that can be cited as rare causes of drug-induced liver disease. This increasing number of potentially causative agents poses several challenges to clinicians,5-10 including concern about what constitutes an adequate level of patient information at the time a drug is prescribed and the reliability of evidence linking an individual agent to a particular type of liver injury.5,11-13 Another development is the appreciation that drug toxicity, in the context of complex medical situations, can interact with other causes of liver injury. Noteworthy examples of such situations are bone marrow transplantation; cancer chemotherapy; highly active antiretroviral therapy (HAART) for human immunodeficiency virus (HIV) infection and the acquired immunodeficiency syndrome (AIDS); use of antituberculosis drugs in patients with chronic viral hepatitis; rifampin hepatitis in patients with primary biliary cirrhosis (Chapter 89); nonalcoholic fatty liver disease (NAFLD)— particularly nonalcoholic steatohepatitis (NASH)—preci­ pitated by tamoxifen; and possibly other drugs in overweight persons with type 2 diabetes mellitus and the metabolic syndrome.

EPIDEMIOLOGY Frequency or risk—the number of adverse reactions for a given number of persons exposed—is the best term for expressing how common a drug reaction is. Timedependent terms such as incidence and prevalence are not appropriate for drug reactions because the frequency is not linearly related to the duration of exposure. For most reactions, the onset occurs within a relatively short exposure time, or latent period, although some rarer types of

chronic liver disease occur after many months or years. The frequency of drug-induced liver disease is usually based on the reported rate of drug reactions; such reports are usually a voluntary part of post-marketing surveillance and are submitted to pharmaceutical companies or adverse drug reaction monitoring bodies. In the United States, following approval by the U.S. Food and Drug Administration (FDA), pharmaceutical companies are required to report serious adverse events (any incident resulting in death, a threat to life, hospitalization, or permanent disability [Code of Federal Regulations]). Surveillance becomes a more passive process, however, when a drug is approved for marketing, and physicians and pharmacists are encouraged to file voluntary written reports through the MediWatch program. Similar systems operate in most industrialized countries. Nevertheless, MediWatch receives reports for fewer than 10% of adverse drug reactions,2 and in France fewer than 6% of hepatic adverse drug reactions are reported.3 The situation may be somewhat better in Sweden, but the annual reported incidence of adverse drug reactions of 2.2 per 100,000 in the population over the age of 15 is still much lower than the predicted incidence of 14 per 100,000.3 A prospective surveillance study in Spain measured the annual incidence of drug-related acute serious liver disease as 7.4 per million inhabitants.4

CASE DEFINITION: WHICH AGENT?

At least 300 agents have been implicated in drug-induced liver injury.10 The evidence for most drugs, however, is confined to individual or small numbers of case reports, especially in letters to scientific journals or to regulatory authorities, or small observational series. Therefore, for most agents, the evidence that they could cause liver injury is circumstantial and incomplete. Reports often lack pathologic definition, full exclusion of other disorders (for older reports), and logistic imputation of causality, especially with respect to temporal associations (see Diagnosis).5,9,10 Overall, probably fewer than 50 agents have been implicated reliably as causes of drug-induced liver disease. In general, agents used most commonly in clinical practice and in the community, including antimicrobials, antineoplastic agents, and NSAIDs, are those that have been implicated most often in drug-induced liver injury in larger series. The challenge of identifying the culprit drug among multiple candidates is discussed later.1,4,5,6,11

FREQUENCIES OF HEPATIC DRUG REACTIONS

Because of incomplete reporting, frequencies of hepatic drug reactions may often be underestimated. These estimated frequencies are also crude indicators of risk because of the inherent inaccuracies of case definitions (see Diagnosis)5,9,10 and because case recognition and reporting depend on the skill and motivation of observers. The increased interest of prescribers when initial cases of drug-induced liver disease have been described, together with inappropriate prescribing (e.g., prolonged use of bromfenac, which was approved only for seven days of use, and overprescribing of flucloxacillin and amoxicillin-clavulanic acid in some countries) can give rise to apparent “mini-epidemics.” More appropriate epidemiologic methods applied to hepatotoxicity have included prescription event monitoring, record linkage, and case-control studies. Prescription event monitoring and record linkage have been used to estimate the frequency of liver injury with some antimicrobials (erythromycins, sulfonamides, tetracyclines, flucloxacillin, amoxicillin-clavulanate) and NSAIDs.11 Epidemiologic studies confirm the rarity of drug-induced liver disease with currently used agents. For NSAIDs, the

Chapter 86  Liver Disease Caused by Drugs risk of liver injury is between 1 and 10 per 100,000 individuals exposed1,4,5,6; amoxicillin-clavulanic acid has been associated with cholestatic hepatitis in 1 to 2 per 100,000 exposed persons1,4,5,6,8; and low-dose tetracyclines have caused hepatotoxicity in less than one case per million persons exposed.1,4-6 The frequency of liver injury may be higher for agents that exert a metabolic type of hepatotoxicity. For example, isoniazid causes liver injury in up to 2% of persons exposed; the risk depends on the patient’s age and gender, concomitant exposure to other agents, and presence of hepatitis B virus (HBV) and possibly hepatitis C virus (HCV) infections.12 For some drugs in which other host factors play an etiopathogenic role, casecontrol studies have been used to define attributable risk. Examples include the implication of aspirin in Reye’s syndrome and oral contraceptives in liver tumors and hepatic vein thrombosis. A relationship may exist between the frequency and severity of serum ALT elevations that indicate liver injury and the risk of severe hepatotoxicity. This relationship was proposed in the 1970s by the late Hyman Zimmerman.6 According to “Hy’s rule,” elevations of serum ALT levels to eight-fold or more above the upper limit of normal or associated increases in the serum bilirubin concentration indicate a potential for the drug to cause acute liver failure at a rate of about 10% of the number of cases of jaundice. Therefore, if two cases of jaundice associated with druginduced liver injury are observed in a total phase 3 clinical trial experience of 2500 patients, approximately one case of acute liver failure would be expected for every 12,500 subjects who were prescribed the drug during the marketing phase.

IMPORTANCE OF DRUGS AS A CAUSE OF LIVER DISEASE

Hepatotoxicity accounts for less than 5% of cases of jaundice or acute hepatitis in the community and for even fewer cases of chronic liver disease5,6; however, drugs are an important cause of more severe types of liver disease and for liver disease in older people. They account for 10% of cases of severe hepatitis admitted to the hospital in France6 and for 43% of cases of hepatitis among patients 50 years of age or older.7 Drugs account for more than half of the cases of acute liver failure referred to special units in the United States7 and between 20% and 75% of cases of acute liver failure in other industrialized countries.4,7 The pattern of agents incriminated varies among countries; for example, herbal medicines are a relatively more common cause in Asian countries than in other countries (Chapter 87). In most cases of drug-related liver injury, drugs are the sole cause of hepatic damage. In other cases, drugs increase the relative risk for types of liver disease that may occur in the absence of drug exposure. Examples include salicylates in Reye’s syndrome, oral contraceptive steroids in hepatic venous thrombosis, methotrexate in hepatic fibrosis associated with alcoholic liver disease and NAFLD, and tamoxifen in NASH. Predisposition of patients with preexisting liver disease to drug-induced injury is minimal, but some interesting potential interactions between chronic HCV infection and several groups of drugs and between chronic HBV infection and antituberculosis chemotherapy are now reasonably established, as discussed later. On the other hand, liver failure may be more likely to develop if the patient with a hepatic drug reaction (e.g., to amoxicillin-clavulanic acid), which usually is associated with a good outcome, has underlying chronic liver disease.

RISK FACTORS

For dose-dependent hepatotoxins such as acetaminophen and methotrexate and for some idiosyncratic reactions that are partly dependent on dose (e.g., bromfenac, tetracyclines, dantrolene, tacrine, oxypenicillins), the factors that influence the risk of drug-induced liver disease include the dose of the drug, blood level of the drug, and duration of intake. For idiosyncratic reactions, however, host determinants appear to be central to liver injury. The most critical determinant is likely to be genetic predisposition, but other “constitutional” and environmental factors can influence the risk of liver injury, as summarized in Table 86-1. The most important factors are age,4 gender, exposure to other substances, a history or family history of previous drug reactions, other risk factors for liver disease, and concomitant medical disorders.

Genetic Factors

Genetic determinants predispose to drug-induced liver disease,13 as they do for other types of drug reaction, such as penicillin allergy. Atopic patients have been thought to have an increased risk of some types of drug hepatitis, but this increase in risk has not been proved. Genetic factors determine the activity of drug-activating and antioxidant pathways, encode pathways of canalicular bile secretion, and modulate the immune response, tissue stress responses, and cell death pathways (see Chapter 72). Documented examples of drugs associated with a familial predisposition to adverse hepatic drug reactions are few and include valproic acid and phenytoin.5,8,13 Inherited mitochondrial diseases are a risk factor for valproic acid–induced hepatotoxicity.14 Some forms of drug-induced liver disease, particularly drug-induced hepatitis and granulomatous reactions, can be associated with the reactive metabolite syndrome (see later). Weak associations have been reported between specific human leukocyte antigen (HLA) haplotypes and some types of drug-induced liver disease. Andrade and colleagues13 found positive associations between the class II HLA haplotype and cholestatic or mixed liver damage for some drugs. They suggested that no specific HLA allele predisposed to the overall risk of drug-induced liver disease but that the pattern of liver injury could be influenced by these genetic determinants. Other investigators have found stronger associations between the HLA haplotype and cholestatic reactions to amoxicillin-clavulanic acid and ticlopidine (see later).13,15

Age

Most hepatic drug reactions are more common in adults than in children. Exceptions include valproic acid hepatotoxicity, which is most common in infants younger than three years of age and rare in adults, and Reye’s syndrome, in which salicylates play a key role.16,17 As discussed later, both may be examples of mitochondrial toxicity.14 In adults, the risk of isoniazid-associated hepatotoxicity is greater in persons older than 40 years of age. Similar observations have been made for nitrofurantoin, halothane, etretinate, diclofenac, and troglitazone.1,4,5,10 The increased frequency of adverse drug reactions in older subjects is largely the result of increased exposure, the use of multiple agents, and altered drug disposition.1,4,5,10 In addition, clinical severity of hepatotoxicity increases strikingly with age, as exemplified by fatal reactions to isoniazid and halothane.

Gender

Women are particularly predisposed to drug-induced hepatitis, a difference that cannot be attributed simply to

1415

1416

Section IX  Liver Table 86-1  Factors Influencing the Risk of Liver Diseases Caused by Drugs FACTOR

EXAMPLES OF DRUGS AFFECTED

INFLUENCE

Age

Isoniazid, nitrofurantoin, halothane, troglitazone Valproic acid, salicylates

Age >60 years: increased frequency, increased severity More common in children

Gender

Halothane, minocycline, nitrofurantoin

More common in women, especially with chronic hepatitis More common in men

Amoxicillin-clavulanic acid, azathioprine Dose

Acetaminophen, aspirin; some herbal medicines Tetracycline, tacrine, oxypenicillins Methotrexate, vitamin A

Blood levels are directly related to the risk of hepatotoxicity Idiosyncratic reactions, but partial relationship to dose Total dose, dosing frequency, duration of exposure is related to the risk of hepatic fibrosis

Genetic factors

Halothane, phenytoin, sulfonamides Amoxicillin-clavulanic acid Valproic acid

Multiple cases in families Strong HLA association Familial cases, association with mitochondrial enzyme deficiencies

History of other drug reactions

Isoflurane, halothane, enflurane Erythromycins Diclofenac, ibuprofen, tiaprofenic acid Sulfonamides, COX-2 inhibitors

Instances of cross-sensitivity have been reported among members of each drug class but are rare

Other drugs

Acetaminophen Valproic acid Anticancer drugs

Isoniazid, zidovudine, and phenytoin lower dose threshold and increase severity of hepatotoxicity Other antiepileptics increase risk of hepatotoxicity Interactive vascular toxicity

Acetaminophen hepatotoxicity Isoniazid, methotrexate

Lowered dose threshold, poorer outcome Increased risk of liver injury, hepatic fibrosis

Halothane, troglitazone, tamoxifen, methotrexate Acetaminophen

Increased risk of liver injury; hepatic fibrosis Increased risk of hepatotoxicity

Hycanthone, pemoline Antituberculosis drugs, ibuprofen

Increased risk of liver injury Increased risk of liver injury with chronic hepatitis B and C

Methotrexate Sulfonamides Tetracycline, methotrexate Azathioprine, thioguanine, busulfan

Increased Increased Increased Increased

Excessive alcohol use Nutritional status:   Obesity   Fasting Preexisting liver disease

Other diseases/conditions:   Diabetes mellitus   HIV infection/AIDS   Renal failure   Organ transplantation

risk of hepatic fibrosis risk of hypersensitivity risk of liver injury, hepatic fibrosis risk of vascular toxicity

AIDS, acquired immunodeficiency syndrome; COX-2, cyclooxygenase-2; HIV, human immunodeficiency virus; HLA, human leukocyte antigen.

increased exposure. Examples include toxicity caused by halothane, nitrofurantoin, sulfonamides, flucloxacillin, minocycline, and troglitazone.5,6 Drug-induced chronic hepatitis caused by nitrofurantoin, diclofenac, or minocycline has an even more pronounced female preponderance.6 Conversely, equal sex frequency or even male preponderance is common for some drug reactions characterized by cholestasis, as for amoxicillin-clavulanic acid. Azathioprine-induced liver disease is more likely to develop in male renal transplant recipients than in female recipients.18

Concomitant Exposure to Other Agents

Patients who are taking multiple drugs are more likely to experience an adverse reaction than those who are taking one agent.5,9,10 The mechanisms include enhanced cytochrome P450 (CYP)-mediated metabolism of the second drug to a toxic intermediate (see later). Examples discussed later include toxicity caused by acetaminophen, isoniazid, valproic acid, other anticonvulsants, and anticancer drugs. Alternatively, drugs may alter the disposition of other agents by reducing bile flow or competing with canalicular pathways for biliary excretion (phase 3 drug elimination) (see later). This mechanism may account for apparent interactions between oral contraceptive steroids and other drugs to produce cholestasis. Drugs or their metabolites may also interact in mechanisms of cellular toxicity and cell death that involve mitochondrial injury, intracellular signaling

pathways, activation of transcription factors, and regulation of hepatic genes involved in controlling the response to stress and injury that triggers pro-inflammatory and cell death processes.19,20

Previous Drug Reactions

A history of an adverse drug reaction generally increases the risk of reactions to the same drug and also to some other agents (see later). Nevertheless, instances of cross-sensitivity to related agents in cases of drug-induced liver disease are surprisingly uncommon. Examples of cross-sensitivity between drugs (or drug classes) include the haloalkane anesthetics (see Chapter 87), erythromycins, phenothiazines and tricyclic antidepressants, isoniazid and pyrazinamide, sulfonamides and other sulfur-containing compounds (e.g., some clyclooxygenase-2 [COX-2] inhibitors), and some NSAIDs. A crucial point is that a previous reaction to the same drug is a major risk factor for an increase in the severity of drug-induced liver injury.6

Alcohol

Chronic excessive alcohol ingestion decreases the dose threshold for, and enhances the severity of, acetaminophen-induced hepatotoxicity and increases the risk and severity of isoniazid hepatitis, niacin (nicotinic acid, nicotinamide) hepatotoxicity, and methotrexate-induced hepatic fibrosis.

Chapter 86  Liver Disease Caused by Drugs Nutritional Status

Obesity is strongly associated with the risk of halothane hepatitis (see Chapter 87) and appears to be an independent risk factor for NASH and hepatic fibrosis in persons taking methotrexate or tamoxifen. Fasting also predisposes to acetaminophen hepatotoxicity,21 and a role for undernutrition has been proposed in isoniazid hepatotoxicity.22

Preexisting Liver Disease

In general, liver diseases such as alcoholic cirrhosis and cholestasis do not predispose to adverse hepatic reactions. Exceptions include toxicity to some anticancer drugs, niacin, pemoline, and hycanthone. Preexisting liver disease is a critical determinant of methotrexate-induced hepatic fibrosis (discussed later). Patients with chronic HBV infection12 and possibly those with chronic HCV infection or HIV/AIDS appear to be at heightened risk of liver injury during anti-tuberculosis or HAART therapy,23 after exposure to ibuprofen and possibly other NSAIDs, after myeloablative therapy in preparation for bone marrow transplantation (resulting in sinusoidal obstruction syndrome [see later]),24 and possibly after taking antiandrogens, such as flutamide and cyproterone acetate.25 A particularly strong association has been observed between HCV infection (present in 33% of patients with HIV/AIDS) and the risk of liver injury during HAART; the risk may be increased 2- to 10-fold.26-30

Other Diseases

Rheumatoid arthritis appears to increase the risk of salicylate hepatotoxicity, and a curious, unexplained observation is that hepatitis associated with sulfasalazine appears to be more common in patients with rheumatoid arthritis than in those with inflammatory bowel disease.8-10,31,32 Diabetes mellitus, obesity, and chronic kidney disease predispose to methotrexate-induced hepatic fibrosis, whereas HIV/AIDS confers a heightened risk of sulfonamide hypersensitivity.31-33 A retrospective cohort study of five health maintenance organizations found that the age- and sex-standardized incidence of drug-induced acute liver failure in patients with diabetes mellitus was 0.08 to 0.15 per 1000 personyears, irrespective of the therapeutic agent used (the number using troglitazone was small); the incidence was highest (approximately 0.3 per 1000) during the first six months of exposure.32 Renal transplantation is a risk factor for azathioprine-associated vascular injury, whereas kidney disease predisposes to tetracycline-induced fatty liver.6 Finally, sinusoidal obstruction syndrome induced by anticancer drugs is more common after bone marrow transplantation24 and in persons with HCV infection.5,6,8-10,26

PATHOPHYSIOLOGY ROLE OF THE LIVER IN DRUG ELIMINATION

By virtue of the portal circulation, the liver is highly exposed to drugs and other toxins absorbed from the gastrointestinal tract. Most drugs tend to be lipophilic compounds that are readily taken up by the liver but that cannot be easily excreted unchanged in bile or urine. The liver is well equipped to handle these agents by an adaptable (inducible) series of metabolic pathways. These pathways include those that alter the parent molecule (phase 1); synthesize conjugates of the drug or its metabolite with a more water-soluble moiety, such as a sugar, amino acid, or sulfate molecule (phase 2); and excrete in an energy-dependent manner the parent molecule, its metabolites, or conjugates into bile

(phase 3). For any given compound, one, two, or all three of these steps may participate in drug elimination. Expression and subcellular location of the proteins (enzymes, membrane transporters) that mediate these steps are controlled by a set of nuclear receptors that function as transcriptional regulators and coregulators, thereby accounting for coordinated regulation among the three phases of hepatic drug elimination.

PATHWAYS OF DRUG METABOLISM

As reviewed elsewhere,5,34 phase 1 pathways of drug metabolism include oxidation, reduction, and hydrolytic reactions. The products can be readily conjugated or excreted without further modification.

Cytochrome P450

Most type 1 reactions are catalyzed by microsomal drug oxidases, the key component of which is a hemoprotein of the CYP gene superfamily. The apparent promiscuity of drug oxidases toward drugs, environmental toxins, steroid hormones, lipids, and bile acids results from the existence of multiple closely related CYP proteins. More than 20 CYP enzymes are present in the human liver.34 The reaction cycle involves binding of molecular oxygen to the iron in the heme prosthetic group, with subsequent reduction of oxygen by acceptance of an electron from nicotinamide-adenine dinucleotide phosphate (NADPH) cytochrome P450 reductase, a flavoprotein reductase. The resulting “activated oxygen” is incorporated into the drug or another lipophilic compound. Reduction of oxygen and insertion into a drug substrate (“mixed function oxidation”) can result in formation of chemically reactive intermediates, including free radicals, electrophilic “oxy-intermediates” (e.g., unstable epoxides, quinone imines), and reduced (and therefore reactive) oxygen species (ROS). The quintessential example is the CYP2E1-catalyzed metabolite of acetaminophen, N-acetyl-p-benzoquinone imine (NAPQI), an oxidizing and arylating metabolite that is responsible for liver injury associated with acetaminophen hepatotoxicity. Other quinone compounds are potential reactive metabolites of troglitazone, quinine, and methyldopa. Epoxide metabolites of diterpenoids may be hepatotoxic products of the hepatic metabolism of some plant toxins (see Chapter 87).35 ROS have broad significance in the production of tissue injury, particularly by contributing to the production of oxidative stress and triggering tissue stress responses and cell death pathways, as discussed later. The hepatic content of CYP proteins is higher in acinar zone 3 (see Chapter 71). Localization of CYP2E1 is usually confined to a narrow rim of hepatocytes 1 to 2 cells thick around the terminal hepatic venule. This finding explains in part the zonality of hepatic lesions produced by drugs and toxins, such as acetaminophen and carbon tetrachloride, which are converted to reactive metabolites.

Genetic and Environmental Determinants of Cytochrome P450 Enzymes

Pharmacogenetics and Polymorphisms of Cytochrome P450 Expression The hepatic expression of each CYP enzyme is genetically determined. This finding largely explains the four-fold or greater differences in rates of drug metabolism among healthy subjects. Some CYPs, particularly minor forms, are also subject to polymorphic inheritance; therefore, occasional persons completely lack the encoded protein.34 One example is CYP2D6, the enzyme responsible for the metabolism of debrisoquine and perhexiline. Poor metabolizers

1417

1418

Section IX  Liver lack CYP2D6 and accumulate perhexiline when given usual doses; lack of CYP2D6 is the critical determinant in serious adverse effects of perhexiline, including chronic hepatitis and cirrhosis.36 Other examples include CYPs 2C9 and 2C19, which affect the metabolism of S-warfarin, omeprazole, tolbutamide, and phenytoin and of S-mephenytoin, respectively34; 3% of white populations and 15% of Asians are poor metabolizers of S-mephenytoin. Developmental Regulation and Constitutive Expression Expression of several CYPs is developmentally regulated. During adult life, the expression of some CYPs declines slightly (by up to 10%) with advancing age, but this change is minor compared with the effects of genetic variation, environmental influences, and liver disease. Differences in the expression of CYPs 3A4 and 2E1 between men and women may explain the slighly enhanced metabolism of certain drugs (erythromycin, chlordiazepoxide, midazolam) in women, but whether this difference contributes to the increased risk of hepatic drug reactions in women remains unclear. Nutrition and Disease-Related Changes A person’s nutritional status influences the expression of certain CYPs, both in health and with liver disease.5,10,20,34 Expression of CYP2E1 is increased by obesity, high fat intake, and fasting.20,34 Diseases that alter the expression of hepatic CYPs include diabetes mellitus (increased CYP2E1), hypothyroidism (decreased CYP1A), and hypopituitarism (decreased CYP3A4).34 Cirrhosis is associated with decreased levels of total cytochrome P450 and also with reduced hepatic perfusion; the result is a decrease in the clearance of drugs such as propranolol that are metabolized rapidly by the liver.34 The effects of cirrhosis vary, however, among individual CYP families (e.g., CYP1A levels are lowered, but CYP2C and CYP2D6 levels often are preserved) and with the type of liver disease (e.g., CYP3A4 levels are preserved with cholestatic liver disease but lowered with hepatocellular liver disease). Adaptive Response and Enzyme Induction Exposure to lipophilic substances results in an adaptive response that usually involves synthesis of new enzyme protein, a process termed enzyme induction. The molecular basis for genetic regulation of constitutive and inducible expression of the major human hepatic cytochrome P450, CYP3A4, has been determined.37 Agents such as rifampin interact with the pregnane X-receptor (PXR), a member of the orphan nuclear receptor family of transcriptional regulators.37 Activated PXR and the analogous constitutive androstane receptor (CAR) in turn bind to cognate nucleotide sequences upstream to the CYP3A4 structural gene within a “xenobiotic-regulatory enhancer module” (XREM). This interaction regulates the CYP3A4 promoter downstream and ultimately the transcription of CYP3A4 protein. Similar control mechanisms apply to several other CYP pathways,37,38 particularly those involved with bile acid synthesis in which the nuclear receptors implicated include the farnesoid X-receptor (FXR), which down-regulates bile acid synthesis and up-regulates bile salt excretory pathways, and liver X receptor, a positive regulator of bile acid synthesis via CYP7A (see also Chapter 64).37 Common examples of the induction of microsomal enzymes by environmental compounds include the effect of smoking cigarettes and cannabis on CYP1A238 and of alcohol on CYP2E1 and possibly CYP3A4.39 Several drugs are potent inducers of CYP enzymes. Isoniazid induces CYP2E1, whereas phenobarbital and phenytoin increase the expression of multiple CYPs.34 Rifampin is a potent inducer of

CYP3A4, as is hypericum,40 the active ingredient of St John’s wort, a commonly used herbal medicine, thereby causing interactions between conventional medicines and a CAM preparation. Further descriptions of the regulation of hepatic drug metabolizing enzymes have been published elsewhere.34,38 The implications for drug-induced liver disease are twofold. First, enzyme induction often involves more than the CYP system, possibly because of activation of PXR and CAR; this observation could account for increases in serum levels of alkaline phosphatase and gamma glutamyl transpeptidase (GGTP), which reflect “hepatic adaptation” to chronic drug ingestion. Second, the influence of one drug on expression and activity of drug metabolizing enzymes and drug elimination (phase 3) pathways can alter the metabolism or disposition of other agents. Such drug-drug interactions are important pharmacologically and may be relevant to mechanisms of drug-induced liver injury. Inhibition of Drug Metabolism Some chemicals inhibit drug metabolism. In persons taking more than one medication, for example, competition for phase 2 pathways such as glucuronidation and sulfation facilitates the presentation of unconjugated drug to the CYP system. This mechanism appears to explain in part why agents such as zidovudine and phenytoin lower the dose threshold for acetaminophen-induced hepatotoxicity.

Other Pathways of Drug Oxidation

In addition to CYP enzymes, electron transport systems of mitochondria may lead to the generation of tissue-damaging reactive intermediates during the metabolism of some drugs. Examples of such reactive intermediates include nitroradicals from nitrofuran derivatives (nitrofurantoin, cocaine). Subsequent electron transfer by flavoprotein reductases into molecular oxygen generates superoxide and other ROS. Some anticancer drugs such as doxorubicin and the imidazole antimicrobial agents can participate in other oxidation-reduction (redox) cycling reactions that generate ROS.

Phase 2 (Conjugation) Reactions

Phase 2 reactions involve formation of ester links to the parent compound or a drug metabolite. The responsible enzymes include glucuronosyl (or glucuronyl) transferases, sulfotrans­ferases, glutathione-S transferases, and acetyl and amino acid N-transferases. The resulting conjugates are highly water soluble and can be excreted readily in bile or urine. Conjugation reactions can be retarded by depletion of their rate-limiting cofactors, such as glucuronic acid and inorganic sulfate, and the relatively low capacity of these enzyme systems restricts the efficacy of drug elimination when substrate concentrations exceed enzyme saturation. In general, drug conjugates are nontoxic, and phase 2 reactions are considered to be detoxification reactions, with exceptions. For example, some glutathione conjugates can undergo cysteine S-conjugate β-lyase-mediated activation to highly reactive intermediates. Little is known about the regulation of such enzymes or their potential significance for drug-induced liver disease or hepatocarcinogenesis.

Phase 3 Pathways

Several transporters secrete drugs, drug metabolites, or their conjugates into bile, and this mechanism is often referred to as phase 3 of hepatic drug elimination. These pathways involve ATP-binding cassette (ABC) proteins, which derive the energy for their transport functions from hydrolysis of ATP. ABC transport proteins are widely distributed in nature and include the cystic fibrosis transmembrane con-

Chapter 86  Liver Disease Caused by Drugs ductance regulator (CFTR) (see Chapter 76) and the canalicular and intestinal copper transporters (see Chapter 75). The role of ABC transport proteins in secretion of bile has been reviewed (see Chapter 64).37,38,41,42 Multidrug resistance protein 1 (MDR1) is highly expressed on the apical (canalicular) plasma membrane of hepatocytes, where it transports cationic drugs, particularly anticancer agents, into bile. Another family of ABC transporters, the multidrug resistance-associated proteins (MRPs), is also expressed in the liver. At least two members of this family excrete drug (and other) conjugates from hepatocytes: MRP-3 on the basolateral surface facilitates passage of drug conjugate into the sinusoidal circulation, and MRP-2, expressed on the canalicular membrane, pumps endogenous compounds (e.g., bilirubin diglucuronide, leukotrieneglutathionyl conjugates, glutathione) and drug conjugates into bile. The bile salt export pump (BSEP) and MDR3 (in humans) and Mdr2 (in mice) are other canalicular transporters concerned, respectively, with bile acid and phospholipid secretion into bile. Genetic polymorphisms of these genes are associated with human cholestatic liver diseases (see Chapters 64 and 76). BSEP interacts with several drugs.42 Regulation of the membrane expression and activity of these drug elimination pathways is complex. The possibility that their altered expression or impaired activity (by competition between agents, changes in membrane lipid composition, or damage from reactive metabolites or covalent binding) could lead to drug accumulation, impairment of bile flow, or cholestatic liver injury has been demonstrated for estrogens,43,44 troglitazone,45 terbinafine,46 and flucloxacillin47 and may have wider mechanistic importance for drug-induced cholestasis and other forms of liver injury.42

TOXIC MECHANISMS OF LIVER INJURY Direct Hepatotoxins and Reactive Metabolites

Highly hepatotoxic chemicals injure key subcellular structures, particularly mitochondria and the plasma membrane. The injury arrests energy generation, dissipates ionic gradients, and disrupts the physical integrity of the cell. This type of overwhelming cellular injury does not apply to currently relevant hepatotoxins, most of which require metabolic activation to mediate damage to liver cells. The resulting reactive metabolites can interact with critical cellular target molecules, particularly those with nucleophilic substituents such as thiol-rich proteins and nucleic acids. Together with ROS, they act as oxidizing species within the hepatocyte to establish oxidative stress, a state of imbalance between pro-oxidants and antioxidants. ROS are also key signaling molecules that mediate biological responses to stress, as discussed later. Alternatively, reactive metabolites bind irreversibly to macromolecules, particularly proteins and lipids. Such covalent binding may produce injury by inactivating key enzymes or by forming protein-drug adducts that could be targets for immunodestructive processes that cause liver injury.

Oxidative Stress and the Glutathione System

The liver is exposed to oxidative stress by the propensity of hepatocytes to reduce oxygen, particularly in mitochondria and also in microsomal electron transport systems (such as CYP2E1), and by NADPH-oxidase-catalyzed formation of ROS and nitroradicals in Kupffer cells, endothelial cells, and stimulated polymorphs and macrophages. To combat oxidative stress, the liver is well-endowed with antioxidant mechanisms, including micronutrients, such as vitamin E and vitamin C, thiol-rich proteins (e.g., metallothionein,

ubiquinone), metal-sequestering proteins (e.g., ferritin), and enzymes that metabolize reactive metabolites (e.g., epoxide hydrolases), ROS (e.g., catalase, superoxide dismutase), and lipid peroxides (e.g., glutathione peroxidases). Glutathione (l-gamma-glutamyl-l-cyteine-glycine) is the most important antioxidant in the mammalian liver.19 Hepatocytes are the exclusive site of glutathione synthesis. Hepatic levels of glutathione are high (5 to 10 mmol/L) and can be increased by enhancing the supply of cysteine for glutathione synthesis; this mechanism is the cornerstone of thiol antidote therapy for acetaminophen poisoning. Hepatocyte glutathione synthesis increases in response to pro-oxidants, as occurs when CYP2E1 is overexpressed as a result of signaling via the redox-sensitive transcription factor Nrf.19,20,48,49 Glutathione is a critical cofactor for several antioxidant pathways, including thiol-disulfide exchange reactions and glutathione peroxidase. Glutathione peroxidase has a higher affinity for hydrogen peroxide than does catalase, and it disposes of lipid peroxides, free radicals, and electrophilic drug metabolites. Reduced glutathione is a cofactor for conjugation reactions catalyzed by the glutathione S-transferases. Other reactions proceed nonenzymatically. In turn, the products include glutathione-protein mixed disulfides and oxidized glutathione. The latter can be converted back to glutathione by proton donation catalyzed by glutathione reductase. Normally, most glutathione within the hepatocyte is in the reduced state, indicating the importance of this pathway for maintenance of the redox capacity of the cell. The reduced form of NADPH is an essential cofactor for glutathione reductase; NADPH formation requires ATP, thereby illustrating a critical link between the energy-generating capacity of the liver and its ability to withstand oxidative stress. Glutathione is also compartmentalized within the hepatocyte, with the highest concentrations found in the cytosol. Adequate levels of glutathione are essential in mitochondria, where ROS are constantly being formed as a minor by-product of oxidative respiration and in response to some drugs or metabolites that interfere with the mitochondrial respiratory chain. Mitochondrial glutathione is maintained by active uptake from the cytosol, a transport system that is altered by chronic ethanol exposure, and is, therefore, another potential target of drug toxicity.19

Biochemical Mechanisms of Cellular Injury

Mechanisms once thought to be central to hepatotoxicity, such as covalent binding to cellular enzymes and peroxidation of membrane lipids, are no longer regarded as exclusive pathways of cellular damage. Rather, oxidation of proteins, phospholipid fatty acyl side chains (lipid peroxidation), and nucleosides appear to be components of the biochemical stress that characterizes toxic liver injury. Secondary reactions also may play a role; these reactions include post-translational modification of proteins via adenosine diphosphate (ADP) ribosylation or protease activation, cleavage of DNA by activation of endogenous endonucleases, and disruption of lipid membranes by activated phospholipases.20 Some of these catabolic reactions could be initiated by a rise in the cytosolic ionic calcium concentration [Ca2+]i, as a result of increased Ca2+ entry or release from internal stores in the endoplasmic reticulum and mitochondria.19,20 The concept that hepatotoxic chemicals cause hepatocyte cell death by a biochemical final common pathway (e.g., activation of catalytic enzymes by a rise in [Ca2+]i) has proved inadequate to explain the diverse processes that can result in lethal hepatocellular injury. Rather, a variety of processes can damage key organelles, thereby causing intracellular stress that activates signaling pathways and tran-

1419

1420

Section IX  Liver scription factors. In turn, the balance between these factors can trigger the onset of cell death or facilitate protection of the cell, as discussed in the following sections.

Types of Cell Death

Apoptosis Apoptosis is an energy-dependent, genetically programmed form of cell death that typically results in controlled deletion of individual cells. In addition to its major roles in developmental biology, tissue regulation, and carcinogenesis, apoptosis is important in toxic, viral, and immunemediated liver injury.50-53 The ultrastructural features of apoptosis are cell and nuclear shrinkage, condensation and margination of nuclear chromatin, plasma membrane blebbing, and ultimately fragmentation of the cell into membrane-bound bodies that contain intact mitochondria and other organelles. Engulfment of these apoptotic bodies by surrounding epithelial and mesenchymal cells conserves cell fragments that contain nucleic acid and intact mitochondria. These fragments are then digested by lysosomes and recycled without release of bioactive substances. As a consequence, apoptosis in it purest form (usually found only in vitro) does not incite an inflammatory tissue reaction. The cellular processes that occur in apoptosis are often mediated by caspases, a family of proteolytic enzymes that contain a cysteine at their active site and cleave poly­ peptides at aspartate residues; non–caspase-mediated programmed cell death has also been described in experimental hepatotoxicity (see also Chapter 72). Apoptosis rarely, if ever, is the sole form of cell death in common forms of liver injury, such as ischemia-reperfusion injury, cholestasis, and toxic liver injury, all of which are typically associated with a hepatic inflammatory response. Whether or not activation of pro-death signals causes cell death depends on several factors, including pro-survival signals, the rapidity of the process, the availability of glutathione and ATP, and the role of other cell types. Some of these issues are discussed briefly here and are reviewed in more detail elsewhere.20,50-53 Hepatocytes undergo apoptosis when pro-apoptotic intracellular signaling pathways are activated, either because of toxic biochemical processes within the cell (intrinsic pathway) or because cell surface receptors are activated to transduce cell death signals (external pathway). Proapoptotic receptors are members of the tumor necrosis factor-α (TNF-α) receptor superfamily, which possess a so-called death domain. These receptors include Fas, for which the cognate ligand is Fas-ligand (Fas-L), TNF-R1 receptor (cognate ligand is TNF), and TNF-related apoptosis-inducing ligand (TRAIL) receptors (cognate ligand is TRAIL). In addition to model hepatotoxins such as the quinone, menadione, and hydrogen peroxide, some drugs (e.g., acetaminophen, plant diterpenoids) have been shown to be converted into pro-oxidant reactive metabolites, thereby initiating the following sequence: CYP-mediated metabolism to form reactive metabolites → glutathione depletion → mitochondrial injury with release of cytochrome c and operation of the mitochondrial membrane permeability transition (MPT) → caspase activation → apoptosis. Mitochondria play a pivotal role in pathways that provoke or oppose apoptosis.50,51,53 In the external pathway, activation of the death domain of pro-apoptotic receptors recruits adapter molecules—Fas-associated death domain (FADD) and TNF receptor-associated death domain (TRADD)— which bind and activate procaspase 8 to form the deathinducing signaling complex (DISC). In turn, caspase 8

cleaves Bid, a pro-apoptotic member of the B cell lymphoma/leukemia (Bcl-2) family, to tBid. tBid causes translocation of Bax to the mitochondria, where it aggregates with Bak to promote permeability of the mito­ chondria.50 Release of cytochrome c and other pro-death molecules, including Smac (which binds caspase inhibitor proteins, such as inhibitor of apoptosis proteins [IAPs]) and apoptosis-inducing factor (AIF, also known as Apaf)51 allows formation of the “aptosome,” which activates caspase 9 and eventually caspase 3 to execute cell death (Fig. 86-1). Intracellular stresses in various sites release other mitochondrial permeabilizing proteins (e.g., Bmf from the cytoskeleton and Bim from the endoplasmic reticulum), whereas members of the Bcl-2 family, Bcl-2 and Bcl-XL, antagonize apoptosis and serve as survival factors by regulating the integrity of mitochondria; the protective mechanism is poorly understood. Stress-activated protein kinases, particularly c-jun N-terminal kinase (JNK) may also be proapoptotic52 by phosphorylating and inactivating the mitochondrial protective protein Bcl-XL. Execution of cell death by apoptosis usually occurs via activation of caspase 3, but more than one caspaseindependent pathway of programmed cell death has been described.53 Stresses to the endoplasmic reticulum can bypass mitochondrial events by activation of caspase 12, which in turn activates caspase 9 independently of the apoptosome. The final steps of programmed cell death are energy dependent. Therefore, depletion of ATP abrogates the controlled attempt at “cell suicide,” resulting instead in necrosis (see later) or an overlapping pattern that has been designated as “apoptotic necrosis” or “necraptosis.”54,55 Furthermore, when apoptosis is massive, the capacity for rapid phagocytosis can be exceeded, and “secondary” necrosis can occur.55 Intracellular processes and activation of pro-apoptotic death receptors are not mutually exclusive pathways of cell death in toxic liver injury. In fact, drug toxicity could predispose the injured hepatocyte to apoptosis mediated by TNF-R or Fas-operated pathways by several mechanisms, including blockade of nuclear factor-kappa B (NF-κB), which usually is a hepatoprotective transcription factor in hepatocytes, and inhibition of purine and protein synthesis. Furthermore, activation of Kupffer cells (e.g., by endotoxin) and recruitment of activated inflammatory cells can increase production of TNF-α. Apoptosis-initiating pathways have broad relevance in several areas of hepatology. Inhibition of caspases is an important protective mechanism against cell death. Such anti-apoptotic pathways include chemical blockade of the cysteine thiol group by nitric oxide (NO) or ROS and cellular depletion of glutathione.20 Protein inhibitors include IAP family members, heat shock proteins (HSPs), and FLICE (caspase-8)inhibitory proteins (FLIP).50-52 FLIP inhibit caspase-8 activation as a decoy for FADD binding. Bcl-2 and Bcl-XL inhibit mitochondrial permeability, whereas phosphatidylinositol 3-kinase/Akt phosphorylates caspase 9 and activates NF-κB. Necrosis In contrast to apoptosis, necrosis has been conceptualized as a relatively uncontrolled process that can result from extensive damage to the plasma membrane with disturbance of ion transport, dissolution of membrane potential, cell swelling, and eventually rupture of the cell. Drug-induced injury to the mitochondrion can impair energy generation, whereas MPT can release stored Ca2+ into the cytosol and perturb other ionic gradients. Mitochondrial enzymes appear to be a particular target of NAPQI, the reactive

Chapter 86  Liver Disease Caused by Drugs Death ligand e.g., TNF, Fas, TRAIL

Death receptor e.g., TNF-R1

Plasma membrane Cytoplasm FLIP

FADD

Caspase 8 (FLICE)

TRADD

RIP DISC IAP

IAP Bid TRAF2 Bax

Bcl-xL Lipid mediators (ceramide)

Effector caspases

Apoptosis

Apoptosome

Mitochondrion MPT

Necrosis Figure 86-1.  Apoptosis and necrosis pathways in mammalian cells. See text for details. Bcl, B-cell lymphoma/leukemia family (Bax, Bid, and Bcl-xL are members); DISC, death-inducing signaling complex; FADD, Fas-associated death domain; FLIP, FLICE-inhibitory proteins; IAP, inhibitor of apoptosis proteins; MPT, mitochondrial permeability transition; RIP, receptor-interacting protein; TNF, tumor necrosis factor; TNF-R1, TNF receptor-1; TRADD, TNF receptor-associated death domain; TRAF2, TNF receptor-associated factor-2; TRAIL, TNF-related apoptosis ligand.

metabolite of acetaminophen. Reye’s syndrome–like disorders (e.g., toxicity caused by valproic acid; some nucleoside analogs, such as fialuridine, didanosine, zidovudine, zalcitabine; and possibly “ecstasy” [see Chapter 87]) may also result from mitochondrial injury. Mitochondrial injury can result in cell death by either apoptosis or necrosis54,55; the type of cell death pathway may depend primarily on the energy state of the cell, as well as the rapidity and severity of the injury process. In the presence of ATP, cell death can proceed by apoptosis, but when mitochondria are deenergized, the mechanism of cell death is necrosis. This apparent dichotomy between cell death processes is probably artificial, and apoptosis and necrosis more likely represent the morphologic and mechanistic ends of a spectrum of overlapping cell death processes.19,55 One important way in which necrosis differs from apoptosis is that uncontrolled dissolution of the cell liberates macromolecular breakdown products, including lipid peroxides, aldehydes, and eicosanoids. These products act as chemoattractants for circulating leukocytes, which enjoin an inflammatory response in the hepatic parenchyma. Even before cell death occurs, oxidative stress produced during drug toxicity can up-regulate adhesion molecules and chemokines that are expressed or secreted by endothelial cells.

These processes contribute to recruitment of the hepatic inflammatory response, which is prominent in some types of drug-induced liver disease. Lymphocytes, polymorphonuclear leukocytes (neutrophils and eosinophils), and macrophages also may be attracted to the liver as part of a cell-mediated immune reaction. Role of Oxidative Stress Although severe oxidative stress in hepatocytes, particularly when focused on mitochondria, is likely to induce necrosis, lesser (or more gradual) exposure can trigger apoptosis because ROS and oxidative stress can activate Fas signaling, JNK and other kinases, p53, and microtubular assembly and impair protein folding, thereby resulting in an unfolded protein response by the endoplasmic reticulum. Oxidative stress also may amplify cell death processes by uncoupling of the mitochondrial respiratory chain, release of cytochrome c, or massive oxidation and export of glutathione (intact glutathione is required for Fas signaling). Conversely, oxidative stress may protect against apoptosis in some circumstances through inhibition of caspase or activation of NF-κB. As a result of these opposing effects, predicting the consequences of hepatic oxidative stress in terms of liver injury is not easy.

1421

1422

Section IX  Liver Role of Hepatic Nonparenchymal Cells and the Innate Immune Response

In addition to migratory cells, activation of nonparenchymal liver cell types is likely to play an important role in drug and toxin-induced liver injury. Kupffer cells function as resident macrophages and antigen-presenting cells. Some of the toxic effects of activated Kupffer cells, as well as of recruited leukocytes, may be mediated by release of cytokines, such as TNF and Fas-L, which under some circumstances can induce cell death in hepatocytes by apoptosis or necrosis.55 In addition, activated Kupffer cells release ROS, nitroradicals, leukotrienes, and proteases. Endothelial cells of the hepatic sinusoids or terminal hepatic veins are vulnerable to injury by some hepatotoxins because of their low glutathione content. Such hepatotoxins include the pyrrolizidine alkaloids, which are an important cause of the sinusoidal obstruction syndrome (hepatic venoocclusive disease).56 Other types of drug-induced vascular injury may be caused primarily by involvement of the sinusoidal endothelial cells (see Chapter 83). Hepatic stellate cells (formerly fat-storing or Ito cells) are the principal liver cell type involved in matrix deposition in hepatic fibrosis. Stellate cells are activated in methotrexate-induced hepatic fibrosis. The possibility that vitamin A, ROS, or drug metabolites can transform stellate cells into collagen-synthesizing myofibroblasts is of considerable interest.

IMMUNOLOGIC MECHANISMS

In addition to the activation of innate inflammatory processes in the liver by toxic mechanisms, (extrinsic) immunologic mechanisms could account for certain aspects of idiosyncratic drug-induced liver disease. Immune attack involves liganding of death receptors, as discussed earlier, or porin-mediated introduction of granzyme.19 The most convincing evidence for drug allergy includes (1) delayed onset after initial exposure and accelerated onset after rechallenge, (2) hepatic inflammatory infiltrates with neutrophils and eosinophils, and (3) fever, rash, lymphadenopathy, peripheral eosinophilia, and involvement of other organs. In some types of drug hepatitis, the liver is clearly implicated as part of a systemic hypersensitivity reaction, as described later for the reactive metabolite syndrome (RMS); sulfonamides, phenytoin, nitrofurantoin, minocycline, nevirapine, and some Chinese herbal medicines are causative agents. Why the liver is the predominant site of injury in some persons whereas other organs are involved in other persons is unclear; genetic factors relevant to tissuespecific gene expression could be involved. One possible immunopathogenic mechanism for druginduced liver disease is the altered antigen concept, in which an initial interaction between drug metabolites and cellular proteins results in the formation of neoantigens (haptens) or drug-protein adducts. An example is the formation of trifluoroacetylated (TFA) adducts after exposure to halothane or other haloalkane anesthetics (see Chapter 87). For these adducts to initiate tissue-damaging immune responses (1) processing should be presented in an immunogenic form (e.g., by Kupffer cells, in association with major histocompatibility complex [MHC] molecules); (2) appropriately responsive CD4+ T cells must be present to provide help to induce an immune response; and (3) the drug-derived antigen, together with a class II MHC molecule, must be expressed on the target cells in order to attract CD8+ (cytotoxic) T cells. That bile duct epithelial cells are more likely to express class II MHC antigens may explain why they are possible targets in drug-induced cholestatic hepatitis.

Although antibodies directed against TFA-protein adducts circulate in the majority of patients following recovery from halothane-induced liver injury,57 the specificity and pathogenicity of these antibodies remain in doubt. Another way in which circulating drug-induced antibodies could result in immune-mediated lysis of hepatocytes is through molecular mimicry of host enzymes.58 Experimental evidence suggests that for diclofenac antibody-dependent cell-mediated immunity could operate as a mechanism for drug-induced liver disease.59 A second type of immunopathogenic mechanism is dysregulation of the immune system, termed drug-induced autoimmunity. This mechanism can lead to the formation of drug-induced autoantibodies (e.g., anti–liver-kidney microsome [LKM] antibodies) directed against microsomal enzymes. For tienilic acid, CYP2C9 is the target of antiLKM, whereas after halothane hepatitis anti-LKM are directed against CYP2E1. Non-tissue specific autoantibodies, such as antinuclear and smooth muscle antibodies, may be detected in patients with nitrofurantoin, methyldopa, or minocycline hepatitis. Like spontaneous autoimmunity, drug-induced autoimmunity may involve genetic predisposition through anomalies of immune tolerance.

CLINICOPATHOLOGIC FEATURES OF DRUG-INDUCED LIVER DISEASE CLASSIFICATION

Hepatic drug reactions mimic all known liver diseases, but classification is often difficult because of overlap among categories. Although a classic (“signature”) syndrome is associated with many individual agents, a given drug may be associated with more than one clinicopathologic syndrome. Furthermore, the clinical and laboratory features of liver disease and the hepatic histologic findings may be discordant. Therefore, although recognition of specific patterns or syndromes is a vital clinical clue to the diagnosis of drug-induced liver disease, the chronologic relationship between administration of the drug and liver injury is a more important clue. Drugs are often divided into dose-dependent, or predictable, hepatotoxins and dose-independent, or unpredictable (idiosyncratic), hepatotoxins. Dose-dependent hepatotoxins generally require metabolic activation to toxic metabolites or interfere with subcellular organelles and biochemical processes at key sites, such as mitochondria or canalicular bile secretion.43 Liver injury produced by dose-dependent hepatotoxins usually occurs after a short latent period (hours), is characterized by zonal necrosis or microvesicular steatosis, and can be reproduced in other species. By contrast, idiosyncratic hepatotoxins cause a wide range of histologic changes and do not reliably cause injury in other species; in addition, the latent period before the onset of injury is variable in duration. The distinction between dosedependent and idiosyncratic hepatotoxins is blurred with agents such as dantrolene, tacrine, perhexiline, flucloxacillin, cyclophosphamide, nucleoside analogs, bromfenac, anticancer drugs, and cyclosporine. Liver injury caused by each of these drugs is partly dose dependent, but reactions occur in only a small proportion of exposed persons. Two general types of mechanisms may account for idiosyncratic hepatotoxicity: metabolic idiosyncrasy and immunoallergy. Metabolic idiosyncrasy refers to the susceptibility of rare persons to hepatotoxicity from a drug that, in conventional doses, is usually safe. Such susceptibility may result from genetic or acquired differences in drug metabo-

Chapter 86  Liver Disease Caused by Drugs lism or canalicular secretion, mitochondrial defects, or cell death receptor signaling. Immunoallergy indicates operation of the immune system in mediating the response to a drug. These two mechanisms may be interrelated (see later). Other pathogenic mechanisms may include indirect mediation of liver injury, as in vascular and possibly hyperthermic changes produced by cocaine, ecstasy, intraarterial fluroxuridine, and possibly anesthetics (see Chapter 87). The most practical classification of drug hepatotoxicity is based on clinical and laboratory features and liver histology, as summarized in Table 86-2. This classification provides a framework for discussing drug-induced hepatic disease in comparison with other hepatobiliary disorders but is imperfect because the clinical and pathologic features are not always congruent. Moreover, much overlap between categories exists, particularly in the spectrum from severe necrosis (which may result from dose-dependent or idiosyncratic hepatotoxicity) to focal necrosis with lobular inflammation (hepatitis) to cholestasis. Many drugs produce a spectrum of syndromes from hepatitis to cholestasis, and some authorities include a further category of mixed cholestatichepatocellular reactions. Granulomatous hepatitis is asso­ ciated with liver biochemical test abnormalities that are usually indistinguishable from those typical of hepatitis, cholestasis, or mixed reactions.

Drugs can alter liver test results without causing significant liver injury. Such adaptive responses include hyperbilirubinemia associated with rifampin, cyclosporine, and indinavir and raised serum GGTP and alkaline phosphatase levels associated with phenytoin and warfarin.5,6 The latter effect is probably attributable to microsomal enzyme induction. For agents such as isoniazid, however, the distinction between adaptation and minor injury is blurred; adaptation in such cases may be a response to oxidative injury. Conversely, liver tumors or hepatic fibrosis may develop insidiously without significant abnormalities of liver biochemical tests—the former in association with sex steroids or vinyl chloride monomer and the latter with methotrexate, arsenic, or hypervitaminosis A. The duration of the disorder is another consideration in classifying drug-induced liver diseases. In general, chronic liver disease is much less commonly attributable to drugs and toxins than are acute reactions,8 but not to consider drugs as a possible eitology of chronic liver disease can lead to a missed diagnosis, with serious clinical consequences.8,9 In contrast to most other types of hepatic pathobiology, drugs and toxins constitute the most important cause of vascular disorders of the liver (see later). Drugs also have been associated with chronic cholestasis, chronic hepatitis, steatohepatitis, hepatic fibrosis, cirrhosis, and benign and malignant liver tumors.

Table 86-2  Clinicopathologic Classification of Drug-Induced Liver Disease CATEGORY

DESCRIPTION

IMPLICATED DRUGS: EXAMPLES

Hepatic adaptation

No symptoms; raised serum GGTP and AP levels (occasionally raised ALT) Hyperbilirubinemia Symptoms of hepatitis; zonal, bridging, and massive necrosis; serum ALT level >5-fold increased, often >2000 U/L Microvesicular steatosis, diffuse or zonal; partially dose dependent, severe liver injury, features of mitochondrial toxicity (lactic acidosis) Symptoms of hepatitis; focal, bridging, and massive necrosis; serum ALT level >5-fold increased; extrahepatic features of drug allergy in some cases Duration >3 months; interface hepatitis, bridging necrosis, fibrosis, cirrhosis; clinical and laboratory features of chronic liver disease; autoantibodies with some types of reaction (see Table 86-6) Hepatic granulomas with varying hepatitis and cholestasis; raised serum ALT, AP, GGTP levels Cholestasis, no inflammation; serum AP levels >twice-normal Cholestasis with inflammation; symptoms of hepatitis; raised serum ALT and AP levels Bile duct lesions and cholestatic hepatitis; clinical features of cholangitis Cholestasis present >3 months Paucity of small bile ducts; resembles primary biliary cirrhosis, but AMA negative Strictures of large bile ducts Steatosis, focal necrosis, Mallory’s hyaline, pericellular fibrosis, cirrhosis; chronic liver disease, portal hypertension Sinusoidal obstruction syndrome, nodular regenerative hyperplasia, others Hepatocellular carcinoma, adenoma, angiosarcoma, others

Phenytoin, warfarin

Dose-dependent hepatotoxicity Other cytopathic toxicity, acute steatosis Acute hepatitis

Chronic hepatitis

Granulomatous hepatitis Cholestasis without hepatitis Cholestatic hepatitis Cholestasis with bile duct injury Chronic cholestasis:   Vanishing bile duct syndrome   Sclerosing cholangitis Steatohepatitis Vascular disorders Tumors

Rifampin, flavaspidic acid Acetaminophen, nicotinic acid, amodiaquine, hycanthone Valproic acid, didanosine, HAART agents, fialuridine, l-asparaginase, some herbal medicines Isoniazid, dantrolene, nitrofurantoin, halothane, sulfonamides, phenytoin, disulfiram, acebutolol, etretinate, ketoconazole, terbinafine, troglitazone Nitrofurantoin, etretinate, diclofenac, minocycline, nefazodone (see also Table 86-6) Allopurinol, carbamazepine, hydralazine, quinidine, quinine (see also Table 86-5) Oral contraceptives, androgens Chlorpromazine, tricyclic antidepressants, erythromycins, amoxicillin-clavulanic acid Chlorpromazine, flucloxacillin, dextropropoxyphene Chlorpromazine, flucloxacillin, trimethoprim-sulfamethoxazole Intra-arterial floxuridine, intralesional scolicidals Perhexiline, amiodarone, others (see Chapter 85) Many (see Table 86-8) Many (see Chapter 94)

ALT, alanine aminotransferase; AMA, antimitochondrial antibodies; AP, alkaline phosphatase; AST, aspartate aminotransferase; GGTP, gamma glutamyl transpeptidase; HAART, highly active antiretroviral therapy.

1423

1424

Section IX  Liver HISTOPATHOLOGIC FEATURES

Although no pathognomonic hallmarks of drug-induced liver disease have been identified, certain histologic patterns suggest drug-induced liver injury. These patterns include zonal necrosis or microvesicular steatosis (which accompanies mitochondrial injury) and mixed histologic features of hepatocellular necrosis and cholestasis. Necrotic lesions that are disproportionately severe compared with the clinical picture also indicate a possible drug cause, whereas destructive bile duct lesions, prominent neutrophils, and eosinophils (at least 25% of inflammatory cells) are suggestive of drug-induced cholestatic hepatitis. Hepatic granuloma formation is another common type of hepatic drug reaction. In cases of steatohepatitis, hepatic fibrosis, or liver tumors, no specific clues to a drug cause have been recognized, although sex steroids increase the vascularity of hepatic tumors and are frequently associated with sinusoidal dilatation or peliosis hepatis. Drug-induced steatohepatitis caused by amiodarone and perhexilene tends to be associated with severe lesions that more closely resemble alcoholic hepatitis than NASH.60 Other drugs (e.g., tamoxifen, methotrexate) cause lesions that are indistinguishable from NASH associated with diabetes mellitus and the metabolic syndrome.6,8,61

CLINICAL FEATURES

The history and physical examination can provide important clues to the diagnosis of hepatic drug reactions. Most important is the temporal pattern of disease evolution in relation to exposure to drugs or toxins. The identification of specific risk factors for hepatotoxicity (e.g., chronic excessive alcohol intake in a person taking acetaminophen) and the presence of systemic features of drug hypersensitivity may indicate the correct diagnosis. Systemic features include fever, rash, mucositis, eosinophilia, lymphadenopathy, a mononucleosis-like syndrome, bone marrow suppression, vasculitis, renal failure, pneumonitis, and pancreatitis. These features may be part of a characteristic syndrome thought to have a genetic basis and likely mediated by formation of drug metabolites that act as haptens to initiate an immunodestructive tissue reaction termed the reactive metabolite syndrome (RMS).62

Reactive Metabolite Syndrome

Drugs implicated as a cause of RMS include sulfonamides, aminopenicillins, fluoroquinolones, clozapine, anticonvulsants (phenytoin, lamotrigine, phenobarbital, carbama­ zepine), minocycline, protease inhibitors (nevirapine, abacavir), some NSAIDs, and Chinese herbal medicines.62 Risk factors for RMS include a family history of an affected first-degree relative (increases the risk to 1 in 4). Use of other drugs, such as glucocorticoids or valproic acid, at the time the new agent is started increases the risk 4- to 10-fold. The presence of a disorder associated with immune dysregulation (e.g., systemic lupus erythematosus) increases the risk 10-fold, whereas HIV/AIDS increases the risk 100-fold. The illness characteristically begins between 1 and 12 weeks (typically 2 to 4 weeks) after the drug is started; “sentinel symptoms” include fever, pharyngitis, malaise, periorbital edema, headache or otalgia, rhinorrhea, and mouth ulcers. A severe drug rash is an essential feature. Erythematous reactions are usual and may evolve to toxic epidermal necrolysis or erythema multiforme, often with mucositis (Stevens-Johnson syndrome). Early abnormalities on blood testing include neutrophilia and elevated levels of acute-phase reactants; atypical lymphocytosis and eosinophilia may be noted later. Hepatic reactions are found in

about 13% of cases. Findings include cholestasis, acute hepatitis, and granulomas. Other features include lymphadenopathy (16%), nephritis (6%), pneumonitis (6%), and more severe hematologic abnormalities (5%).

Latent Period to Onset

For idiosyncratic reactions, a latent period occurs between the commencement of drug intake and the onset of clinical and laboratory abnormalities. The duration of the latent period is commonly 2 to 8 weeks for immunoallergic types of drug hepatitis (such as the RMS) and often 6 to 20 weeks or longer for agents such as isoniazid, dantrolene, and troglitazone. Occasionally, liver injury may become evident after discontinuation of the causative agent; for oxypenicillins and amoxicillin-clavulanate, the onset of hepatotoxicitiy may occur as long as two weeks after the end of therapy. In other cases, hepatotoxicity is rare after the first exposure to a drug but more frequent and more severe after subsequent courses. Examples include halothane, nitrofurantoin, and dacarbazine. A history of a previous reaction to the drug in question (inadvertent rechallenge) may, therefore, be the key to the diagnosis of drug-induced liver disease.

Dechallenge and Rechallenge

Another aspect of the temporal relationship between ingestion of a drug and hepatotoxicity is the response to discontinuation of the drug, or dechallenge. Dechallenge should be accompanied by discernible and progressive improvement within days or weeks of stopping the incriminated agent. Exceptions occur with ketoconazole, troglitazone, coumarol, etretinate, and amiodarone; with these agents, reactions may be severe, and clinical recovery may be delayed for months. Although some types of drug-induced cholestasis also can be prolonged, failure of jaundice to resolve in a suspected drug reaction most often is indicative of an alternative diagnosis. Rarely, deliberate rechallenge may be used to confirm the diagnosis of drug-induced liver disease or to prove the involvement of one particular agent when the patient has been exposed to several drugs; however, this approach is potentially hazardous and should be undertaken only with a fully informed and consenting (in writing) patient and preferably with the approval of an institutional ethics committee.

A PRACTICAL APPROACH TO DIAGNOSIS In the absence of specific diagnostic tests, diagnosis of druginduced liver disease requires clinical suspicion, a thorough drug history, careful consideration of the temporal relationships between drug ingestion and liver disease, and exclusion of other disorders. The objective weighing of evidence for and against an individual agent—causality assessment— is a probabilistic form of diagnosis.63,64 Several clinical scales that incorporate and weigh various features of hepatic adverse drug reactions have been described.9,65-67 A liver biopsy may be indicated in some cases to exclude other diseases and to provide further clues to a drug etiology.68 In the future, in vitro tests may provide confirmatory evidence for particular drugs,57-65 but rechallenge is currently the standard test for drug-induced liver disease.66

PHYSICIAN AWARENESS

Physician awareness is critical for the diagnosis of druginduced liver disease. The sources of potential hepatotoxins include not only prescribed medications, but also over-the-

Chapter 86  Liver Disease Caused by Drugs counter drugs (e.g., ibuprofen), CAM preparations (see Chapters 87 and 127), substances taken for recreational use (e.g., cocaine, ecstasy) or self-poisoning, and environmental contaminants in food and water supplies, the home, the workplace, and the community. Unfortunately, patients and physicians do not always heed early nonspecific symptoms associated with reactions to hepatotoxic drugs. For example, preventable deaths from liver failure still occur more than 40 years after the recognition that isoniziad can cause drug hepatitis.69 Although continuing education and availability of information about potentially hepatotoxic drugs are important issues, physicians have a professional and legal obligation to inform patients about possible adverse drug reactions. A study from Switzerland found that the frequency of new cases of drug-induced liver injury among over 4000 hospital admissions was 1.4% (57 cases). Nevertheless, the drug reaction was not mentioned as a diagnosis in the physicians’ discharge note in 52% to 67% of cases.66 Drug toxicity should be considered a possibility in cases of obscure or poorly explained liver disease, particularly in cases in which mixed or atypical patterns of cholestasis and hepatitis, cholestasis in which common causes have been excluded, especially in the elderly, and histologic features suggestive of a drug etiology are observed. In such cases, the drug history must be addressed as a special investigation, with attention paid to additional sources of information (household members, primary care providers), household drugs, non-prescribed medications, and environmental toxins (see Chapter 87).

EXCLUSION OF OTHER DISORDERS

Other diseases must be excluded before hepatobiliary disease can be ascribed to a drug. For acute and chronic hepatocellular reactions, viral and autoimmune causes of hepatitis and vascular and metabolic disorders must be considered. Some types of drug-induced chronic hepatitis are associated with autoantibodies and superficially resemble autoimmune hepatitis. An approach to the correct diagnosis is described later (see Nitrofurantoin). Drug-induced cholestasis should be considered only when dilatation of the bile duct has been excluded by imaging. In older patients, and particularly when drug exposure does not include agents known to cause cholestasis, cholangiography (e.g., magnetic resonance imaging [MRCP], endoscopic retrograde cholangiography [ERCP]) is obligatory, as is liver biopsy. Drugs and metabolic factors may interact to cause steatohepatitis, as discussed later.

EXTRAHEPATIC FEATURES

The constellation of rash, eosinophilia, and other organ involvement is relatively specific for an adverse drug reaction as a cause of liver disease (see earlier). These findings, however, are present in only a minority of cases, so their absence is not helpful. In particular, drugs that cause idiosyncratic liver injury by nonimmunologic mechanisms are not usually associated with extrahepatic features. Specific diagnostic tests for individual drug-induced liver diseases have been described57 but are not generally accepted or available. In the case of dose-dependent hepatotoxins, blood levels may be helpful (see later).

CHRONOLOGIC RELATIONSHIPS

For most drugs, the chronologic relationship among drug ingestion, onset, and resolution of liver injury remains the most important consideration in diagnosis. The criteria for temporal eligibility include the relationship of drug ingestion to onset, course of the reaction after discontinuation of

the drug, and response to readministration of the drug.5,6-9 Deliberate rechallenge can be hazardous and is rarely indicated for logistic and ethical reasons, but inadvertent rechallenge may have occurred already. The rechallenge is regarded as positive if the serum ALT or alkaline phosphatase level increases at least two-fold.1,6,9 Deliberate rechallenge may be considered to ascertain whether a drug that is important for an individual patient is responsible for hepatotoxicity (e.g., amiodarone needed for refractory ventricular tachycardia). In other cases, documenting the propensity of newer agents, hitherto unrecognized as hepatotoxins, to cause liver injury may be desirable. Written informed consent is required for a deliberate rechallenge.

WHICH DRUG?

New and nonproprietary compounds should arouse particular suspicion. For patients who are taking multiple drugs, the agent started most recently before the onset of liver injury is most likely to be responsible. If that agent is unlikely to be the culprit and another well-known hepatotoxin is being taken, the latter is the more likely culprit. When possible, the most likely hepatotoxin or all therapeutic agents should be discontinued. If the patient improves, the drugs that are unlikely to be responsible can be carefully reintroduced.

INDICATIONS FOR LIVER BIOPSY

Liver biopsy may be helpful in difficult cases, especially when temporal relationship between the ingestion of a known hepatotoxic agent and the onset of liver injury is unclear. In practice, for example, the onset of jaundice within two to six weeks of starting an agent such as amoxicillin-clavulanic acid or of acute hepatitis in the presence of other features of RMS in a person taking nevirapine as part of HAART would be contexts in which the suspicion of a drug etiology is so strong that liver biopsy is unnecessary. Conversely, substantially abnormal liver biochemical test levels (e.g., a serum ALT level elevated more than five-fold) in a person who has one or more autoantibodies suggestive of autoimmune hepatitis and has been taking a statin or other cardiovascular drug for three to six months may constitute a clinical challenge that often can be resolved only by liver biopsy. The community may benefit when new instances or patterns of drug-induced liver disease are adequately defined; this benefit may persuade the clinician (but not always the informed patient) to proceed with a liver biopsy in equivocal cases.

CONSIDERATIONS IN PATIENTS WITH VIRAL HEPATITIS

Patients with chronic hepatitis B or C may be at higher risk of liver injury from antituberculosis chemotherapy, ibuprofen and possibly other NSAIDs, anti-cancer drugs, and HAART compared with persons without viral hepatitis.26-30 A more common clinical problem is the finding of a serum ALT level greater than 300 U/L at a routine office visit in a patient with previous levels less than 150 U/L. In patients with hepatitis C, the rise in serum ALT is more likely the result of drug toxicity than a spontaneous change in the activity of the hepatitis C, particularly when the ALT level is greater than 1000 U/L.68 The most commonly implicated agents are acetaminophen taken in moderate doses under conditions of increased risk (e.g., fasting, alcohol excess, use of other medication) and CAM preparations, typically Chinese herbal remedies (see Chapter 87). Clinical suspicion is essential for recognizing the drug cause of liver

1425

1426

Section IX  Liver injury so that appropriate advice can be given. Determination of blood levels of acetaminophen also may be useful in difficult cases, but levels (particularly undetectable levels) can be difficult to interpret in the context of regular ingestion, as opposed to a single episode of self-poisoning.

PREVENTION AND MANAGEMENT With the exception of acetaminophen hepatotoxicity (discussed later), little effective treatment for drug-induced liver disease is available. Special emphasis, therefore, must be placed on prevention and early detection of liver injury as well as on prompt withdrawal of the offending agent. Safe use of self-medication of agents such as acetaminophen, NSAIDs, and CAM preparations is important. Clear and open communication between the physician and patient and appropriate recommendations about dose limitations could prevent most instances of liver injury from these agents. The majority of drugs associated with drug-induced liver disease are idiosyncratic hepatotoxins, for which liver injury occurs rarely. The overall frequency of adverse hepatic reactions can be minimized by avoiding overuse of these drugs; antibiotics such as amoxicillin-clavulanic acid and flucloxacillin are pertinent examples. Similarly, polypharmacy should be avoided when possible. The rarity of adverse drug reactions also means that the hepatotoxic potential of new agents may not be recognized until after their introduction. Therefore, all physicians share the responsibility to report suspected adverse effects to monitoring agencies during postmarketing surveillance of new drugs. For dose-dependent hepatotoxins, prevention depends on adherence to dosage guidelines or use of blood levels. This approach has virtually abolished some forms of druginduced liver injury, such as tetracycline-induced fatty liver, aspirin hepatitis, and methotrexate-induced hepatic fibrosis. In cases with specific risk factors, strategies to prevent toxicity are essential (e.g., avoid use of valproic acid with other drugs in the very young; do not prescribe methotrexate to persons who consume alcohol in excess). Moderate doses of acetaminophen are contraindicated in heavy drinkers and after fasting,21 and administration of halothane should not be repeated within 28 days or in persons suspected of previous sensitivity to a haloalkane anesthetic. Early detection is also critical. Patients should be warned to report any untoward symptoms, particularly unexplained nausea, malaise, right upper quadrant abdominal pain, lethargy, or fever. These nonspecific features may represent the prodrome of drug-induced hepatitis. They are an indication for liver biochemical testing and, if the results suggest liver injury, for cessation of treatment. A more difficult issue is whether regular (protocol) screening with liver biochemical tests should be performed when a drug is prescribed. Although such screening often is recommended by authors and drug manufacturers, the efficiency and cost-effectiveness of this approach are unknown. The onset of liver injury is often rapid, rendering once-a-month or every-second-week screening futile. Furthermore, 7.5% of persons who receive placebo in clinical trials have persistently raised serum ALT levels.70 If liver biochemical test levels are monitored, the level of abnormality at which a drug should be discontinued is uncertain, as illustrated by isoniazid, which causes some liver biochemical test abnormality in 30% of exposed sub-

jects. Generally, the recommendation is that isoniazid be stopped if serum ALT levels exceed 250  U/L or more than five times the upper limit of normal, but elevation of the serum bilirubin or albumin concentration or prolongation of the prothrombin time provides a clearer indication to stop the drug. Conversely, a rise in the serum GGTP level or a minor elevation of serum alkaline phosphastase level usually indicates hepatic adaptation rather than liver injury. We do not routinely recommend protocol screening, but this approach could be useful for agents such as valproic acid, isoniazid, pyrazinamide, ketoconazole, dantrolene, tacrine, thiazolidinediones, and synthetic retinoids, either because the onset of liver injury may be delayed and gradual in some cases or because such screening can emphasize the hepatotoxic potential of these drugs to patients and physicians. Liver biopsy has a role in the assessment of hepatic fibrosis in patients who take methotrexate (see later). Highly toxic solvents should be avoided in the workplace, and such agents have been abandoned. Adequate ventilation and use of masks and protective clothing are vital to prevent occupational exposure to hepatotoxic chemicals. In some cases, liver biochemical tests are performed routinely in exposed persons, but abnormalities are more likely to reflect diseases such as chronic hepatitis C, alcoholism, and NAFLD than toxic liver injury. In the case of vinyl chloride exposure, periodic physical examination (for hepatomegaly) and hepatic imaging with ultrasonography may be useful (see Chapter 87). Active management of drug-induced liver injury includes removal of the drug and administration of antidotes and anti-inflammatory and cytoprotective agents. In practice, treatment usually is confined to discontinuation of the hepatotoxic drug. Failure to discontinue a drug that is the cause of liver injury is the single most important factor leading to poor outcomes, such as acute liver failure and chronic liver disease.8,9 For ingested toxins such as metals, poisonous mushrooms, and acetaminophen, removal of the unabsorbed drug through the aspiration of stomach contents may be appropriate. Methods to remove absorbed toxins, such as hemodialysis through a charcoal column and forced diuresis, are not effective for hepatotoxins. For chlordecone, an organochlorine insecticide that is lipid-soluble and excreted in bile, oral administration of cholestyramine enhances removal of the agent from the body by interrupting the enterohepatic cycle.71 Thiol replacement therapy, usually with N-acetylcysteine (NAC), is indicated as an antidote for acetaminophen poisoning. Whether NAC or other antioxidants have a role in other types of acute hepatotoxicity is unclear, but the flavonoid, silybin (silymarin), is traditionally used for Amanita phalloides toxicity72 and tocopherol analogs show promise in experimental hepatotoxicity (see Chapter 87). Beyond discontinuation of the offending agent, the management of drug hepatitis and cholestasis is symptomatic and supportive. In cases of acute liver failure, hepatic transplantation should be considered (see Chapters 93 and 95).7 Ursodeoxycholic acid has shown some promise in the management of chronic cholestasis and pruritus caused by drug hepatotoxicity. Glucocorticoids have little role in the management of drug-induced cholestasis or hepatitis and are ineffective in chlorpromazine-, methyldopa-, and isoniazidinduced hepatitis and in drug-induced fulminant hepatic failure. Case reports attest to the occasional effectiveness of glucocorticoids in protracted cases of hepatitis caused by etretinate, allopurinol, diclofenac, or ketoconazole.5 Glucocorticoids should be reserved for atypical and refractory cases, particularly those associated with vasculitis. Clinical

Chapter 86  Liver Disease Caused by Drugs evidence of the effectiveness of putative hepatoprotective agents, such as prostaglandin analogs, is lacking.

DOSE-DEPENDENT HEPATOTOXICITY Few dose-dependent hepatotoxins are clinically important today. Examples include acetaminophen, some herbal medicines (CAM preparations), plant and fungal toxins, amodiaquine, hycanthone, vitamin A, methotrexate, cyclophosphamide, anti-cancer drugs, carbon tetrachloride, phosphorus, and metals (especially iron, copper, and mercury). Acetaminophen is by far the most important of these; hepatotoxicity caused by CAM preparations is discussed in Chapter 87.

Table 86-3  Risk Factors for Acetaminophen-Induced Hepatotoxicity FACTOR

RELEVANCE

Age

Children may be more resistant than adults Minimal hepatotoxic dose: 7.5g (≈100 mg/kg) in adults, 150 mg/kg in children Severe toxicity possible with dose >15 g Influenced by dose, time after ingestion, gastric emptying Best indicator of risk of hepatotoxicity (see text and Fig. 86-2) Toxic dose threshold lowered; worsens prognosis (also related to late presentation); nephrotoxicity common Toxic dose threshold lowered— therapeutic misadventure (see text) Toxic dose threshold lowered— therapeutic misadventure; worsens prognosis (e.g., isoniazid, phenytoin, zidovudine) Late presentation or delayed treatment (>16 hr) predicts worse outcome

Dose

Blood level

Chronic excessive alcohol ingestion Fasting

ACETAMINOPHEN General Nature, Frequency, and Predisposing Factors

Acetaminophen (paracetamol) is a widely used analgesic available without prescription. It is safe when taken in the recommended therapeutic dose of 1 to 4 g daily, but hepatotoxicity produced by self-poisoning with acetaminophen has been recognized since the 1960s. Despite the effectiveness of thiol-based antidotes, acetaminophen remains the most common cause of drug-induced liver injury in most countries and an important cause of acute liver failure.7,73 Parasuicide and suicide are the usual reasons for overdose.73,74 Although controversial,75,76 hepatologists and pediatricians see cases of acetaminophen poisoning that have arisen through what Zimmerman and Maddrey termed therapeutic misadventure.77 This occurrence is especially common in persons who habitually drink alcohol to excess and has also been recognized after daily ingestion of moderate therapeutic doses (10 to 20 g over three days) of acetaminophen in adults and children who are fasting or malnourished21 or who are taking drugs that interact with the metabolism of acetaminophen.77 Single doses of acetaminophen that exceed 7 to 10 g (140 mg/kg body weight in children) may cause liver injury, but this outcome is not inevitable. Severe liver injury (serum ALT level greater than 1000 U/L) or fatal cases usually involve doses of at least 15 to 25 g, but because of interindividual variability, survival is possible even after ingestion of a massive single dose of acetaminophen (greater than 50 g).78 Among persons with an untreated acetaminophen overdose, severe liver injury occurred in only 20%, and among those with severe liver injury, the mortality rate was 20%.78 Conversely, among heavy drinkers, daily acetaminophen doses of 2 to 6 g have been associated with fatal hepatotoxicity.75-78 Risk factors for acetaminophen-induced hepatotoxicity are summarized in Table 86-3. Children are relatively resistant to acetaminophen-induced hepatotoxicity,79 possibly because of their tendency to ingest smaller doses, greater likelihood of vomiting, or biological resistance. Therapeutic misadventure after multiple doses, especially during fasting and when weight-based recommendations have been exceeded, has a high mortality rate. By contrast, the presence of underlying liver disease does not predispose to acetaminophen hepatotoxicity. Self-poisoning with acetaminophen is most common in young women, but fatalities are most frequent in men, possibly because of alcoholism and late presentation.73-75 The time of presentation is critical because thiol therapy given within 12 hours of acetaminophen poisoning virtually abolishes significant liver injury (see later). Therapeutic

Concomitant medication

Time of presentation

misadventure is also associated with a worse outcome.76 Concomitant use of agents such as phenobarbital, pheny­ toin, isoniazid, and zidovudine is another risk factor for acetaminophen hepatotoxicity. These drugs may promote the oxidative metabolism of acetaminophen to NAPQI by inducing CYP2E1 (for isoniazid) or CYP3A4 (for phenytoin) or by competing with glucuronidation pathways (for zidovudine). Alcohol and fasting have dual effects by enhancing expression of CYP2E1 and by depleting hepatic glutathione. Fasting also may impair acetaminophen conjugation by depleting cofactors for the glucuronidation and sulfation pathways.21 Acetaminophen hepatotoxicity produces zone 3 hepatic necrosis, with extension into submassive (bridging) or panacinar (massive) necrosis in severe cases. Inflammation is minimal, and recovery is associated with complete resolution without fibrosis. The zonal pattern of acetaminophen-induced necrosis is related to the mechanism of hepatotoxicity, particularly the role of CYP2E1, which is expressed in zone 3, and to lower levels of glutathione in zone 3 hepatocytes than in hepatocytes in the other zones.

Clinical Course, Outcomes, and Prognostic Indicators

In the first two days after acetaminophen self-poisoning, features of liver injury are not present. Nausea, vomiting, and drowsiness are often caused by concomitant ingestion of alcohol and other drugs. After 48 to 72 hours, serum ALT levels may be elevated, and symptoms such as anorexia, nausea and vomiting, fatigue, and malaise may occur. Hepatic pain may be pronounced. In severe cases, the course is characterized by repeated vomiting, jaundice, hypoglycemia, and other features of acute liver failure, particularly coagulopathy and hepatic encephalopathy. The liver may shrink as a result of severe necrosis. Serum levels of ALT are often between 2000 and 10,000 U/L. These high levels, together with high levels of other intracellular proteins (ferritin, glutathione S-transferases), may provide a clue to the diagnosis in complex settings, as may occur with alcoholic patients and those with viral hepatitis.77 Indicators of a poor outcome73-76 include grade IV hepatic coma, acidosis, severe and sustained impairment of coagu-

1427

Section IX  Liver lation factor synthesis, renal failure, and a pattern of falling serum ALT levels in conjunction with a worsening prothrombin time (see also Chapter 93). Renal failure reflects acute tubular necrosis or the hepatorenal syndrome. Myocardial injury also has been attributable to acetaminophen toxicity.78 Death occurs between 4 and 18 days after the overdose and generally results from cerebral edema and sepsis complicating hepatic and multiorgan failure. A majority of patients recover completely. Cases of apparent chronic hepatotoxicity rarely have been attributed to continued ingestion of acetaminophen (2 to 6 g/day), usually in a susceptible host, such as a heavy drinker or a person with preexisting, unrecognized liver disease.5,6 Rare cases of acetaminophen hypersensitivity, typically involving skin or lung, have been reported in association with liver injury.80,81

Management

In patients who present within four hours of ingesting an excessive amount of acetaminophen, the stomach should be emptied with a wide-bore gastric tube. Osmotic cathartics or binding agents have little if any role in management. Charcoal hemoperfusion has no established role. The focus of management is on identifying patients who should receive thiol-based antidote therapy and, in those with established severe liver injury, assessing the patient’s candidacy for liver transplantation. Blood levels of acetaminophen should be measured at the time of presentation. Because of delayed gastric emptying, however, blood levels within four hours of ingestion may underestimate the extent of exposure. After four hours, acetaminophen blood levels give a reliable indicator of the risk of liver injury in patients with an acute overdose (not in those with a therapeutic misadventure). The risk of liver injury is then estimated by reference to the Prescott nomogram (Fig. 86-2).78 Indications for antidote therapy include a reliable history of major poisoning (more than 10 g) or blood acetaminophen levels in the moderate or high-risk bands on the monogram, or both.74,78 At-risk patients should be hospitalized for monitoring. Hepatic necrosis occurs only when glutathione concentrations fall below a critical level, thereby allowing NAPQI to produce liver injury. Administration of cysteine donors stimulates hepatic synthesis of glutathione. Many cysteine precursors or thiol donors could be used, but NAC has become the agent of choice. Oral administration is preferred in the United States,73,78 with a loading dose of 140 mg/kg followed by administration of 70 mg/kg every 4 hours for 72 hours. This regimen is highly effective, despite the theoretical disadvantage that delayed gastric emptying and vomiting may reduce intestinal absorption of NAC. In Europe and Australia, NAC is administered by slow bolus intravenous injection followed by infusion (150 mg/kg over 15 minutes in 200 mL of 5% dextrose, with a second dose of 50 mg/kg 4 hours later, if the blood acetaminophen levels indicate a high risk of hepatoxicity, and a total dose over 24 hours of 300 mg/kg).78 The intravenous route may be associated with a higher rate of hypersensitivity reactions because of the higher systemic blood levels achieved.5 Adverse reactions to NAC may be severe, with rash, angioedema, and shock, which occasionally is fatal.5 Therefore, NAC must be administered under close supervision and only for appropriate indications. In patients known to be sensitized to NAC, methionine is probably just as effective but is not available in a commercial preparation; it must be made up fresh and often causes vomiting.78 Cases of acetaminophen-induced severe liver injury are virtually abolished if NAC is administered within 12 hours and possibly within 16 hours of acetaminophen inges-

µg/mL µmol/L

Acetaminophen plasma concentration

1428

300

2000

200 150

1300 1000 900 800 700 600 500 400

100 90 80 70 60 50 40 30

300 250 200

Potential for toxicity

20

10 9 8 7 6 5 4 3

Toxicity unlikely

100 90 80 70 60 50 40 30

Treatment recommended if level is above broken line

20

2 10

4 Measure level at least 4 hours post-ingestion

8

12

16

20

24

28

32

36

Hours post-ingestion

Figure 86-2.  Acetaminophen toxicity nomogram. The risk of hepatotoxicity correlates with the plasma acetaminophen level and the time after ingestion. (From Smilkstein MJ, Knapp GL, Kulig KW, et al. Efficacy of oral N-acetylcysteine in the treatment of acetaminophen overdose. Analysis of the National Multicenter Study [1976-1985]. N Engl J Med 1988; 319:1557-62.)

tion.73,74,78 After 16 hours, thiol donation is unlikely to affect the development of liver injury because oxidation of acetaminophen to NAPQI with consequent oxidation of thiol groups is complete and mitochondrial injury and activation of cell death pathways are likely to be established. Nevertheless, NAC has been reported to decrease the mortality associated with acetaminophen-induced hepatotoxicity when administered 16 to 36 hours after self-poisoning,73,74,78 possibly because NAC stabilizes vascular reactivity in patients with liver failure. Therefore, administration of NAC is recommended for patients with a late presentation after acetaminophen overdose. Other strategies to protect the liver against acetaminophen poisoning, such as inhibition of CYP-dependent metabolism through the use of cimetidine or administration of prostaglandin analogs, which are efficacious in rats, have not been established as clinically useful. The constitutive androstane receptor (CAR) has been identified as a regulator of acetaminophen metabolism and hepatotoxicity in mice.82 Inhibition of CAR activity by administration of androstanol one hour after acetaminophen dosing blocks liver injury.83 Liver transplantation has been advocated as a therapeutic option for select patients in whom liver failure develops after acetaminophen poisoning.73,74 The selection of cases is based on the prognostic indicators discussed earlier and is strongly influenced by the prospects for successful psychological rehabilitation (see Chapter 95).74 In several series,

Chapter 86  Liver Disease Caused by Drugs about 60% of listed patients have been transplanted, and survival rates have exceeded 70%.74

Prevention

Safe use of acetaminophen involves adherence to the recommended maximum dose for healthy adults and children and education about the risk factors that lower the toxic dose threshold. Acetaminophen doses of more than 2 g a day are contraindicated in heavy drinkers, in those taking other medications (particularly phenytoin, zidovudine, and isoniazid), and during fasting. Prolonged use of acetaminophen requires caution in patients with severe cardiorespiratory disease or advanced cirrhosis. Use of acetaminophen for self-poisoning continues despite attempts at public education about the risks involved. The chances of harm from a suicidal gesture may be reduced by the sale of acetaminophen in smaller package sizes and in blister packs, which hamper ready access to the tablets or capsules.84,85

OTHER TYPES OF CYTOPATHIC LIVER INJURY

Some hepatotoxins are not as clearly dose dependent as acetaminophen but cause cytopathic or cytotoxic changes, such as extensive hydropic change, diffuse or zonal microvesicular steatosis, and zonal necrosis.5,6 Injury likely represents metabolic idiosyncrasy, in which the drug or one of its metabolites accumulates and interferes with protein synthesis or intermediary metabolism, or both. The mitochondrion often appears to be the main subcellular target, and other metabolically active tissues can be involved. Pancreatitis and renal tubular injury may accompany severe liver injury caused by valproic acid, tetracycline, and HAART, and metabolic acidosis with a shock-like state is common. The first agent recognized to cause this clinicopathologic syndrome was tetracycline administered in high doses (greater than 2 g/day for more than four days, usually intravenously) to pregnant women, to men taking estrogens, or to patients with renal failure.6 With appropriate dose limitations, this reaction is entirely preventable.

Niacin (Nicotinic Acid)

Hepatotoxicity associated with use of niacin, or nicotinic acid (3-pyridinecarboxylic acid), has been noted since the 1960s. When used to treat hypercholesterolemia, niacin has been an important cause of liver injury.86 It is a dosedependent hepatotoxin; liver injury usually occurs at doses that exceed 2 g/day, but in rare instances, low-dose (500 mg/ day) sustained-release niacin has been implicated in fulminant hepatic failure.87 Patients who are taking sulfonylurea drugs and those with preexisting liver disease, particularly alcoholic hepatitis, are at increased risk. No association with age, diet, or insulin-managed diabetes mellitus has been recognized. The symptoms of niacin hepatotoxicity begin as early as one week to as long as four years after the drug is started. The clinicopathologic spectrum encompasses mild and transient increases in serum ALT levels, jaundice, acute hepatitis, and cholestasis. Liver injury resolves completely when the drug is stopped. Liver biopsy specimens show hepatic necrosis and centrilobular cholestasis. Well-documented cases of fulminant hepatitis, some necessitating liver transplantation, also have been attributed to niacin. Substitution of one niacin preparation for another without a dose adjustment should be avoided; switching from immediate- to sustained-release preparations requires a 50% to 70% reduction in the dose of niacin.

Valproic Acid (Sodium Valproate)

Valproic acid-associated hepatic injury occurs almost exclusively in children, particularly those younger than three

years of age. Also at risk are persons with a family history of a mitochondrial enzyme deficiency (particularly involving the urea cycle or long-chain fatty acid metabolism), Friedreich’s ataxia, or Reye’s syndrome or with a sibling affected by valproic acid hepatotoxicity. Another risk factor is multiple drug therapy. Cases in adults have been described rarely. The overall risk of liver injury among persons taking valproic acid varies from 1 per 500 persons exposed among high-risk groups (children under age 3, polypharmacy, genetic defects of mitochondrial enzymes) to less than 1 in 37,000 in low-risk groups.88 No relationship exists between valproic acid toxicity and dose, but blood levels of valproic acid tend to be high in one half of affected persons. The metabolite 4-en-valproic acid, produced by CYP-catalyzed metabolism of valproic acid, is a dose-dependent hepatotoxin in animals and in vitro. The concept has emerged that valproic acid is an occult dose-dependent toxin in which accumulation of a hepatotoxic metabolite (favored by coexposure to CYPinducing antiepileptic agents) produces mitochondrial injury in a susceptible host (e.g., young children, especially those with partial deficiencies of mitochondrial enzymes).89 Symptoms begin 4 to 12 weeks after treatment with valproic acid is started and are often nonspecific, including lethargy, malaise, poor feeding, somnolence, worsening seizures, muscle weakness, and facial swelling. In typical cases, features of hepatotoxicity follow, including anorexia, nausea, vomiting, abdominal discomfort over the liver, and weight loss.88,89 When jaundice ensues, hypoglycemia, ascites, coagulation disorders, and encephalopathy indicate liver failure with imminent coma and death. In some cases, a neurologic syndrome characterized by ataxia, mental confusion, and coma predominates, with little evidence of hepatic involvement. In other cases, fever and tender hepatomegaly suggestive of Reye’s syndrome may be present (see later); such cases tend to have a better prognosis. Additional extrahepatic features may include alopecia, hypofibrinogenemia, thrombocytopenia, and pancreatitis. The terminal phase is often indicated by renal failure, hypoglycemia, metabolic acidosis, and severe bacterial infection. Laboratory features include modest elevations of serum bilirubin and aminotransferase levels; the aspartate aminotransferase (AST) level is usually higher than the ALT level. A profound decrease in clotting factor levels, hypoalbuminemia, and raised serum ammonia levels are common. A small liver with increased echogenicity suggestive of steatosis or extensive necrosis is seen on hepatic imaging. Histologic examination of the liver shows submassive or massive hepatic necrosis in two thirds of cases with either zonal or generalized microvesicular steatosis.89 Ultrastructural studies indicate conspicuous abnormalities of the mitochondria. Treatment is supportive. Prevention depends on careful adherence to prescribing guidelines, including the avoidance of valproic acid in combination with other drugs in the first three years of life and in children who may have mitochondrial enzyme abnormalities. Elevations in liver biochemical test levels develop in at least 40% of patients who take valproic acid and therefore are an unreliable predictor of severe hepatotoxicity. Warning patients and parents about the need to report any adverse symptoms during the first six months of valproic acid therapy is most important.

Antiretroviral Agents

Elevated liver biochemical test levels and clinical evidence of liver disease are common in patients with HIV/AIDS. Reasons for this finding include HBV or HCV infection,

1429

1430

Section IX  Liver other hepatobiliary infections, lymphoma and other tumors, and possibly effects of HIV infection itself. The frequency of hepatic injury with HAART (which often includes three or four agents) is at least 10%.23,27,28,90 The agents used can be broadly categorized as nucleoside (or nucleotide) reverse transcriptase inhibitors, non-nucleoside reverse transcriptase inhibitors, and protease inhibitors. Because HIV coinfection with HBV or HCV increases the risk of toxicity, all patients should be screened for viral hepatitis before starting HAART.28 Nucleoside (or Nucleotide) Reverse Transcriptase Inhibitors Nucleosides and nucleotides that block HIV reverse transcriptase are also weak inhibitors of mitochondrial DNA polymerase gamma in vitro; the order of their potency is: zalcitabine > didanosine > stavudine > lamivudine > zi­ dovudine > abacavir.91 The mechanisms of hepatotoxicity may also involve oxidative stress, resulting in further deletion of mitochondrial DNA, and the consequences of impaired oxidative phosphorylation, fatty acyl β-oxidation, and insulin resistance. In clinical studies, zidovudine, didanosine, and stavudine are the agents implicated most often in liver injury.90-93 Risk factors for mitochondrial drug toxicity among persons with HIV infection include obesity, female gender, and absence of an AIDS-defining illness.90-94 Hallmarks of mitochondrial hepatotoxicity include extensive microvesicular or macrovesicular steatosis (or both), lactic acidosis, and liver biochemical test abnormalities with progression to acute liver failure. Asymptomatic hyperlactatemia is common (especially with stavudine) among persons treated with HAART,94 but life-threatening lactic acidosis with hepatic steatosis is rare, with an estimated risk of 1.3 per 1000 person-years of antiretroviral use. The onset is a median of 6 months (with a range of 3 to 17 months) after treatment is started. Patients present with symptoms that are nonspecific and include nausea, vomiting, diarrhea, dyspnea, lethargy, and abdominal pain. Extrahepatic manifestations, such as myopathy or peripheral neuropathy, and in severe cases pancreatitis and renal failure, may follow the onset of the lactic acidosis and liver injury. Discontinuation of the drug is mandatory but does not prevent fatalities. Nevertheless, the overall mortality rate is low. One suggested approach to prevention is to monitor therapy with nucleos(t)ide reverse transcriptase inhibitors by coupling serum ALT and AST testing with serial measurements of the HIV load and CD4 count. Any new aminotransferase elevation should be followed immediately by measurement of serum lactate, muscle, and pancreatic enzyme levels.92 Non-nucleoside Reverse Transcriptase Inhibitors Non-nucleoside reverse transcriptase inhibitors occasionally may cause hepatitis as part of a hypersensitivity reaction within the first six weeks of use.90,95,96 Reactions are usually accompanied by peripheral and tissue eosinophilia, skin rash, and lymphadenopathy. Resolution occurs within four weeks of discontinuing the drug.95 Nevirapine also has been implicated in several instances of severe hepatotoxicity,97,98 including cases among healthcare workers in whom nevirapine was used for post-exposure prophylaxis against HIV infection.97 The FDA received 12 reports of such hepatotoxic reactions between 1997 and 2000; liver failure requiring hepatic transplantation developed in one person, seven had clinical features of hepatitis (jaundice, fever, nausea, vomiting, abdominal pain, and hepatomegaly), and four others had elevated serum aminotransferase levels without symptomatic illness. The recommended two-week

dose escalation regimen was not adhered to in some of the cases.99 Sequential toxicity with nevirapine followed by efavirenz has been reported in an HIV-HCV coinfected person.100 Protease Inhibitors Elevation of liver enzymes occurs commonly with protease inhibitors, but clinical hepatitis is infrequent. The agents most often implicated in liver injury are ritonavir and indinavir. The latter also may be associated with unconjugated hyperbilirubinemia in 7% of treated persons, a finding that is of no clinical consequence.9 Severe acute hepatitis may occur rarely. The association with peripheral or tissue (in liver biopsy specimens) eosinophilia in some cases suggests an immunoallergic basis for liver injury.101,102 Acute hepatitis also has been reported in 2.9% to 30% of persons who take ritonavir.103 The course of the illness is generally mild, and the liver injury responds favorably to drug withdrawal. Rarely, acute liver failure may develop; in these cases, liver histologic examination has shown severe microvesicular steatosis, cholestasis, and extensive fibrosis. Several studies have addressed the potential influence of underlying chronic viral hepatitis on the toxicity of protease inhibitors. Although hepatotoxicity appeared to be more common, liver injury was rapidly reversible in most cases; this observation suggests that the overall effect of protease inhibitors in coinfected persons is not detrimental.104 Many protease induce or inhibit CYP3A4, thereby causing important drug-drug interactions.105 Furthermore, the immune reconstitution that can follow successful HAART may cause a flare-up of previously quiescent chronic hepatitis B (see Chapter 33).

Aspirin

Aspirin occasionally has been associated with major increases in serum ALT levels suggestive of drug hepatitis, but hepatotoxicity occurs only when blood salicylate concentrations exceed 25 mg/100 mL.106 In addition, individual susceptibility factors include hypoalbuminemia, active juvenile rheumatoid arthritis, and systemic lupus erythematosus. Most cases of aspirin-induced hepatotoxicity have been identified by biochemical testing, rather than clinical features. If present, symptoms usually begin within the first few days or weeks of high-dose aspirin therapy. Acute liver failure and fatalities have been rare. Resolution occurs rapidly after drug withdrawal, and salicylates can be reintroduced at a lower dose. All salicylates appear to carry hepatotoxic potential so there is no advantage to replacing aspirin with another salicylate. Liver biopsy specimens reveal a nonspecific focal hepatitis with hepatocellular degeneration and hydropic changes. Steatosis is not usually present, and the absence of steatosis distinguishes aspirin hepatotoxicity from Reye’s syndrome. Reye’s syndrome has been linked with use of aspirin in febrile children. Although Reye’s syndrome is not simply a form of drug-induced liver disease, aspirin plays an important role in its multifactorial pathogenesis. Reye’s syndrome usually occurs between three and four days after an apparently minor viral infection. It is characterized by acute encephalopathy and hepatic injury, the latter documented by a three-fold or greater rise of serum aminotransferase or ammonia levels and by characteristic histologic findings. Because of effective public health campaigns against the use of aspirin in young febrile children, the incidence of Reye’s syndrome has declined markedly.107 Patients with juvenile rheumatoid arthritis (Still’s disease) or systemic lupus erythematosus appear to be at particular risk of Reye’s syndrome. Clinical and laboratory features of

Chapter 86  Liver Disease Caused by Drugs Table 86-4  Types of Drug-Induced Acute Hepatitis: Immunoallergic Reaction versus Metabolic Idiosyncrasy CHARACTERISTIC

IMMUNOALLERGIC reaction

METABOLIC IDIOSYNCRASY

Frequency Gender predilection Latent period to onset Relationship to dose Interactions with other agents

50 yr; 80% of cases in women; repeated courses, continued ingestion in sensitized patient Age >65 yr; most cases in women

Clinical features of chronic hepatitis, liver failure; some cases with features of cholestasis; 20% with pneumonitis; hyperglobulinemia usual, ANA, SMA Jaundice, diarrhea, liver failure; hyperglobulinemia, ANA, SMA positive; protracted course Clinical features of chronic hepatitis, liver failure; ANA, SMA, hyperglobulinemia Often part of drug-induced systemic lupus erythematosus syndrome (arthritis, rash, rarely nephritis); ANA, hyperglobulinemia

Diclofenac Minocycline

Young women; prolonged use of drug

Isoniazid

Age >50 yr; continued drug ingestion after onset; duration of therapy Age >30 yr; dose, duration of therapy Age >50 yr; two thirds in women

Dantrolene Etretinate Acetaminophen

Regular intake at moderate doses (2 to 6 g/day); alcohol, fasting, other drugs

High mortality rate

Severe and fatal cases with cirrhosis; no immune phenomena

Response to glucocorticoids in a few cases Cases may be severe, with fatal outcome or need for liver transplantation; glucocorticoid treatment may be indicated High mortality rate or need for liver transplantation

Jaundice, liver failure; no immune phenomena

High mortality rate

Jaundice, weight loss, liver failure; deterioration after stopping drug No features of chronic liver disease, no autoimmune phenomena; there are cases of chronic toxicity

Response to glucocorticoids in two reported cases Rapid normalization of liver biochemical test levels after drug is stopped

*Other drugs include oxyphenisatin and tienilic acid, which are now of historical interest,5,6 and clometacin, for which many presumably affected patients have now been shown to have had hepatitis C.100 Several other agents, including sulfonamides, aspirin, halothane, cimetidine, methotrexate, trazadone, fluoxetine, fenfibrate, and germander, have been mentioned as associated with chronic hepatitis, but details are not always convincing.5 ANA, antinuclear antibodies; SMA, smooth muscle antibodies.

Chapter 86  Liver Disease Caused by Drugs alleles, are not found. Treatment with immunosuppressants usually is not indicated; the clinical condition improves spontaneously after withdrawal of the causative drug. In individual cases, however, glucocor­ticoids occasionally appear to hasten recovery (see later).

DICLOFENAC

Diclofenac is one of the world’s most prescribed NSAIDs and appears to be at least as safe as comparable agents. Significant hepatotoxicity occurs in about 1 to 5 per 100,000 persons exposed, or 0.4 per 1 million defined daily doses; the latter rate is minimally greater than that for phenylbutazone (0.2 per 1 million) and piroxicam (0.3 per 1 million) but less than that for benoxaprofen (12.6 per 1 million) and bromfenac, which was withdrawn from the market (see earlier). More than 200 cases of diclofenac hepatitis have been reported,295 including several proved by indavertent rechallenge. Only four cases have been fatalities, and five cases can reasonably be regarded as chronic hepatitis. Genetic susceptibility to diclofenac hepatotoxicity has been documented.296,297 In these cases, polymorphisms have been observed within genes that affect metabolic pathways that lead to formation of reactive metabolites of the drug; biliary excretion and the immune response to drug metaboliteprotein adducts have been identified.296,297 The risk of diclofenac hepatitis is increased in women and with aging. A prodromal illness characterized by anorexia, nausea, vomiting, and malaise heralds the onset of liver injury, which usually occurs within 3 months (range, 1 to 11 months) of the start of treatment. Fever and rash occur in 25% of patients.295 Liver biochemical test results reflect acute hepatitis with or without cholestasis. Reactions tend to be severe, with jaundice occurring in 50% of cases. Liver biopsy specimens reveal acute lobular hepatitis, and in severe cases, bridging or confluent necrosis, interface hepatitis, and fibrous expansion of the portal tracts had been noted.298 The prognosis is usually good; resolution occurs after discontinuation of the drug. Cases of druginduced chronic hepatitis have been described in which the clinical and laboratory features (ascites, hypoalbuminemia, hyperglobulinemia, jaundice) suggested autoimmune hepatitis, although the frequency of autoantibodies is unclear. These cases usually improve spontaneously after discontinuation of the drug, but glucocorticoids have been used successfully in a few protracted cases.299 Cross-sensitivity with other NSAIDs seems to be rare, but one patient with diclofenac hepatitis also had an adverse reaction to ibuprofen and another had an adverse reaction to tiaprofenic acid.299 The rarity of severe diclofenac-induced hepatotoxicity makes liver biochemical monitoring unrealistic. Patients need to be advised to report adverse effects, and clinicians must be aware that diclofenac can cause both acute and chronic hepatitis.

MINOCYCLINE

Minocycline has been associated with rare cases of druginduced systemic lupus erythematosus syndrome (rash, polyarthritis, hyperglobulinemia, and antinuclear antibodies), chronic hepatitis with autoimmune features, and both syndromes in the same patient.300,301 The onset often occurs after treatment with minocycline for more than six months, and young women appear to be particularly affected. The reactions are severe; some patients have died or required liver transplantation.302 Progression to cirrhosis has been reported.303 The course may be prolonged after the drug is discontinued; several patients have been treated with glucocorticoids.301

DRUG-INDUCED ACUTE CHOLESTASIS IMPORTANCE, TYPES OF REACTIONS, AND DIAGNOSIS

Drugs are an important cause of acute cholestasis, with or without hepatitis.304 The full spectrum of drug-related disorders includes cholestatic hepatitis with cholangitis and chronic cholestasis, either with a vanishing bile duct syndrome resembling primary biliary cirrhosis or with biliary strictures reminiscent of sclerosing cholangitis.305,306 The clinical and biochemical features of drug-induced cholestasis resemble those of several other hepatobiliary disorders, and clinicians must take a thorough drug history from all patients with cholestasis. The prompt discontinuation of exposure to a causative agent prevents an adverse outcome and avoids unnecessary invasive investigations or surgery. The clinical syndrome of cholestasis is characterized by pruritus, dark urine, pale stools, and, in more serious cases, jaundice. Liver biochemical test results show a predominant elevation of the serum alkaline phosphastase level, with a lesser increase in the serum ALT level, elevation of GGTP and 5′-nucleotidase levels, raised serum bile acid levels, and conjugated hyperbilirubinemia. The serum ALT level may be elevated up to eight-fold, as a result of either the toxic effects of acute bile retention on hepatocellular integrity or concomitant “hepatitis.” In such cases, the ratio of the relative increases in serum ALT and alkaline phosphatase levels (based on multiples of the upper limits of normal) is typically less than 2 : 1 in patients with cholestasis.305 Cases of mixed cholestasis and hepatitis are highly suggestive of a drug reaction. Hepatobiliary imaging is essential for excluding dilatation of bile ducts produced by biliary obstruction and for detecting a hepatic or pancreatic mass lesion. In the absence of such findings, drug-induced cholestasis is more likely, and a liver biopsy is often advisable. Certain histologic features suggest a hepatic drug reaction, whereas others (e.g., edema of the portal tracts) suggest biliary obstruction. When the temporal relationship to drug ingestion indicates a high probability of a drug reaction, particularly when the agent is known to be potentially hepatotoxic, the incriminated drug should be discontinued and the patient observed for improvement. Management should focus on symptom relief; with particular attention to pruritus (see Chapters 20 and 89).305-307 Pruritus is often ameliorated with cholestyramine. In intractable cases, ursodeoxycholic acid has shown promise.307,308 Rifampin can be tried, and phototherapy, plasmapheresis, and morphine receptor antagonists (e.g., naloxone, naltrexone, nalmefene) have been used as third-line therapies.307 Glucocorticoids have no role. Phenobar­bital and antihistamines are usually ineffective or cause oversedation.

CHOLESTASIS WITHOUT HEPATITIS

Cholestatic reactions are characterized by the retention of bile in canaliculi, Kupffer cells, and hepatocytes, with minimal inflammation or hepatocellular necrosis. Synonyms are pure, canalicular, or bland cholestasis. Cholestasis without hepatitis reflects a primary disturbance in bile flow. Sex steroids are the typical causative agents. Some drugs generally associated with cholestatic hepatitis occasionally produce bland cholestasis, (e.g., amoxicillinclavulanic acid, sulfonamides, griseofulvin, ketoconazole, tamoxifen, warfarin, ibuprofen).305,306 Cyclosporine is associated with liver biochemical test abnormalities; the fea-

1439

1440

Section IX  Liver tures resemble those of cholestasis, but hyperbilirubinemia usually is predominant.5 The reaction is mild and reverses rapidly with a reduction in dose. Cholestasis also has been documented with tacrolimus,309 whereas sirolimus has been implicated in cases of mild acute hepatitis.310

oxypenicillins,319 ketoconazole [see earlier],211 sulfonylureas, sulindac,320 ibuprofen, phenylbutazone, piroxicam,321 captopril,322 flutamide,323 enalapril,153 pravastatin,324 ator­ vastatin,325 ticlopidine,326 ciprofloxacin,327 norfloxacin,328 and metformin.329)

Steroids

Chlorpromazine

Oral Contraceptive Steroids The frequency of cholestasis with oral contraceptive steroids (OCS) is 2.5 per 10,000 women exposed. The occurrence of cholestasis with OCS is partly dose dependent and less likely with low-dose than high-dose estrogen preparations.311 Genetic factors influence the frequency of this complication, with a particularly high rate observed among women in Chile and Scandinavia.306 Persons with a history of cholestasis of pregnancy are also at risk (50%). The estrogenic component is most likely responsible. Symptoms develop two to three months, rarely as late as nine months, after OCS are started. A mild transient prodrome of nausea and malaise may occur and is followed by pruritus and jaundice. The serum alkaline phosphatase level is moderately elevated, and serum aminotransferase levels are increased transiently, occasionally to levels exceeding 10 times the upper limit of normal. The serum GGTP level is often normal. Recovery is usually prompt, within days to weeks after cessation of the drug. Chronic cholestasis is rare.306 Hormonal replacement therapy (HRT) is safe in patients with liver disease. Jaundiced patients, however, may experience an increase in serum bilirubin levels, and liver biochemical test levels should be monitored in HRT users with liver disease.306 Anabolic Steroids At high doses, anabolic steroids often produce reversible bland cholestasis, usually within one to six months of the start of treatment. Recovery usually follows drug withdrawal, but protracted cholestasis with biliary ductopenia can occur.312,313 Rarely, anabolic steroid use can also cause a predominantly acute hepatocellular pattern of injury.314 Both OCS and the 17-alkylated anabolic steroids are associated with cholestasis, vascular lesions, and hepatic neoplasms (see later). The strength of these associations with individual lesions varies. Benign hepatic neoplasms, except hemangiomas,315 are clearly associated with use of OCS, whereas their association with hepatocellular carcinoma is controversial.316 By contrast, hepatocellular carcinoma is well documented in users of anabolic steroids. Likewise, hepatic and portal vein thrombosis is an established adverse effect of OCS but not of anabolic steroids. Other vascular lesions, such as peliosis hepatis (see Chapter 83), are observed more often with anabolic steroids than with OCS.

CHOLESTASIS WITH HEPATITIS

Cholestasis with hepatitis is a common type of hepatic drug reaction and is characterized by conspicuous cholestasis and hepatocellular necrosis. Histologic lesions in the liver include lobular and portal tract inflammation, often with neutrophils and eosinophils, as well as mononuclear cells. This type of reaction overlaps with drug-induced acute hepatitis (occasionally resulting in acute liver failure), cholestasis without hepatitis, and cholestasis with bile duct injury. Causative agents include chlorpromazine (see later), antidepressants and other psychotropic agents, erythromycins and other macrolides, and related ketolide anti­biotics (telithromycin,317 clindamycin,318 sulfonamides,

Chlorpromazine hepatitis, the prototypical drug-induced cholestatic hepatitis,330 has been recognized since the 1950s, and cases still occur. The full spectrum of hepatic reactions to chlorpromazine includes asymptomatic liver biochemical test abnormalities in 20% to 50% of persons exposed to the drug and rare cases of fulminant hepatic necrosis. The frequency of cholestatic hepatitis varies from 0.2% to 2.0%, depending on the type of study; the lower value probably is representative of the risk in the general population. No relationship to dose or to underlying liver disease has been recognized. Female predominance is evident. Reactions do not appear to be more common with increasing age but are rare in children. The onset of cholestatic hepatitis is generally 1 to 6 weeks after the start of chlorpromazine and occasionally 5 to 14 days after its discontinuation. Accelerated onset occurs with rechallenge. A prodromal illness of fever and nonspecific symptoms is usual and is followed by gastrointestinal symptoms and jaundice. Pruritus is common and occurs later with chlorpromazine hepatitis than with druginduced cholestasis without hepatitis. In a small proportion of affected patients, right upper quadrant abdominal pain is severe. Rash is infrequent. Liver biochemical test results show elevation of both serum ALT and alkaline phosphatase levels and hyperbilirubinemia. Eosinophilia is present in 10% to 40% of patients. Most patients with chlorpromazine hepatitis recover completely—one third within four weeks, another third between four and eight weeks, and the remainder after eight weeks.308,330 In about 7% of cases, full recovery has not occurred by six months (see later).

Amoxicillin-Clavulanic Acid

At least 150 cases of cholestatic hepatitis have been attributed to the use of amoxicillin-clavulanic acid (e.g., Augmentin), a commonly prescribed antibiotic. The overall frequency is 1.7 cases per 10,000 prescriptions19; male gender, increasing age (older than 55 years), and possibly use of the antibiotic for a prolonged period of time are risk factors.331 The clavulanic acid component has been implicated because similar lesions have been noted with ticarcillin-clavulanic acid,332 whereas amoxicillin rarely causes liver disease.333 The onset of symptoms is within 6 weeks (mean 18 days) of the start of drug therapy, although rarely the onset is delayed up to 6 weeks after discontinuation of the drug. Features of hypersensitivity such as fever, skin rash, and eosinophilia are seen in 30% to 60% of patients. Liver biopsy specimens show cholestasis with mild portal inflammation. Bile duct injury (usually mild) and perivenular bilirubinostasis with lipofuscin deposits are often present. Other histologic features include hepatic granulomas, biliary ductopenia, and cirrhosis.334 Most patients recover completely in 4 to 16 weeks; fatal outcomes are rare. A strong association with the HLA-DRB1*1501-DRB5*0101DQB1*062 haplotype, supports the view that an immunologic idiosyncrasy mediated through HLA class II antigens could play a pathogenic role in this form of drug-induced cholestatic hepatitis. The presence of this haplotype, however, has no influence on the clinical characteristics, severity, and outcome of the disease.

Chapter 86  Liver Disease Caused by Drugs CHOLESTATIC HEPATITIS WITH BILE DUCT INJURY

Bile duct (cholangiolytic) injury is observed with several drugs that cause cholestatic hepatitis, such as chlorpromazine308 and flucloxacillin.319 The severity of bile duct injury may be a determinant of the vanishing bile duct syndrome (see later).335,336 The clinical features may resemble those of bacterial cholangitis, with upper abdominal pain, fever, rigors, tender hepatomegaly, jaundice, and cholestasis. Liver biochemical test levels are typical of cholestasis. Compounds associated with this syndrome include arsphenamine, carbamazepine,337 dextropropoxyphene,338 and methylenediamine, an industrial toxin responsible for Epping jaundice, an outbreak of jaundice associated with the ingestion of bread made from contaminated flour (see Chapter 87).339

Dextropropoxyphene

Dextropropoxyphene, an opioid analgesic used alone or in compound analgesics, has caused cholestasis with bile duct injury in at least 25 reported cases,338 some proved by inadvertent rechallenge. A female predominance has been recognized. The onset of symptoms is usually within two weeks of the start of dextropropoxyphene. Illness is often heralded by abdominal pain, which may be severe and simulate other causes of cholangitis. Jaundice is usual. ERCP shows normal bile ducts. Liver biopsy specimens demonstrate cholestasis with expansion of the portal tracts by inflammation and mild fibrosis; portal tract edema also may be present. Other features include irregularity and necrosis of the biliary epithelium, together with an infiltrate of neutrophils and eosinophils on the outer surface of bile ducts. Bile ductule proliferation is universal. Recovery has occurred in all reported cases.338 Liver biochemical test levels normalize between one and three months after discontinuation of the drug. This type of hepatic drug reaction must be distinguished from bile duct obstruction and bacterial cholangitis.

DRUG-INDUCED CHRONIC CHOLESTASIS Drug-induced liver disease is considered to be chronic when typical liver biochemical changes last longer than 3 months306; earlier definitions required the presence of jaundice for more than 6 months or anicteric cholestasis (raised alkaline phosphatase and GGTP levels) for more than 12 months after the implicated agent was stopped.305 Drug-induced chronic cholestasis is uncommon but has been ascribed to more than 45 compounds.305-307,330,340-342 Chronicity complicates approximately 7% of cases of chlorpromazine hepatitis308 and is a feature in 10% to 30% of cases of flucloxacillin hepatitis.306 Chronicity has been reported in less than 5% of cases of erythromycin hepatitis341 and in only isolated instances for other agents, such as tetracycline,342 amoxicillin-clavulanic acid,343 ibuprofen,344,345 trimethoprim-sulfamethoxazole,346 and ciprofloxacin.347 Chronic cholestasis always is preceded by an episode of acute cholestatic hepatitis. The episode of acute cholestatic hepatitis tends to be severe and occasionally is associated with the Stevens-Johnson syndrome.345 One study indicated that the severity of bile duct lesions at the time of the initial hepatic reaction is a critical determinant of a chronic course.336 Other possible mechanisms include continuing toxic or immunologic destruction of the biliary epithelium.340 The hepatic histologic lesion is characterized by a

paucity of smaller (septal, interlobular) bile ducts and ductules, often with residual cholestasis, and portal tract inflammation directed against injured bile ducts. This process may lead to an irreversible loss of biliary patency and the vanishing bile duct syndrome.348 The clinical features are those of chronic cholestasis. Pruritus is the dominant symptom and is often severe. Continuing jaundice, dark urine, and pale stools are possible but not invariable findings and may resolve despite persistence of liver biochemical abnormalities. In severe cases, intestinal malabsorption, weight loss, and bruising caused by vitamin K deficiency may occur; xanthelasma, tuberous xanthomata, and other complications of severe hypercholesterolemia also have been noted. Firm hepatomegaly may be found on physical examination, but splenomegaly is unusual unless portal hypertension develops. Antimitochondrial antibodies are not a feature of drug-induced chronic cholestasis. Cases usually have a favorable outcome, with resolution of jaundice in most instances. Progression to biliary cirrhosis is rare305,306 and is associated with a severe reduction in the number of bile ducts.

FLUCLOXACILLIN

Flucloxacillin is one of the most important causes of druginduced hepatitis in Europe, Scandinavia, and Australia.319,349 Flucloxacillin-induced hepatotoxiciy is usually severe, and several fatalities have resulted from the systemic features and associated cholestatic hepatitis. The course is prolonged, and a high proportion of cases result in chronic cholestasis and the vanishing bile duct syndrome.349 Other oxypenicillins appear to be less prone to cause this complication, but cholestasis has been reported with cloxacillin and dicloxacillin.319,350 Acute hepatocellular injury has been reported with oxacillin.351

FIBROTIC BILE DUCT STRICTURES

Chronic cholestasis caused by some drugs may result from the development of fibrotic strictures of the larger bile ducts. This complication has been associated with intralesional therapy of hepatic hydatids with formalin352 and intraarterial infusion of floxuridine for metastatic colorectal carcinoma.353 After several months of floxuridine infusion, the frequency of toxic hepatitis or bile duct injury, or both, is 25% to 55%. Acalculous cholecystitis also may occur. ERCP shows strictures, typically in the common, left, and right hepatic ducts. Unlike primary sclerosing cholangitis, the bile duct and the smaller intrahepatic bile ducts are spared. Ischemia has been suspected, and toxicity to biliary epithelial cells is another possibility. Recovery may occur after floxuridine is discontinued. Some patients require dilation or stenting of biliary strictures.

DRUG-INDUCED STEATOHEPATITIS AND HEPATIC FIBROSIS Drug-induced liver disease can produce cirrhosis by a variety of processes; chronic hepatitis and chronic cholestasis with the vanishing bile duct syndrome have already been discussed. Steatohepatitis is a form of chronic liver disease in which fatty change is associated with focal liver cell injury, Mallory’s hyaline, focal inflammation of mixed cellularity, including neutrophils, and progressive hepatic fibrosis in a pericentral (zone 3) and pericellular distribution.354 Alcohol is a common etiologic factor. NASH is associated with insulin resistance, diabetes mellitus, obesity, and several drugs (e.g., perhexiline maleate and amioda-

1441

1442

Section IX  Liver rone).355 In addition to causing steatohepatitis or chronic injury to liver cells or bile ducts, some exogenous compounds appear to promote hepatic fibrogenesis directly, most likely through effects on hepatic nonparenchymal cells; stellate cells are central to this process (see Chapter 71). Compounds that stimulate hepatic fibrosis include arsenic, vitamin A, and methotrexate.

AMIODARONE

Amiodarone is an iodinated benzofuran derivative used for therapy-resistant ventricular tachyarrhythmias. Adverse effects lead to discontinuation of therapy in 25% of patients. These adverse effects include pulmonary infiltrates, worsening cardiac failure, hypothyroidism, peripheral neuropathy, nephrotoxicity, and corneal deposits, but liver disease is one of the most serious. The spectrum of abnormalities includes abnormal liver biochemical test levels in 15% to 80% of patients and clinically significant liver disease, including rare cases of acute liver failure, in 0.6%. Liver disease also has been reported in patients who receive an intravenous loading dose of amiodarone; the toxic ingredient is likely to be the vehicle (polysorbate 80) rather than amiodarone because oral amiodarone has been used subsequently in these cases without a problem.356-361 The most typical hepatic lesion is steatohepatitis; cirrhosis develops in 15% to 50% of patients with hepatoxicity.357,358 A notable feature of amiodarone-induced liver disease is that progression of the disease may occur despite discontinuation of amiodarone.358,361 Amiodarone is highly concentrated in the liver, and after a few weeks of treatment, the drug accounts for as much as 1% of the wet weight of the liver. The iodine content absorbs radiation, so that the liver appears opaque on computed tomography scans.361 Although odd, this appearance is not clinically significant. Hepatic storage of amiodarone also produces phospholipidosis, a storage disorder characterized by enlarged lysosomes stuffed with whorled membranous material (myeloid bodies). In animals fed amiodarone, development of phospholipidosis is time and dose dependent.359 Phospholipi­ dosis may result from the direct inhibition of phospholipase or from the formation of nondegradable drug-phospholipid complexes and appears to have no relationship to the development of NASH and hepatocyte injury. Other occasional hepatic abnormalities include granuloma formation and acute liver failure, apparently caused by severe acute hepatitis or a Reye’s syndrome-like illness.362 Amiodarone, by virtue of its physicochemical properties, is concentrated in mitochondria and may interrupt mitochondrial electron transport.363 In rats and mice, treatment with amiodarone produces microvesicular steatosis, augments mitochondrial production of ROS, and causes lipid peroxidation.364,365 Chronic liver disease is detected only 1 year or more (median, 21 months) after amiodarone is started. The duration of amiodarone therapy and possibly the total dose,362,366 but not the incremental dose, are related to the development of chronic liver disease. Cases of cirrhosis with low-dose amiodarone have been documented.367 The frequency of other toxic effects of amiodarone (most of which are thought to be dose dependent) is increased in patients with liver disease.366 Patient complaints include fatigue, nausea and vomiting, malaise, weight loss, and abdominal swelling as a result of ascites. Hepatomegaly, jaundice, bruising, and other features of chronic liver disease may be present. Liver biochemical test abnormalities include increased aminotransferase levels, often to at least five times the upper limit

of normal, and minor increases in the serum alkaline phosphatase level. The ratio of serum AST to ALT levels is close to unity and thus differs from the ratio seen in patients with alcoholic hepatitis. In severe cases, hyperbilirubinemia, a low serum albumin level, and prolongation of the prothrombin time are evident. Diagnosing the cause of abnormal liver biochemical test results and hepatomegaly is often difficult in patients taking amiodarone, and a liver biopsy may be indicated. The histologic changes in the liver include phospholipidosis, steatosis, focal necrosis with Mallory’s hyaline, infiltration with neutrophils, and pericellular fibrosis.358 Cirrhosis is often present. Prevention and management of amiodarone-induced liver disease are problematic because liver biochemical test abnormalities are common in patients who take amiodarone, particularly in those with congestive heart failure. Furthermore, the frequency of amiodarone hepatotoxicity does not appear to differ between patients with and without baseline serum ALT elevations, and amiodarone should not be withheld in patients with an elevated serum ALT level.368 In asymptomatic or less severe cases of amiodarone hepatotoxicity, abnormalities resolve in two weeks to four months after amiodarone is discontinued. In cases of severe liver disease, the mortality rate is high.358,366 Cessation of amiodarone therapy does not always result in clinical improvement, presumably because of prolonged hepatic storage of amiodarone, and in one study, outcome was worse (usually from fatal arrhythmias) in patients who discontinued amiodarone than in those who did not.358 Although serial liver biochemical test measurements are recommended in patients who take amiodarone,366 whether such testing is adequate to prevent serious hepatotoxicity and reduce the overall mortality rate is unknown.

TAMOXIFEN AND OTHER CAUSES OF DRUG-INDUCED STEATOHEPATITIS

For agents reported to be associated with steatohepatitis during the 1990s, causality has been difficult to prove,369 particularly because NASH is a common disorder among patients with obesity, insulin resistance, or metabolic syndrome (see Chapter 85). Calcium channel blockers, used to treat hypertension and cardiac arrhythmias, have rarely been associated with steatohepatitis,370,371 and methyldopa has been reported to be associated with cirrhosis in obese middle-aged women372; however, these associations may have been fortuitous. Other drugs, including estrogens373 and glucocorticoids,374 may precipitate NASH in predisposed persons because of their effects on the risk factors for NASH—insulin resistance, type 2 diabetes mellitus, obesity, and hypertriglyceridemia. On the other hand, the association between NASH and tamoxifen appears to be much stronger. Tamoxifen is an estrogen-receptor ligand with both agonist and antagonist actions. It is widely used for the prevention and treatment of breast cancer. Several forms of liver injury have been attributed to tamoxifen375: cholestasis,376 hepatocellular carcinoma,377 peliosis hepatis,378 acute hepatitis, massive hepatic necrosis,375 steatosis, and steatohepatitis—occasionally with cirrhosis.379-385 In one series of 66 women with breast cancer who had received tamoxifen for three to five years, 24 showed radiologic evidence of hepatic steatosis.382 Seven other patients have been diagnosed with NASH (proved by liver biopsy) after taking tamoxifen for 7 to 33 months.380,381,383 The metabolic profile of women with radiologic evidence of hepatic steatosis (or histologic proof of steatohepatitis) during tamoxifen therapy appears similar to that of most

Chapter 86  Liver Disease Caused by Drugs patients with NASH; one half have been obese, and the increase in body mass index has correlated with hepatic steatosis.386 Tamoxifen can induce hypertriglyceridemia, another risk factor for NASH. Treatment with bezafibrate, a PPAR-α stimulator, has been reported to decrease the severity of hepatic steatosis, as assessed radiologically, markedly.387 Therefore, tamoxifen may play a synergistic role with other factors such as insulin resistance, hyperli­ pidemia, and obesity in causing steatohepatitis. This hypothesis is supported by results of an Italian tamoxifen chemoprevention trial in which development of fatty liver or steatohepatitis was confined mainly to overweight or obese women with the metabolic syndrome.388 Physicians need to be aware of the high frequency (approximately 30%) of hepatic steatosis, as determined by hepatic imaging, or steatohepatitis in women who receive tamoxifen. Patients who take tamoxifen should be monitored for this adverse effect by physical examination (to detect hepatomegaly) and liver biochemical testing; some authors also advocate annual hepatic imaging (by ultrasonography or computed tomography).389 Liver biopsy may be indicated to establish the severity of the disorder, particularly if liver biochemical test abnormalities do not resolve after tamoxifen is discontinued, or, in some cases, to exclude metastatic breast cancer. Many patients appear to improve after tamoxifen is discontinued, but whether treatment should always be withdrawn permanently is not clear, particularly because the effect of tamoxifen on survival from breast cancer is impressive. An alternative option to bezafibrate (see earlier) is to consider the use of exemstane (or other aromatose inhibitor), which can lower serum triglyceride levels.390 Toremifene, an analog of tamoxifen, also has been reported to cause steatosis or steatohepatitis, but with a lower frequency (less than 10%) than that reported for tamoxifen.389 Raloxifene, a selective estrogen receptor modulator, has been implicated in a case of steatohepatitis391 and a case of acute hepatocellular injury accompanied by eosinophilia.392 Although evidence of acute hepatocellular injury was provided in the first case, causality could not be established in

the second case because preexisting liver disease (NASH) could not be excluded. Perhexiline maleate and coralgil (4,4′-diethylaminoethoxyhexestrol) are definite toxic causes of steatohepatitis and have been withdrawn from the U.S. market.369

METHOTREXATE

Methotrexate is a dose-dependent toxin. In high doses, methotrexate can result in bone marrow suppression, mucocutaneous reactions, pneumonitis, and hepatotoxicity. In the 1950s, previous methotrexate treatment of acute childhood leukemia was shown to be complicated by severe hepatic fibrosis and cirrhosis; a few cases were complicated by hepatocellular carcinoma.393 In the 1960s, the use of methotrexate for psoriasis was found to be associated with the development of hepatic fibrosis and cirrhosis in as many as 25% of cases.394 Since then, a clearer picture of methotrexate as a dose-dependent promoter of hepatic fibrosis has emerged, particularly in persons who drink alcohol excessively or have preexisting liver disease. Guidelines have been instituted for scheduled pretreatment and interval liver biopsies to monitor the safety of methotrexate therapy. Methotrexate is now used more often as a low-dose weekly regimen in the management of rheumatoid arthritis, psoriasis, and other immunologic conditions, including inflammatory bowel disease. The problem of hepatotoxicity has largely been overcome by the avoidance of daily dosing with methotrexate and a reduction in the weekly dose to 5 to 15 mg (see also Chapter 22).395-397

Risk Factors

Risk factors for methotrexate-induced hepatic fibrosis are listed in Table 86-7; dose, alcohol intake, and preexisting liver disease are the most important.396,397 Total dose, incremental dose, dose interval, and duration of methotrexate therapy each influence the risk of hepatic fibrosis. After the cumulative ingestion of 3 g of methotrexate, the chance of histologic progression is 20%, but only 3% of patients are found to have advanced hepatic fibrosis.398 Obesity and diabetes mellitus may be important risk factors for hepatic

Table 86-7  Risk Factors for Methotrexate-Induced Hepatic Fibrosis* RISK FACTOR

IMPORTANCE

IMPLICATIONS FOR PREVENTION

Age

Increased risk >60 yr; possibly related to reduced renal clearance and/or biologic effect on fibrogenesis Incremental dose Dose frequency Duration of therapy Cumulative (total) dose Increased risk with daily levels >15 g (1-2 drinks)

Care in use of methotrexate in older people

Dose

Alcohol consumption Obesity Diabetes mellitus Preexisting liver disease

Increased risk Increased risk in obese persons (type 2 diabetes mellitus) Greatly increased risk, particularly related to alcohol, obesity, and diabetes mellitus (NASH)

Systemic disease

Possibly risk greater with psoriasis than rheumatoid arthritis (may depend on preexisting liver disease, alcohol intake) Increased risk because of reduced clearance NSAIDs, vitamin A, and arsenic may increase risk

Impaired renal function Other drugs

NASH, nonalcoholic steatohepatitis; NSAIDs, nonsteroidal anti-inflammatory drugs. *See Chapter 22.

5-15 mg/wk is safe Weekly bolus (pulse) safer than daily schedules Consider liver biopsy every 2 years Consider liver biopsy after each 2 g of methotrexate Avoid methotrexate use if intake not curbed Consider pretreatment liver biopsy with relevant history Consider pretreatment and interval liver biopsies Consider pretreatment and interval liver biopsies Pretreatment liver biopsy mandatory Avoid methotrexate, or schedule interval biopsies according to severity of hepatic fibrosis, total dose, and duration of methotrexate therapy Monitor liver biochemical tests during therapy None Reduce dose; greater caution with use of methotrexate Greater caution with use; monitor liver biochemical tests

1443

1444

Section IX  Liver fibrosis because they predispose to NASH and are associated with induction of CYP2E1; the strong association between NASH and methotrexate in causing liver injury during long-term, low-dose methotrexate treatment has been highlighted,399 as has the possibility that methotrexate itself can cause a pattern of injury resembling steatohepatitis. Increasing age, impaired renal function, and concomitant use of certain drugs decrease the elimination of methotrexate or facilitate tissue uptake by displacing metho­ trexate from plasma-protein binding sites. Psoriasis and rheumatoid arthritis are associated with hepatic abnormalities that range from abnormal liver biochemical test levels (25% to 50% of cases) and minor histologic changes (50% to 70%) to fibrosis (11%) and cirrhosis (1%). In patients with psoriasis, alcoholism often is a complicating factor, and in a meta-analysis,397 alcohol consumption was the most important determinant of advanced hepatic fibrosis in patients treated with methotrexate; the risk of progressive hepatic fibrosis was 73% in persons who drank more than 15 g of alcohol daily, compared with 26% in those who did not. The possibility that low-dose (5 to 15 mg) methotrexate given as a single weekly dose can cause hepatic fibrosis has been debated.395-397 The available data are limited by a lack of controlled studies in which liver histologic findings were evaluated; the lack of pretreatment evaluation of liver histology is a particularly serious deficiency in view of the high frequency of liver abnormalities among patients with rheumatoid arthritis and psoriasis. The conclusion has been reached that current regimens that are in use can promote hepatic fibrosis, at least at the ultrastructural level, but that cases of clinically significant liver disease are now virtually unknown. Indeed, repeat liver biopsies in some series have shown a reduction in fibrosis despite continuation of methotrexate in lower doses.398 Therefore, although methotrexate remains a potential cause of liver disease, advanced hepatic fibrosis is in large part preventable.

Clinicopathologic Features

Liver biochemical test abnormalities are common among patients who take methotrexate, but advanced hepatic fibrosis occasionally can develop in the absence of such abnormalities. Likewise, nausea, fatigue, and abdominal pain are common adverse effects of methothrexate, but patients with hepatic fibrosis are typically asymptomatic unless complications of liver failure or portal hypertension, such as bleeding esophageal varices, develop. A firm liver edge, hepatomegaly, splenomegaly, and ascites may be noted. Liver biochemical test levels are either normal or show nonspecific changes, including minor elevations of serum ALT and GGTP levels. In more advanced cases, hypoalbuminemia is present, but elevation of the serum bilirubin level and coagulation disturbances are rare. Thrombocytopenia may be present in patients with cirrhosis. Liver histologic findings often are graded according to the system of Roenigk, which has been useful in analyzing the published literature.397 In this system, grades I and II indicate a variable amount of steatosis, nuclear pleomorphism, and necroinflammatory activity but no fibrosis. Higher grades reflect increasing degrees of fibrosis, as follows: grade IIIa, few septa; grade IIIb, bridging fibrosis; and grade IV, cirrhosis. The pattern of hepatic fibrosis includes pericellular fibrosis, a feature of both alcoholic steatohepatitis and NASH; the possibility that methotrexate itself causes steatohepatitis or accentuates fibrogenesis among persons with underlying “primary NASH” has been suggested.399 Cases of hepatic fibrosis in livers with a relative paucity (or complete

absence) of portal and lobular inflammation have been reported.

Outcome and Prevention

Serious clinical sequelae (portal hypertension, liver failure, hepatocellular carcinoma) resulting from methotrexateinduced liver disease are now rarely seen. In a study of 32 patients with inflammatory bowel disease who received long-term methotrexate (mean dose of 2.6 g; follow-up period of 131 weeks), minor histologic changes in the liver were common, but advanced hepatic fibrosis was rare. Cases that have come to liver transplantation generally have been associated with suboptimal supervision of methotrexate therapy.400 Cases of severe hepatic fibrosis (Roenigk grades IIIb and IV) are often associated with lack of progression and even improvement after discontinuation of methotrexate or a reduction in the dose.398 In less severe cases, a balanced judgment must be made about the appropriateness of continuing or discontinuing methotrexate. An interval liver biopsy after an additional two years or 2 g of methotrexate may be judicious in a patient who is found to have minor hepatic fibrosis on a liver biopsy specimen. Recommendations for preventing methotrexate-induced hepatic fibrosis have been made (see Chapter 35).396,401 If possible, methotrexate should be avoided when the risk of liver injury is high. Persons treated with methotrexate should abstain from alcohol, and those who drink more than 100 g of ethanol per week should not be given methotrexate.396,397,401 A pretreatment liver biopsy is indicated only if the liver biochemical test levels are abnormal or if the history (e.g., alcoholism) and clinical features (e.g., hepatomegaly, risk factors for NASH) indicate possible underlying liver disease.402 The use of liver biochemical testing to monitor the patient’s progress during treatment with methotrexate is recommended but is problematic because of the lack of specificity and sensitivity of the tests; four to six sets of liver biochemical tests are often performed each year in patients undergoing treatment with methotrexate. Persistent or recurrent elevations of the serum AST or ALT levels, any decrease in the serum albumin level, or the development of hepatomegaly warrants investigation by liver biopsy. Scheduled liver biopsies are recommended after a cumulative dose of 4 g of methotrexate or duration of therapy of two years.397 Other authorities have suggested that a liver biopsy be performed after 5 g of methotrexate.403 Whether a liver biopsy is necessary in patients with normal liver biochemical test levels and without major risk factors for hepatic fibrosis remains unclear.398,401 Some authors have suggested a lower threshold for liver biopsy (cumulative dose of 1.5 g and after every 1 g of methotrexate thereafter) for patients with no risk factors for hepatoxicity.404 Serum biochemical tests that indicate progressive hepatic fibrosis, such as procollagen peptide-3 levels, have not proved sufficiently accurate for monitoring patients taking methotrexate.

DRUG-INDUCED VASCULAR TOXICITY Vascular injury to the liver may give rise to several unusual types of liver disease, including sinusoidal obstruction syndrome (formerly veno-occlusive disease, a form of hepatic venous outflow obstruction), peliosis hepatis (dilatation and destruction of hepatic sinusoids), noncirrhotic portal hypertension, and nodular regenerative hyperplasia. Drugs and chemical toxins are the most common causes of hepatic

Chapter 86  Liver Disease Caused by Drugs Table 86-8  Types of Drug-Induced Hepatic Vascular Disorders: Clinicopathologic Features, Outcome, and Implicated Etiologic Agents DISORDER

CLINICOPATHOLOGIC FEATURES

OUTCOMEs

IMPLICATED ETIOLOGIC AGENTS

Sinusoidal obstruction syndrome (veno-occlusive disease) Nodular regenerative hyperplasia

Abdominal pain, tender hepatomegaly, ascites, liver failure; occasionally chronic liver disease, other signs of portal hypertension Portal hypertension, encephalopathy— especially after variceal bleeding; diagnosed by histology Splenomegaly, hypersplenism, varices; ascites if associated hepatocellular disease Incidental finding, hepatomegaly, hepatic rupture, liver failure; diagnosed from appearances at surgery, vascular imaging Hepatomegaly, abdominal pain

High mortality rate; some cases may evolve into nodular regenerative hyperplasia Relatively good prognosis

Especially in bone marrow transplantation: 6-thioguanine, busulfan; dactinomycin, azathioprine, mitomycin; pyrrolizidine (e.g., in comfrey) Anticancer drugs: busulfan, dactinomycin; azathioprine

Prognosis depends on cause and associated liver injury

Vitamin A, methotrexate, azathioprine, arsenic, vinyl chloride, anticancer drugs

Prognosis depends on cause and complications

Anabolic steroids, azathioprine, 6-thioguanine

May regress after stopping oral contraceptives

Oral contraceptive steroids

Noncirrhotic portal hypertension Peliosis hepatis

Sinusoidal dilatation

vascular injury.405 The mechanism of injury is primarily dose-dependent toxicity to sinusoidal and other vascular endothelial cells, particularly when drugs are used in combination or concurrently with radiotherapy. Activation of inflammatory cells also may be important. Individual drugs (e.g., azathioprine) have been associated with more than one vascular syndrome, and the individual disorders overlap and may evolve from one type to another. Vascular toxicity may give rise to a continuum of disorders, each resulting from injury to different components of the hepatic vasculature. The essential features of these disorders are summarized in Table 86-8, and the more important conditions are discussed in Chapter 83. Hepatic imaging and measurement of portal pressure play a role in the diagnosis of these conditions; some, particularly nodular regenerative hyperplasia, are difficult to confirm in needle biopsy specimens.

AZATHIOPRINE

Hepatic complications of azathioprine, although rare, may be severe, diverse, and often late in onset. The overall frequency of azathioprine hepatotoxicity is less than 0.1%. Many cases occur in complex medical situations, particularly organ transplantation, in which activation of the immune system, viral infections, and other agents may increase the risk of hepatotoxicity (see Chapter 34). The central role of azathioprine has been confirmed in some cases that resolved after discontinuation of the drug and others in which a positive rechallenge was documented.406,407 Disturbances associated with azathioprine include bland cholestasis, cholestatic hepatitis with bile duct injury,407,408 zonal necrosis, and vascular toxicity.406 Vascular toxicity may give rise to the diverse syndromes of sinusoidal obstruction syndrome, peliosis hepatis, nodular regenerative hyperplasia, and noncirrhotic portal hypertension.406-409 Hepatocellular carcinoma with focal glycogenosis also has been reported in a long-term recipient of azathioprine.410 Azathioprine is associated with an extraordinary range of hepatic disorders, including liver biochemical test abnormalities in asymptomatic patients, bland cholestasis, cholestatic hepatitis, bile duct injury, and vascular injury. Cholestatic hepatitis is probably the most common; several cases have been associated with zone 3 necrosis and congestion, suggesting acute vascular injury, and azathioprine shares the vascular toxicity of other thiopurines. All hepatic

syndromes that result from vascular injury have been associated with azathioprine, particularly after organ transplantation. Cases of azathioprine-induced nodular regenerative hyperplasia and sinusoidal obstruction syndrome also have been reported in patients treated with azathioprine for other medical conditions, including inflammatory bowel disease.405,411 No relation between toxicity and the dose or duration of azathioprine therapy has been observed, but men are almost exclusively involved in cases of hepatic vascular injury following renal transplantation. The onset of cholestatic reactions is 2 weeks to 22 months after the start of azathioprine, but vascular toxicity is recognized later, typically 3 months to 3 years, and occasionally more than 9 years, after transplantation.409,410 The presentation and clinical features depend on the type of reaction. Cases of later onset are the result of delayed recognition and tend to be associated with complications of portal hypertension, including ascites, and liver failure. Recovery can occur in such cases,407 but the overall mortality rate is high. Azathioprine is used as an immunosuppressive agent following solid organ transplantation and as a glucocorticoid-sparing agent in some autoimmune diseases, including autoimmune hepatitis (see Chapter 88). In contrast to azathioprine, 6-mercaptopurine is a dose-dependent cause of hepatocellular necrosis and has been fatal in a few cases and, even more rarely, associated with cholestasis.5

LIVER TUMORS Several associations between pharmacologic and environmental agents and benign and malignant liver tumors have been described, but causality has been difficult to prove because of the rarity of these associations. For some sex steroid-related tumors, as well as for vinyl chloride-induced angiosarcoma, the relative risk attributable to the causative agent has been determined. Prevention and early detection are critical for improving outcomes. The major tumors of interest include cavernous hemangioma, focal nodular hyperplasia, hepatic adenoma, hepatocellular carcinoma, angiosarcoma, hepatoblastoma, cholangiocarcinoma, mixed carcinosarcoma, and epithelioid hemangioendothelioma (see Chapters 69 and 94).

1445

1446

Section IX  Liver KEY REFERENCES

Biour M, Poupon R, Grange JD, et al. Drug-induced hepatotxicity. The 13th updated edition of the bibliographic database of drug-related liver injuries and responsible drugs. Gastroenterol Clin Biol 2000; 24:1052-91. (Ref 10.) Björnsson E, Olsson R. Outcome and prognostic markers in severe druginduced liver disease. Hepatology 2005; 42:481-9. (Ref 8.) Bruno S, Maisonneuve P, Castellana P, et al. Incidence and risk factors for non-alcoholic steatohepatitis: Prospective study of 5408 women enrolled in Italian tamoxifen chemoprevention trial. BMJ 2005; 330:932-5. (Ref 388.) Chalasani N, Aljadhey H, Kesterson J, et al. Patients with elevated liver enzymes are not at higher risk for statin hepatotoxicity. Gastroenterology 2004; 128:1287-92. (Ref 162.) Daly AK, Aithal GP, Leathart JB, et al. Genetic susceptibility to diclofenac-induced hepatotoxicity: Contribution of UGT2B7, CYP2C8, and ABCC2 genotypes. Gastroenterology 2007; 132:272-81. (Ref 297.) Fontana RJ, McCashland TM, Benner KG, et al. Acute liver failure associated with prolonged use of bromfenac leading to liver transplantation. The Acute Liver Failure Study Group. Liver Transpl Surg 1999; 5:480-4. (Ref 282.) Ibanez L, Perez E, Vidal X, et al. Prospective surveillance of acute serious liver disease unrelated to infectious, obstructive, or metabolic

diseases: Epidemiological and clinical features, and exposure to drugs. J Hepatol 2002; 37:592-600. (Ref 4.) Kosters A, Karpen SJ. Bile acid transporters in health and disease. Xenobiotica 2008; 38:1043-71. (Ref 38.) Lewis JH, Mortensen ME, Zweig S, et al. Efficacy and safety of high-dose pravastatin in hypercholesterolemic patients with well-compensated chronic liver disease. Hepatology 2007; 46:1453-63. (Ref 160.) Lucena MI, Andrade RJ, Fernández MC, et al. Determinants of the clinical expression of amoxicillin-clavulanate hepatotoxicity: A prospective series from Spain. Hepatology 2006; 44:850-6. (Ref 334.) Meier Y, Cavaliaro M, Roos M, et al. Incidence of drug-induced liver injury in medical patients. Eur J Clin Pharmcol 2005; 61:135-43. (Ref 66.) Rostom A, Goldkind L, Laine L. Nonsteroidal anti-inflammatory drugs and hepatic toxicity: A systematic review of randomized controlled trials in arthritis patients. Clin Gastroenterol Hepatol 2005; 3:489-98. (Ref 285.) Zimmerman HJ. Hepatotoxicity. The Adverse Effects of Drugs and Other Chemicals on the Liver, 2nd ed. Philadelphia: Lippincott, Williams and Wilkins; 1999. (Ref 6.) Zollner G, Marschall HU, Wagner M, et al. Role of nuclear receptors in the adaptive response to bile acids and cholestasis: Pathogenetic and therapeutic considerations. Mol Pharm 2006; 3:231-51. (Ref 37.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

87 Liver Disease Caused by Anesthetics, Toxins, and Herbal Preparations James H. Lewis

CHAPTER OUTLINE Anesthetic Agents  1447 Halothane  1447 Other Anesthetic Agents  1449 Jaundice in the Postoperative Period  1449 Chemicals  1450 Commercial and Industrial Chemicals  1450 Pesticides  1452 Metals  1453 Iron  1453 Phosphorus  1453 Copper Salts  1453 Thorium Dioxide  1453 Other Metals  1453

Although halothane hepatitis is now largely of historical interest, it holds an important place in the annals of causality assessment in drug-induced liver disease.1 In contrast with the largely unpredictable hepatotoxicity seen with more modern anesthetics and most other medicinal agents (as discussed in Chapter 86), liver damage caused by occupationally and environmentally encountered chemical compounds and other toxins often is more predictable, dose related, and predominantly cytotoxic in nature.1-4 Industrial exposure to hepatotoxic chemicals is a less frequent occupational hazard today than in the past, but reports of toxicity from chemical agents, as well as metals, adulterated cooking oils, and botanical toxins, have not disappeared.3,4 Additionally, the use of complementary and alternative medicine (CAM) preparations continues to increase, and reports of liver injury from potentially hepatotoxic herbal and weight loss products continue to appear (see Chapter 127).5,6 Anesthetics, herbal products, mushrooms, and other toxins continue to account for a substantial percentage of emergency liver transplants for acute liver failure.7

ANESTHETIC AGENTS The volatile inhalational anesthetics in current use are derivatives of some of the most potent chemical hepato­ toxins developed for medicinal purposes. Chloroform, the original haloalkane anesthetic, has long been abandoned but remains an important experimental hepatotoxin, as does carbon tetrachloride (another chlorinated aliphatic hydro-

Adulterated Cooking Oils and Contaminated Foods  1453 Drugs of Abuse  1454 Cocaine  1454 Others  1454 Botanical and Environmental Hepatotoxins  1454 Mushrooms  1454 Other Foodstuffs  1455 Vitamins and Herbal Preparations  1455 Hypervitaminosis A  1455 Herbal Remedies and Nutritional Supplements  1456

carbon), which found use as an early vermifuge and is still employed as a household reagent in some parts of the world.1,8 Halothane (fluothane), introduced in the 1950s as a safer, nonexplosive alternative to ether, is a haloalkane compound that produced a well-described but rare syndrome of acute hepatotoxicity, usually after repeat exposure.9 The anesthetics that followed-methoxyflurane, enflurane, isoflurane—all have been implicated as a cause of similar injury, albeit less commonly for enflurane and isoflurane than for halothane; even fewer instances have been reported for the newest agents, sevoflurane and desflurane,10,11 because of their proportionally lower degree of metabolism.12 Halothane is no longer being produced in the United States but continues to be employed in other countries13 and is a case study in the elucidation of immunologic-mediated liver injury.14

HALOTHANE

The retrospective National Halothane Study, cited in the past as the basis for exonerating halothane as a cause of hepatotoxicity,15 is now considered flawed.1 Nearly 1000 cases of halothane hepatotoxicity were reported worldwide during the 1960s and 1970s.1,9,16 A fairly uniform clinical picture of postoperative fever, eosinophilia, jaundice, and hepatic necrosis occurred a few days to weeks after administration of anesthesia, usually after repeat exposure to halothane, and the case-fatality rate was high (Table 87-1). Rare cases of halothane-induced liver injury were reported after workplace exposure among anesthesiologists, surgeons, nurses, and laboratory staff and after halothane sniffing for “recreational” use; in affected persons, antibodies to trifluo-

1447

1448

Section IX  Liver Table 87-1  Clinicopathologic Features of Halothane Hepatitis Estimated incidence After first exposure: 0.3 to 1.5 per 10,000 After multiple exposures: 10 to 15 per 10,000 Female-to-male ratio 2 : 1 Latent period to first symptom After first exposure: 6 days (11 days to jaundice) After multiple exposures: 3 days (6 days to jaundice) Jaundice as presenting symptom in 25% (range of serum bilirubin: 3-50 mg/dL) Fever in 75% (precedes jaundice in 75%); chills in 30% Rash in 10% Myalgias in 20% Ascites, renal failure, and/or gastrointestinal hemorrhage in 20%-30% Eosinophilia in 20%-60% Serum ALT and AST levels: 25-250 × ULN Serum alkaline phosphatase level: 1-3 × ULN Histologic features Zone 3 massive hepatic necrosis in 30%; submassive necrosis in 70% (autopsy series) Inflammation usually less marked than in viral hepatitis Eosinophilic infiltrate in 20% Granulomatous hepatitis occasionally Course and outcome Mortality rate (pretransplantation era): 10%-80% Symptoms can resolve within 5-14 days Full recovery can take 12 weeks or longer Chronic hepatitis not well documented Adverse prognostic findings Age >40 years Obesity Short duration to the onset of jaundice Serum bilirubin level >20 mg/dL Coagulopathy ALT, alanine aminotransferase; AST, aspartate aminotransferase; ULN, upper limit of normal.

roacetylated (TFA) proteins were demonstrated, indicating previous exposure.17 Two types of postoperative liver injury have been asso­ ciated with halothane. A minor form is seen in 10% to 30% of patients, in whom mild asymptomatic elevations in serum alanine aminotransferase (ALT) levels develop between the first and tenth postoperative days; the risk of hepatotoxicity is higher after two or more exposures to halothane than with subsequent use of alternative agents such as enflurane, isoflurane, and desflurane. Evidence of immune activation is lacking in these patients,18 in whom the ALT elevations generally are rapidly reversible. The major form of halothane-induced hepatotoxicity is a rare, doseindependent, severe hepatic drug reaction with elements of immunoallergy and metabolic idiosyncrasy (see Table 87-1). After an initial exposure to halothane, the frequency of this form of toxicity is only approximately 1 per 10,000,19 but the rate increases to approximately 1 per 1000 after two or more exposures, especially when the anesthetic agent is readministered within a few weeks.1 Typically, zone 3 (centrilobular) hepatic necrosis is seen histologically.20 The case-fatality rate ranged from 14% to 71% in the pre-liver transplantation era.1

Risk Factors for Halothane Hepatitis

Host-related risk factors for halothane hepatitis are listed in Table 87-2. The reaction is rare in childhood11; patients younger than 10 years of age represent only about 3% of the total, and cases in persons younger than 30 years account for less than 10%.11,16 The disease tended to be more severe in persons older than 40 years of age. Two thirds of cases

Table 87-2  Risk Factors for Halothane Hepatitis Older age (>40 years) Female gender Two or more exposures (documented in 80%-90% of cases) Obesity Familial predisposition Induction of CYP2E1 by phenobarbital, alcohol, or isoniazid CYP2E1, cytochrome P450 2E1.

have been in women, and repeat exposure to halothane (especially within a few weeks or months) is documented in as many as 90% of cases.1 The time between exposures can be as long as 28 years.21 After repeat exposure, hepatitis is earlier in onset and more severe. Obesity is another risk factor, possibly because of storage of halothane in body fat. The induction of cytochrome P450 (CYP) enzymes (especially CYP2E1) that metabolize halothane to its toxic intermediate has been produced experimentally with phenobarbital, alcohol, and isoniazid; valproate inhibits and phenytoin has no specific effect on halothane hepatotoxicity.1 Inhibition of CYP2E1 by administration of a single dose of disulfiram has been suggested as a means of preventing halothane hepatitis—by inhibiting the production of the metabolite responsible for neoantigen formation.22 Familial predisposition to halothane-induced liver injury has been reported in closely related family members.23 Serum antibodies to volatile anesthetics have been found in pediatric anesthesiologists,17 who, like patients with halothane hepatitis, had higher levels of serum autoantibodies to CYP2E1 and to endoplasmic reticulum protein (ERp58) than those found in general anesthesiologists and control subjects who had never been exposed to inhalational anesthetics. The autoantibodies are not thought to have a role in pathogenesis.1

Pathology

In a study of 77 cases of halothane hepatitis reviewed by the Armed Forces Institute of Pathology,20 various degrees of liver injury were seen, depending on the severity of the reaction. Massive or submassive necrosis involving zone 3 was present in all autopsy specimens, whereas biopsy material revealed a broader range of injury—from spotty necrosis in about one third of cases to zone 3 necrosis in two thirds. The zone 3 injury is sharply demarcated, and the inflammatory response is less severe than in acute viral hepatitis.

Pathogenesis

Halothane injury occurs by one or more of three potential mechanisms: hypersensitivity, production of hepatotoxic metabolites, and hypoxia, in decreasing order of importance.1 Evidence for the role of hypersensitivity is found in the increased susceptibility and shortened latency after repeat exposure, the hallmark symptoms and signs of drug allergy (fever, rash, eosinophilia, and granuloma formation), and the detection of neoantigens and antibodies. Halothane oxidation yields trifluoroacetylchloride, which acts on hepatocyte proteins to produce neoantigens that are responsible for the major form of injury. By contrast, reductive pathways produce free radicals that can act as reactive metabolites that may have a role in causing minor injury. A unifying hypothesis set forth by Zimmerman1 suggests that halothane injury most likely is the result of immunologic enhancement of zone 3 necrosis produced by the reductive metabolite(s). Accordingly, the hepatotoxic potential of halothane depends on the susceptibility of the patient and on factors that promote production of hepatotoxic or immu-

Chapter 87  Liver Disease Caused by Anesthetics, Toxins, and Herbal Preparations Table 87-3  Hepatotoxic Anesthetics Other Than Halothane anesthetic

PERCENT METABOLIZED

INCIDENCE OF hepatotoxicity

CROSS-REACTIVITY WITH OTHER HALOALKANES

OTHER clinical FEATURES

Methoxyflurane Enflurane Isoflurane Desflurane

>30% 2% 0.2% 40 years F > M 2 : 1 Obese May or may not be present

Any F=M Any Documented in 50%

Any F=M Any Absent

ALT, alanine aminotransferase; F, female; M, male; ULN, upper limit of normal.

Table 87-6  Phases of Illness after Ingestion of Various Hepatotoxins Toxin phase Phase I (1-24 hours) Onset of toxicity A,N,V,D Shock Neurologic symptoms Phase II (24-72 hours) Asymptomatic latent period Phase III (>72 hours) Jaundice Hepatic failure Renal failure Maximum AST and ALT (× ULN) Zonal necrosis Steatosis Case-fatality rate

ACETAMINOPHEN

PHOSPHORUS

AMANITA PHALLOIDES

CARBON TETRACHLORIDE

Immediate + − −

Immediate ++++ + +

Delayed 6-20 hr ++++ ± ±

Immediate + − −

+

±

+

+

+ + + 1000 3 − 5%-15%

+ + + 3

AMA negative Ceruloplasmin normal Normal α1-antitrypsin phenotype Normal or near-normal serum iron level HBsAg, anti-HCV, IgM anti-HAV negative Liver biopsy Interface hepatitis ± lobular hepatitis (see Figs. 88-1 and 88-2)

Definite Figure 88-1.  Histopathology of interface hepatitis. The limiting plate of the portal tract is disrupted by a lymphoplasmacytic infiltrate. This histologic pattern is the hallmark of autoimmune hepatitis, but it is not disease specific. (Hematoxylin and eosin, ×200.)

Gamma globulin level ≥1.5 normal ANA, SMA, or anti-LKM1 ≥1:80 No exposure to drugs or blood products Alcohol intake 17 12-17

*Times upper limit of normal. AMA, antimitochondrial antibodies; ANA, antinuclear antibodies; anti-LC1, antibodies to liver cytosol type 1; anti-LKM1, antibodies to liver-kidney microsome type 1; anti-SLA, antibodies to soluble liver antigen; AP/AST (or AP/ALT), ratio of serum alkaline phosphatase level to serum aspartate aminotransferase (or serum alanine aminotransferase) level; HLA, human leukocyte antigen; IgG, immunoglobulin G; pANCA, perinuclear anti-neutrophil cytoplasmic antibodies; SMA, smooth muscle antibodies. Adapted from Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group report: Review of criteria for diagnosis of autoimmune hepatitis. J Hepatol 1999; 31:929-38.

95%), but the simplified system has greater specificity (90% vs. 73%) and predictability (92% vs. 82%).15 Whereas the original scoring system is useful for evaluating patients in whom every component must be assessed because of few or atypical findings, the simplified scoring system is useful for excluding AIH in patients with other conditions and concurrent immune features.

PATHOGENESIS The pathogenic mechanisms of AIH are unknown. The most popular hypotheses invoke a constellation of interactive factors that include a triggering agent, genetic predisposition, and various determinants of autoantigen display, immunocyte activation, and effector cell expansion (Fig.

CATEGORY

VARIABLE

Autoantibodies† Antinuclear antibodies or smooth muscle antibodies

1 : 40

+1

Antibodies to liver-kidney microsome type 1 Antibodies to soluble liver antigen Immunoglobulin Level Immunoglobulin G

≥1 : 80 ≥1 : 40

+2 +2

Positive

+2

>Upper limit of normal >1.1 times upper limit of normal

+1

Compatible with autoimmune hepatitis Typical of autoimmune hepatitis

+1

No viral markers

+2

Histologic Findings Morphologic features

Viral Disease Absence of viral hepatitis Pretreatment Aggregate Score Definite diagnosis Probable diagnosis

SCORE

+2

+2

≥7 6

*Adapted from Hennes EM, Zeniya M, Czaja AJ, et al. Simplified diagnostic criteria for autoimmune hepatitis. Hepatology 2008; 48:169-76. † Autoantibody titers as determined by indirect immunofluorescence.

88-4).16-18 Proposed triggering factors include infectious agents, drugs, and toxins. The lag time between exposure to the trigger and onset of the disease can be long, and the triggering factor may not be needed for perpetuation of the disorder. The CD4+ helper T cell is the principal effector cell, and its activation is the initial step in the pathogenic pathway. Molecular mimicry of a foreign antigen and a self-antigen is the most common explanation for the loss of selftolerance, but this mechanism has not been established for any autoimmune disease.16-18 Genetic factors influence autoantigen presentation and CD4+ helper T cell recognition. The antigen-binding groove of the class II molecule of the major histocompatibility complex (MHC) is encoded by alleles that determine the groove’s configuration and ability to activate immunocytes. The susceptibility alleles of AIH in white North Americans and northern Europeans reside on the DRB1 gene and are DRB1*0301 and DRB1*0401 (see Fig. 88-4).19-21 Different ethnic groups have different susceptibility alleles, a finding that supports a “shared motif hypothesis” of pathogenesis.19-21 According to this hypothesis, the risk of disease relates to amino acid sequences in the antigenbinding groove of the class II MHC molecule, and multiple alleles encode the same or similar sequence (“shared motif”). The critical shared motif in white North Americans and northern Europeans with AIH is a six-amino-acid sequence represented by the code LLEQKR.19-21 This sequence is located between positions 67 and 72 of the DRβ polypeptide chain of the class II MHC molecule, and lysine (K) in position 71 is the critical determinant of susceptibility. DRB1*0301 and DRB1*0401 encode identical amino acid sequences in the DRβ67-72 region and affect susceptibility similarly. DRB1*0404 and DRB1*0405 are the susceptibility alleles in Mexican, Japanese, mainland Chinese, and Argentine

1463

1464

Section IX  Liver 1st signal

B7 2nd signal CD28

APC P4

P4 Aspartic acid Glutamic acid

Lysine Arginine

DRB1*0401

Antigen-binding groove

DRB1*1301 Diverse autoantibodies DRB1*03 DRB1*07

T cell

DRβ71 DRB1*0301

Antigenic peptide

CD4+

CTLA4 Type 1 cytokine response TNFA*2

Autoimmune promoters Cytokine Polymorphisms Increased bcl-2 expression Cytotoxic FAS Granzyme T cells Polymorphism perforin

Activated CD4+ T-helper cell

Deficient T-reg cells

FAS ligand

Type 2 cytokine response

Plasma cells (from B cells)

FAS Host DNA cytosolic fragments

FAS Polymorphism Hepatocyte apoptosis

Cell-mediated cytotoxity

Fc receptor Hepatocyte NK Antibody-mediated cellular cytoxicity

Figure 88-4.  Interactive mechanisms that contribute to the development of autoimmune hepatitis in white North American and northern European adults. The initial stimulus for immune activity is an antigenic peptide (upper left corner) that has a negatively charged aspartic acid or glutamic acid at a position within its structure (P4) that can form a salt bridge with a positively charged lysine or arginine residue at position 71 within the DR beta polypeptide chain (DRb71) of the antigen binding groove of the class II DR molecule of the major histocompatibility complex (top center). The DR molecule-antigen complex of the antigen-presenting cell (APC) then interacts with the antigen receptor of a CD4+ T-helper cell (interaction not shown), and the first co-stimulatory signal is completed (1st signal). The CD28 molecule on the surface of the CD4+ T-helper cell ligates with the B7 ligand on the surface of the APC, and the second co-stimulatory signal (2nd signal) is completed (upper right corner). The activated CD4+ T-helper cell can then differentiate and proliferate along type 1 and type 2 cytokine pathways (middle right). Deficiencies in the function or amount of cytotoxic T lymphocyte antigen 4 (CTLA4) can enhance the strength of the 2nd signal. Differentiation along the type 1 cytokine pathway can be promoted by polymorphisms of the tumor necrosis factor gene (TNFA*2) and tumor necrosis factor receptor superfamily gene (FAS), resulting in cell-mediated cytotoxicity by sensitized liver-infiltrating cytotoxic T cells and increased hepatocyte apoptosis (middle bottom). The apoptosis of hepatocytes can, in turn, release DNA cytosolic fragments that contribute to the production of diverse collateral autoantibodies (middle left). Autoantibody expression is, in part, host-dependent and influenced by the susceptibility alleles DRB1*03 and DRB1*07. Host genetic predispositions are also important in encoding the antigen-binding groove of the class II DR molecule through the actions of DRB1*0301, DRB1*0401, and DRB1*1301 alleles and in generating autoimmune promoters (cytokine and FAS polymorphisms) that enhance cell-mediated cytotoxicity and hepatocyte apoptosis. The enhanced expression of the anti-apoptotic protein (bcl-2) on the surface of cytotoxic T cells can protect them from programmed cell death and perpetuate their autoreactivity (middle). The cytokine pathways can be enhanced by deficiencies in the actions of T-regulatory cells (T-reg cells), which have suppressive effects that can be reversed by glucocorticoid treatment. Differentiation of B cells into plasma cells, via the type 2 cytokine response of activated CD4+ T-helper cells, can result in immunoglobulin production that, in turn, generates an antibodymediated cellular toxicity. Natural killer (NK) cells with Fc receptors are directed against complexes of immunoglobulin with normal hepatocyte membrane constituents. (Adapted from Czaja AJ. Autoimmune hepatitis—Part A: Pathogenesis. Expert Rev Gastroenterol Hepatol 2007; 1:113-128.)

adults and encode a similar sequence, except for arginine (R) instead of lysine (K) at the DRβ71 position.20,21 Arginine is a positively charged amino acid that is structurally similar to lysine, and its substitution for lysine should not greatly alter the antigen-binding properties of the class II MHC molecule. By contrast, DRB1*1501 protects against AIH in white North Americans and northern Europeans, and this allele encodes isoleucine (I) instead of leucine (L) at position DRβ67 and alanine (A) instead of lysine (K) at position DRβ71. Alanine is a neutral, nonpolar amino acid that, when substituted for lysine, should greatly affect antigen presentation and immunocyte activation. Antigenic peptides are selected for display by the nature of the amino acids that interact with residues within the antigen-binding groove (see Fig. 88-4).21,22 The critical sixamino-acid motif in AIH restricts the range of peptides that can be accommodated. Multiple self-antigens or foreign

antigens may satisfy the minimal structural requirements and serve as immunogenic peptides. The ideal triggering epitope must have a negatively charged amino acid residue (aspartic acid or glutamic acid) at peptide position P4 to form a salt bridge with the positively charged lysine or arginine at DRβ71.21 Molecular modeling indicates that a negatively charged P4 residue in the antigenic peptide and the positively charged lysine or arginine at DRβ71 can form a P4-DRβ71 immunoreactive unit that is independent of the other residues within the antigen and antigen-binding groove. This minimal immunoreactive unit can be created by multiple antigenic peptides and class II MHC molecules, and the number of these units may affect susceptibility by a “dose effect.” DRB1*1301 is associated with AIH in Argentine children22 and Brazilian patients23,24 and encodes ILEDER at positions DRβ67-72. Glutamic acid (E), aspartic acid (D),

Chapter 88  Autoimmune Hepatitis and glutamic acid (E) are at positions DRβ69, DRβ70, and DRβ71, respectively, in the class II MHC molecule, and the presence of these critically located but negatively charged amino acid residues argues against the “shared motif” hypothesis of pathogenesis. The “molecular footprint” hypothesis of pathogenesis holds that susceptibility to AIH in different regions and ethnic groups relates to indigenous factors or agents favored by certain genetic phenotypes.20,21 In South America, DRB1*1301 is associated with protracted hepatitis A virus infection, and persons with this allele may be “selected” from their environment to have prolonged exposure to viral and hepatic antigens that favor the development of AIH.25 An understanding of the individual susceptibility allele in different geographic regions may allow use of this “footprint” to track the cause of the disease. The “autoimmune promoter hypothesis” of pathogenesis complements the “shared motif” and “molecular footprint” hypotheses by proposing that genetic promoters inside and outside the MHC can affect disease occurrence, either in synergy (epistasis) with the principal susceptibility factors or in lieu of them.18,20,21 Polymorphisms of the tumor necrosis factor (TNF)-α gene (TNFA*2),26 the cytotoxic T lymphocyte antigen 4 gene (CTLA4),27 and Fas gene promoter at position -670 (TNFRSF6)28 have been associated with increased immunoreactivity, disease severity, and early progression to cirrhosis in white North American and northern European patients. Constellations of autoimmune promoters, as yet undefined, may individualize the disease by affecting its occurrence, clinical phenotype, and outcome. Liver cell destruction is accomplished by either cellmediated cytotoxicity, antibody-dependent cell-mediated cytotoxicity, or a combination of both mechanisms (see Fig. 88-4).16-18 Cell-mediated cytotoxicity depends on the clonal expansion of CD8+ cytotoxic T cells that accomplish liver cell injury through the release of lymphokines. This mechanism is regulated by type 1 cytokines, and the −308 polymorphism of TNFA*2 may facilitate this pathway.26 Antibody-dependent cell-mediated cytotoxicity is regulated by type 2 cytokines, and the natural killer cell accomplishes liver cell destruction by binding of its Fc receptor with an antigen-antibody complex on the hepatocyte surface.16-18 The predominant mechanism depends on the phenotypic differentiation of the CD4+ helper T cell, which in turn reflects the cytokine milieu. The cytokine milieu may reflect polymorphisms of the cytokine genes that favor excessive production of some modulators, such as TNF-α, or deficient production of others. Defects in the counter-regulatory cytokine milieu may also reflect reduced numbers of intrahepatic natural killer T (NKT) cells and the failure of T-regulatory (T-reg) cells (CD4+CD25+ cells) to modulate CD8+ T cell proliferation and cytokine production.18,29 Increased numbers of gd T cells may also contribute to the cytodestructive process by recognizing antigens presented by the non-classic MHC molecules, and the recruitment and intrahepatic trafficking of cytotoxic T lymphocytes may be enhanced by the upregulation of chemokines, such as CXCL16.18 Fibrogenesis, in turn, is fueled by the resulting inflammatory activity as perivascular hepatic stellate cells transform into myofibroblasts. The matrix proteins accumulate as tissue inhibitors retard the counteractive degradative actions of matrix metalloproteinases, and stellate cells continue to be activated in an autocrine fashion by transforming growth factor-β (TGF-β).30 Glucocorticoid therapy can favorably alter the cytokine milieu, improve the number and function of the T-reg cells, impair activation of TGF-β, promote disappear-

ance of the metalloproteinase inhibitors, enhance degradation of the fibrotic liver matrix, and foster apoptosis of the hepatic stellate cells (see also Chapter 2).31

CLASSIFICATION Two types of AIH have distinctive serologic profiles. Neither has been ascribed a unique cause, specific treatment strategy, or special type of behavior (Table 88-3). The terms are useful as clinical descriptors and as research designations to ensure homogeneous study populations.

TYPE 1 AUTOIMMUNE HEPATITIS

Type 1 AIH is characterized by SMA, ANA, or both (see Table 88-3).1,7 Antibodies to actin have greater specificity, but less sensitivity, for the diagnosis of AIH than SMA.6 Atypical pANCA are found in as many as 90% of patients with type 1 AIH and typically are absent in type 2 AIH.6,7 Type 1 AIH can occur at any age and in either gender (see Table 88-3).7 Initial studies that suggested a bimodal age distribution probably reflected referral biases to tertiary medical centers. The disease has been described in infants and probably is underdiagnosed in the elderly.32 Seventyeight percent of patients are women (female-to-male ratio 3.6 : 1), and 38% have concurrent extrahepatic immunologic diseases.7 Autoimmune thyroiditis (occurring in 12% of the cases), Graves’ disease (6%), and ulcerative colitis (6%) are the most common associated immune disorders. Rheumatoid arthritis, pernicious anemia, scleroderma, Coombs-positive hemolytic anemia, autoimmune throm­ bocytopenic purpura, symptomatic cryoglobu­linemia, leukocytoclastic vasculitis, nephritis, erythema nodosum, systemic lupus erythematosus, and fibrosing alveolitis also may occur (less than 1% each). Cholangiography is warranted to exclude PSC in all patients who have concurrent ulcerative colitis (see Chapter 68).33 Type 1 AIH is associated with an abrupt onset of symptoms in 40% of cases and may manifest in a fulminant fashion.2 The acute presentation frequently reflects preexisting subclinical disease that is unmasked by progression or represents a spontaneous exacerbation of inflammatory activity. Features of chronicity are lacking in 8% of patients, in whom the presentation of the disorder is indistinguishable from that of acute viral or toxic hepatitis. The target autoantigen of type 1 AIH is unknown. Human leukocyte antigen (HLA)-DR3 (DRB1*0301) and HLA-DR4 (DRB1*0401) are independent risk factors for the disease in white North Americans and northern Europeans.19-21 More than 80% of white patients in Great Britain and the United States possess either DRB1*0301, DRB1*0401, or both, compared with 42% of the unaffected white population. In South America, especially in children, DRB1*1301 is the principal susceptibility allele. These findings indicate that type 1 AIH is a complex polygenic disorder.

TYPE 2 AUTOIMMUNE HEPATITIS

Type 2 AIH is characterized by the expression of anti-LKM1 (see Table 88-3).1,7,34 Most affected persons are children (ages 2 to 14 years), but in Europe, especially in Germany and France, 20% of patients are adults.34 In the United States, type 2 AIH is rare, and only 4% of patients older than 18 years have anti-LKM1.35 The regional differences in prevalence may relate to ethnic differences in the genetic predisposition to the disease.36 Type 2 patients are younger than type 1 patients and may have different clinical and laboratory features (see Table

1465

1466

Section IX  Liver Table 88-3  Classification of Autoimmune Hepatitis Based on Autoantibodies CLINICAL FEATURE

TYPE 1

TYPE 2

Signature autoantibodies

Smooth muscle Nuclear Atypical pANCA Actin* Asialoglycoprotein receptor* Chromatin* Soluble liver antigen* Unknown Infants to elderly 78% 38% Autoimmune thyroiditis Graves’ disease Ulcerative colitis

Liver-kidney microsome type 1

Associated autoantibodies

Putative autoantigen Age Female Concurrent immune diseases Typical concurrent autoimmune diseases

Organ-specific antibodies Serum gamma globulin elevation HLA associations Allelic risk factors

Glucocorticoid responsive

4% +++ B8, DR3, DR4, DR13 DRB1*0301 and *0401   (white North Americans and northern Europeans) DRB1*1301   (South Americans, especially children) +++

Liver cytosol type 1* Soluble liver antigen*

CYP2D6 Children (2-14) 89% 34% Autoimmune thyroiditis Vitiligo Type 1 diabetes mellitus APECED 30% + B14, DR3, C4A-Q0, DR7 DQB1*0201, DRB1*07, DRB1*03   (Germans and Brazilians) ++

*Autoantibodies with an asterisk are investigational only and not available for routine clinical use. APECED, autoimmune polyendocrinopathy–candidiasis–ectodermal dystrophy; CYP2D6, cytochrome P450 2D6; pANCA, perinuclear anti-neutrophil cytoplasmic antibodies.

88-3).1,7,34 As with type 1 AIH, an acute or fulminant presentation is possible and important to recognize and treat early. Susceptibility to type 2 AIH has been associated with DQB1*0201, DRB1*07, and DRB1*03.37 DQB1*0201 is in strong linkage disequilibrium with DRB1*07 and DRB1*03 and has been proposed as the principal genetic determinant of the disease. The expression of anti-LKM1 has been associated with DRB1*07, and various aspects of type 2 AIH may have different genetic determinants.38 The target antigen of type 2 AIH is the cytochrome P450 2D6 mono-oxygenase (CYP2D6).39 This protein is a 50-kd microsomal drug-metabolizing enzyme, and its expression on the hepatocyte surface can be modulated by interleukins and TNF-α. Antibodies to LKM1 inhibit the activity of CYP2D6 in vitro but not in vivo, and lymphocytes extracted from the liver tissue of patients who have the disease exhibit immunoreactivity specific to the antigen. CYP2D6 has been sequenced, cloned, and mapped, and five epitopes are recognized by anti-LKM1.40 The amino acid sequence spanning 193-212 of the CYP2D6 molecule is the target of anti-LKM1 in 93% of patients. Homologies exist between CYP2D6 and the genomes of the hepatitis C virus, cytomegalovirus, and herpes simplex type 1 virus. These molecular mimicries may result in cross-reacting antibodies and support the hypothesis that repeated viral infections may break self-tolerance and cause the disease.17

VARIANT FORMS Patients who have atypical features of AIH currently lack an official designation and confident treatment strategy.41 They may have manifestations of AIH and another type of chronic liver disease or findings that are incompatible with AIH by current diagnostic criteria (Table 88-4).42,43

VARIANT WITH PRIMARY BILIARY CIRRHOSIS

AIH in patients who also have antimitochondrial antibodies (AMA) and histologic features of cholangitis constitutes an

overlap syndrome with PBC (see Table 88-4).41-44 Typically, affected patients have low titers of AMA and concurrent features of bile duct injury or loss.45 Antibodies against the PBC-specific M2 mitochondrial antigens may be present46; histologic features of cholangitis, including destructive cholangitis, may be seen45; and copper staining of hepatic tissue indicative of cholestasis may be observed.45 Occurrence rates range from 5% of patients initially diagnosed as having AIH to 19% of patients initially diagnosed as having PBC (see also Chapter 89).42 The clinical course of the disease and response to treatment depend mainly on the predominant component of the disease. Patients who have high serum aspartate aminotransferase (AST) levels, serum alkaline phosphatase levels less than twice the upper limit of normal, moderate to severe interface hepatitis on histologic examination, and high diagnostic scores for AIH commonly respond to glucocorticoid therapy.42,45 By contrast, patients who have serum alkaline phosphatase levels greater than twice the upper limit of normal, serum gamma glutamyl transpeptidase levels at least five times the upper limit of normal, and florid bile duct lesions on histologic examination mainly have PBC and commonly respond to ursodeoxycholic acid in combination with glucocorticoids (prednisone, prednisolone, or budesonide).44 AMA can be detected in 18% of patients with AIH in the absence of cholestatic features and can appear and disappear during the course of the disease without evolution to a different syndrome or the need for different therapy.47,48 Histologic changes of bile duct injury are required in addition to AMA and other classic features of AIH to designate a variant form of AIH.

VARIANT WITH PRIMARY SCLEROSING CHOLANGITIS

Histologic changes of lymphocytic, pleomorphic, or fibrous cholangitis; cholestatic laboratory findings; concurrent inflammatory bowel disease; or failure to respond to glucocorticoids constitute indications for cholangiography in

Chapter 88  Autoimmune Hepatitis Table 88-4  Variant Forms of Autoimmune Hepatitis AIH + PBC

AIH + PSC

Clinical and laboratory features

AIH features AMA +

Histology

Cholangitis Cholestasis Prednisone if AP ≤ 2 × ULN Prednisone and UDCA if AP >2 × ULN and/or florid duct lesions

AIH features Ulcerative colitis AMA − Abnormal cholangiogram Normal cholangiogram (small duct disease) Cholangitis Cholestasis Prednisone and UDCA

Treatment

AIH + CHOLESTATIC FEATURES

autoantibodynegative AIH

ANA and/or SMA + AMA − No ulcerative colitis Normal cholangiogram

AIH features No autoantibodies HLA-DR3 or DR4

Cholangitis Cholestasis Prednisone and/or UDCA depending on AP level and histologic features

Interface hepatitis Conventional regimens for AIH

AIH, autoimmune hepatitis; AMA, antimitochondrial antibodies; ANA, antinuclear antibodies; AP, serum alkaline phosphatase level; HLA, human leukocyte antigen; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; SMA, smooth muscle antibodies; UDCA, ursodeoxycholic acid; ULN, upper limit of normal.

patients who have AIH.41-43 As many as 41% of persons with one of these features have cholangiographic changes of PSC and are classifiable as having a variant syndrome of AIH (see Table 88-4).33 Furthermore, 54% of patients who have PSC have features that support a probable or definite diagnosis of AIH.42 The absence of characteristic cholangiographic changes does not preclude the diagnosis of PSC because small-duct disease may be present (see Table 88-4) (see Chapter 68).33,49 Magnetic resonance cholangiography has demonstrated unsuspected PSC of the large bile ducts in 8% of adults with AIH, and the possibility of PSC must be evaluated in all individuals with AIH and disease refractory to glucocorticoid therapy regardless of other features.50 Children with AIH may also have unsuspected bile duct changes. Autoimmune sclerosing cholangitis is a disorder described in children who have the clinical phenotype of AIH but abnormal findings on cholangiographic studies.51 Inflammatory bowel disease is frequently absent, and these children can respond to glucocorticoid therapy; in this respect, they may differ from adults with AIH and PSC. Treatment is empirical and typically ineffective in adults.33,42,43 Glucocorticoids and ursodeoxycholic acid (13 to 15 mg/kg orally daily), alone or in combination, can be considered, depending on whether hepatitic or cholestatic features predominate. Some studies have suggested that high-dose ursodeoxycholic acid (20 to 25 mg/kg daily) has some value in typical PSC, and this treatment can be considered in conjunction with glucocorticoids.

sent collateral damage associated with severe inflammatory activity; they do not constitute a variant syndrome, nor do they change the diagnosis or affect the treatment strategy.

AUTOANTIBODY-NEGATIVE AUTOIMMUNE HEPATITIS

Thirteen percent of adults with chronic hepatitis of undetermined cause satisfy international criteria for the diagnosis of AIH but lack the characteristic autoantibodies (see Table 88-4).53 These patients commonly are designated as having cryptogenic chronic hepatitis and may be excluded inappropriately from therapies of potential benefit. Autoantibody-negative patients are similar in age, female predominance, frequency of concurrent immunologic diseases, histologic features, and laboratory findings to patients with classic AIH.53 Furthermore, they have similar frequencies of HLA-B8, HLA-DR3, and HLA-A1-B8-DR3,53 and they respond as well to glucocorticoid treatment as do their autoantibody-positive counterparts.53,54 These persons probably have a form of AIH that has escaped detection by conventional serologic assays, and they are candidates for a closely monitored treatment trial of glucocorticoids. Assays for atypical pANCA, anti-SLA, and IgA endomysial antibodies or antibodies to tissue transglutaminase occasionally yield positive results in these patients,6,7 and successive testing for conventional autoantibodies may demonstrate the late appearance of typical autoimmune markers in some cases (see Table 88-4).55

VARIANT WITH CHOLESTATIC FEATURES

Eight percent of patients with AIH have histologic features of bile-duct injury and laboratory changes of cholestasis in the absence of AMA and cholangiographic changes of largeduct PSC.42,52 These individuals may have AMA-negative PBC, small-duct PSC, or a separate syndrome (“autoimmune cholangitis”) (see Table 88-4 and Chapter 89).42,52 This variant is probably a heterogeneous category that encompasses patients with atypical, early, or transitional features of classic disease.52 Persons who have this variant are inconsistently responsive to glucocorticoids and ursodeoxycholic acid.42,52 Preliminary experience suggests that these therapies can help to improve the clinical and laboratory abnormalities but not the histologic changes. Histologic findings of bile duct injury, including destructive cholangitis, may be seen in classic AIH in the absence of other cholestatic features.4,5 These changes are transient and probably repre-

AUTOIMMUNE HEPATITIS AND CHRONIC HEPATITIS C Eight percent of white North American adults with classic AIH have concurrent infection with hepatitis C virus (HCV), and 52% of patients with chronic hepatitis C have autoantibodies, concurrent immune diseases, or both.56 Identification of these patients as belonging to one or the other group is important because interferon therapy can enhance the immune manifestations of persons with AIH and concurrent HCV infection, and immunosuppressive treatment can increase serum viral levels in persons with chronic hepatitis C and background autoimmune features. The nature and degree of the associated immune mani­ festations distinguish AIH with coincidental HCV infection

1467

1468

Section IX  Liver

Autoimmune predominant

Autoab titers ≥1:320 Multiple autoab species Autoag-driven disease Interface hepatitis Plasma cells

Viral predominant

Autoab titers 3 6

Total Numerical Score 5-6 7-9 10-15 *Or Child-Pugh class.

Child Class* A B C

Chapter 90  Portal Hypertension and Gastrointestinal Bleeding Devascularization Procedures Devascularization procedures typically have been used to prevent recurrent variceal bleeding in patients with extensive splenic and portal vein thrombosis when a suitable vein is not available for creation of a portosystemic shunt.169 In the original operation described by Sugiura and Futagawa, both a thoracotomy and a laparotomy were carried out.170 Subsequently, the operation has been carried out through an abdominal approach and combined with a splenectomy. The procedure consists of total devascularization of the greater curvature of the stomach combined with devascularization of the upper two thirds of the lesser curvature of the stomach and circumferential devascularization of the lower 7.5 cm of the esophagus. The rate of recurrent bleeding following this procedure is variable but may be as high as 40%, depending on the population being treated and duration of follow-up.

Portosystemic Shunts

With the increasing availability of TIPS, the use of surgical shunts for refractory variceal bleeding has declined markedly. Surgical portosystemic shunts are categorized as selective shunts such as distal splenorenal shunts, partial shunts such as the side-to-side calibrated portacaval shunt, and total portosystemic shunts such as the side-to-side portacaval shunt or end-to-side portacaval shunt. Selective Shunts The most widely used selective shunt is the distal splenorenal shunt, originally described by Warren and colleagues.171 With this shunt, only varices at the gastroesophageal junction and spleen are decompressed, and portal hypertension is maintained in the superior mesenteric vein and portal vein; therefore, variceal bleeding is controlled, but ascites persists. The shunt procedure involves a portalazygous disconnection and subsequent anastomosis between

the splenic vein and left renal vein in an end-to-side fashion (Fig. 90-11). The entire length of the pancreas must be mobilized, and the left adrenal vein must be ligated. The distal splenorenal shunt has been associated with control of variceal bleeding in approximately 90% of patients and a lower rate of hepatic encephalopathy than that reported for total shunts.172 Partial Portosystemic Shunts A partial portosystemic shunt is carried out using a synthetic interposition graft between the portal vein and the inferior vena cava. When the shunt diameter is 8 mm, portal pressure is reduced below 12 mm Hg, and antegrade flow to the liver is maintained in most patients.173 Rates of preventing variceal rebleeding and encephalopathy following the shunt are similar to those seen with a distal splenorenal shunt. As in patients who have had a distal splenorenal shunt, ascites may occur in approximately 20% of patients who have had a partial portosystemic shunt because hepatic sinusoidal pressure is not reduced.174,175 Portacaval Shunts End-to-side and side-to-side portacaval shunts have been described, but nowadays only the side-to-side portacaval shunt is in use.176 Any portacaval shunt that is greater than 12 mm in diameter is likely to result in a total shunting of portal blood. A shunt with a diameter 18, require >4 units of packed red blood cell transfusions,185 and in whom renal failure develops have the highest risk of death. Alcohol as the cause of cirrhosis, a higher serum bilirubin level, a lower serum albumin level, hepatic encephalopathy, and hepatocellular carcinoma are additional factors associated with an increased six-week mortality rate. Treatment of esophageal variceal bleeding is classified as either primary prophylaxis, that is, prevention of variceal hemorrhage in patients who have never bled; control of acute variceal bleeding; or secondary prevention of rebleeding in patients who have survived an initial bleeding episode. Effective treatments to prevent the development of

varices and ascites in patients with cirrhosis are not yet available, although beta blockers may slow enlargement of small varices into large varices.

Prevention of Bleeding

The utility of pre-primary prophylaxis, that is, the efficacy of beta blockers to prevent the formation of varices, has not been demonstrated.79,186 Patients with Class C cirrhosis who have small varices may be considered for treatment with a beta blocker. All patients with large varices (diameter greater than 5 mm) should be considered for prophylactic therapy (“primary prophylaxis”) to prevent variceal bleeding. The presence of additional endoscopic signs such as red wales does not influence the decision regarding prophylactic therapy. Twelve trials have addressed the use of a nonselective beta blocker for primary prophylaxis of variceal bleeding and have demonstrated a decrease in the risk of variceal bleeding from 25% in patients in the control group to 15% in patients taking a beta blocker. The absolute risk reduction is thus approximately 10%, and the number needed to treat to prevent one variceal bleed is approximately 10 patients. The mortality rate is reduced from 28.4% in control patients to 23.9% in patients taking a beta blocker; the absolute risk reduction is 4.5%. The number of patients needed to be treated to prevent one death is approximately 22. In patients who do not bleed during therapy and who do not experience side effects, treatment should be continued indefinitely because withdrawal of a beta blocker can result in an increased risk of bleeding.187,188 The side effects of beta-blocker treatment are probably overemphasized because only approximately 15% of patients need to discontinue the drug.189 A baseline heart rate and blood pressure recording will help determine whether a patient is a candidate for pharmacologic treatment with a beta blocker. A resting heart rate of less than 55 to 60 beats per minute or a systolic blood pressure less than 90 mm Hg indicates that the patient is likely to be intolerant of beta blockers. In other patients, the HVPG ideally should be measured at baseline (Fig. 90-12). A longacting preparation of propranolol or nadolol may be started; the usual starting dose of long-acting propranolol is 60 mg once daily and that of nadolol is 20 mg once daily. Because the risk of bleeding is greatest at night, the beta blocker should probably be administered in the evening.108 The dose of propranolol or nadolol can be increased gradually every three to five days until the target heart rate of 25% below baseline or 55 to 60 beats per minute or the maximum tolerated dose is reached, provided that the systolic blood pressure remains above 90 mm Hg. The daily dose of long-acting propranolol or nadolol required to reach the target heart rate ranges from 40 to 160 mg. Patients with a decrease in systolic blood pressure below 90 mm Hg are most likely to experience side effects. In patients on pharmacologic therapy, follow-up endoscopy is unnecessary unless gastrointestinal bleeding occurs. When the target heart rate is reached, however, a repeat HVPG measurement may be carried out as close to one month as possible. In patients in whom the HVPG has decreased to less than 12 mm Hg, the risk of bleeding is virtually eliminated. Patients in whom the HVPG decreases by at least 20% have a risk of variceal bleeding of less than 10%. Unfortunately, only 30% to 40% of patients respond to a beta blocker; those with better liver function show the best response.189 In patients who are intolerant of or who have contraindications to beta blockers, isosorbide mononitrate has been tried but is no better than placebo in preventing variceal bleeding.190 In these patients, endoscopic prophylaxis should be pursued. Unfortunately,

Chapter 90  Portal Hypertension and Gastrointestinal Bleeding ligation to prevent primary prophylaxis is not currently recommended.

Upper endoscopy

No varices

Large esophageal varices

Small esophageal varices

Repeat endoscopy in 2–3 years

Consider HVPG measurement

Repeat endoscopy in 1–2 years

Nonselective β-adrenergic blockers

Consider HVPG remeasurement

Contraindication to β-adrenergic blockers Intolerance to β-adrenergic blockers Target HVPG not reached Patient preference

Perform EVL Figure 90-12.  Algorithm for primary prophylaxis of esophageal variceal hemorrhage. The hepatic venous pressure gradient (HVPG) may be measured before a nonselective b-adrenergic blocking agent is started and one month after the maximum tolerated dose of the beta blocker is reached. The goal of treatment is to reduce the HVPG to 3.0 mmol/L at 12 hr* or INR > 6.5 (or PT > 100 sec); and Serum creatinine > 3.4 mg/dL; and Stage 3 or 4 encephalopathy

INR > 6.5 (PT > 100 sec) or Any three of the following five: Age 40 yr Duration of jaundice > 7 days Cause: non-A, non-B hepatitis, halothane hepatitis, idiosyncratic drug reaction, indeterminate INR > 3.5 (or PT > 50 sec) Serum bilirubin > 17.5 mg/dL

*Measured after fluid resuscitation. INR, international normalized ratio; PT, prothrombin time.

requires only brief history taking, routine laboratory studies, and serologic testing for HAV and HBV.86 Another study from the U.S. Acute Liver Failure Study Group has demonstrated a lower sensitivity rate and negative predic­ tive value for these prognostic factors in 108 patients with acetaminophen-induced acute liver failure.15 Other groups have found that markers of organ failure, such as the Acute Physiology and Chronic Health Evaluation (APACHE) II score and Sequential Organ Failure Assessment (SOFA) index, are better predictors of outcome in patients with acetaminophen-related acute liver failure than the King’s College criteria and Model for End-stage Liver Disease (MELD) score (Table 93-6).16,87,88 Liver histologic evaluation in acute liver failure is associated with substantial sampling error and potential complications and does not reliably predict outcome.89 Therefore, percutaneous or transjugular liver biopsy is not recommended for prognosis or staging purposes but can be helpful in confirming a diagnosis of malignant infiltration, autoimmune hepatitis, and nonhepatotropic viral infection (Fig. 93-2). The predictive value of serum Gc-globulin levels is comparable to the King’s College criteria, but the assay is technically difficult and not generally available.90 Other investigators have examined the prognostic useful­ ness of measuring plasma factor V levels (see Table 95-5)91 and hepatic volumetry,92 but these parameters do not appear to add significantly to the assessment of outcome. The U.S. Acute Liver Failure Study Group and others have reported on the potential use of elevated serum phosphate levels (>3.7  mg/dL) as a marker of impaired liver regenera­ tion and poor prognosis in patients with acetaminopheninduced acute liver failure.93,94 Other groups, however, have failed to confirm the clinical usefulness of serum phosphate measurements in acute liver failure.95 Serial assessment of serum alpha fetoprotein levels, which cor­ relate with hepatic regeneration, have been reported to be of value in predicting prognosis in patients with acute liver failure.96

TREATMENT A variety of therapies have been proposed and studied in patients with acute liver failure, including glucocorticoid therapy, prostaglandin infusions, and exchange transfu­

1563

1564

Section IX  Liver Table 93-6  Criteria Used to Predict the Prognosis of Patients with Acetaminophen-Induced Acute Liver Failure SOURCE

CRITERIA

U.S. ALFSG, N = 275

Royal Free Hospital, N = 100

SENSITIVITY (%)

SPECIFICITY (%)

PPV

NPV

APACHE II score ≥ 20 King’s College criteria ≥ 1 (see Table 93-5) MELD score ≥ 35

68 26

87 92

0.77 0.63

0.81 0.69

61

71

0.54

0.76

APACHE II score ≥ 12 Arterial lactate level ≥ 3.3 mmol/L King’s College criteria ≥ 1 (see Table 93-5) SOFA score ≥ 12

67 68 47

76 73 83

0.69 0.68 0.70

0.75 0.73 0.65

67

80

0.74

0.74

ALFSG, Acute Liver Failure Study Group; APACHE, Acute Physiology and Chronic Health Evaluation; MELD, Model for End-stage Liver Disease; PPV, positive predictive value; NPV, negative predictive value; SOFA, sequential organ failure assessment. Adapted from Larson AM, Polson J, Fontana RJ, et al. Acetaminophen–induced acute liver failure: Results of a United States multicenter prospective study. Hepatology 2005; 43:1364-72; and Cholongitas EB, Betrossian A, Leandro G, et al. King’s criteria, APACHE II, and SOFA scores in acute liver failure. Hepatology 2006; 43:881.

Patient with acute liver failure

Intensive medical management Cause-specific treatment (if appropriate) Immediate consultation with liver transplantation center

Contraindications to transplantation

Yes

Continued intensive support

No Transfer to liver transplantation center Place on waiting list for emergency transplantation

Donor organ available Figure 93-2.  Liver histopathology in acute liver failure. The specimen shown is from a 59-year-old man who underwent a transjugular liver biopsy for presumed amoxicillin-clavulanate liver toxicity. Findings include severe acute hepatitis with confluent areas of necrosis predominating in the central zones, consistent with, but not diagnostic of, an idiosyncratic drug reaction. Despite supportive care, multiorgan failure developed, culminating in death. (Hematoxylin and eosin, ×40.)

Yes Assess prognosis

Recovery likely

Yes

No

sions. Only liver transplantation, however, has permitted salvage of patients with irreversible liver failure. Unfortu­ nately, many patients with irreversible acute liver failure do not undergo liver transplantation because of late referral, contraindications, or the lack of a donor liver. Therefore, patients with acute liver failure should be evaluated for liver transplantation as soon as possible and, if no contra­ indications are identified, placed on a liver transplant waiting list. If and when a donor organ becomes available, a patient listed for transplantation should be reassessed for the continued need for transplantation. An algorithm depicting the management of acute liver failure is shown in Figure 93-3.

INITIAL EVALUATION AND MANAGEMENT

The initial management of a patient with acute liver failure should include rapid identification of the cause of acute liver failure, with an emphasis on treatable conditions. For example, acetaminophen toxicity is treated initially with

No

Contraindications to transplantation

Ongoing intensive support Reassess for recovery and presence of contraindications to transplantation Yes

No Liver transplantation Figure 93-3.  Algorithm for the management of acute liver failure. The initial approach to management includes rapid identification and treatment of reversible causes of acute liver failure. Supportive care with careful monitoring for complications of the disorder in an intensive care unit is recommended. Contact with a liver transplantation center should be established promptly for potential transfer and transplantation evaluation as early as possible.

gastric lavage, oral charcoal, and prompt administration of oral N-acetylcysteine (see Chapter 86).97 For patients with severe nausea and vomiting, an approved intravenous for­ mulation of N-acetylcysteine can be administered safely in a monitored setting.98 Similarly, patients with Amanita

Chapter 93  Acute Liver Failure mushroom poisoning should be treated with immediate gastric lavage and instillation of charcoal in an attempt to reduce the toxin load.99 In addition, hemodialysis can remove toxin from the serum, and intravenous penicillin, milk thistle (silymarin), and cytochrome c may further lower the enterohepatic toxin load (see Chapter 87); however, the clinical benefit of these measures is uncer­ tain.100 HSV-induced acute liver failure has been reported to respond to intravenous acyclovir.101 Rapid delivery and supportive care constitute the treatment of choice for preg­ nant women with acute fatty liver of pregnancy, the he­ molysis, elevated liver enzyme levels, and low platelet (HELLP) syndrome, and preeclampsia (see Chapter 38). The benefit of oral antiviral agents in patients with fulminant hepatitis B remains unproven; many experts advise pre­ scribing a nucleoside analog, such as entecavir or lamivu­ dine, in light of their favorable safety profiles, in the hope of salvaging liver function and reducing the level of viremia prior to liver transplantation (see Chapter 78).102-104 N-acetylcysteine has been proposed as a potential treat­ ment for non–acetaminophen-related acute liver failure on the basis of studies demonstrating improvements in tissue oxygenation and systemic hemodynamics and the drug’s antioxidant properties.105 The U.S. Acute Liver Failure Study Group reported on a multicenter, randomized, con­ trolled trial that compared a 72-hour infusion of intravenous N-acetylcysteine and placebo in 173 adult patients with acute liver failure.106 The overall patient survival at three weeks was similar in the two groups (70% N-acetylcysteine vs. 66% placebo, P = 0.28), but transplant-free survival was significantly better in patients who received N-acetylcyste­ ine (40% vs. 27%, P = 0.04). The benefit of N-acetylcysteine appeared to be limited to the subgroup of patients with grade 1 or 2 encephalopathy at entry (52% vs. 31%, P = 0.021). In addition, a single-center retrospective study of 170 children with nonacetaminophen acute liver failure demonstrated that treatment with N-acetylcysteine was associated with a shorter length of hospital stay and a higher rate of spontaneous recovery.107 In this study, however, the untreated controls were not contemporaneous and had more severe illness at presentation. In both studies,106,107 N-acetylcysteine was generally well tolerated, with a low rate of side effects (e.g., rash, bronchospasm, arrhyth­ mia). Additional studies of N-acetylcysteine for non– acetaminophen-related acute liver failure are in progress to identify which patients may benefit from this treatment. All patients with acute liver failure should be cared for in an intensive care unit because they can deteriorate rapidly.104 Serial laboratory studies, including acid-base status, arterial ammonia levels, and INR, should be carried out to monitor the patient’s condition. Urgent transfer to a liver transplantation center is advisable early in the course, prior to the development of advanced encephalopathy or complications.108

ENCEPHALOPATHY AND CEREBRAL EDEMA

Encephalopathy associated with acute liver failure tends to be progressive, unless liver failure is reversed. Sedativehypnotic drugs, which may exacerbate encephalopathy, should be avoided unless patients require mechanical ven­ tilation. Lactulose is of uncertain benefit and may be asso­ciated with bowel ischemia. Reversible conditions that may contribute to altered mental status (e.g., hypogly­ cemia, hypoxemia) should be treated immediately. Hypoglycemia generally responds to parenteral administra­ tion of glucose. Similarly, underlying infection and sepsis should be treated aggressively with fluids and anti­

biotics, because systemic cytokines may alter brain function (see Table 93-4). Patients with stage 3 or 4 encephalopathy should undergo elective endotracheal intubation and mechanical ventila­ tion for protection of the airway, particularly before being transported to a liver transplantation center. Many mechani­ cally ventilated patients are also deeply sedated or para­ lyzed, and evidence of generalized seizure activity that can worsen encephalopathy may be concealed. Therefore, continuous electroencephalographic monitoring of deeply sedated or paralyzed patients with acute liver failure has been proposed. Treatment of subclinical seizures with phe­ nytoin or other antiepileptic medications is appropriate, but the efficacy of prophylactic therapy to prevent seizure activity has not been established.109 Intracranial hypertension can be suspected on the basis of noninvasive assessment or direct measurement, but non­ invasive assessment by physical examination and radiologic imaging has important limitations. Impaired pupillary responses, posturing, or seizures, which may suggest the presence of intracranial hypertension, are not sensitive signs for intracranial hypertension, particularly when seda­ tives or neuromuscular blocking agents are used in mechan­ ically ventilated patients. CT of the head is useful for identifying mass lesions, intracranial hemorrhage, and evi­ dence of brainstem herniation because these conditions may affect clinical decision making. CT scans of the head should be obtained in all patients with advanced encepha­ lopathy. Nevertheless, the correlation between CT evidence of cerebral edema and measured ICP is imperfect, with a sensitivity varying from 60% to 75%.110,111 Transcranial Doppler measurements of middle cerebral artery blood flow continue to improve, but further refinements are needed to make this noninvasive modality reliable enough for clinical decision-making.112 Monitoring of ICP represents the most accurate way to detect intracranial hypertension but has several potential limitations. First, placement of an ICP transducer requires correction of underlying coagulopathy. Second, the ICP transducer represents a potential portal of entry for infec­ tious organisms. Third, placement of the transducer can precipitate intracranial hemorrhage, which can be fatal. The frequency of serious complications ranges from 4% to 20%; parenchymal catheters are associated with a higher rate of complications than subdural or epidural transducers.113 Nevertheless, ICP transducers can provide invaluable physi­ ologic data that influence management and decisions regard­ ing liver transplantation. In one study, 92 patients with advanced encephalopathy had ICP monitors placed and were compared with 239 unmonitored patients. The moni­ tored patients tended to have more severe liver failure and multiorgan failure.114 In subjects listed for liver trans­ plantation, ICP monitoring was associated with a greater number of medical treatments and procedural interven­ tions. Although overall and post-transplantation survival rates were similar in the two groups, the monitored patients had a low rate of intracranial bleeding (i.e., 10%). The data suggest that future studies of therapies for acute liver failure should include ICP monitoring. For patients with refractory intracranial hypertension, technetium perfusion scans are useful to detect irreversible brain death, which may other­ wise be unrecognized in a sedated patient. Elevation of the head of the bed to at least 30 degrees from horizontal (and avoidance of the head-down position) is a simple measure to reduce ICP. If this maneuver fails, spe­ cific treatment is required. Osmotherapy and barbiturates are two options for treating intracranial hypertension. Osmotherapy with intravenous mannitol (0.5 to 1 g/kg)

1565

1566

Section IX  Liver requires preserved renal function (or concomitant hemofil­ tration, if necessary) and effectively controls intracranial hypertension in approximately 60% of cases.115 Hypertonic saline may also be of value in patients with cerebral edema, but controlled trials are needed to confirm a benefit.116 Uncontrolled data have supported the use of intravenous thiopental, a barbiturate; its efficacy is similar to that of mannitol.117 Thiopental has two relative advantages—its onset of action is rapid, and its use does not require pre­ served renal function. Potential drawbacks of thiopental are a risk of hypotension and, of greater importance, the poten­ tial to mask clinical indicators of neurologic recovery or deterioration. In general, it is reasonable to use mannitol as first-line therapy and to reserve a barbiturate for patients with renal insufficiency or refractory intracranial hyperten­ sion. Glucocorticoids are of no benefit.113 Pilot studies have suggested that moderate hypothermia (e.g., 32° C to 33° C), achieved by the use of an external cooling blanket, may be of benefit in patients with acute liver failure and refractory cerebral edema.118,119 All patients treated with hypothermia require the placement of an ICP monitor because paralytic agents are required to prevent shivering. Moreover, the effect of moderate hypothermia on the risk of bleeding and infection require further study. A prophylactic role for hypothermia also requires further study.

COAGULOPATHY AND BLEEDING

Placement of a nasogastric tube to monitor gastrointestinal bleeding and gastric pH is recommended for intubated patients with acute liver failure. The risk of upper gastroin­ testinal hemorrhage can be reduced by intravenous admin­ istration of a histamine H2 receptor antagonist,120 and proton pump inhibitors probably have a similar benefit. Adminis­ tration of subcutaneous vitamin K to attempt to reverse hypoprothrombinemia is also reasonable. Coagulation parameters, including the INR, plasma factor V level, plate­ let count, and plasma fibrinogen level, should be assessed serially in all patients with acute liver failure. The decision to administer clotting factors prophylacti­ cally in nonbleeding patients should be tempered by the recognition that infusion of plasma will mitigate the value of the INR as a prognostic factor. Furthermore, infusion of plasma can lead to volume overload and respiratory failure, particularly in patients with renal failure, and prophylactic administration of plasma has not been shown to improve the clinical outcomes of patients with acute liver failure. Therefore, unless a patient is actively bleeding or an inva­ sive procedure is planned, prophylactic infusions of plasma are not recommended. For patients with acute liver failure who undergo an inva­ sive procedure and whose INR fails to improve with plasma, recombinant factor VIIa has been shown to be efficacious, but the optimal dose has not been established and use of this agent carries a risk of thrombosis.121 Cryoprecipitate, plasma, and platelets should be given to patients with hypo­ fibrinogenemia, DIC, and active bleeding.

INFECTION

Clinical recognition of infection may be difficult because signs such as hypotension, leukocytosis, and acidosis may reflect the underlying liver failure. Therefore, daily surveil­ lance cultures of blood, urine, and ascitic fluid are recom­ mended in patients with acute liver failure. The advisability of prophylactic antibiotics in the setting of acute liver failure is debatable. On one hand, prophylactic antibiotics may delay the development of infections that limit the

applicability of liver transplantation. On the other hand, antibiotics may increase the risk of superinfection with resistant bacteria or fungi. This issue has been addressed in a small randomized trial.122 Patients treated with prophylac­ tic intravenous cefuroxime had a significant reduction in the rate of documented infections (from 61% to 32%) com­ pared with those treated conservatively and a modest (but statistically insignificant) increase in the rate of survival (from 45% to 67%). Enteral decontamination with orally administered antibiotics (as well as systemic antibiotics) does not appear to alter the clinical outcome of patients with acute liver failure, compared with that observed with systemic antibiotics alone.123 The usefulness of systemic prophylactic antibiotics warrants further investigation. At the least, a high level of suspicion for infection and a low threshold for administering antibiotics are required in man­ aging patients with acute liver failure. If infection is sus­ pected, the choice of antibiotics should be based on the spectrum of likely bacterial pathogens (e.g., Staphylococcus, gram-negative aerobes) and local hospital patterns of microbial sensitivity. A reasonable empirical regimen is intravenous vancomycin and a third-generation cephalo­ sporin or fluoroquinolone.

MULTIPLE ORGAN FAILURE SYNDROME

The fundamental goal of management of multiple organ failure syndrome in patients with acute liver failure is similar to that in patients with other causes of multiple organ failure—to optimize arterial pressure and tissue oxy­ genation. Ideally, the mean arterial pressure (MAP) should be kept above 60 mm Hg to maintain cerebral perfusion.64 A central venous or right heart catheter may be useful for monitoring the patient’s intravascular volume status. Hypo­ tension resulting from intravascular volume depletion should be corrected with blood or colloids. If hypotension is caused by reduced vascular resistance, administration of an α-adrenergic agonist may be useful. Although pressors can be used to maintain MAP within a physiologic range, they have the potential to impair tissue oxygenation further; terlipressin (a long-acting vasopressin analog not available in the United States) may worsen cerebral edema (see Chapters 90 and 92).124,125 Most experts recommend using norepinephrine or dopamine rather than vasopressin because of adverse effects of the latter on intracranial hyper­ tension.125 In small short-term studies, N-acetylcysteine has been shown to improve tissue oxygenation without adverse effects on hemodynamics126; however, the impact of this agent on overall patient outcome has not yet been determined. Endotracheal intubation and mechanical ventilation are frequently necessary for patients with acute liver failure. Hypoxemia can result from respiratory depression caused by coma or impaired gas exchange caused by ARDS or superimposed pneumonia. Vigorous suctioning and Val­ salva maneuvers should be avoided to prevent surges in ICP. Patients with acute liver failure tolerate volume overload poorly in light of their propensity to develop ARDS and cerebral edema. Early measurement of the central venous or pulmonary capillary wedge pressure is preferable in oligu­ ric patients to empirical administration of fluid boluses. If oliguria persists in the face of adequate central filling pres­ sures, continuous renal replacement therapy should be ini­ tiated. Continuous venovenous hemofiltration has been shown to be superior to intermittent hemodialysis, with less hemodynamic instability and improved tissue oxygen deliv­ ery, in oliguric patients with acute liver failure.127 Nephro­ toxic drugs such as aminoglycosides and nonsteroidal

Chapter 93  Acute Liver Failure Table 93-7  Results of Liver Transplantation for Acute Liver Failure in the United States LIVER TRANSPLANTATION CENTER LOCATION* Chicago132 Michigan131 Nebraska134 Pittsburgh133 Philadelphia129 San Francisco132 United States (multicenter)130 United States (multicenter)8 Total

NO. OF PATIENTS

STUDY PERIOD

EARLY PATIENT SURVIVAL (%)†

ONE-YEAR PATIENT SURVIVAL (%)

19 19 30 42 18 35 121 89 373

1984-1988 1985-1990 1986-1991 1980-1987 1985-1990 1988-1992 1994-1996 1998-2001 —

74 NR 75 74 65 94 NR 84 80 (187/233)

58 68 42 59 65 92 76 NR 70 (198/284)

*Superscript numbers indicate references. † Early patient survival reflects discharge from the hospital following transplantation. NR, not reported.

anti-inflammatory drugs should be avoided in all patients with acute liver failure, and appropriate precautions should be taken if intravenous contrast dye is required.

LIVER TRANSPLANTATION

Liver transplantation has transformed the management of patients with acute liver failure and is discussed in greater detail in Chapter 95. Before the advent of liver transplanta­ tion, less than 30% of patients with acute liver failure sur­ vived. By contrast, survival rates for patients with acute liver failure who undergo liver transplantation have been substantially higher, with a short-term survival rate of 80% and a one-year survival rate of 70% when the results of several major transplantation centers are combined (Table 93-7).8,128-134 The decision to perform transplantation in a patient with acute liver failure must balance the likelihood of spontaneous recovery with the risks of surgery and longterm immunosuppression. Furthermore, contraindications to transplantation, particularly irreversible brain damage, active extrahepatic infection, or multiple organ failure syn­ drome, must be considered. Although changes in the rules governing allocation of donor livers have shortened waiting times for patients with acute liver failure in the United States, the decision to place a patient on the waiting list for transplantation must still be made promptly.135,136 A patient’s clinical status needs to be assessed frequently to determine whether the patient is likely to recover or has developed a contraindication to transplantation. In one series, a contraindication that pre­ cluded transplantation developed in 22% of liver transplan­ tation candidates, whereas 12% improved and were removed from the waiting list.8 The shorter median waiting times among liver transplant recipients compared with the times to exclusion for nontransplanted patients (three versus five days) highlights the critical, ongoing shortage of donor organs. Because of the shortage of donor organs, patients with acute liver failure are more likely to receive an ABOincompatible rather than an ABO-compatible or ABOidentical graft.137 In addition, marginal donor grafts that are older or steatotic are used more frequently in transplant recipients with acute liver failure than in recipients with other indications for transplantation. These factors may explain in part the higher rate of primary nonfunction and rejection among transplant recipients with acute liver failure compared with other recipients.138 Posttransplantation seronegative chronic hepatitis is also more common in transplant recipients with acute liver failure than in recipients with cirrhosis (41% versus 14% at one

year).139 The long-term functional and cognitive outcomes of transplant recipients with acute liver failure have not been well studied but may be inferior to those of transplant recipients with cirrhosis.140

Live Donor Liver Transplantation

In countries in which cadaveric livers are not readily avail­ able, live donor liver transplantation has been performed successfully in highly selected patients with acute liver failure (see Chapter 95).141,142 In the United States, adult live donor liver transplantation has rarely been undertaken for acute liver failure, and a moratorium has been placed on using this procedure for patients with acute liver failure in New York. One study, however, has demonstrated that seven of ten live donor organ recipients with acute liver failure survived, compared with two of three cadaveric transplant recipients with acute liver failure.143 In addition, the frequency of complications in live donors to a patient with acute liver failure (50%) was similar to that reported in donors who were evaluated electively for donation to a transplant recipient with cirrhosis (33%). Nevertheless, because of concerns regarding the ability to evaluate donors safely in an accelerated time frame, the potential for coer­ cion, and the potential for poorer outcomes, most experts recommend a whole-sized cadaveric transplant whenever possible in a patient with acute liver failure.

INVESTIGATIONAL APPROACHES Treatment strategies, such as charcoal hemoperfusion and administration of prostaglandin E1, which showed promise in uncontrolled trials, have not been shown to be superior to standard care when studied in randomized studies.144,145 Plasmapheresis and hepatectomy have been suggested as a possible bridge to liver transplantation, but prospective trials have yet to be performed.146,147 Three additional forms of therapy may provide a bridge to liver transplantation or to regeneration of the native liver with spontaneous recovery—auxiliary liver transplantation, extracorporeal liver support devices, and hepatocyte transplantation.

AUXILIARY LIVER TRANSPLANTATION

Auxiliary liver transplantation, in which the donor graft is implanted orthotopically beside the surgically reduced native liver or heterotopically inferior to the native liver, has been investigated by a number of centers.148,149 The

1567

1568

Section IX  Liver advantage of this procedure is that by providing a temporary auxiliary liver, the severely diseased native liver may be allowed to regenerate. Ideally, immunosuppression may then be gradually withdrawn, thereby allowing the trans­ planted liver to involute or be surgically removed. The usefulness of this operation is limited by technical compli­ cations as well as by the difficulty in predicting which patients with acute liver failure are likely to experience hepatic regeneration. This approach should only be under­ taken in centers with specialized expertise.

EXTRACORPOREAL LIVER SUPPORT

Extracorporeal liver support devices fall into two broad categories, hemodiadsorption systems and bioartificial livers.150 Hemodiadsorption systems use hemodialysis in combination with perfusion of the patient’s plasma or blood through three hollow fiber filters impregnated with char­ coal, resins, and albumin, respectively. Although these devices may remove circulating toxins, they do not replace other liver functions. Albumin dialysis (e.g., the molecular adsorbent recirculating system [MARS]) uses hemodialysis of whole blood in series with an albumin dialyzer and char­ coal filter. This technology is simpler to use than plasma­ pheresis and has shown some promise in trials of cirrhotic patients with hepatic encephalopathy,151,152 but information regarding the safety and efficacy of these systems in patients with acute liver failure is limited.153,154 Bioartificial liver devices contain liver cells grown within specialized hollow fiber cartridges through which the patient’s plasma is perfused. The success of such devices depends largely on the mass of cells they contain, the extent to which these cells maintain liver-specific functions, and the duration for which these functions are maintained. Because the devices under clinical investigation contain only hepatocytes, derangements attributable to nonparen­ chymal cells, such as Kupffer cells and biliary epithelia, are not replaced. The results of the HepatAssist bioartificial liver device trial were reported in 2004.155 This device uses a dialysis cartridge loaded with approximately 100 g of cryopreserved porcine hepatocytes, or 7 billion cells, and also has a charcoal filter. The 85 patients with acute liver failure who were treated with the HepatAssist device did not experience an improvement in 30-day survival com­ pared with the 86 patients with acute liver failure who received standard care (71% versus 62%, P = 0.26). Treat­ ment with the device was well tolerated, and the rate of thrombocytopenia, hypotension, and other adverse events was not significantly greater in treated patients. Further­ more, there were no reports of inadvertent transmission of porcine retroviruses or development of xenogenic antibod­ ies in treated patients.156 Although this pioneering trial failed to demonstrate a significant benefit in outcome with the HepatAssist device, the proof of concept that an extra­ corporeal device with porcine hepatocytes can be used and can lead to a trend toward improvement in metabolic, hemodynamic, and clinical parameters was realized. The development of devices with a larger hepatocyte mass, sim­ plified circuitry, and differentiated hepatocyte function is eagerly awaited.

HEPATOCYTE TRANSPLANTATION

The potential role of hepatocyte transplantation in patients with acute liver failure is likely to be as a bridge to liver transplantation or regeneration.157 Human hepatocyte trans­ plantation has demonstrated efficacy in preliminary studies of patients with metabolic disorders and decompensated

cirrhosis.158,159 Stable expression of transplanted hepato­ cytes has been difficult to achieve. In one trial, three of six patients with acute liver failure survived 14, 20, and 52 days after transplantation of 109 to 1010 hepatocytes, representing 1% to 10% of normal liver cell mass.121 Although metabolic parameters improved within 72 hours of transplantation, transient respiratory insufficiency was observed in several patients. This report demonstrated that transplanted human hepatocytes can engraft into a regenerating liver, but further work on enhancing graft function, native liver regeneration, and cell delivery is needed. Pluripotent hepatocyte stem cells derived from bone marrow may prove useful for hepatocyte transplantation. In one study, liver biopsy specimens from human recipients of gender-discordant bone marrow or liver transplants were analyzed for marrow-derived hepatocytes and cholangio­ cytes.160 Differences in the extent of engraftment were asso­ ciated with the degree of allograft injury, thereby suggesting a possible role for bone marrow–derived stem cells in the treatment of severe acute hepatitis, acute liver failure, and metabolic defects. Further studies are needed to identify and isolate pluripotent liver stem cells from human bone marrow and clarify the factors that govern cellular differen­ tiation and liver regeneration.161

KEY REFERENCES

Bernal W, Donaldson N, Wyncoll D, et al. Blood lactate as an early predictor of outcome in paracetamol-induced acute liver failure: A cohort study. Lancet 2002; 359:558-63. (Ref 85.) Bernal W, Hall C, Karvellas CJ, et al. Arterial ammonia and clinical risk factors for encephalopathy and intracranial hypertension in acute liver failure. Hepatology 2007; 46:1844-52. (Ref 67.) Davern TJ, James LP, Hinson JA, et al. Measurement of serum acetaminophen-protein adducts in patients with acute liver failure. Gastroenterology 2006; 130:687-94. (Ref 62.) Demetriou AA, Brown RS, Busuttil RW, et al. Prospective, randomized, multicenter controlled trial of a bioartificial liver in treating acute liver failure. Ann Surg 2004; 239:660-70. (Ref 155.) Fontana RJ. Acute liver failure due to drugs. Sem Liv Dis 2008; 28:17588. (Ref 27.) Houlihan DD, Newsome PN. Critical review of clinical trials of bone marrow stem cells in liver Disease. Gastroenterology 2008; 135:43850. (Ref 161.) Kumar M, Satapathy S, Monga R, et al. A randomized controlled trial of lamivudine to treat acute hepatitis B. Hepatology 2007; 45:97-101. (Ref 102.) Larson AM, Polson J, Fontana RJ, et al. Acetaminophen-induced acute liver failure: Results of a United States Multicenter prospective study. Hepatology 2005; 43:1364-72. (Ref 16.) McDiarmid SV, Goodrich NP, Harper AM, et al. Liver transplantation for status 1: The consequences of good intentions. Liver Transpl 2007; 13:699-707. (Ref 136.) O’Grady JG, Alexander GJ, Hayllar KM, et al. Early indicators of prog­ nosis in fulminant hepatic failure. Gastroenterology 1989; 97:439-45. (Ref 83.) Ostapowicz G, Fontana RJ, Schiodt FV, et al. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med 2002; 137:947-54. (Ref 8.) Stravitz RT, Kramer AH, Davern TJ, et al. Intensive care of patients with acute liver failure: Recommendations of the U.S. Acute Liver Failure Study Group. Crit Care Med 2007; 35:2498-508. (Ref 104.) Vaquero J, Fontana RJ, Larson AM, et al. Complications and use of intracranial pressure monitoring in patients with acute liver failure and severe encephalopathy. Liver Transpl 2005; 11:1581-9. (Ref 114.) Wade J, Rolando N, Philpott-Howard J, et al. Timing and a cause of bacterial infections in a liver intensive care unit. J Hosp Infection 2003; 53:144-6. (Ref 72.) Watkins PB, Seeff LB. Drug-induced liver injury: Summary of a singletopic clinical research conference. Hepatology 2006; 43:618-31. (Ref 36.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

94 Tumors and Cysts of the Liver Adrian M. Di Bisceglie and Alex S. Befeler

CHAPTER OUTLINE Primary Malignant Tumors  1569 Hepatocellular Carcinoma  1569 Intrahepatic Cholangiocarcinoma  1580 Hepatoblastoma  1582 Hemangiosarcoma  1582 Epithelioid Hemangioendothelioma  1583 Other Primary Malignant Tumors of the Liver  1583 Hepatic Metastases  1583 Benign Tumors  1584 Hepatocellular Adenoma  1584

Mass lesions of the liver occur sufficiently often that clinicians interested in liver diseases should have a thorough understanding of their presentations, diagnosis, and treatment. Hepatic mass lesions include tumors, tumor-like lesions, abscesses, cysts, hematomas, and confluent granulomas. The frequency with which each is seen varies in different geographic regions and different populations. The more common hepatic tumors and cysts and those important for other reasons are reviewed in this chapter. Hepatic tumors may originate in the liver—from hepatocytes, bile duct epithelium, or mesenchymal tissue—or spread to the liver from primary tumors in remote or adjacent organs. In adults in most parts of the world, hepatic metastases are more common than primary malignant tumors of the liver, whereas in children, primary malignant tumors outnumber both metastases and benign tumors of the liver. Except for cavernous hemangiomas, benign hepatic tumors are rare in all geographic regions and in all age groups.

PRIMARY MALIGNANT TUMORS Among primary malignant tumors of the liver, hepatocellular carcinoma is by far the most common.

HEPATOCELLULAR CARCINOMA Epidemiology

Hepatocellular carcinoma is the commonest primary malignant tumor of the liver. It is the fifth most common cancer in men and the eighth most common in women, and it ranks fourth in annual cancer mortality rates.1,2 Information on incidence is derived from an increasing but still limited number of cancer registries, and it is possible to classify countries into broad risk categories only. Moreover, in

Cavernous Hemangioma  1586 Infantile Hemangioendothelioma  1587 Other Benign Tumors of the Liver  1587 Tumor-like Hepatic Lesions  1587 Focal Nodular Hyperplasia  1587 Other Nodular Disorders  1588 Hepatic Cysts  1589 Fibrocystic Diseases of the Liver  1589 Approach to the Patient with a Hepatic Mass Lesion  1590

low-income (developing) countries, especially in subSaharan Africa, hepatocellular carcinoma is underdiagnosed and underreported, in some cases by as much as 50%. Despite these sources of inaccuracy, hepatocellular carcinoma clearly has an unusual geographic distribution (Fig. 94-1). Moreover, the tumor is not necessarily uniformly common throughout countries with a high incidence, such as China3 and Mozambique.4 The incidence of hepatocellular carcinoma has increased considerably in Japan since the 1980s, and lesser increases have been recorded in developed Western countries, including North America and Western Europe.5 Interestingly, a study from Japan has shown that the rate of hepatocellular carcinoma began to decline in 2000, presumably because of the aging of the cohort of persons infected with hepatitis C virus (HCV).6 A similar downward trend has been noted in some European countries, including France and Italy.7 By contrast, in the United States, hepatocellular carcinoma is the cancer that has been increasing in incidence most rapidly since 2000, at a time when other major cancers such as cancers of the lung, breast, prostate, and colon are decreasing.8 Considerable racial and ethnic variation exits in the incidence of hepatocellular carcinoma in the United States. The incidence among Asians is highest, almost double that of white Hispanics and more than four times higher than that of whites.9 Migrants from countries with a low incidence to areas with a high incidence of hepatocellular carcinoma usually retain the low risk of their country of origin, even after several generations in the new environment. The consequences for migrants from countries with a high incidence to those with a low incidence differ, depending on the major risk factors for the tumor in their country of origin and whether chronic hepatitis B virus (HBV) infection, if this is the major risk factor, is acquired pre­ dominantly by the perinatal or horizontal route (see later and Chapter 78).2,10,11

1569

1570

Section IX  Liver

High Intermediate Low Figure 94-1.  Incidence of hepatocellular carcinoma in different parts of the world. High, age-adjusted rate of more than 15 cases/100,000 population/ year; intermediate, age-adjusted rate of 5 to 15 cases/100,000/year; low, age-adjusted rate of fewer than 5 cases/100,000/year.

Men are generally more susceptible than women to hepatocellular carcinoma. Male predominance is, however, more obvious in populations at high risk for the tumor (mean male-to-female ratio, 3.7 : 1.0) than in those at low or intermediate risk (2.4 : 1.0).1,2 In industrialized countries, the number of men and number of women with hepatocellular carcinoma in the absence of cirrhosis are almost equal. The incidence of hepatocellular carcinoma increases progressively with advancing age in all populations, although it tends to level off in the oldest age groups.1,2 In Chinese and particularly in black African populations, however, the mean age of patients with the tumor is appreciably younger than in other populations. This finding is in sharp contrast to the age distribution in Japan, where the incidence of hepatocellular carcinoma is highest in the cohort of men ages 70 to 79 years.6 Hepatocellular carcinoma is rare in children.12,13

Clinical Features

Although the typical clinical features of hepatocellular carcinoma are well recognized (including abdominal pain and weight loss in patients with cirrhosis), more patients are now being diagnosed at an early stage, when they have no specific symptoms or signs. This trend toward earlier diagnosis is probably the result of surveillance programs in patients with chronic liver disease (see later). In faradvanced disease, patients with hepatocellular carcinoma usually present with typical symptoms and signs, and diagnosis is easy. In addition, hepatocellular carcinoma often coexists with cirrhosis,14 and the onset of hepatocellular carcinoma is marked by a sudden unexplained change in the patient’s condition. Patients with hepatocellular carcinoma often are unaware of its presence until the tumor has reached an advanced stage. The most common, and frequently first, symptom is right hypochondrial or epigastric pain. Other symptoms are listed in Table 94-1. Physical findings vary with the stage of disease. Early in the course, evidence of cirrhosis alone may be present, or abnormal findings may be absent (see Table 94-1). When the

Table 94-1  Symptoms and Signs of Hepatocellular Carcinoma SYMPTOM Abdominal pain Weight loss Weakness Abdominal swelling Nonspecific gastrointestinal symptoms Jaundice

FREQUENCY (%) 59-95 34-71 22-53 28-43 25-28 5-26

SIGN Hepatomegaly Ascites Fever Splenomegaly Wasting Jaundice Hepatic bruit

54-98 35-61 11-54 27-42 25-41 4-35 6-25

tumor is advanced at the time of the patient’s first medical visit, the liver is almost always enlarged, sometimes massively. Hepatic tenderness is common and may be profound, especially in the later stages. The surface of the enlarged liver is smooth, irregular, or frankly nodular. An arterial bruit may be heard over the tumor15; the bruit is heard in systole, rough in character, and not affected by changing the position of the patient. Although not pathognomonic, a bruit is a useful clue to the diagnosis of hepatocellular carcinoma. Less often, a friction rub may be heard over the tumor, but this sign is more characteristic of hepatic metastases or abscesses. Ascites may be present when the patient is first seen or may appear with progression of the tumor. In most patients, ascites is the result of long-standing cirrhosis and portal hypertension (see Chapter 91), but in some cases it is caused by invasion of the peritoneum by the primary tumor or

Chapter 94  Tumors and Cysts of the Liver Table 94-2  Paraneoplastic Syndromes Associated with Hepatocellular Carcinoma Carcinoid syndrome Hypercalcemia Hypertrophic osteoarthropathy Hypoglycemia Neuropathy Osteoporosis Polycythemia (erythrocytosis) Polymyositis Porphyria Sexual changes—isosexual precocity, gynecomastia, feminization Systemic arterial hypertension Thyrotoxicosis Thrombophlebitis migrans Watery diarrhea syndrome

metastases. The ascitic fluid may be blood-stained. In a small proportion of patients, hepatocellular carcinoma invades the hepatic veins, thereby causing Budd-Chiari syndrome, and tense ascites results (see Chapter 83).16 Splenomegaly, if present, reflects coexisting cirrhosis and portal hypertension. Physical evidence of cirrhosis may also be noted. Severe pitting edema of the lower extremities extending up to the groins occurs when hepatocellular carcinoma has invaded the hepatic veins and propagates into and obstructs the inferior vena cava.16 A Virchow-Trosier (supraclavic­ ular) node, Sister Mary Joseph’s (periumbilical) nodule, or enlarged axillary lymph node is rarely present. Paraneoplastic Manifestations Some of the deleterious effects of hepatocellular carcinoma are not caused by local effects of the tumor or metastases (Table 94-2). Each of the paraneoplastic syndromes in hepatocellular carcinoma is rare or uncommon. One of the more important is type B hypoglycemia, which occurs in less than 5% of patients, manifests as severe hypoglycemia early in the course of the disease,16 and is believed to result from the defective processing by malignant hepatocytes of the precursor to insulin-like growth factor II (pre-IGF II).17 By contrast, type A hypoglycemia is a milder form of glycopenia that occurs in the terminal stages of hepatocellular carcinoma (and other malignant tumors of the liver). It results from the inability of a liver extensively infiltrated by tumor, and often cirrhotic, to satisfy the demands for glucose by a large, often rapidly growing tumor and by the other tissues of the body. Another important paraneoplastic syndrome is polycythemia (erythrocytosis), which occurs in less than 10% of patients with hepatocellular carcinoma.18 This syndrome appears to be caused by the synthesis of erythropoietin or an erythropoietin-like substance by malignant hepatocytes. Patients with hepatocellular carcinoma, especially the sclerosing variety, may present with hypercalcemia in the absence of osteolytic metastases. When hypercalcemia is severe, it may result in the typical complications of hypercalcemia, including drowsiness and lethargy. The probable cause is secretion of parathyroid hormone–related protein (PTHrP) by the tumor.19 Cutaneous paraneoplastic manifestations of hepatocellular carcinoma are rare except for pityriasis rotunda (circumscripta), which may be a useful marker of the tumor in black Africans. The rash consists of single or multiple, round or oval, hyperpigmented, scaly lesions on the trunk and thighs that range in diameter from 0.5 to 25 cm.20

Diagnosis

The gold standard for the diagnosis of hepatocellular carcinoma is pathology. For practical purposes (i.e., to apply treatment), hepatocellular carcinoma can only be diagnosed in the presence of an abnormality on imaging of the liver. The development of hepatocellular carcinoma is thought to occur as a result of a multistep sequential process from a dysplastic focus of hepatocytes to a low-grade dysplastic nodule to a high-grade dysplastic nodule to early welldifferentiated hepatocellular carcinoma, and then to less differentiated states.21,22 In early hepatocellular carcinoma, particularly when a needle biopsy specimen is examined, controversy may exist among pathologists as to whether a particular specimen is consistent with dysplasia or carcinoma. Dysplastic nodules and even regenerative cirrhotic nodules can be seen on imaging studies and are potentially confused with hepatocellular carcinoma. Although there are specific imaging features based on the enhancement patterns with dynamic imaging of dysplastic nodules and hepatocellular carcinoma (see later), some overlap occurs.23,24 Nevertheless, there is a growing consensus, based on guidelines from the major European and American liver societies and now backed up by published experience, that the diagnosis of hepatocellular carcinoma can be made in the appropriate clinical setting based on specific imaging characteristics, with or without an elevated serum alpha fetoprotein (AFP) level.24-27 Serum Tumor Markers Serum tumor markers generally are not diagnostic for hepatocellular carcinoma by themselves but can be used in conjunction with imaging findings to diagnose hepatocellular carcinoma. Additionally, they may raise the suspicion for hepatocellular carcinoma and lead to more sensitive and serial imaging of the liver. Conventional liver biochemical tests do not distinguish hepatocellular carcinoma from other hepatic mass lesions or cirrhosis. Many of the substances synthesized and secreted by hepatocellular carcinoma are not biologically active. Nevertheless, a few are produced by a sufficiently large proportion of tumors to warrant their use as serum markers for hepatocellular carcinoma. The most helpful of these markers is AFP (Table 94-3). Alpha Fetoprotein.  AFP is an α1-globulin normally present in high concentrations in fetal serum but in only minute amounts thereafter. Reappearance of high serum levels of AFP strongly suggests the presence of hepatocellular carcinoma (or hepatoblastoma [see later]),28 especially in populations in which hepatocellular carcinoma is most prevalent: The great majority of Chinese and black African patients have a raised serum concentration of AFP (>10 ng/mL), and approximately 75% have a diagnostic level (>500 ng/mL). These percentages are lower in populations at low or intermediate risk for the tumor, in which the sensitivity ranges from 25% to 65%, with a specificity of 79% to 95% and cutoff values for an elevated and diagnostic level of 16 and 200 ng/mL, respectively.29-35 With higher levels of AFP, the confidence in the diagnosis of hepatocellular carcinoma is greater. Although levels higher than 500 ng/mL usually indicate hepatocellular carcinoma, they sometimes can be seen in patients with active viral hepatitis. In the setting of a cirrhotic patient with a hepatic mass lesion larger than 2 cm in diameter and suggestive features of hepatocellular carcinoma, an AFP level higher than 200 ng/mL is con­ sidered diagnostic for hepatocellular carcinoma.25,26,33,36,37 The mean serum value of AFP in affected patients in regions with a high incidence of hepatocellular carcinoma is

1571

1572

Section IX  Liver Table 94-3  Tumor Markers of Hepatocellular Carcinoma* MARKER

SENSITIVITY (%)

SPECIFICITY (%)

CUTOFF†

Alpha fetoprotein (AFP) High-incidence populations Low-incidence populations Des-γ-carboxyprothrombin AFP-L3

80-90 25-65 28-89 71

90 79-95 87-95 63

>10 ng/mL 16-200 ng/mL 10-125 mAU/mL >10%

COMMENTS Most widely available; most extensively studied Precise cutoff value not determined Adds specificity to AFP if AFP level is 10-200 ng/mL

*Note that sensitivity and specificity rates vary both with the population under study and with the absolute level of the marker. Thus, the specificity of a markedly elevated serum AFP level in patients with cirrhosis greatly exceeds the sensitivity of mildly elevated levels in persons without cirrhosis (and similarly for the other markers). † Cutoff is the value above which the level is considered abnormal. AFP-L3, Lens culinaris agglutinin reaction fraction of alpha fetoprotein; mAU, milliarbitrary units.

60,000 to 80,000 ng/mL, compared with approximately 8,000 ng/mL in regions with a low or intermediate incidence of the tumor. Raised serum values range over six orders of magnitude, although an AFP concentration higher than 1 million ng/mL is rare. False-positive results also may occur in patients with tumors of endodermal origin, nonseminomatous germ cell tumors, and pregnancy. A progressively rising serum AFP concentration is highly suggestive of hepatocellular carcinoma. AFP is not essential to hepatocarcinogenesis, and thus not all hepatocellular carcinomas produce AFP. The levels of AFP appear to be affected by ethnicity, underlying cause of liver disease, and tumor stage.30,33 Synthesis of AFP by a tumor is permanent and age-related; the younger the patient, the more likely the serum value is to be raised and the higher the level attained. According to the American Association for the Study of Liver Diseases (AASLD) guidelines, hepatocellular carcinoma can be diagnosed with confidence in patients with a serum AFP level higher than 200 ng/mL and a mass in the liver.25 An AFP level higher than about 500 ng/mL predicts worse outcomes with liver transplantation compared with lower levels.38 Attempts to correlate the degree of differentiation of hepatocellular carcinoma with production of AFP have produced conflicting results. Because both false-positive and false-negative results are obtained when AFP is used as a serum marker for hepatocellular carcinoma, the search for an ideal marker continues. Alternative markers have not proved to be more useful than AFP. Fucosylated Alpha Fetoprotein.  AFP is heterogeneous in structure. Its microheterogeneity results from differences in the oligosaccharide side chain and accounts for the differential affinity of the glycoprotein for lectins. AFP secreted by malignant hepatocytes contains unusual and complex sugar chains that are not found in AFP produced by nontransformed hepatocytes. One variant, Lens culinaris agglutinin reactive fraction (AFP-L3), appears to improve the specificity of AFP, particularly AFP serum levels from 10 to 200 ng/mL.39,40 The recommended cutoff value for AFP-L3 to diagnose hepatocellular carcinoma is higher than 10%, although the specificity varies depending on the absolute level of AFP. A Western series has suggested that a cutoff value of 35% is necessary to achieve 100% specificity.40 Therefore, AFP-L3 is not sufficiently validated to confirm the diagnosis of hepatocellular carcinoma without other supporting findings, such as suggestive imaging. Des-γ-Carboxy Prothrombin.  Serum concentrations of desγ-carboxy prothrombin (DCP) (also known as prothrombin

produced by vitamin K absence or antagonist II [PIVKA II]) are raised in most patients with hepatocellular carcinoma.41 DCP is an abnormal prothrombin that is thought to result from a defect in the post-translational carboxylation of the prothrombin precursor in malignant cells.42 In Western populations, DCP may be a better marker than, or at least a complementary marker to, AFP.43-45 In black Africans, however, DCP is less sensitive and less specific than AFP.46 The appropriate cutoffs are not well established, and thus the precise role of DCP in the diagnosis of hepatocellular carcinoma requires validation. Other Markers.  Multiple other potential serum markers for hepatocellular carcinoma are in the exploratory phase of evaluation, including glypican 3, Golgi protein 73, hepatocyte growth factor, insulin growth factor 1, transforming growth factor-β1, and proteomic profiling using surfaceenhanced laser desorption/ionization time-of-flight (SELDITOF) mass spectrometry.47-51 All these novel markers have been shown to be elevated in patients with hepatocellular carcinoma compared with those with only chronic liver disease, but clear cutoff values and comparisons with other markers have not been established. Some of these markers may be complementary to established markers, although none of them has an established high throughput method of measurement, as required for a clinical test. The roles of these markers in the diagnosis of hepatocellular carcinoma await further study.

Imaging

The diagnosis of hepatocellular carcinoma generally requires imaging evidence of a focal lesion in the liver, although large infiltrating lesions can also be diagnostic. Arterial hyperenhancement, particularly seen on dynamic contrast imaging of the liver, is observed because the blood supply of hepatocellular carcinoma comes from newly formed abnormal arteries (neoangiogenesis).23,52,53 As a nodule transforms from low- to high-grade dysplasia and then to hepatocellular carcinoma, the primary blood supply shifts from portal to arterial—especially new abnormal arterial branches that produce characteristic findings on dynamic contrast imaging of the liver.27 Ultrasonography Ultrasonography detects most hepatocellular carcinomas but may not distinguish this tumor from other solid lesions in the liver. As with all imaging methods, the sensitivity increases with increasing size of the lesion. A systematic review of eight studies using histologic reviews of liver explants has shown that ultrasound has fair sensitivity

Chapter 94  Tumors and Cysts of the Liver (pooled estimate, 48%; 95% confidence interval [CI], 34% to 62%) with good specificity, estimated at 97% (95% CI, 95% to 98%).24 Advantages of ultrasonography include safety, availability, and cost-effectiveness, although it has the drawbacks of being nonstandardized and examinerdependent. Body habitus, particularly obesity, may limit the sensitivity of this test. Approximately two thirds of symptomatic hepatocellular carcinomas are uniformly hyper­ echoic, whereas the remainder are partly hyperechoic and partly hypoechoic.54 Small tumors are uniformly hypoechoic. The ultrasonographic appearance is influenced by the presence of fat, calcium, and necrosis. Tumors located immediately under the right hemidiaphragm may be difficult to detect. In Japanese patients in particular, hepatocellular carcinoma may have a well-defined, even thick capsule, which can be seen on ultrasonography. Ultrasonography with Doppler technology is useful for assessing the patency of the inferior vena cava, portal vein and its larger branches, hepatic veins, and biliary tree. Dynamic contrast-enhanced Doppler ultrasonography with intra-arterial infusion of CO2 microbubbles and intravenous enhanced color Doppler ultrasonography are refinements that, by characterizing hepatic arterial and portal venous flow in tumorous nodules, facilitate the diagnosis of malignant and benign hepatic nodules.55 These techniques are not often performed in the United States. Computed Tomography Multiphase, also called dynamic, helical computed tomography (CT) is the imaging technique of choice for the diagnosis of hepatocellular carcinoma.24,54,55 CT during arterial portography is also helpful but rarely done because it is invasive. Phases in dynamic contrast-enhanced CT can include noncontrast, arterial, portal venous, and delayed phases. The classic and most diagnostic pattern for hepatocellular carcinoma is a combination of enhancement in the arterial phase (with the uninvolved liver lacking enhancement), loss of central nodule enhancement compared with the uninvolved liver (washout), and capsular enhancement in the portal-venous and delayed phases (Fig. 94-2).25,56

Noncontrast

Arterial

Portal venous

Delayed

When the lesion is larger than 2 cm in diameter, this pattern has almost 100% specificity for hepatocellular carcinoma.36,37,56 When the nodule is 1 to 2 cm, guidelines recommend a second type of dynamic imaging (magnetic resonance imaging [MRI] or contrast ultrasonography) to confirm the diagnosis of hepatocellular carcinoma, although the specificity of one dynamic study is higher than 90%.57 CT often finds so-called hypervascular-only lesions, which enhance in the arterial phase and become isodense to the surrounding liver in the portal-venous and delayed phases. These lesions may be dysplastic nodules, arterial portal shunts, atypical hemangiomas, hepatocellular carcinoma, confluent fibrosis, or aberrant venous drainage. When less than 2 cm in diameter, only about 30% are hepatocellular carcinomas, which grow over time. Other causes disappear or remain stable on follow-up studies. Current guidelines recommend biopsy of lesions larger than 1 cm if the serum AFP level is less than 200 ng/mL and serial imaging for lesions smaller than 1 cm.58 Hepatocellular carcinoma may also have other patterns on CT, such as washout only on delayed imaging, a hypovascular nodule, or a fat-containing nodule.27,58 Overall, the pooled estimate of sensitivity and specificity for detecting hepatocellular carcinoma by CT is 67.5% (95% CI, 55% to 80%) and 92.5% (95% CI, 89% to 96%), respectively. Dynamic CT is also useful for detecting invasion into the portal or hepatic veins and identifying the location and number of tumors; these findings are critical for planning treatment. Magnetic Resonance Imaging Dynamic MRI using gadolinium contrast agents provides another way of distinguishing hepatocellular carcinoma from normal liver tissue. The performance of MRI and the findings on multiphase contrast enhancement are similar to those described for CT (Fig. 94-3). Typically, the signal intensity on T1-weighted images is low.27,54 The pooled estimate of sensitivity and specificity for detecting hepatocellular carcinoma by MRI is 80.6% (95% CI, 70% to 91%) and 84.8% (95% CI, 77% to 93%), respectively.24 MRI may be slightly superior overall to CT, although local expertise should dictate the choice of imaging technique. Hepatic Angiography Since the advent of CT and MRI, the diagnostic role of hepatic angiography has decreased. Digital subtraction angiography is helpful for recognizing small hyper­vascular hepatocellular carcinomas but may miss early, welldifferentiated hypovascular tumors. Hepatocellular carcinomas often are densely vascular, although multinodular tumors may be relatively avascular.59 The arteries in the tumor are irregular in caliber and do not taper in the usual way, and the smaller branches may show a bizarre pattern. The hepatic veins fill early, and retrograde filling of the portal veins results from the presence of arteriovenous anastomoses within the tumor. An additional finding is a delay in capillary emptying, which is seen as a blush. The center of some large tumors may be avascular as a result of necrosis or, less often, hemorrhage. Angiography is essential for delineating the hepatic arterial anatomy in planning embolization or chemoembolization of the tumor or infusion of cytotoxic drugs directly into the hepatic artery or its branches (see later).

Laparoscopy Figure 94-2.  Dynamic computed tomography scan of a patient with hepatocellular carcinoma showing no lesion in the noncontrast phase, an enhancing lesion in the arterial phase of contrast administration, and a faint lesion in the portal venous phase seen better in the delayed phase.

Laparoscopy can be used to detect peritoneal and other extrahepatic spread, ascertain whether the nontumorous part of the liver is cirrhotic, and obtain biopsies under direct vision.

1573

1574

Section IX  Liver T2

T1

Delayed venous

Arterial Venous Figure 94-3.  Multiphasic magnetic resonance imaging of the liver showing hepatocellular carcinoma with characteristic features, including hyperintensity (arrow) on T2-weighted image (top left panel) but not on T1-weighted image (top right panel), enhancement during the arterial phase of contrast administration (bottom left panel), with central washout of contrast and capsular enhancement during the venous and delayed phases (bottom middle and right panels).

Pathology

Definitive diagnosis of hepatocellular carcinoma depends on demonstrating the typical histologic features. Suitable samples generally can be obtained by percutaneous biopsy or fine-needle aspiration. The yield and safety of the procedure can be increased by directing the needle under ultrasonographic or CT guidance. Laparoscopically directed biopsy is an alternative approach. Needle biopsy of the tumor carries a small but definite risk of spread along the needle track. Gross Appearance Hepatocellular carcinoma may take one of three forms— nodular, massive, or diffusely infiltrating. The nodular variety of hepatocellular carcinomas is most common and usually coexists with cirrhosis. It is characterized by numerous round or irregular nodules of various sizes scattered throughout the liver; some of the nodules are confluent. The massive type is characterized by a large circumscribed mass, often with small satellite nodules. This type of tumor is most prone to rupture and is more common in younger patients with a noncirrhotic liver. In the rare diffusely infiltrating variety, a large part of the liver is infiltrated homogeneously by indistinct minute tumor nodules, which may be difficult to distinguish from the regenerating nodules of cirrhosis that are almost invariably present. The portal vein and its branches are infiltrated by tumor in up to 70% of cases seen at autopsy; the hepatic veins and bile ducts are invaded less often. Microscopic Appearance Hepatocellular carcinoma is classified histologically into well-differentiated, moderately differentiated, and undifferentiated (pleomorphic) forms.60

Well-Differentiated Appearance.  Despite the aggressive nature and poor prognosis of hepatocellular carcinoma, most tumors are well differentiated. Trabecular and acinar (pseudoglandular) varieties occur, sometimes in a single tumor. In the trabecular variety, the malignant hepatocytes grow in irregular anastomosing plates separated by often inconspicuous sinusoids lined by flat cells resembling Kupffer cells. The trabeculae resemble those of normal adult liver but often are thicker and may be composed of several layers of cells. Scanty collagen fibers may be seen adjacent to the sinusoid walls. The malignant hepatocytes are polygonal, with abundant, slightly granular cytoplasm that is less eosinophilic than that of normal hepatocytes. The nuclei are large and hyperchromatic, with prominent nucleoli. Bile production is the hallmark of hepatocellular carcinoma, regardless of the pattern. Gland-like structures are present in the acinar variety. The structures are composed of layers of malignant hepatocytes surrounding the lumen of a bile canaliculus, which may contain inspissated bile. A tubular or pseudopapillary appearance may be produced by degeneration and loss of cells, or cystic spaces may form in otherwise solid trabeculae. The individual cells may be more elongated and cylindrical than in the trabecular variety. Moderately Differentiated Appearance.  Solid, scirrhous, and clear cell varieties of hepatocellular carcinoma are described. In the solid variety, the cells usually are small, although they vary considerably in shape. Pleomorphic multinucleated giant cells occasionally are present. The tumor grows in solid masses or cell nests. Evidence of bile secretion is rare, and connective tissue is inconspicuous. Central ischemic necrosis is common in larger tumors. In the scirrhous variety, the malignant hepatocytes grow in

Chapter 94  Tumors and Cysts of the Liver narrow bundles separated by abundant fibrous stroma. Duct-like structures occasionally are present. In most tumors, the cells resemble hepatocytes. In an occasional tumor, the malignant hepatocytes are predominantly or exclusively clear cells. More often, tumors contain areas of clear cells. The appearance of these cells results from a high glycogen or, in some cases, fat content. Undifferentiated Appearance.  The cells are pleomorphic, varying greatly in size and shape. The nuclei also are extremely variable. Large numbers of bizarre-looking giant cells are present. The cells may be spindle-shaped, resembling those of sarcomas. Globular hyaline structures may be seen in all types of hepatocellular carcinoma. These structures reflect the presence of AFP, α1-antitrypsin, or other proteins. Mallory’s hyaline occasionally is present. Progenitor Cell Hepatocellular Carcinoma.  A class of primary liver cancer appears to have its origins in progenitor cells, the stem cells of the liver, located in association with the canals of Hering. Progenitor cell activation is seen in association with chronic viral hepatitis and cirrhosis, presumably relegated to senescence of hepatocytes. These tumors may appear morphologically like typical hepatocellular carcinoma or mixed cholangiohepatocellular carcinoma. Tumor cells stain positively for cytokeratin 19, and the tumor appears to have a more aggressive course than typical hepatocellular carcinoma.61

Metastases

Extrahepatic metastases are present at autopsy in 40% to 57% of patients with hepatocellular carcinomas.62 The most common sites are the lungs (up to 50% in some reports) and regional lymph nodes (approximately 20%). The adrenal glands are frequently involved.

Fibrolamellar Hepatocellular Carcinoma

The fibrolamellar variant of hepatocellular carcinoma typically occurs in young patients, has an approximately equal gender distribution, does not secrete AFP, is not caused by chronic hepatitis B or C, and almost always arises in a noncirrhotic liver.63-65 Fibrolamellar hepatocellular car­ cinoma is more often amenable to surgical treatment and therefore generally carries a better prognosis than that for conventional hepatocellular carcinoma. It does not, however, respond to chemotherapy any better than other forms of hepatocellular carcinoma. The hepatocytes are characteristically plump, deeply eosinophilic, and encompassed by abundant fibrous stroma composed of thin, parallel fibrous bands that separate the cells into trabeculae or nodules. The cytoplasm is packed with swollen mitochondria and, in approximately half of the tumors, contains pale or hyaline bodies. Nuclei are prominent, and mitoses are rare.

Staging

Accurate staging of hepatocellular carcinoma is necessary for prognostication and also to assist with selection of therapy. Determining the optimal staging system for hepatocellular carcinoma has been controversial, in part because it needs to take into account both the severity of the underlying liver disease and the size and degree of spread of the tumor. As with all cancers, the TNM (tumornode-metastasis) system can be used to stage hepatocellular carcinoma, but this system does not factor in the underlying liver disease. A study66 comparing the usefulness of seven staging systems, including the Okuda, TNM, Cancer of the Liver Italian Program (CLIP), Barcelona Clinic Liver Cancer

(BCLC), Chinese University Prognostic Index (CUPI), Japanese Integrated Staging (JIS), and Group d’Etude et Traitement du Carcinome Hépatocellulaire (GETCH) systems in a cohort of patients from the United States, has found the BCLC staging system to have the best independent predictive power for survival. The BCLC system has been adopted by the AASLD for use in its practice guidelines on management of hepatocellular carcinoma.25 This staging classification also includes a treatment schedule based on stage (Fig. 94-4).67

Causes and Pathogenesis

In contrast to many other malignancies, for which risk factors can only sometimes be identified, the immediate cause of hepatocellular carcinoma can usually be identified and is most commonly chronic viral hepatitis or cirrhosis. Hepatocellular carcinoma is multifactorial in cause and complex in pathogenesis. Four major causative factors have been identified (Table 94-4). The differing blend of risk factors in various parts of the world may explain, in part, the diverse biologic characteristics of hepatocellular carcinoma in various populations.68 Hepatitis B Virus Some 387 million carriers of HBV exist in the world today, and hepatocellular carcinoma will develop in as many as 25% of them (see Chapter 78). HBV accounts for up to 80% of hepatocellular carcinomas, which occur with high frequency in East Asian and African populations.68,69 Persistent HBV infection antedates the development of hepatocellular carcinoma by several to many years, an interval commensurate with a cause and effect relationship between the virus and the tumor. Indeed, in at-risk populations, the HBV carrier state is largely established in early childhood by perinatal or horizontal infection.70,71 Approximately 90% of children infected at this stage of life become chronic carriers of the virus, and these early-onset carriers face a lifetime relative risk for developing hepatocellular carcinoma of more than 100, compared with uninfected controls.72 An effective vaccine against HBV has been available since the early 1980s and, in countries in which this vaccine has been included in the expanded program of immunization for a sufficient length of time, the HBV carrier rate among children has decreased by 10-fold or more. Studies in Taiwan, where universal immunization was started in 1984

Table 94-4  Risk Factors for Hepatocellular Carcinoma Major Risk Factors Chronic hepatitis B virus infection Chronic hepatitis C virus infection Cirrhosis Dietary exposure to aflatoxin B1 Other Liver Conditions α1-Antitrypsin deficiency Hemochromatosis Membranous obstruction of the inferior vena cava Type 1 and type 2 glycogen storage disease Type 1 hereditary tyrosinemia Wilson disease Inherited Conditions Not Associated with Liver Disease Ataxia-telangiectasia Hypercitrullinemia Other Factors Cigarette smoking Diabetes mellitus Oral contraceptive steroids

1575

1576

Section IX  Liver HCC

Stage 0 PST 0 and Child-Pugh A

Stage D PST >2 or Child-Pugh C

Stage A–C PST 0-2 and Child-Pugh A-B

Single nodule 20 yr Asian men, age > 40 yr Asian women, age > 50 yr Family history of hepatocellular carcinoma Patients with cirrhosis Patients with high serum HBV DNA level and ongoing hepatic injury Alcoholic cirrhosis α1-Antitrypsin deficiency* Autoimmune hepatitis* Hemochromatosis Hepatitis B Hepatitis C Nonalcoholic steatohepatitis* Primary biliary cirrhosis

Patients with cirrhosis

*No data on efficacy of surveillance available. HBV DNA, hepatitis B viral deoxyribonucleic acid. Adapted from Bruix J, Sherman M. Management of hepatocellular carcinoma. Hepatology 2005; 42:1208-36.

universal infant vaccination against HBV in many countries, the full impact of universal HBV vaccination on the occurrence of the tumor will not be realized for many years. A significant reduction has already been noted in childhood hepatocellular carcinoma in Taiwan, where universal infant vaccination was adopted in the mid-1980s.138 In the meantime, the huge numbers of existing HBV carriers worldwide remain at risk for hepatocellular carcinoma, and little progress has been made in preventing malignant transformation in persons with chronic viral hepatitis, nor has much progress has been made on other fronts. A vaccine against HCV will not be available in the near future, and prevention of aflatoxin-induced tumors is far from a becoming a reality, despite ongoing trials of chemopreventive agents. Considerable interest has been expressed in the impact of antiviral therapy against HBV and HCV in reducing the incidence of hepatocellular carcinoma. One randomized controlled trial of long-term therapy of lamivudine versus placebo in patients with chronic hepatitis B has shown a significant decrease in the frequency of clinical events in the treated group, including a decrease in the frequency of hepatocellular carcinoma.77 Several large retrospective studies have shown a decrease in frequency of hepatocellular carcinoma in patients successfully treated for chronic hepatitis C with interferon-based regimens.139

INTRAHEPATIC CHOLANGIOCARCINOMA

Cholangiocarcinoma is a malignant neoplasm arising from the biliary duct epithelium. It often carries different names based on the particular portion of the biliary tree involved—small intrahepatic bile ducts (peripheral cholangiocarcinoma), hepatic duct bifurcation (perihilar cholangiocarcinoma, or Klatskin tumor), and extrahepatic bile ducts (bile duct carcinoma). The location of the tumor has a major impact on the presenting symptoms and treatment approach. Perihilar cholangiocarcinoma is classified with the intrahepatic group based on International Classification of Diseases, 9th revision (ICD-9) codes even though it is extrahepatic in origin and is the most common form.140,141 This section will be limited to a discussion of intrahepatic cholangiocarcinoma; extrahepatic cholangiocarcinoma is discussed in Chapter 69.

Epidemiology

Intrahepatic cholangiocarcinoma represents approximately 10% to 20% of all primary liver cancers and 20% to 25% of cholangiocarcinomas. The geographic variation in prevalence rates is marked, ranging from 0.2 to 96/100,000 in men and from 0.1 to 38/100,000 in women, because of differences in the frequencies of known risk factors in various populations.142 The highest prevalence rates are found in parts of Asia, most notably certain regions of Thailand, Hong Kong, China, Japan, and Korea. Chronic infestation of the biliary tree with one of the liver flukes is thought to be the cause of these high rates.143 The overall prevalence rate in the United States is 0.85/100,000, with a 1.5-fold higher rate in men than women. The rate in whites is about equal to that in African Americans and about half that in Asians. Although the underlying predisposing factor for most cases of cholangiocarcinoma is unknown, a number of risk factors have been recognized. The strongest association is with Opisthorchis viverrini, a liver fluke endemic in parts of Southeast Asia and acquired by ingestion of raw or uncooked fish.142,144,145 The association with Clonorchis sinensis, a related liver fluke, is weaker.146 An association with the radiographic contrast agent thorium dioxide (Thorotrast), which was banned in the 1950s, has been well established.147 Primary sclerosing cholangitis is linked to a diagnosis of cholangiocarcinoma at a young age, with a lifetime risk of 8% to 20% (see Chapter 68).148-150 Congenital and acquired abnormalities of the biliary tract that may result in bile stasis, chronic inflammation, and infection, as in biliary atresia,151 von Meyenburg complexes,152 Caroli’s disease,153 choledochal cyst,153 and intrahepatic cholelithiasis, have been associated with the development of cholangiocarcinoma. Cirrhosis, particularly caused by HCV, also has been associated with cholangiocarcinoma.154 Intrahepatic cholangiocarcinoma is rare before the age of 40 years, and historically the worldwide approximate average age at presentation is 50 years. Epidemiologic data indicate that the age at presentation has shifted to more than 65 years. Additionally, the incidence and mortality rates are increasing worldwide.140 Surveillance, Epidemiology and End Results (SEER) registry data from the United States have shown a 165% increase between the late 1970s and the late 1990s.142 This increase may be a result, in part, of the increased prevalence of cirrhosis, particularly HCVassociated cirrhosis.154

Molecular Pathogenesis

Malignant transformation of the bile duct cells generally occurs in an environment of inflammation or cholestasis (or both), usually as a result of one of the known risk factors. The proposal has been made that a combination of these environmental factors and genetic predisposition—for example, defects in oncogenes or bile salt transporters— leads to an accumulation of genetic defects that results in carcinoma.140,155 A polymorphism in the gene for the natural killer cell receptor G2D (NKG2D) has been associated with an increased risk of cholangiocarcinoma in patients with primary sclerosing cholangitis.156 At the molecular level, numerous changes have been described, including mutations of the K-ras gene, the gene for interleukin-6, and allelic loss or mutations of TP53 and p16, as well as many others (see Chapter 69).

Clinical Features

Peripheral cholangiocarcinoma seldom produces symptoms until the tumor is advanced. The clinical features are then similar to those of hepatocellular carcinoma, including

Chapter 94  Tumors and Cysts of the Liver malaise, weight loss, abdominal pain, and jaundice, which may be more frequent and prominent than with hepatocellular carcinoma.153,157 The clinical presentation of perihilar and extrahepatic cholangiocarcinoma is with progressive painless jaundice, acholic stools, pruritus with or without weight loss, and, rarely, cholangitis.158 Patients with primary sclerosing cholangitis may present with worsening jaundice caused by a dominant bile duct stricture, weakness, and weight loss.

Diagnosis

In patients with perihilar cholangiocarcinoma and extrahepatic cholangiocarcinoma, obstructive jaundice is evident, with elevated serum levels of bilirubin, alkaline phosphatase, gamma glutamyl transpeptidase (GGTP), and often aminotransferases. In patients with peripheral cholangiocarcinoma, often only the alkaline phosphatase level is elevated. CA 19-9 is the most frequently used serum tumor marker for cholangiocarcinoma but has significant limitations because CA 19-9 levels are also elevated in pancreatic, colorectal, gastric, and gynecologic cancers and in acute bacterial cholangitis.159 CA 19-9 is always undetectable in the 7% of the population that is Lewis blood group– negative. In patients with unexplained biliary obstruction without primary sclerosing cholangitis, the sensitivity of CA 19-9 is 53%, and the negative predictive value is 72% to 92%, for a cutoff value of 100 U/mL. In patients with primary sclerosing cholangitis, the sensitivity ranges from 38% to 89% and specificity from 50% to 98%. The addition of carcinoembryonic antigen (CEA) probably does not improve the performance of CA 19-9 in the setting of primary sclerosing cholangitis. Initial imaging with ultrasound helps identify biliary obstruction. Dynamic contrast-enhanced CT or MRI further aids in localizing the lesion and determining the possibility of resection.140,160 MRI with magnetic resonance cholangiography (MRCP) is a superior modality because of a higher sensitivity than CT for detecting lesions and localizing biliary obstruction. The tumor is hypodense on T1-weighted images and moderately intense on T2-weighted images. Endoscopic retrograde cholangiopancreatography (ERCP) or transhepatic cholangiography allows for localization of the tumor, sampling of tissue and bile, and relief of biliary obstruction if the tumor is unresectable. A perihilar tumor may demonstrate a classic appearance on ERCP, but for other biliary strictures, particularly in patients with primary sclerosing cholangitis, determining whether cholangiocarcinoma is present may be difficult. Cytology has a 30% sensitivity, which improves to 40% to 70% with the addition of brushings and biopsies. Newer cytologic techniques that assess the cells for aneuploidy and chromosomal aberrations may improve the diagnostic yield but are not widely available (see Chapter 69). Endoscopic ultrasound (EUS) with fine-needle aspiration (FNA) in patients without primary sclerosing cholangitis has the advantage of improving sensitivity and specificity for diagnosis of the primary lesion and nodal metastasis but the disadvantage of causing peritoneal seeding, and thus should be avoided if surgical resection is contemplated. Percutaneous biopsies also carry the risk of peritoneal seeding and are generally avoided if the tumor is potentially resectable.

Pathology

Peripheral cholangiocarcinoma usually is a large and solitary tumor, but it may be multinodular.161 It is grayishwhite, firm, and occasionally umbilicated and can produce a focal hepatic mass, a tumor growing along and infiltrating the bile ducts, or an intraductal papillary lesion.160 The

tumor is poorly vascularized and rarely bleeds internally or ruptures. Perihilar cholangiocarcinoma may take the form of a firm intramural tumor that encircles the bile duct, a bulky mass centered on the duct or hilar region that radiates into the hepatic tissue, or a spongy friable mass within the lumen of the duct. Metastatic nodules may be distributed irregularly throughout the liver. The bile ducts peripheral to the tumor may be dilated, resulting in some cases in biliary cirrhosis. Metastases in regional lymph nodes occur in about 50% of cases. Microscopically, cholangiocarcinoma exhibits acinar or tubular structures that resemble those of other adenocarcinomas.161 Most tumors are well differentiated. Secretion of mucus may be demonstrable, but bile production is not seen. The tumor cells provoke a variable desmoplastic reaction, and, in many tumors, the collagenized stroma may be the most prominent feature. Distinguishing the tumor from metastatic adenocarcinoma may be difficult, and some experts have advocated assuming that an adenocarcinoma in the liver is cholangiocarcinoma if no primary tumor can be found elsewhere.162

Treatment and Prognosis

Early diagnosis of intrahepatic cholangiocarcinoma is unusual, and the annual mortality rate is almost identical to the annual incidence of the tumor.153,157 Long-term survival after diagnosis in the United States based on the SEER database is dismal, with a one-year survival rate of 28% and a five-year survival rate less than 5%. The five-year survival rate has not improved since the late 1980s.142 In a person with suspected or proven intrahepatic cho­ langiocarcinoma, staging is recommended to determine surgical resectability, which is the only opportunity for cure. The staging evaluation usually includes dynamic MRI and MRCP of the abdomen (or dynamic helical CT, if MRI is unavailable) and a chest x-ray or CT.140 Positron emission tomography (PET) has been assessed in small series and does not clearly add to other modalities. EUS with FNA of suspicious lymph nodes may detect otherwise unrecognized metastasis in up to 20% of cases.160 Surgical resectability of intrahepatic cholangiocarcinoma should be determined in conjunction with an experienced hepato­ biliary surgeon and requires the ability to achieve clear surgical margins, which usually necessitates a major hepatectomy. Criteria for resection include absence of all the following: evidence of extrahepatic metastasis; main portal vein or hepatic artery invasion or encasement; bilateral segmental bile duct involvement; and contralateral hepatic lobar atrophy. Additionally, the patient must be medically fit to undergo surgery and have sufficient hepatic reserve. Without clear margins of resection, surgery provides benefits similar to those of endoscopic or biliary drainage. Patients well selected for surgical resection achieve a one- to two-year median survival and a 29% to 36% five-year survival rate. If resection is not possible and major biliary obstruction is present, biliary drainage, either endoscopic or percutaneous, should be performed, because drainage appears to improve symptoms and survival (see Chapter 70).162 Placement of an expandable metal stent is generally preferred to plastic stents if the expected survival of the patient is more that three to six months.140 The rates of response and survival following radiation therapy and chemotherapy are modest, as suggested by predominantly small uncontrolled series, so these modalities generally should be restricted to clinical trials. Photodynamic therapy in addition to biliary stent placement may provide benefit. Liver transplantation alone results in unacceptably high recurrence rates and

1581

1582

Section IX  Liver limited survival.160 In a single center report, aggressive preoperative therapy of unresectable hilar cholangiocarcinoma with external beam radiation, brachytherapy, and chemosensitization, followed by liver transplantation in patients who survived the treatment and had contained disease, produced a five-year survival rate of 82%.163

HEPATOBLASTOMA Epidemiology

In children, hepatoblastoma is the third most common malignant tumor and the most common malignant hepatic tumor. It occurs almost exclusively in the first three years of life; boys are affected twice as often as girls.164,165

Clinical Features

Most children with hepatoblastoma come to medical attention because of abdominal swelling.166 Other reasons include failure to thrive, weight loss, poor appetite, abdominal pain, irritability, and intermittent vomiting and diarrhea. The tumorous liver almost always is enlarged and firm and may be tender. Its surface is smooth or nodular. Hepatoblastomas rarely rupture. Distant metastases are evident, usually in the lung, in 20% of patients at the initial visit.167 The tumor occasionally causes isosexual precocity in boys as a result of the ectopic production of human chorionic gonadotropin.168

Diagnosis

AFP is present in high concentrations in the serum of 80% to 90% of patients with hepatoblastoma and is a useful clue to diagnosis.169 The few patients with a low serum AFP level appear to have a worse prognosis.170 Anemia is common, as is thrombocytosis, which is attributed to raised serum thrombopoietin levels. Pulmonary metastases and, rarely, mottled calcification in the tumor may be seen on plain radiography. Ultrasonography is the most widely used initial imaging technique, although the findings are not specific. CT and MRI are used to define the extent of the tumor and plan definitive surgery. The tumor is seen as an avascular mass on hepatic arteriography.171

Pathology

Hepatoblastomas are the malignant derivatives of incompletely differentiated hepatocyte precursors. Their constituents are diverse, reflecting both the multipotentiality of their mesodermal origin and the progressive stages of embryonic and fetal development. Hepatoblastomas are classified morphologically into an epithelial type, composed predominantly of epithelial cells of varying maturity, and a mixed epithelial and mesenchymal type, which contains, in addition, tissues of mesenchymal derivation.166,169 The tumors usually are solitary, ranging in diameter from 5 to 25 cm, and always well circumscribed (about half are encapsulated). They vary in color, ranging from tan to grayish-white, and contain foci of hemorrhage, necrosis, and calcification. Vascular channels may be prominent on the capsular surface. Epithelial hepatoblastomas are solid, whereas tumors of the mixed variety often are separated into lobules by white bands of collagen tissue. Two types of epithelial cells are present in the tumor.172 Cells of the first type resemble fetal hepatocytes and are arranged in irregular plates, usually two cells thick, with bile canaliculi between individual cells and sinusoids between plates. Cells of the second type are embryonal and are less differentiated than the fetal type. Mixed hepatoblastomas contain mesenchymal tissue consisting of areas of a highly cellular primitive type of mesenchyme intimately admixed with epithelial elements. Cartilage and striated

muscle may be present. Hepatoblastomas may show foci of squamous cells, with or without keratinization, and foreign body–type giant cells. Vascular invasion may be evident. Metastases most commonly involve lung, abdominal lymph nodes, and brain.

Cause and Pathogenesis

Hepatoblastoma may occur sporadically or in association with hereditary syndromes such as familial adenomatous polyposis (FAP) and Beckwith-Wiedemann syndrome, suggesting a possible role for chromosomes 5 and 11 in the genesis of the tumor. The FAP tumor suppressor gene downregulates β-catenin. Sporadic hepatoblastoma is not associated with any known environmental risk factor, and its pathogenesis is unclear. Most patients with hepatoblastoma have mutations of the FAP gene, and a similar number have activating mutations of the β-catenin gene, raising the possibility that the wnt signaling pathway plays a role in the development of the tumor.173

Treatment and Prognosis

Hepatoblastomas are rapidly progressive. If the lesion is solitary and sufficiently localized to be resectable, surgery often is curative, with five-year survival rates as high as 75%.166 The current practice is to pretreat the patient with cisplatin and doxorubicin. When the tumor is judged to be inoperable, neoadjuvant chemotherapy may reduce the size of the tumor sufficiently to permit resection. Encouraging results also have been obtained with liver transplantation in patients with bilobar multifocal tumors without extrahepatic extension.174 If surgery is not possible or the tumor recurs after surgery, the prognosis generally is poor.

HEMANGIOSARCOMA Epidemiology

Although rare, angiosarcoma is the most common malignant mesenchymal tumor of the liver.175,176 It occurs almost exclusively in adults and is most prevalent in the sixth and seventh decades of life.177,178 Men are affected four times as often as women.

Pathogenesis

Despite its rarity, hepatic angiosarcoma is of special interest because specific risk factors have been identified, although no cause is discerned in most cases. In early reports, the tumor became evident approximately 20 years after the patient had been exposed to thorium dioxide (see Chapter 87).179 Angiosarcoma also has occurred in German vintners who used arsenic-containing insecticides and drank wine adulterated with arsenic.180 A few patients with angiosarcoma had taken potassium arsenite (Fowler’s solution) for many years to treat psoriasis.181 Hepatic angiosarcoma in workers exposed to vinyl chloride monomer (VCM) was first reported in 1974.177,182,183 The monomer is converted by enzymes of the endoplasmic reticulum to reactive metabolites that form DNA adducts and guanosine-to-adenine transitions in the K-ras and TP53 genes. Angiosarcomas have occurred after exposures of 11 to 37 years (or after shorter periods with a heavy initial exposure). The mean age of patients at diagnosis is 48 years. In addition to angiosarcoma, persons exposed to VCM may be at increased risk of hepatocellular carcinoma and soft tissue sarcoma.

Clinical Features

The most common presenting symptom is upper abdominal pain. Other frequent complaints are abdominal swelling, rapidly progressing liver failure, malaise, weight loss, poor appetite, and nausea.176,177 Vomiting occurs

Chapter 94  Tumors and Cysts of the Liver occasionally. The duration of symptoms generally ranges from one week to six months, but a few patients have had symptoms for as long as two years before seeking medical attention. The liver almost always is enlarged and usually is tender. Its surface may be irregular, or a definite mass may be felt. An arterial bruit occasionally is heard over the enlarged liver. Splenomegaly may be present and is attributed to the hepatic fibrosis and consequent portal hypertension that also may complicate exposure to VCM. Ascites is frequent, and the fluid may be blood-stained. The patient often has jaundice. Fever and dependent edema are less common. Approximately 15% of patients present with acute hemoperitoneum following tumor rupture. Rarely, pulmonary or skeletal metastases are present.

Diagnosis

A rising serum bilirubin level and other evidence of pro­ gressive hepatic dysfunction may be present, especially in the later stages of the tumor. Plain radiography may show pulmonary metastases, a raised right hemidiaphragm, or, rarely, skeletal metastases. In patients who received thorium dioxide, radiopaque deposits of the material may be evident in the liver and spleen.179 One or more mass lesions may be demonstrated on ultrasonography, CT, or MRI, but diffusely infiltrating tumor may not be visualized. Hepatic arteriography reveals a characteristic appearance.184 The hepatic arteries are displaced by the tumor, which shows a blush and “puddling” during the middle of the arterial phase that persist for many seconds, except in the central area, which may be hypovascular.

Complications and Prognosis

Hepatic angiosarcomas grow rapidly, and the prognosis is poor; death ensues within six months. Patients may have thrombocytopenia resulting from entrapment of platelets within the tumor (Kasabach-Merritt syndrome), disseminated intravascular coagulation with secondary fibrinolysis,185 or microangiopathic hemolytic anemia as a result of fragmentation of erythrocytes within the tumor circulation.186

Pathology

Angiosarcomas usually are multicentric.187 Their hallmark is the presence of blood-filled cysts, although solid growth also is seen. The lesions are fairly well circumscribed but not encapsulated. Larger masses are spongy and bulge beneath Glisson’s capsule. The earliest microscopic change is the presence of hypertrophic sinusoidal lining cells with hyperchromatic nuclei in ill-defined loci throughout the liver. With progression of the lesion, sinusoidal dilatation and disruption of hepatic plates occur, and the malignant cells become supported by collagen tissue. Enlarging vascular spaces lined by malignant cells cause the tumor to become cavernous. The malignant endothelial cells usually are multilayered and may project into the cavity in intricate fronds and tufts supported by fibrous tissue. The fronds commonly are elongated, with ill-defined borders. The cytoplasm is clear and faintly eosinophilic. Nuclei are hyperchromatic and vary greatly in size and shape; some cells are multinucleated. Evidence of phagocytosis may be seen. Foci of extramedullary hematopoiesis are common, and invasion of the portal and central veins occurs in most cases. Distant metastases are present in 50% of tumors.

Treatment

Operative treatment usually is precluded by the advanced stage of the tumor. Even when surgery is undertaken, the

patient commonly survives only one to three years, although long-term survival may be achieved in the few patients with a solitary tumor.176 The results of irradiation and chemotherapy are poor.

EPITHELIOID HEMANGIOENDOTHELIOMA Epidemiology

Epithelioid hemangioendothelioma is a rare tumor whose incidence is not known. A case series of 137 cases has been collected at a specialized referral center.188 Two thirds of patients were female, and the tumor occurred at all ages in adulthood.

Clinical Features

Patients typically present with nonspecific symptoms, such as abdominal pain and weight loss.

Diagnosis

Imaging studies show a characteristically highly vascular mass, which may infiltrate throughout the liver. Case reports indicate that the tumor can be visualized on PET. Correct diagnosis requires histologic examination of tissue obtained by biopsy.

Complications and Prognosis

The tumor has low-grade malignant potential and must be distinguished from hemangiosarcoma, because it has a much better prognosis if treated appropriately and aggressively. Epithelioid hemangioendothelioma may metastasize, both within and beyond the liver.

Pathology

Tumors are often multiple and may be diffuse throughout the liver. Histologically, they are characterized by the presence of dendritic and epithelioid cells that contain vacuoles, representing intracellular lumina. These cells stain positively for endothelial markers, such as factor VIII– related antigen, CD34, or CD31.

Treatment

The primary treatment modality for epithelioid hemangioendothelioma is surgical, including resection or liver transplantation. Transplantation appears to be effective for this tumor, even in the presence of advanced or even metastatic disease. The tumor does not appear to be sensitive to radiation or chemotherapy.

OTHER PRIMARY MALIGNANT TUMORS OF THE LIVER

Undifferentiated (embryonal) sarcoma is a rare primary malignancy of the liver that occurs in both children and adults.189,190 The tumor tends to be aggressive, but long-term survival can be achieved with radical surgery and chemotherapy. Other rare sarcomas arising in the liver include liposarcoma,191 lymphoma,176,192 and rhabdomyosarcoma.190

HEPATIC METASTASES The liver is the most frequent target for metastatic spread of tumors. Hepatic metastases occur in 40% to 50% of adult patients with extrahepatic primary malignancies.193 Foremost among the reasons for the high frequency of hepatic metastases are the double blood supply of the liver and the presence of fenestrations in the sinusoidal endo­thelium that facilitate penetration of malignant cells

1583

1584

Section IX  Liver into the hepatic parenchyma.194 Hepatic metastases commonly originate from primary sites in the distribution of the portal venous system, including the pancreas, stomach, and colon. Outside this distribution, tumors of the lung and breast are the most common origins of hepatic metastases.

Clinical Features

Symptoms resulting from hepatic metastases often are absent or overshadowed by those of the primary tumor. Occasionally, the symptoms and signs attributable to metastases are the presenting manifestations of an asymptomatic primary tumor. In such cases, the likely symptoms are malaise, weight loss, and upper abdominal pain. Jaundice, when present, is seldom attributable to replacement of hepatic tissue by metastases. Depending on the extent of the metastatic disease, the liver may be enlarged, sometimes markedly. Its surface may be irregular, and umbilicated nodules may be felt by the examiner. A friction rub may be heard over hepatic metastases.

Diagnosis

CT is the most useful imaging technique.195 Multiphase helical CT and CT during arterial portography are more sensitive than conventional CT. Dynamic contrast-enhanced Doppler ultrasonography with intra-arterial infusion of CO2 microbubbles also is useful for the diagnosis of hepatic metastases.54 T1-weighted MRI also may be helpful, and iron oxide–enhanced MRI is even better.

Pathology

Macroscopic Appearance Hepatic metastases almost always are multiple.193 Their pathologic features vary, depending on the site of origin. Metastases are expansive, when they are discrete, or infiltrative. Individual metastases may reach a large size, and, with multiple metastases, the liver may be greatly enlarged. Metastases commonly are gray-white and may show scattered hemorrhages or central necrosis. Individual metas­ tases may be surrounded by a zone of venous stasis. Subcapsular lesions often are umbilicated. The dictum that cirrhotic livers are less likely than noncirrhotic livers to harbor metastatic deposits remains to be verified. Microscopic Appearance The microscopic features, including the degree of stromal growth, of most hepatic metastases duplicate those of the tumor of origin. Metastatic deposits usually are easily delineated from the surrounding liver tissue. Invasion of portal or hepatic veins may be seen, although less often than with hepatocellular carcinoma.193 It may be difficult to distinguish metastatic adenocarcinoma from primary cholangiocarcinoma.162

Treatment and Prognosis

The extent of replacement of liver tissue by metastases generally determines the patient’s prognosis. The greater the tumor burden, the worse the outlook, with only approximately 50% of patients surviving three months after the onset of symptoms and less than 10% surviving more than one year.196 Improved imaging modalities, advances in sur­ gical techniques for resection and transplantation, and new chemotherapeutic agents and regional therapies have made it possible to achieve long-term survival in individual patients. Long-term survival has been accomplished most often by resection of hepatic metastases in patients with colorectal cancer, a substantial number of whom have been cured or have obtained up to 20 years of disease-free sur-

vival.196-198 Survival for five years can be achieved in up to 60% who undergo resection of a solitary colon cancer metastasis to the liver.199 If the primary tumor has been removed completely and metastases are confined to the liver, resection of hepatic metastases should be considered. Liver transplantation, with or without chemotherapy, has been undertaken in a few patients but is generally not considered. RFA is a valid therapy for colorectal metastases in patients who are unable to tolerate or refuse surgical resection. Other invasive methods of destroying metastases, such as ethanol injection, freezing with cryoprobes, and laser vaporization, warrant further study. Radiation therapy and intra-arterial infusion of cytotoxic drugs have limited roles.

BENIGN TUMORS HEPATOCELLULAR ADENOMA Epidemiology and Pathogenesis Hepatocellular adenomas (or hepatic adenomas) were extremely rare before the use of oral contraceptive steroids became widespread. They are still rare in men, and the development of this tumor in women who are taking or have taken contraceptive steroids strongly implies a cause and effect relationship.200,201 Nevertheless, in light of the large number of women who use this form of contraception, the risk of hepatocellular adenoma is small, and its occurrence implies some form of genetic predisposition (see later). The association is particularly strong with prolonged use of oral contraceptive steroids; the estimated risk for women who use oral contraceptive agents continuously for five to seven years is five times the normal rate, and this risk increases to 25 times with use for longer than nine years. Using preparations with high hormone potency further increases the risk. Contraceptive pill–associated hepatocellular adenomas are more likely to develop in older than younger women. Hepatocellular adenomas have been linked to both types of synthetic estrogen and all forms of progestogen contained in oral contraceptive preparations.202 Current evidence favors estrogens as the culprit, although progestogens may contribute through their enzyme-inducing properties. The growth of hepatocellular adenomas appears to be hormone-dependent, as evidenced by an increase in size during pregnancy and occasional cases of regression (and even disappearance) of the tumor after cessation of oral contraceptive use. Although hormonal replacement therapy has not proved to be associated with the development of hepatic adenoma, avoidance of this form of therapy in women with a history of oral contraceptive–related hepatic adenoma is prudent, unless compelling reasons to the contrary exist. Hepatocellular adenomas and adenomatosis also may occur in those receiving long-term anabolic androgenic steroids63 and in those with certain inherited metabolic disturbances, especially type 1 glycogen storage disease, in which one or more hepatocellular adenomas occur in approximately 60% of patients (see Chapter 76).201 Genetic alterations have been identified in hepatocellular adenomas. Bilallelic mutations of the TCF1 gene that codes for hepatocyte nuclear factor 1α (HNF-1α) have been identified in up to 60% of patients with adenoma.203 A second pathway, the wnt pathway, which also is activated in about 25% of patients with hepatocellular carcinoma, is activated in at least some cases of adenoma.204 β-Catenin activation via this pathway appears to confer a higher risk of malignant transformation.205 The third identified pathway for formation of hepatocellular adenomas includes acute inflamma-

Chapter 94  Tumors and Cysts of the Liver Genetic Altered molecular mutations and pathway and risk factors its frequency TCF1 mut. MODY3

Obesity Alcohol

b-catenin mut. Glycogenolysis

Oral contraceptive

CYP1B1 mut.

Main clinical and pathologic characteristics

HNF1α inactivation 35%-45%

Familial adenomatosis (MODY3) Marked steatosis

β-catenin activation* 15%-19%

Males Cytologic abnormalities Pseudoglandular formation Higher risk of HCC

Acute phase inflammation* 30%-35%

Inflammatory infiltrates Dystrophic vessels Telangiectasia

Unidentified 10%-20%

A

B

Figure 94-5.  Schematic representation of the different molecular pathways altered in hepatocellular adenoma. Left, Main risk factors and known genetic predispositions. Center, Altered molecular pathways and their frequencies. Right, Principal clinical and pathologic features of the subtypes of adenomas. Arrows indicate the significant relationships. *Some tumors may be simultaneously inflammatory and β-catenin activated. CYP1B1, cytochrome P450 1B1; HCC, hepatocellular carcinoma; HNF1α, hepatocyte nuclear factor 1α (gene symbol TCF1); MODY3, maturity-onset diabetes of the young type 3; mut, mutation. (Adapted from Rebouissou S, Bioulac-Sage P, Zucman-Rossi J. Molecular pathogenesis of focal nodular hyperplasia and hepatocellular adenoma. J Hepatol 2008; 48:163-70.)

tory responses demonstrable by histologic examination of the tumor205 and associated with obesity and alcohol (Fig. 94-5).

Clinical Features

Hepatocellular adenomas manifest in a number of ways. They may produce no symptoms and be discovered during routine physical examination, if large, or during imaging of the upper abdomen for other reasons, if small. Approximately 25% of patients experience pain in the right hypochondrium or epigastrium. The pain usually is mild and ill-defined but may be severe as a result of bleeding into or infarction of the tumor. If the liver is enlarged, the surface usually is smooth, and the liver may be slightly tender. The most alarming presentation is with an acute hemoperitoneum following rupture of the adenoma. This complication is not uncommon, especially with tumors linked to oral contraceptive use, and carries an appreciable mortality rate.206,207 Tumors that rupture generally are large and solitary, although the most important determinant of rupture is a superficial location. Often, the affected woman is menstruating at the time; rupture also may occur during pregnancy.208

Diagnosis

Serum AFP concentrations are normal. Ultrasonography is used for initial imaging. Multiphase helical CT or MRI may also be used.209,210 The tumor has a clearly defined margin and, often, has almost parallel vessels entering it from the periphery (spoke wheel appearance). Alternatively, the lesion may contain tortuous vessels coursing irregularly through it. The adenoma may have focal avascular areas as a result of hemorrhage or necrosis. Because adenomas mimic normal liver tissue microscopically, needle biopsy and FNA may be of limited diagnostic value because the

Figure 94-6.  A, Surgical specimen of a large hepatocellular adenoma. The tumor is yellowish and slightly lobular, with a pseudocapsule and areas of necrosis and hemorrhage. B, Photomicrograph of a hepatocellular adenoma showing the resemblance to normal liver tissue, with cords of normal-looking, although generally slightly larger, hepatocytes, as well as Kupffer cells (but fewer in number than normal) lining the sinusoids. Bile ducts and central veins are not seen, but the presence of abnormal vascular structures is evident. (Hematoxylin and eosin.) (A courtesy of Elizabeth Brunt, MD, St. Louis, MO; B courtesy of Professor A.C. Paterson, Johannesburg, South Africa.)

cytologic appearance of the cells is similar to those of normal hepatocytes, although the tumor does not have normal portal structures.

Pathology

Hepatocellular adenoma generally occurs as a solitary, relatively soft, light brown to yellow tumor. It is sharply circumscribed but does not have a true capsule, although a pseudocapsule is formed by compression of the surrounding liver tissue (Fig. 94-6A).211 Hepatocellular adenomas arise in an otherwise normal liver. Occasionally, two or more tumors are present. Adenomas range in diameter from 1 to 30 cm and are commonly 8 to 15 cm in diameter. They are larger on average in women taking contraceptive steroids than in those not taking contraceptive steroids; they usually occupy a subcapsular position and project slightly from the surface of the liver. A pedunculated variety is seen occasionally. The cut surface of the tumor may show illdefined lobulation but is never nodular or fibrotic. Foci of

1585

1586

Section IX  Liver hemorrhage or necrosis are frequent, and bile staining may be evident. Microscopically, hepatocellular adenoma may mimic normal liver tissue to an astonishing degree (see Fig. 94-6B).211 The tumor is composed of sheets or cords of normal-looking or slightly atypical hepatocytes that show no features of malignancy. Few or no portal tracts or central veins are present, and bile ducts are conspicuously absent. Only an infrequent fibrous or vascular septum traverses the lesion. An essentially normal reticulin pattern is demonstrable throughout the adenoma. The walls of arteries and veins are thickened. Some areas with vascular abnormalities are infarcted, and thrombi may be seen. Peliosis hepatis may be found in relation to the tumor.

Treatment and Prognosis

Because of the danger that a hepatocellular adenoma may rupture, surgical treatment is recommended.211,212 Resection usually is feasible in an uncomplicated case. When rupture has occurred, emergency resection should be performed if possible. If resection cannot be accomplished, the hepatic artery should be ligated. Arterial embolization has also been used successfully to control hemorrhage from a ruptured adenoma, either in preparation for surgery or when surgery is not possible.206 Whether or not the tumor is removed, the patient must refrain from taking oral contraceptive steroids. If the adenoma is not resected, pregnancy should be avoided. Hepatocellular carcinoma occurs in a small number of women taking oral contraceptive steroids, and speculation has been raised that hepatocellular adenomas might undergo malignant transformation. Indeed, this sequence has been documented in a few instances. Therefore, managing hepatocellular adenomas merely by discontinuing the use of contraceptive steroids carries the risk that malignant transformation may still occur. The management of hepatic adenomatosis is problematic.213 Often, in these cases, the number of tumors is large, and they cannot be resected entirely. The role of liver transplantation for adenomatosis is not clear at present.

CAVERNOUS HEMANGIOMA Epidemiology

Cavernous hemangioma is the most common benign tumor of the liver and is found in as many as 7% of autopsies.63,211 The lesion is thought to be a congenital malformation or hamartoma that increases in size, initially with growth of the liver and thereafter by ectasia. Cavernous hemangiomas affect persons of all ages, although they manifest most often in the third, fourth, and fifth decades of life. Women are predominantly affected (4 : 1 to 6 : 1) and often present at a younger age and with larger tumors in comparison with men. Cavernous hemangiomas may increase in size with pregnancy or the administration of estrogens and are more common in multiparous than in nulliparous women.

Occasionally, an arterial bruit is heard over the tumor. Arteriovenous shunting has been described with cavernous hemangiomas.

Diagnosis

The ultrasonographic appearance is variable and nonspecific, although the lesion usually is echogenic.215,216 Provided that the cavernous hemangioma is larger than 3 cm in diameter, single photon emission computed tomography (SPECT) with colloid 99mTc-labeled red blood cells shows the tumor to be highly vascular and has a sensitivity and accuracy similar to those of MRI.217 Almost all cavernous hemangiomas can be diagnosed by bolus-enhanced CT with sequential scans.218 The center of the lesion remains hypodense, whereas the peripheral zone, which varies in thickness and may have a corrugated inner margin, is enhanced. MRI has a high degree of specificity and a central role in the diagnosis of small hemangiomas (Fig. 94-7).219,220 With small hemangiomas, the contrast material may assume a ring-shaped or C-shaped configuration, with an avascular center resulting from fibrous obliteration; this appearance is pathognomonic. Thrombocytopenia resulting from sequestration and destruction of platelets in large hemangiomas (KasabachMerritt syndrome) is seen occasionally in infants but rarely in adults.214,221 Malignant transformation has not been reported.

A

Clinical Features

The great majority of cavernous hemangiomas are small and asymptomatic and are discovered incidentally during imaging of the liver for another reason, at autopsy, or at laparotomy. Larger or multiple lesions produce symptoms.214 Those larger than 4 cm in diameter are called giant cavernous hemangiomas, which may be as large as 27 cm. Upper abdominal pain is the most common complaint and results from partial infarction of the lesion or pressure on adjacent tissues. Early satiety, nausea, and vomiting also may occur. Cavernous hemangiomas occasionally rupture. The only physical finding may be an enlarged liver.

B Figure 94-7.  Magnetic resonance imaging of a small cavernous hemangioma in the liver (arrow). A, T1-weighted image showing a rounded mass with a uniform increase in T1 signal intensity (low signal). B, Heavily T2-weighted image showing a mass with a uniform increase in signal intensity (bright signal relative to the water signal of cerebrospinal fluid). (Courtesy of Dr. P. Sneider, Johannesburg, South Africa.)

Chapter 94  Tumors and Cysts of the Liver Because of the risk of severe bleeding, percutaneous needle biopsy should not be performed if a cavernous hemangioma is suspected. Moreover, a needle biopsy is of limited diagnostic value.

Pathology

Cavernous hemangiomas usually are solitary lesions, although multiple tumors occur in 10% of patients.63,211 Reddish-purple or bluish masses are seen under Glisson’s capsule or deep in the substance of the liver. The larger lesions may be pedunculated. Cavernous hemangiomas are well circumscribed but seldom encapsulated. They may show central necrosis and, in some cases, the whole tumor is firm in consistency and grayish-white in appearance. Microscopically, hemangiomas are composed of multiple vascular channels of varying sizes lined by a single layer of flat epithelium and supported by fibrous septa. The vascular spaces may contain thrombi. The demonstration of mast cells within hemangiomas suggests that they may have a role in pathogenesis.222 Sclerosing cavernous hemangiomas may sometimes be seen and probably represent natural involution of these lesions. Occasionally, cavernous hemangiomas are associated with hemangiomas in other organs. They also may coexist with cysts in the liver or pancreas,223 von Meyenburg complexes (see later),224 or focal nodular hyperplasia (see later).225

Treatment

The great majority of cavernous hemangiomas can safely be left untreated. A cavernous hemangioma that is large but localized, and the cause of incapacitating symptoms, should be resected.214 If resection is not feasible, reduction in the size of the tumor with relief of symptoms is rarely achieved with irradiation, arterial ligation, arteriographic embolization, or systemic glucocorticoids.226,227 RFA has been used with some success. If a cavernous hemangioma has ruptured, it may be necessary to embolize or clamp the hepatic artery to stop bleeding before proceeding with resection, although this complication is exceedingly rare.

INFANTILE HEMANGIOENDOTHELIOMA Epidemiology and Clinical Features

Although rare, infantile hemangioendothelioma is the most common tumor of the liver in infants. Its importance stems from the high incidence of congestive heart failure in infants with this tumor and the resulting high mortality rate. The tumor almost invariably manifests in the first six months of life and is twice as common in girls as in boys.164,228 Hepatic hemangioendothelioma often coexists with hemangiomas in other organs, especially the skin (in approximately 50% of patients). Small hemangioendotheliomas are usually asymptomatic. The presence of a large lesion is recognized clinically by the diagnostic triad of an enlarged liver, high-output cardiac failure, and multiple cutaneous hemangiomas.201,228 The liver is larger than expected on the basis of the severity of the cardiac failure, and hepatomegaly persists after the heart failure has been treated successfully. When hemangioendotheliomas occur diffusely throughout the liver, as they usually do, their combined effect is to act as a large peripheral arteriovenous shunt. Shunts of this size are responsible for the cardiac failure. Approximately one third of patients have jaundice. Patients may be anemic, partly because of the dilutional effect of the increased circulating plasma volume that develops with large peripheral arteriovenous fistulas. A microangiopathic hemolytic anemia may con­tribute. In addition, thrombocytopenia may be present (KasabachMerritt syndrome). Malignant change is a rare complication.

Diagnosis

Ultrasonography may show one or more echogenic masses in the liver. Hepatic angiography is particularly helpful in diagnosis and shows stretching, but not displacement, of the intrahepatic arteries.229 Abnormal vessels arise from the hepatic arteries and promptly opacify the liver, thereby giving rise to the characteristic blush of an arteriovenous shunt. The circulation time through the liver is short. Focal avascular areas may be evident when hemorrhage into or necrosis of the tumor has occurred. CT with enhancement and MRI are as specific as hepatic arterio­graphy for the diagnosis of hemangioendotheliomas.230 Percutaneous biopsy is contraindicated because of the danger of bleeding.

Pathology

Two types of infantile hemangioendothelioma are recognized. Type I lesions are often calcified and have a fibrous stromal separation (with bile ductules) between channels. Type II lesions have a more malignant and disorganizedappearing endothelial cell lining and no stromal bile ductules.164,231 Infantile hemangioendotheliomas typically are multifocal and produce a nodular deformity of the entire liver. The nodules range in size from a few millimeters to many centimeters and are well demarcated but not encapsulated. At laparotomy, the nodules can be seen to pulsate. They are reddish purple, although large tumors are gray to tan. They may show hemorrhages, fibrosis, or calcification. Microscopically, infantile hemangioendothelioma is composed of layers of plump endothelial cells. A single layer characterizes the type I pattern, whereas several layers characterize the type II pattern. In some areas of the tumor, solid masses of mesoblastic primordial cells that differentiate early into vascular structures are observed. Fibrous septa may be prominent, and extramedullary hematopoiesis occurs frequently. Thrombosis may be followed by scarring and calcification.

Treatment and Prognosis

The course of infantile hemangioendothelioma is characterized by tumor growth during the early months of life, followed by gradual involution.228 If the child survives, the tumor involutes completely. Life-threatening aspects of the disorder are intractable congestive heart failure and, to a lesser extent, consumptive coagulopathy or rupture of the tumor. Cardiac failure should be treated by conventional means initially, but if these measures fail, more aggressive treatment of the tumor, such as embolization, ligation of the hepatic artery, surgical resection, or liver transplantation, should be considered.232,233 Use of glucocorticoids has been successful in many (but not all) patients,234 whereas irradiation has seldom been beneficial. When the tumor is confined to one lobe, surgical resection is curative, even in the presence of cardiac failure.228

OTHER BENIGN TUMORS OF THE LIVER

Other rare benign tumors of the liver include angiomyolipoma,235 bile duct adenoma, biliary cystadenoma, and biliary adenofibroma.236,237

TUMOR-LIKE HEPATIC LESIONS FOCAL NODULAR HYPERPLASIA

Focal nodular hyperplasia is a circumscribed, usually solitary lesion composed of nodules of benign hyperplastic hepatocytes surrounding a central stellate scar.238

1587

1588

Section IX  Liver Epidemiology and Pathogenesis

Focal nodular hyperplasia is more common than hepatocellular adenoma. The lesion is seen more often in women than in men, although the gender difference is much less striking than that for hepatocellular adenoma. Focal nodular hyperplasia occurs at all ages, but most patients present in the third and fourth decades of life201; the age distribution is similar to that of hepatocellular adenomas and the two lesions may coexist. The cause of focal nodular hyperplasia is unknown. Abnormalities in arteries in small and medium-sized portal tracts have been described, suggesting a role of vascular malformation in its pathogenesis.239,205 It has been described to occur with other vascular lesions, such as cavernous hemangioma, epithelioid hemangioendothelioma, and hereditary hemorrhagic telangiectasia.240,241 A role for oral contraceptive steroids in the development of the lesion was suggested but has been disputed.201 Nevertheless, some evidence suggests that focal nodular hyperplasia may be hormone-dependent.242,243 Contraceptive steroids may accentuate the vascular abnormalities in focal nodular hyperplasia and cause the lesion to enlarge, become more symptomatic, and, rarely, rupture.

Clinical Features

Most of these lesions do not produce symptoms and are often discovered during upper abdominal imaging for other reasons or because an enlarged liver is felt on routine examination or found during abdominal surgery or at autopsy.201,244,245 Patients may experience mild pain, particularly with bleeding into or necrosis of the lesion.

Diagnosis

Serum AFP levels are normal. The mass lesion seen on ultrasonography and CT is not specific for focal nodular hyperplasia,246,247 unless the central scar and feeding artery are seen (Fig. 94-8). MRI may be useful for the diagnosis of focal nodular hyperplasia.248

Pathology

Focal nodular hyperplasia manifests as a firm, coarsely nodular, light brown or yellowish-gray mass of variable size with a dense, central stellate scar and radiating fibrous septa that divide the lesion into lobules.249 The nodule may be small, resembling a cirrhotic nodule, or extremely large.

Figure 94-8.  Contrast-enhanced computed tomography scan of the liver during arterial phase showing a typical focal nodular hyperplasia (arrow) with enhancement of the mass lesion and the central stellate scar apparent by its lack of enhancement.

The lesion of focal nodular hyperplasia usually occupies a subcapsular position and may be pedunculated. It generally is solitary. Larger lesions may show foci of hemorrhage or necrosis, although these features are seen less frequently than in hepatocellular adenomas. The fibrous septa sometimes are poorly developed, and the central scar may be absent. The lesion is sharply demarcated from the surrounding liver tissue, which is normal, but a true capsule is absent. Focal nodular hyperplasia is associated with hepatic hemangiomas in as many as 20% of cases. Microscopically, focal nodular hyperplasia closely resembles a focal form of inactive cirrhosis. Individual hepatocytes are indistinguishable from those of normal liver but lack the normal cord arrangement in relation to sinusoids, central veins, and portal tracts. Kupffer cells are present. Characteristically, the fibrous septa contain numerous bile ductules and vessels. Other features include heavy infiltrations of lymphocytes and, to a lesser extent, plasma cells and histiocytes. Bile duct proliferation in portal tracts also may be evident. Branches of the hepatic artery and portal vein show various combinations of intimal and smooth muscle hyperplasia, subintimal fibrosis, thickening of the wall, occlusive luminal lesions, and, at times, occluding thrombosis. Whether these vascular changes are primary or secondary is not known. Peliosis hepatis may be an associated lesion. The histologic features almost always make it possible to distinguish focal nodular hyperplasia from hepatocellular adenoma, although the distinction may be extremely difficult to make, particularly in small biopsy specimens.

Treatment

Large symptomatic or complicated lesions should be resected, usually by segmental resection or enucleation. Recurrence after resection is rare. These lesions may also be treated with RFA. Otherwise, focal nodular hyperplasia should be left alone. If the lesion is not resected, discontinuation of contraceptive steroids is recommended and may result in regression of the lesion. Periodic ultrasonography should be performed if a firm diagnosis of focal nodular hyperplasia has not been made, and a lesion seen to increase substantially in size should be resected. The available evidence argues against the notion that focal nodular hyperplasia is a premalignant condition.

OTHER NODULAR DISORDERS

Nodular regenerative hyperplasia is characterized by nodularity of the liver without fibrosis (see Chapter 35).250 This rare condition may be associated with a number of diseases, such as rheumatoid arthritis and Felty’s syndrome. Although generally diffuse, the nodularity occasionally is focal, in which case the lesion may be mistaken for a tumor. Patients with nodular regenerative hyperplasia typically present clinically with portal hypertension. Partial nodular transformation is characterized by nodules that are limited to the perihilar region of the liver. These patients also present with portal hypertension. Macroregenerative nodules may occur in advanced cirrhosis or after massive hepatic necrosis. In the presence of cirrhosis, they are believed to be premalignant and may, in addition, be mistaken for hepatic tumors during hepatic imaging.60 Inflammatory pseudotumor is a rare entity, resulting from focal infection, that may be mistaken for a hepatic tumor (see Chapter 82).251 It occurs particularly in young men, who present with intermittent fever, abdominal pain, jaundice, vomiting, and diarrhea. Leukocytosis, an elevated erythrocyte sedimentation rate, and polyclonal hyperglobulinemia are present in approximately 50% of patients. The lesion

Chapter 94  Tumors and Cysts of the Liver may be solitary or multiple and shows a mixture of chronic inflammatory cells, with plasma cells predominating. Focal fatty infiltration, or focal fatty sparing in the presence of diffuse fatty infiltration, may also be mistaken for a hepatic tumor (see Chapter 85).252

HEPATIC CYSTS Hepatic cysts are abnormal fluid-filled spaces in the hepatic parenchyma and biliary tree. They are categorized into three main types—fibrocystic diseases of the liver, cystadenomas and cystadenocarcinomas, and hydatid cysts. Cystadenomas and cystadenocarcinomas are discussed in Chapter 69. Hydatid cysts are discussed in Chapter 82.

FIBROCYSTIC DISEASES OF THE LIVER

Fibrocystic diseases of the liver originate from abnormal persistence or defects in the progressive remodeling of the ductal plate during development, resulting in dilated fluid-filled spaces, including hepatic and choledochal cysts, portal fibrosis, and ductal plate malformations (see Chapter 62).253,254 Fibrocystic disorders of the liver described here include simple hepatic cysts, polycystic liver disease (PCLD), von Meyenburg complexes, and Caroli’s disease (type V choledochal cyst). (The other diseases are congenital hepatic fibrosis and type IV choledochal cysts; see Chapter 62.)

Simple Hepatic Cysts

Simple hepatic cysts are thought to be congenital in origin and have a frequency of about 2.5% of the population.255 They generally are smaller than 5 cm in diameter and can number up to three before being considered part of PCLD.256 The cysts usually are asymptomatic and discovered incidentally during upper abdominal imaging. They occur more often in women than in men, and their prevalence increases with age. When symptomatic, they can produce complications similar to those of PCLD, including intracystic bleeding, infection, rupture, or compression of adjacent organs. Typically, initial imaging with ultrasound, CT, or MRI provides an accurate diagnosis and distinguishes a simple cyst from a hydatid cyst and cystadenoma. Septations, papillary projects, or calcification should raise the suspicious for an alternative diagnosis.257 Asymptomatic solitary hepatic cysts should be left alone. If intervention is required because of symptoms, percutaneous aspiration and sclerosis with alcohol or doxycycline will almost always ablate the cyst, but recurrence is frequent.256 An alternative approach is laparoscopic (or, rarely, open surgical) fenestration, which is seldom followed by recurrence but has greater morbidity.

Polycystic Liver Disease

PCLD is a rare condition in which multiple cysts form in the hepatic parenchyma; usually it comes to clinical attention in adulthood (Fig. 94-9). PCLD usually presents in association with autosomal dominant polycystic kidney disease (ADPKD)258,259 but can appear as isolated polycystic liver disease.260,261 The cysts range in diameter from a few millimeters to 10 cm or more. They contain clear, colorless, or strawcolored fluid and are lined by a single layer of cuboidal or columnar epithelium, resembling that of bile ducts.258-262 Rarely, the cysts may be lined by squamous epithelium; these cysts may be complicated by the development of squamous cell carcinoma. In addition to the nature of the lining

Figure 94-9.  Magnetic resonance imaging of the abdomen in a patient with severe polycystic liver disease. This coronal T2-weighted image shows a massively enlarged liver with numerous bright fluid-filled cysts. (Courtesy of Dr. N. Cem Balci, St. Louis, Mo.)

epithelium, evidence for a biliary origin of these cysts is suggested by the composition of the cystic fluid, which has a low glucose content and contains secretory immunoglobulin A and gamma glutamyl transpeptidase. The cysts are thought to arise as a result of ductal plate malformation. This process gives rise to von Meyenburg complexes (see later), which become disconnected from the biliary tree during development and growth and dilate progressively to form cysts. PCLD is fairly common in patients with ADPKD. It occurs in approximately 24% of patients in the third decade of life to 80% in the sixth decade of life, but the kidney disease usually dominates the clinical course.263 Cysts also may be present in the pancreas, spleen and, less often, other organs. Symptomatic liver disease correlates with advancing age, severity of renal cysts, and renal dysfunction.264 Women tend to have larger and more numerous cysts, and a correlation with the number of pregnancies has been found. The use of exogenous female sex hormones may accelerate the rate of growth and size of the cysts. PCLD may coexist with other fibrocystic liver diseases, such as congenital hepatic fibrosis (in which the patient is likely to present with portal hypertension), Caroli’s disease, or von Meyenburg complexes.258-262 PCLD also is associated with other conditions such as berry aneurysms, mitral valve prolapse, diverticular disease, and inguinal hernias. Isolated PCLD not associated with ADPKD is rare, representing 7% of all PCLD in autopsy series.265 It usually is asymptomatic.266 Like ADPKD-associated PCLD, isolated PCLD is associated with pregnancy and appears to be more symptomatic in women than men. ADPKD is a common genetic disease with a frequency of 1:1,000 in whites.267 Two genes are responsible. The gene affected in ADPKD1 is PKD-1, which is located on chromosome 16q13-q23 and expresses a ubiquitous protein, polycystin-1.268,269 The gene responsible for ADPKD2 is PKD-2, which is located on chromosome 4 and expresses polycystin-2. The two polycystins are transmembrane

1589

1590

Section IX  Liver glycoproteins that complex and localize in the primary cilium, a microtubule-based structure found on renal and biliary tubule epithelium and thought to act as a flow sensor and regulator of Ca2+ influx.270 Although the mutation is inherited as an autosomal dominant trait, a second somatic mutation is thought to be necessary to produce the monoclonally derived cysts.271 Isolated PCLD has been shown in North American and Finnish families to be linked to the gene PRKCSH (also known as protein kinase C substrate 80K-H) on chromosome 19 locus p13.2-13.1 and to SEC63 on chromosome 6q21, although other associated genes undoubtedly exist.268,272,273 The gene products hepatocystin and SEC63p are thought to be involved in the folding and quality control of glycoproteins and protein translocation in the endoplasmic reticulum, respectively.274 The genes appear to be autosomal dominant, and a second somatic mutation is thought to be needed to cause disease. The hepatic cysts in polycystic liver disease, whether or not they occur in association with renal cysts, rarely cause morbidity, and many affected persons are asymptomatic.258-262 The livers of these patients contain only a few cysts or cysts smaller than 2  cm in diameter. Symptoms occur in patients with more numerous and larger cysts (10% to 15% of patients, usually women), generally with markedly enlarged livers. Abdominal discomfort or pain, postprandial fullness, awareness of an upper abdominal mass, a protuberant abdomen, and shortness of breath may be present. Severe pain may be experienced with rupture or infection of a cyst, bleeding into a cyst, or torsion of a pedunculated cyst. Jaundice is evident in approximately 5% of patients and is caused by compression of the major intrahepatic or extrahepatic bile ducts. Ascites, if present, is the result of portal hypertension, which generally is caused by associated congenital hepatic fibrosis but occasionally by compression of the hepatic veins by the cysts. Gastroesophageal variceal bleeding has rarely been reported.275 Liver biochemical test results generally are normal, with intact hepatic function, although serum alkaline phosphatase and GGTP levels may be increased. A raised right hemidiaphragm may be evident on a plain x-ray of the chest in patients with severe PCLD. The diagnosis of PCLD is confirmed by ultrasound, CT, or MRI (see Fig. 94-9). On the rare occasions when cysts require treatment, fenestration (unroofing) should be performed.258,276 Cyst fenestration originally was done at laparotomy but is now performed laparoscopically. A high recurrence rate is observed for cysts treated in this way. Cysts also have been treated by percutaneous injection of a sclerosing substance such as alcohol or doxycycline, but most patients have too many small cysts to warrant this approach. Patients who fail to respond to cyst fenestration may be considered for partial hepatic resection. Liver transplantation (sometimes combined with renal transplantation) is generally reserved for patients with hepatic failure or severe symptoms that interfere with activities of daily living.277

Fibrocystic Disease Associated with Autosomal Recessive Polycystic Kidney Disease

Fibrocystic liver disease may manifest in childhood as an autosomal recessive disorder that is usually rapidly fatal as a consequence of associated autosomal recessive polycystic kidney disease (ARPKD).258,259 A proportion of patients maintain renal function into adulthood, however, and complications of the associated liver disease then predominate. The liver cysts are microscopic, rather than macroscopic,

and present a clinical picture of congenital hepatic fibrosis. Complications of portal hypertension are the usual hepatic manifestations of the disease. The gene responsible for this disease, PKHD1, has been identified at chromosomal locus 6p21-cen, and the ARPKD protein, fibrocystin, is predicted to be an integral receptor-like protein.278 Many different mutations throughout the gene have been identified in patients with ARPKD.

Von Meyenburg Complexes

Von Meyenburg complexes (also known as biliary microhamartomas) are common and do not produce symptoms; they are small and usually multiple. Each complex is composed of cystically dilated intra- and interlobular bile ducts embedded in a fibrous stroma.175,279 The cysts are lined by cuboidal or flat epithelium. They occur in almost all patients with congenital hepatic fibrosis and may coexist with Caroli’s disease or ADPKD. Von Meyenburg complexes are found in or adjacent to portal tracts and are believed to arise as a result of malformation of the ductal plate (see Chapter 62), and they may be complicated by the development of peripheral cholangiocarcinoma.280

Caroli’s Disease

Caroli’s disease is a rare disorder characterized by congenital nonobstructive gross dilatation of the segmental intrahepatic bile ducts.256 The disease has been included in the classification of choledochal cysts (as type V)256,258 and may occur in association with medullary sponge kidney (in 60% to 80% of patients) or congenital hepatic fibrosis (see Chapter 62). Caroli’s disease is believed to be caused by an intrauterine event that arrests ductal plate remodeling at the level of the larger intrahepatic bile ducts.253 The resulting bile duct ectasia may be diffuse or localized. Both autosomal recessive and autosomal dominant modes of inheritance have been proposed. Caroli’s disease affects men and women equally and usually becomes symptomatic in early adulthood; more than 80% of patients present with symptoms before the age of 30 years. Patients typically present with recurrent episodes of fever and abdominal pain caused by cholangitis. The liver often is enlarged. Ductal ectasia predisposes to bile stagnation, which in turn may lead to cholangitis, abscess formation, and septicemia.281 Gallstones form in the ectatic ducts in one third of patients. The result of these complications may be cholangiocarcinoma, which develops in less than 10% of patients. Caroli’s disease usually is discovered when the liver is imaged during investigation of suspected cholangitis. Irregular dilatations of the larger intrahepatic bile ducts are seen. Attacks of cholangitis require treatment with antibiotics. Endoscopic retrograde cannulation of the biliary system may be used to facilitate removal of sludge or stones from the accessible part of the biliary system, and the cysts may be drained by an endoscopic or percutaneous route. Liver resection for unilobar Caroli’s disease and liver transplantation for diffuse Caroli’s disease are associated with excellent long-term patient survival and a low rate of complications.282

APPROACH TO THE PATIENT WITH A HEPATIC MASS LESION The approach to the diagnosis of a mass lesion in the liver will be influenced by the age and gender of the patient and the presence or absence of symptoms (Fig. 94-10). Making

Chapter 94  Tumors and Cysts of the Liver A

Ultrasound

Possible abscess (see Chapter 82)

Solid

Cystic

Simple benign cyst (single) Polycystic disease (multiple) Echinococcal cyst (daughter cysts) Biliary cystadenoma

Suspicious for hemangioma

Not hemangioma

MRI or

Characterize with further imaging

99mTc-labeled

Evaluate further and treat if symptomatic or if echinococcosis or malignancy is suspected

red blood cell study

Consider needle biopsy

B

Ultrasound, CT, or MRI Not suspicious for HCC

Suspicious for HCC

2 cm

Repeat imaging in 3-4 mo

Two dynamic imaging studies

Dynamic imaging Serum AFP

Typical of HCC

Not typical of HCC

Biopsy

Typical of HCC or Serum AFP >200 ng/mL

Cyst Hemangioma Metastases

Not typical of HCC

Biopsy

a definitive diagnosis of a mass lesion in the liver solely on clinical grounds is seldom possible. Nevertheless, detailed history taking will provide important clues about the prob­ able benign or malignant nature of the lesion. The approach to a mass in the liver differs, depending on whether or not cirrhosis is present. In a noncirrhotic liver, masses or tumors may be found in the liver incidentally or because of symptoms; the main concern is cancer metastatic from elsewhere. Initial imaging studies such as ultrasonography, CT, or MRI will indicate if the lesion is cystic. Cystic lesions should be investigated further and treated only if echinococcal cysts are suspected (see Chapter 82). Possible solid lesions in a noncirrhotic liver include a hemangioma, which can be confirmed by MRI or a 99mTc-labeled red blood cell scan. Alternatively, a biopsy should be considered to exclude malignancy, unless the lesion has the characteristic radiologic appearance of focal nodular hyperplasia. In a patient known to have cirrhosis, the presence of a nodular or mass lesion should be presumed to be hepatocellular carcinoma until proven otherwise. The AASLD practice

Figure 94-10.  A, Algorithm depicting the approach to management of a patient, not known to have cirrhosis, with a hepatic mass (often incidental, possibly symptomatic). B, Algorithm depicting the approach to management of a patient with known or suspected cirrhosis and a hepatic mass (found on routine surveillance because of symptoms or increasing AFP level). AFP, alpha fetoprotein; CT, computed tomography; HCC, hepatocellular carcinoma; MRI, magnetic resonance imaging.

guidelines provide criteria for the noninvasive diagnosis of hepatocellular carcinoma based on the vascularity of the tumor (see earlier). Contrast enhancement during the arterial phase of a multiphase CT or MRI study with subsequent washout during the venous phase is considered diagnostic of hepatocellular carcinoma if the lesion is larger than 2 cm in diameter. For lesions between 1 and 2 cm in diameter, this characteristic vascularity should be demonstrated on two imaging modalities; biopsy of the lesion may be required. For a lesion smaller than 1 cm, biopsy may be difficult to accomplish technically, and the lesion should be observed; if it becomes larger than 1 cm in diameter, the diagnosis of hepatocellular carcinoma is essentially confirmed. If ascertaining whether a patient has underlying cirrhosis is difficult on clinical and imaging grounds, a biopsy of the nontumorous liver may be done. Ultrasonography generally is the first imaging study performed to evaluate a hepatic mass to determine whether the lesion is cystic or solid and whether the intrahepatic bile ducts are dilated. Ultrasonography is the cheapest and

1591

1592

Section IX  Liver most widely available hepatic imaging modality but is operator-dependent. Ultrasonography also is an effective way to demonstrate the presence of daughter cysts within hydatid cysts. Some tumors, such as biliary cystadenoma and cystadenocarcinoma, may be partly cystic. In addition, solid tumors may show hypoechoic areas interspersed with the expected hyperechoic picture. The hypoechoic areas result from the histologic characteristics of the tumor or from necrosis or hemorrhage in the lesion. Focal fatty change produces focal hypoechoic areas that also are hypodense on CT. Hemangiomas typically are hyperechoic. Ultrasonography does not permit diagnosis of a specific hepatic tumor to be made, and additional information needs to be obtained from dynamic contrast-enhanced ultrasonography, CT, or MRI. These dynamic studies demonstrate the vascular nature of certain tumors and can distinguish the arterial vascularity of hepatocellular carcinomas from the general increase in vascularity of hepatic adenomas. Metastases are generally much less vascular, with the exception of neuroendocrine tumors that have spread to the liver. Hemangiomas, although very vascular, enhance slowly with these dynamic studies because the vascularity is of capillary origin. Rapid washout of contrast dye from lesions that enhance during the arterial phase is a feature of hepatocellular carcinoma. Each technique has limitations, however, and, at present, an ideal method for imaging the liver is not available. Cavernous hemangiomas larger than 3 cm in diameter may be recognized on scintigraphy following the injection of radiolabeled red blood cells. Hepatic arteriography may show the typical features of vascular malignant tumors but is not always helpful and is required less frequently than in the past. If hepatocellular carcinoma or hepatoblastoma is suspected, the serum AFP concentration should be measured. Laparoscopy may demonstrate unsuspected tumor seeding or the presence of cirrhosis. Definitive diagnosis of a solid hepatic lesion depends on demonstrating the typical histologic features of the tumor. Histologic confirmation usually can be achieved by percutaneous needle biopsy or FNA, often with CT or ultrasound guidance.

KEY REFERENCES

Bismuth H, Chiche L, Castaing D. Surgical treatment of hepatocellular carcinomas in noncirrhotic liver: Experience with 68 liver resections. World J Surg 1995; 19:35-41. (Ref 101.) Bruix J, Sherman M. Management of hepatocellular carcinoma. Hepatology 2005; 42:1208-36. (Ref 25.) Colli A, Fraquelli M, Casazza G, et al. Accuracy of ultrasonography, spiral CT, magnetic resonance, and alpha-fetoprotein in diagnosing hepatocellular carcinoma: A systematic review. Am J Gastroenterol 2006; 101:513-23. (Ref 24.) Desmet VJ. Congenital diseases of intrahepatic bile ducts: Variations on the theme “ductal plate malformation.” Hepatology 1992; 16:1069-83. (Ref 253.) El-Serag H, Mason A. Rising incidence of hepatocellular carcinoma in the United States. N Engl J Med 1999; 340:745-50. (Ref 5.) Everson GT. Hepatic cysts in autosomal dominant polycystic kidney disease. Am J Kidney Dis 1993; 22:520-5. (Ref 262.) Khan SA, Thomas HC, Davidson BR, Taylor-Robinson SD. Cholangiocarcinoma. Lancet 2005; 366:1303-14. (Ref 162.) Lazaridis KN, Gores GJ. Cholangiocarcinoma. Gastroenterology 2005; 128:1655-67. (Ref 160.) Lencioni R, Cioni D, Crocetti L, et al. Early-stage hepatocellular carcinoma in patients with cirrhosis: Long-term results of percutaneous image-guided radiofrequency ablation. Radiology 2005; 234:961-7. (Ref 117.) Liaw YF, Sung JJ, Chow WC, et al. Lamivudine for patients with chronic hepatitis B and advanced liver disease. N Engl J Med 2004; 351:152131. (Ref 77.) Llovet J, Bruix J. Systematic review of randomized trials for unresectable hepatocellular carcinoma: Chemoembolization improves survival. Hepatology 2003; 37:429-42. (Ref 126.) Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008; 359:378-90. (Ref 136.) Marrero J, Fontana R, Fu S, Conjeevaram H, Su G, Lok A. Alcohol, tobacco and obesity are synergistic risk factors for hepatocellular carcinoma. J Hepatol 2005; 42:218-24. (Ref 93.) Nguyen MH, Garcia RT, Simpson PW, et al. Racial differences in effectiveness of alpha-fetoprotein for diagnosis of hepatocellular carcinoma in hepatitis C virus cirrhosis. Hepatology 2002; 36:410-17. (Ref 33.) Ribero D, Curley SA, Imamura H, et al. Selection for resection of hepatocellular carcinoma and surgical strategy: Indications for resection, evaluation of liver function, portal vein embolization, and resection. Ann Surg Oncol 2008; 15:986-92. (Ref 105.) Zucman-Rossi J. Genetic alterations in hepatocellular adenomas: Recent findings and new challenges. J Hepatol 2004; 40:1036-9. (Ref 204.) Full references for this chapter can be found on www.expertconsult.com.

ACKNOWLEDGMENT

The authors gratefully acknowledge the contributions of Professor Michael C. Kew, who authored this chapter in previous editions of the textbook.

CHAPTE R

95 Liver Transplantation Paul Martin and Hugo R. Rosen

CHAPTER OUTLINE Indications  1594 Listing Criteria and Policies of the United Network for Organ Sharing  1595 Absolute and Relative Contraindications  1595 Transplantation Evaluation and Listing  1598 Disease-Specific Indications  1598 Alcoholic Liver Disease  1598 Hepatitis B  1599 Hepatitis C  1600 Acute Liver Failure  1601 Cholestatic Liver Disease  1601 Hepatic Malignancy  1602 Metabolic Disorders  1603 Nonalcoholic Fatty Liver Disease  1603 Vascular Disorders  1603

Despite continued advances in the treatment of chronic liver disease, most notably antiviral therapy, and management of complications of chronic liver disease, such as transjugular intrahepatic portosystemic shunt (TIPS) placement, liver transplantation remains the only prospect for long-term survival in patients with advanced liver disease who have reached the limits of purely medical interventions. The major indications for liver transplantation in adults are decompensated cirrhosis, unresectable primary hepatic malignancies, and acute liver failure in which spontaneous recovery is not anticipated.1 Liver transplantation has continued to evolve in response to the shortage of deceased-donor organs and the frequency of recurrent disease. The availability of a wider array of immunosuppressive agents has made graft rejection a lesser threat than disease recurrence.2 Recurrence of hepatitis C virus (HCV) infection, because of its frequency as an indication for liver transplantation, high rate of graft reinfection and failure, and lack of effective prophylaxis, has become a major challenge in the care of liver transplant recipients.3 By contrast, modern regimens of prophylaxis to prevent hepatitis B virus (HBV) reinfection now allow liver transplantation to be undertaken with a low likelihood of recurrence.4 Recurrence of nonviral liver disease is now recognized as a threat to the graft, albeit of an order of magnitude less frequent than that for HCV reinfection.5 Because immunosuppression has a proviral effect, a number of liver transplantation centers have attempted earlier reduction or withdrawal of glucocorticoids in recipients with chronic HBV or HCV infection.6 By contrast, more intensive immunosuppression may be necessary in liver transplant recipients with autoimmune liver diseases.7 The apparently intractable shortage of deceased donors is reflected in continuing deaths of potential recipients listed for liver transplantation. A major change in the allocation of available organs has occurred with the introduction of

Autoimmune Hepatitis  1603 Other Indications  1603 Surgical Aspects of Liver Transplantation  1604 Native Hepatectomy  1604 Live-Donor Liver Transplantation  1605 Immunosuppression  1605 Postoperative Course  1606 Initial Phase to Discharge from Hospital  1606 Following Discharge from Hospital  1608 Long-Term Management  1609 General Preventive Medicine  1609 Immunizations and Antibiotic Prophylaxis  1610 When to Call the Transplantation Center  1611 Hepatic Retransplantation  1611

the Model for End-stage Liver Disease (MELD) score, which has achieved its stated aim of reducing the number of deaths on the liver transplantation waiting list.8 This method of organ allocation assigns organs to recipients on the basis of an objective, continuous measure of severity of liver disease, thereby removing time spent on the waiting list as a determining factor. Extension of live-donor liver transplantation (LDLT) to adult recipients has increased the organ pool despite the tempering of enthusiasm following an increased appreciation of potential risks to the donor.9 Other innovations such as splitting deceased-donor grafts to benefit two recipients and use of so-called marginal or extended criteria grafts, including those from older and non–heart-beating donors, also have expanded the organ supply modestly.10 An increased frequency of complications, most notably of the biliary tract, has been a consequence of expanding the donor criteria, however.11 Efforts to expand the deceaseddonor supply by public education programs have succeeded in enhancing organ donation, although the major dis­ crepancy between the number of potential recipients and the number of available organs persists, resulting in continuing attrition of listed patients who succumb to com­ plications of decompensated liver disease while awaiting liver transplantation.12 Although the organ allocation system will undoubtedly continue to evolve and recurrence of disease will remain a threat, the prospects for long-term survival are very good to excellent for most liver transplant recipients who otherwise would succumb to their liver disease. For instance, the likely one-year survival rate for patients with decompensated cirrhosis is less than 10% without liver transplantation but approximately 85% to 90% at one year and 75% at five years after transplantation for most indications.13 Access to liver transplantation has transformed the management of advanced liver disease but has resulted in an expanding cohort of decompensated potential recipients

1593

1594

Section IX  Liver who require frequent medical attention.14 Because the best outcomes following liver transplantation are obtained in patients who have not already experienced multiple complications of liver disease,15 referral for transplantation evaluation is appropriate when a cirrhotic patient has had an index complication, such as the onset of ascites. For at least some potential recipients, access to LDLT may avoid a lengthy waiting period with the risk of further and potentially life-threatening complications of liver disease. In parallel with the evolution of liver transplantation, the care of transplantation candidates with advanced disease has become a major clinical challenge. The transplantation hepatologist must combine the skills necessary to practice gastroenterology, multidisciplinary internal medicine, and intensive care. This skill set has been formally recognized by the development of a secondary subspecialty in transplantation hepatology by the American Board of Internal Medicine.16

INDICATIONS The major indications for liver transplantation in adults reflect the most frequent causes of adult liver disease, notably chronic hepatitis C, alcoholic liver disease, and, to a lesser extent, chronic hepatitis B, primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), autoimmune hepatitis, and hemochromatosis (Fig. 95-1). Many liver transplantation candidates previously diagnosed as having “cryptogenic” cirrhosis are now considered to have underlying nonalcoholic fatty liver disease (NAFLD). An uncommon but important indication for liver transplantation is acute liver failure, which has a high mortality rate in the absence of liver transplantation. The role of liver transplantation in primary hepatic malignancy has become better defined; a subset of patients with primary hepatocellular carcinoma (HCC) have a high likelihood of cure with transplantation, although the roles of adjunctive therapies such as transarterial chemoembolization (TACE), radiofrequency ablation, and the oral chemotherapy agent sorafenib need to be defined (see Chapter 94).17 The other major primary adult hepatic malignancy, cholangiocarcinoma, had been regarded as a contraindication to liver transplantation because of its rapid and almost invariable recurrence post–liver transplantation, although acceptable outcomes have been reported in a subset of patients with hilar tumors who receive adjuvant external beam radiation and chemo-

10%

33%

7% 4% 1% 6%

4%

14% 9%

12%

Hepatitis C Hepatitis B Alcohol Cryptogenic/nonalcoholic steatohepatitis Cholestatic disorders Hepatocellular carcinoma Other malignancies Metabolic disorders Pediatric diseases Miscellaneous causes

Figure 95-1.  Proportion of liver transplants performed for specific indications, 1992 to 2007. (From O’Leary JG, Lepe R, Davis GL. Indications for liver transplantation. Gastroenterology 2008; 134:1764-76, with permission.)

sensitization (see Chapter 69).18 The major indication for pediatric liver transplantation is biliary atresia following a failed Kasai procedure (portoenterostomy) or delayed recognition of the diagnosis (see Chapter 62). Other major pediatric indications include α1-antitrypsin deficiency and other metabolic disorders (see Chapter 76). Recognition of cirrhosis per se is not an indication for liver transplantation, although a key issue in managing a cirrhotic patient is deciding whether transplantation will be needed in the future and when referral for evaluation is appropriate (Table 95-1). Another important aspect of the management of a compensated cirrhotic patient is the anticipation of complications. Endoscopic surveillance of patients with cirrhosis for gastroesophageal varices, for instance, to offer prophylaxis, either pharmacologic or endoscopic, to prevent initial or recurrent gastrointestinal bleeding is now routine practice (see Chapter 90).14 Surveillance for HCC is also regarded as the standard of care in a cirrhotic patient (see Chapter 94). Discovery of a hepatic mass suggestive of HCC in a cirrhotic patient should prompt evaluation to determine whether hepatic resection or liver transplantation is the most appropriate curative approach. Otherwise, liver transplantation should normally be a consideration only when the limits of medical therapy for complications of cirrhosis have been reached, particularly if LDLT is an option. The risk of surgery must always be weighed against a realistic assessment of the potential recipient’s prognosis in the absence of liver transplantation. For example, in a patient with decompensated cirrhosis caused by HBV infection, effective suppression of viral replication by an effective antiviral agent may result in impressive clinical improvement, thereby delaying or even obviating the need for liver transplantation (see Chapter 78). Similarly, abstinence from alcohol can result in resolution of signs of hepatic decompensation in a patient with alcoholic liver disease (see Chapter 84). The course of chronic liver disease remains unpredictable, however, and observing an apparently well-compensated patient deteriorate dramatically because of an intercurrent complication such as variceal bleeding is sobering. Although recognition of cirrhosis implies a risk of major complications and diminished life expectancy, a cirrhotic patient can remain stable for a protracted period of time. For example, Fattovich and colleagues19 observed that in patients with well-compensated

Table 95-1  Indications for Liver Transplantation Acute liver failure Complications of cirrhosis Ascites Chronic gastrointestinal blood loss due to portal hypertensive gastropathy Encephalopathy Liver cancer Refractory variceal hemorrhage Synthetic dysfunction Liver-based metabolic conditions with systemic manifestations α1-Antitrypsin deficiency Familial amyloidosis Glycogen storage disease Primary oxaluria Tyrosemia Urea cycle enzyme deficiencies Wilson disease Systemic complications of chronic liver disease Hepatopulmonary syndrome Portopulmonary hypertension

Chapter 95  Liver Transplantation Probability of decompensation (%)

50

28% 10 yrs

25

0 0

12

24

36

48

60

72

84

96

108 120

355 336 304 253 205 133 104

65

41

29

Months 18

Patients at risk Figure 95-2.  Probability of development of major complications (decompensation) in patients with well-compensated cirrhosis due to hepatitis C. (From Fattovich G, Giustina G, Degos F, et al. Morbidity and mortality in compensated cirrhosis type C: A retrospective follow-up study of 384 patients. Gastroenterology 1999; 112:463-72, with permission.)

cirrhosis caused by HCV infection, major complications of portal hypertension such as ascites and variceal hemorrhage occurred in less than 30% at 10 years; in the absence of an index complication, survival was excellent (Fig. 95-2). Once a complication supervenes, however, survival diminishes rapidly. For example, after the development of ascites refractory to diuretics, only 25% of patients survive beyond 1 year.20 A prospective study of more than 200 Italian patients with HCV-related compensated cirrhosis followed for up to 17 years found that HCC was the most common complication of cirrhosis detected, occurring in 32%, followed by ascites. These findings imply that HCC, rather than complications of portal hypertension, is the most frequent reason for a worsening prognosis in a patient with cirrhosis.21 The development of predictive models based on the natural history of PBC (see Chapter 89) and PSC (see Chapter 68)22 has helped clinical decision making for patients with these cholestatic disorders, which tend to progress in a fairly stereotypical fashion. Before the introduction of the MELD score, analogous models had not been available for the noncholestatic forms of cirrhosis, and the decision to refer a patient for liver transplantation generally was based on an estimate of disease severity using objective parameters such as the serum albumin level as well as more subjective factors such as the presence of hepatic encephalopathy, as in the Child-Turcotte-Pugh score (see Chapter 90). On clinical rather than biochemical grounds, important indications for liver transplantation remain disease severity reflective of hepatocellular failure, reflected by coagu­ lopathy and jaundice; complications of portal hypertension, such as refractory ascites and recurrent variceal bleeding; or the combination of portosystemic shunting and diminished hepatocellular function, as in hepatic encephalopathy (see Table 95-1). Following validation of predictive models for the natural history of PBC and PSC, prediction of an individual patient’s course has been possible on the basis of simple clinical and laboratory parameters, the most ominous of which is a rising serum bilirubin level. Although

deterioration in a patient’s quality of life typically may not be reflected adequately in predictive models, including MELD, the presence of potentially disabling symptoms such as pruritus and osteopenia in patients with cholestatic and other forms of cirrhosis, as well as recurrent bacterial cholangitis in those with PSC, are important considerations in deciding when to refer a patient for liver transplantation. MELD exceptions, such as adding points to the so-called biological MELD score in order to increase the likelihood of liver transplantation, require approval by a local regional review board, which includes representatives from local transplantation programs. The awarding of extra MELD points recognizes that although this system is a major advance in organ allocation, at least some patients may be disadvantaged by its use of purely objective parameters and exclusion of factors such as intractable ascites or encephalopathy that were incorporated into older allocation schemes. Ideally, liver transplantation should occur before a protracted period of disability reduces the likelihood that post-transplantation rehabilitation will lead to full employment and normal social functioning.

LISTING CRITERIA AND POLICIES OF THE UNITED NETWORK FOR ORGAN SHARING Organ allocation within the United States is administered by the United Network for Organ Sharing (UNOS), which now considers only disease severity (and not waiting time, as in the past) to determine a patient’s priority for liver transplantation. A variety of organ allocation systems have been used since the 1980s to allocate the limited number of deceased-donor organs in an equitable manner. Older systems were based on a combination of clinical and biochemical parameters such as the Child-Turcotte-Pugh score. The MELD score is a mathematical formula (available at www.unos.org) that incorporates the serum bilirubin level, creatinine level, and international normalized ratio (INR) and provides a more objective and accurate way to stratify liver transplantation candidates for organ allocation and to eliminate time waiting for a donor organ as a determining factor.23 The MELD score overcomes some of the inherent limitations of the Child-Turcotte-Pugh score, including limited discriminatory ability, subjective interpretation of parameters such as presence or absence of ascites on the basis of the physical examination, and the “ceiling effect” of the Child-Turcotte-Pugh score (e.g., no greater weight is given to a serum bilirubin level of 35 mg/dL than to a level of 3.5 mg/dL, even though a patient with the markedly higher bilirubin level clearly has more advanced liver disease). Inclusion of the serum creatinine level reflects the major prognostic importance of renal dysfunction in patients with advanced liver disease. Adoption of the MELD score has been a major step in achieving an equitable organ allocation system in the United States, although it will undoubtedly continue to undergo refinement.

ABSOLUTE AND RELATIVE CONTRAINDICATIONS Contraindications to liver transplantation have also evolved, reflecting innovations in surgical technique and postoperative care. As an example of the latter, effective prophylaxis against HBV recurrence now allows excellent graft and patient survival rates in patients with chronic hepatitis B.24 On the other hand, greater experience has highlighted the

1595

1596

Section IX  Liver Table 95-2  Absolute Contraindications to Liver Transplantation Acquired immunodeficiency syndrome Active alcoholism or substance abuse Advanced cardiac or pulmonary disease Anatomic abnormality that precludes liver transplantation Child-Turcotte-Pugh score 50 mm Hg or CPP 200 mm Hg).38 In the majority of patients with HPS, hypoxemia resolves within several months of liver transplantation, although a protracted period of ventilatory support in the immediate postoperative period may be required. Because of the potential for improvement with liver transplantation, extra MELD points may be allocated to a patient with HPS. HPS must be distinguished from portopulmonary hypertension, because the latter is associated with high perioperative mortality and frequently unchanged pulmonary hemodynamics despite liver transplantation. Specifically, documentation of a mean pulmonary arterial pressure greater than 35 mm Hg, pulmonary vascular resistance greater than 300 dynes • s • cm−5, and cardiac output less than 8 L/minute are indicative of a high perioperative risk because the patient will be unable to increase the cardiac output appropriately in response to altered intra- and postoperative hemodynamics. Vasodilator therapy may reduce pulmonary arterial pressure and permit liver transplantation.39 Hepatic hydrothorax is a frequent manifestation of portal hypertension characterized by collection of a transudate in the pleural cavity, usually on the right side and often with relatively little ascites remaining in the abdominal cavity (see Chapter 91). Hepatic hydrothorax can be particularly difficult to manage and often requires repeated tho­ racentesis or placement of a TIPS pre–liver transplantation.40 The temptation to insert a permanent chest tube needs to be resisted because a chest tube can lead to infection in the pleural cavity with the risk of fistula formation. Similarly, interventions such as pleurodesis or pleural decortication should be avoided because hepatic hydrothorax can be controlled only by a reduction in portal pressure. Active uncontrolled extrahepatic infection is an absolute contraindication to liver transplantation, which should be deferred if sepsis is suspected. In a patient with decompensated cirrhosis, an unexplained clinical deterioration such as altered mental status or systemic hypotension in the absence of gastrointestinal hemorrhage must be presumed to reflect sepsis and is an indication to start broad antibiotic coverage while culture results are awaited. Liver transplantation, however, may be the only option for a patient with recurrent bacterial cholangitis in the setting of PSC (see Chapter 68). Repeated bouts of spontaneous bacterial peritonitis need to be controlled by antibiotic therapy prior to attempting liver transplantation (see Chapter 91). A particularly ominous finding is fungemia, which is typically impossible to eradicate in a debilitated cirrhotic patient and precludes liver transplantation.

An important technical consideration in the liver transplantation candidate is the presence of vascular abnor­ malities that may increase the difficulty of surgery. With increased surgical experience, however, such abnormalities, most notably portal vein thrombosis, are less likely to be obstacles to liver transplantation. More extensive vascular thrombosis with involvement of the superior mesenteric vein may require extensive vascular reconstruction.41 The presence of a prior portosystemic shunt, particularly a nonselective (side-to-side or end-to-side) portocaval shunt, increases the technical complexity of liver transplantation but is no longer regarded as a contraindication. With the widespread use of TIPS to control manifestations of portal hypertension, including variceal hemorrhage, intractable ascites, and hydrothorax, without disrupting the vascular anatomy, TIPS is now the most frequent shunt encountered in liver transplant recipients and does not usually present an additional operative challenge during liver transplantation. Age restrictions have been relaxed in liver transplantation candidates, although close attention must be paid to comorbid conditions in older patients that not only increase the risk of perioperative mortality, but also may decrease the likelihood that the liver transplant recipient will be able to return to an active lifestyle, particularly because severe liver disease may cause more debility in older than in younger patients.42 Because a subset of robust older recipients have good outcomes after liver transplantation, candidates in their late 60s or even older who are otherwise in good health should not be precluded a priori from liver transplantation. The differential diagnosis of renal insufficiency in patients with advanced liver disease includes hepatorenal syndrome, which is potentially reversible. Renal failure remains an important predictor of a poor outcome post–liver transplantation (see Chapter 92).43 Typically, renal dysfunction in patients with decompensated cirrhosis reflects a variety of insults, including sepsis, hypotension, and use of nephrotoxic medications. In liver transplant recipients with decompensated cirrhosis, renal insufficiency severe enough to require dialysis has been associated consistently with poorer patient outcomes. Assessment of the potential for improvement following liver transplantation is critical. A rule of thumb is that return of adequate renal function is unlikely after liver transplantation if dialysis has been required for more than one month prior to liver transplantation. Inclusion of the serum creatinine level in the MELD score reflects the major prognostic importance of renal insufficiency in patients with advanced liver disease. A consequence of this recognition has been an increased rate of combined liver-kidney transplantation in the MELD era of liver transplantation, with a consequent depletion in the supply of kidneys for patients awaiting isolated deceaseddonor renal transplantation.44 An important reflection of impaired free-water handling in patients with decompensated cirrhosis is dilutional hyponatremia. Consequences of hyponatremia include altered mental status with more profound degrees of hyponatremia and an increased risk of calcineurin-induced neurotoxicity after liver transplantation (see later). Incorporation of the serum sodium level into the MELD formula (MELDNa) may increase the prognostic accuracy of the MELD score, particularly in patients with relatively low MELD scores.45 A major systemic manifestation of decompensated cirrhosis is malnutrition. Loss of muscle mass increases the likelihood of perioperative morbidity, with a need for more protracted ventilator support and poorer patient survival. Peripheral edema and ascites make changes in body weight

1597

1598

Section IX  Liver or anthropometric measurements unreliable for assessing nutritional status in a patient with advanced cirrhosis. More profound nutritional deficiencies may reflect the specific cause of cirrhosis, as in a malnourished alcoholic person with multiple vitamin and electrolyte deficiencies or a patient with cholestatic liver disease and depletion of fatsoluble vitamins as a result of small intestinal malabsorption. Evaluation by a dietitian is an integral part of the assessment of the liver transplantation candidate. Attempts to improve the nutritional status of liver transplantation candidates have included enteral and parenteral feeding, which may result in improvement, albeit modest, in some patients.46 An increasingly obese liver transplantation candidate pool is raising concerns about the role of obesity in the pathogenesis of NAFLD and in postoperative mortality resulting from cardiovascular events, as well as postoperative complications such as wound infections.47

TRANSPLANTATION EVALUATION AND LISTING The details of the formal liver transplantation evaluation vary from center to center, but the essentials are to establish that liver transplantation is indicated in the management of the potential recipient’s liver disease, the patient has no comorbidities severe enough to preclude transplantation, and the patient has adequate emotional and social resources to undergo a major surgical procedure and continue on long-term immunosuppression afterward (Table 95-3). Generally in the United States, clearance is needed from the patient’s insurance carrier before the extensive testing necessary for liver transplantation evaluation is undertaken. The patient is seen by a transplantation surgeon, hepatologist, psychiatrist, dietitian, and social worker, with additional consultations as clinically indicated. As increasingly frailer and older candidates are evaluated, identifying potential causes of perioperative morbidity, such as carotid artery stenosis, is imperative. Detailed abdominal imaging is performed not only to screen for HCC, but also to uncover vascular abnormalities such as portal vein thrombosis that may make surgery technically challenging. Disease-specific issues need to be addressed, such as the likelihood of recidivism in the alcoholic patient or management of a large

tumor burden in the patient with HCC. The appropriateness of liver transplantation for each candidate is then discussed formally at a meeting of the patient selection committee with input from the members of the transplantation team. If the patient’s candidacy is deemed to be appropriate, formal listing is undertaken with UNOS by matching of the recipient by blood type and weight with potential deceased donors. Once listed, the patient’s priority for organ allocation is determined by the MELD score, based either on the pure score consisting of the serum bilirubin level, serum creatinine level, and INR or that score plus additional points awarded for specific indications, such as HCC. Waiting time on the list is no longer a determining factor. With the critical and seemingly intractable shortage of deceaseddonor organs, a major challenge for UNOS and organ retrieval agencies elsewhere in the world has been to develop an equitable system of allocation in an effort to ensure that hepatic allografts are not used for patients whose prognosis without liver transplantation remains good. Patients with a MELD score of less than 15 appear have a better survival without transplantation than with transplantation. As shown in Figure 95-3, the MELD score has been shown to correlate with the three-month survival rate. Patients with a MELD score of less than 10 are ineligible for active listing with UNOS unless they are eligible to receive extra points because of an additional complication of liver disease, such as HCC or HPS. Once the evaluation process is complete and the patient is accepted for liver transplantation, financial clearance is sought from the patient’s private, state, or federal insurer to fund the procedure. Unfortunately, criteria for coverage for liver transplantation vary among insurers; however, in the United States, if Medicare, the major federal payor, funds a particular indication for liver transplantation, the other insurance carriers generally follow suit.

DISEASE-SPECIFIC INDICATIONS ALCOHOLIC LIVER DISEASE

Despite the high frequency of chronic HCV infection as an indication for liver transplantation (see later), alcoholic liver disease (ALD) remains the most frequent cause of

Table 95-3  Transplantation Evaluation Process Financial screening Medical evaluation Hepatology evaluation Laboratory testing Cardiac evaluation Hepatic imaging General health assessment Transplantation surgery evaluation Anesthesia evaluation Psychiatry or psychology consultation Social work evaluation Financial and insurance counseling Nutritional evaluation

Secure approval for evaluation As discussed in text Confirm diagnosis and optimize management Assess hepatic synthetic function, serum electrolytes, renal function, viral serologies, markers of other causes of liver disease, tumor markers, ABO-Rh blood typing; 24-hour urine for creatinine clearance; urinalysis and urine drug screen Electrocardiography and two-dimensional echocardiography, stress testing and cardiology consult if risk factors are present and/or patient is age 40 years or older Ultrasonography with Doppler to document portal vein patency, triple-phase computed tomography or gadolinium magnetic resonance imaging for tumor screening Chest film, prostate specific antigen level (men), Pap smear and mammogram (women), colonoscopy if patient is age 50 years or older or has primary sclerosing cholangitis Assess technical issues and discuss risks of procedure Required if unusually high operative risk, e.g., patient has portopulmonary hypertension, hypertrophic obstructive cardiomyopathy, previous anesthesia complications If history of substance abuse, psychiatric illness, or adjustment difficulties Address potential psychosocial issues and possible effect of transplantation on patient’s personal and social system Itemize costs of transplantation and post-transplantation care; help develop financial management plans Assess nutritional status and patient education

From O’Leary JG, Lepe R, Davis GL. Indications for liver transplantation. Gastroenterology 2008; 134:1764-76, with permission.

Chapter 95  Liver Transplantation pathologic drinking after liver transplantation appear to have a higher rate of medical problems, including pneumonia, cellulitis, and pancreatitis, that require hospital admission and occasionally lead to graft loss and death. In addition, alcoholic liver transplant recipients are prone to develop de novo oropharyngeal and lung tumors, likely reflecting other aspects of an alcoholic lifestyle, most notably cigarette smoking.48

100

3-month survival (%)

80

60

HEPATITIS B

40

20

0 0

20

40 50 30 MELD score Figure 95-3.  Relationship between the three-month survival rate and Model for End-stage Liver Disease (MELD) score in patients with cirrhosis. 10

decompensated chronic liver disease (see Chapter 84). Decompensated alcoholic cirrhosis is now firmly established as a legitimate indication for liver transplantation despite some lingering controversy.48 Concerns in the past included the risk of recidivism following liver transplantation as well as potentially poor compliance by transplant recipients with a history of alcoholism. In addition, the large number of patients with ALD was thought to have the potential to outstrip the donor supply. These fears have not been confirmed, and even patients with evidence of acute alcoholic hepatitis in the explant do not appear to have inferior post–liver transplantation survival rates despite some earlier reports to the contrary. Excellent graft and patient survival rates are the norm following liver transplantation for ALD. Key factors in determining an alcoholic patient’s suitability for liver transplantation include recognition by the patient of the key role alcohol has played in the genesis of the liver disease, participation in some form of alcohol rehabilitation, such as attendance at Alcoholics Anonymous, stable social support, and a defined period of abstinence from alcohol before transplantation. Conventionally this period of abstinence has been six months, although rigorous studies have failed to confirm that this duration of abstinence per se confers a high likelihood of continued sobriety but have emphasized the importance of factors such as a lack of either isolation or depression. Despite these strategies, however, a substantial proportion of alcoholic recipients resume drinking after liver transplantation, although surprisingly, graft loss or early death attributable to alcohol abuse has been uncommon. A higher rate of return to alcohol use is elicited by use of anonymous questionnaires or toxicology screening than by direct questioning of patients. With longer-term follow-up, as many as 40% of alcoholic recipients resume alcohol use.49 A particularly difficult dilemma arises in the alcoholic patient with severely decompensated liver disease and alcohol use until the time of hepatic decompensation in whom the likelihood of surviving without prompt liver transplantation is low. Clearly enunciated criteria, including a contractual commitment by the patient to sobriety and active involvement in some form of alcohol rehabilitation such as participation in Alcoholics Anonymous, ensure that the selection process is equitable under these circumstances. Patients who return to

Effective prevention of graft reinfection in the HBV-infected candidate has been a major triumph of liver transplantation. HBV recurrence was frequent and resulted in reduced patient and graft survival rates during the 1980s; as a result, Medicare refused to fund liver transplantation in HBVinfected persons. Key factors in improving outcomes included recognition of the key role of pre–liver transplantation active viral replication, as demonstrated by detection in serum of hepatitis B e antigen or HBV deoxyribonucleic acid (DNA) by molecular hybridization techniques as a predictor of recurrent HBV infection in the graft and the protective effect of long-term use of high-dose hepatitis B immunoglobulin (HBIG). In a seminal study50 Samuel and colleagues observed that patients with a fulminant presentation of acute HBV infection or hepatitis D virus coinfection had a reduced risk of recurrent hepatitis B post–liver transplantation in the absence of immunoprophylaxis, reflecting the lower level of HBV replication in patients with fulminant hepatitis B than in those with chronic hepatitis B. Long-term administration of high-dose HBIG, initially administered intravenously, resulted in markedly reduced rates of recurrence of hepatitis B post-transplantation; HBIG administered in combination initially with the nucleoside analog lamivudine further decreased the rate of HBV recurrence. Lamivudine had been used as monotherapy to prevent recurrent HBV infection post-transplantation but was limited by frequent mutations in HBV polymerase gene, leading in turn to graft reinfection.51 The optimal dosing regimen for HBIG has been difficult to establish in the absence of controlled clinical data. Some groups have titrated HBIG doses according to trough serum levels of antibody to hepatitis B surface antigen (anti-HBs). Sub­ sequently, intramuscular administration of HBIG has been confirmed as an efficacious, and less expensive, alternative to intravenous regimens, when used in combination with lamivudine.52 HBV infection that recurs despite administration of HBIG may reflect inadequate doses of HBIG or a genomic mutation in the “a” moiety of HBsAg that results in less avid binding of the virus to HBIG.53 Lamivudine resistance, acquired before liver transplantation, has also been implicated in HBV recurrence despite apparently adequate immunoprophylaxis after liver transplantation. The availability of additional oral antiviral agents with efficacy against HBV has expanded options for preventing graft reinfection (see Chapter 78). A large, multicenter study, for example, has demonstrated that adefovir dipivoxil taken for 48 weeks results in significant virologic, biochemical, and clinical improvement in patients with chronic HBV infection both pre– and post–liver transplantation and may obviate the need of liver transplantation because of improved hepatocellular function.54 Despite the complexities of managing HBV infection in the liver transplantation patient, excellent graft and patient survival rates are now routine, in contrast to the gloomy picture for HBV-infected liver transplantation candidates prior to the use of HBIG. The ever-expanding list of oral antiviral agents promises to improve outcomes further, even in patients in whom antiviral resistance develops. With the licensing of several

1599

1600

Section IX  Liver effective oral agents to treat HBV, HBIG use will likely be superseded by the use of combinations of oral antiviral agents. The liver transplantation candidate with chronic HBV infection is now readily accepted for liver transplantation, albeit with the requirement for long-term antiviral immunoprophylaxis.

HEPATITIS C

A major concern in the liver transplantation community, apart from the shortage of donor organs, is the burden of recurrent HCV infection after liver transplantation. After HCV was initially identified, early reports suggested that recurrent HCV infection with detection of viremia post– liver transplantation was frequent and did not have an adverse effect on overall patient or graft survival rates in the first several years after liver transplantation. Extensive experience now indicates, however, that HCV-infected liver transplant recipients have inferior long-term outcomes, compared with patients transplanted for other causes of cirrhosis, because of graft injury from recurrent HCV infection. Analysis of serial liver biopsy specimens from liver transplant recipients with recurrent HCV infection has identified accelerated fibrosis and progression to cirrhosis compared with immunocompetent patients with HCV infection. A particular challenge is to identify liver transplant recipients with recurrent HCV infection at increased risk of rapidly progressive graft injury. Less than 10% of patients with mild recurrent hepatitis C at one year post–liver transplantation appear to progress to allograft cirrhosis by five years. By contrast, two thirds of patients with at least moderately severe hepatitis C at one year post-transplantation progress to cirrhosis by 5 years.55 Concern has been raised, however, that with prolonged follow-up, some patients with initially mild recurrent hepatitis C after liver transplantation may experience a more aggressive course. A study by Berenguer and colleagues56 evaluated serial protocol liver biopsy specimens to assess the histologic outcome of 57 HCV genotype 1b-infected liver transplant recipients with an initially mild recurrence, defined as no or minimal fibrosis (fibrosis stage F0 or F1) during the first three years posttransplantation (see Chapter 79).56 Progression to bridging fibrosis or cirrhosis (stage F3 or F4) occurred in 35% (n = 20) of such patients, with 12 recipients progressing to stage F3 and 8 recipients progressing to stage F4. Variables associated with progressive fibrosis on univariate analysis were the baseline fibrosis stage and activity grade (P < 0.0001), recipient female gender (P = 0.04), the serum alanine aminotransferase (ALT) level at one year post-transplantation (P = 0.02), and the aspartate aminotransferase (AST) and ALT levels at baseline (P = 0.008 and P = 0.005, respectively). By multivariate analysis, the only variable that was significant was fibrosis stage at baseline (relative risk, 11; 95% CI, 3 to 41; P = 0.0007). Therefore, delayed HCVrelated severe liver damage is frequent, even in transplant recipients with an initially mild recurrence, and is seen in approximately one third of patients. Some degree of fibrosis at baseline appears to predict accelerated recurrent HCV infection. A particularly ominous finding is prominent biochemical and histologic cholestasis that frequently is a precursor to rapid allograft failure. Antiviral therapy for patients with a cholestatic HCV recurrence may need to be continued indefinitely because discontinuation leads to a rapid relapse of the cholestatic syndrome and death.3 Reliable predictors of severe recurrent HCV infection have been difficult to identify, although a number of viral and nonviral factors have been implicated (Table 95-4). Infection with viral genotype 1b had been suggested to be a

Table 95-4  Factors Associated with Severe Hepatitis C Virus Recurrence Following Liver Transplantation Viral Factors Absence of pretransplantation hepatitis B viral coinfection Cytomegalovirus coinfection High serum HCV RNA levels before transplantation and within two weeks after transplantation Viral genotype 1b Immunosuppression Multiple episodes of rejection (indicating a high cumulative prednisone dose) Use of OKT3 to treat rejection Other Factors High tumor necrosis factor-α production in the graft Impaired HCV-specific CD4+ T-cell responses Ischemic-preservation injury Nonwhite recipient HCV, hepatitis C virus; RNA, ribonucleic acid.

key predictor, but this observation has not been universal. Higher serum levels of HCV RNA before liver transplantation and immediately after liver transplantation and possibly more rapid evolution of HCV quasispecies have been described in patients with more aggressive recurrent HCV56 (see also Chapter 79). Older deceased-donor age has been implicated repeatedly. Episodes of acute cellular rejection, particularly if multiple, lead to a greater likelihood of severe recurrent hepatitis C. A major challenge is to distinguish recurrent HCV infection from graft rejection, particularly because many of the histologic hallmarks of acute rejection, including bile duct injury, are also consistent with recurrent HCV infection. Examination of serial liver biopsy specimens may help clarify this issue and allow avoidance of inappropriate additional immunosuppression in the recipient with recurrent HCV infection rather than graft rejection. Once recurrent HCV infection in the graft progresses to cirrhosis, overt hepatic decompensation is frequent. In contrast to recurrent HBV infection, effective prophylaxis against recurrent HCV infection has not been possible (Fig. 95-4).3 Current treatment strategies generally fall into three categories: (1) pretransplantation antiviral therapy; (2) preemptive therapy started in the early post–liver transplantation period before the development of clinically apparent acute hepatitis C; and (3) post-transplantation therapy at the time of diagnosis of acute hepatitis C or for established or severe chronic hepatitis C.3 The approach followed by most transplantation centers is to initiate antiviral therapy when clinically significant evidence of recurrent HCV infection is identified (as defined by either grade 3 or 4 (of 4) hepatic inflammation or a fibrosis stage of F2 or more). Accurate assessment of the efficacy of treating recurrent HCV infection is difficult, however, because most reported studies have been uncontrolled, single-centered, and small in size. Interferon alpha monotherapy generally has been ineffective in treating established recurrent hepatitis C (see Chapter 79). When interferon is combined with ribavirin, the rate of virologic response is increased, but at the cost of frequent and potentially severe side effects.3 Leukopenia is a par­ ticularly vexing problem in patients undergoing such treatment, and adjunctive administration of granulocyte colony– stimulating factor (G-CSF) may permit continuation of interferon therapy. The long-acting pegylated interferons have been evaluated in a number of clinical trials. Generally, prophylactic interferon-based regimens administered

Chapter 95  Liver Transplantation .8

Table 95-5  Criteria for Liver Transplantation in Acute Liver Failure +++

Probability (%)

.6 +++

++

+ +

.4 +++

.2

+

++

++ 0.0 0

90 180 270 360 450 540 630 720 810 900 9901080 Follow-up in days

Figure 95-4.  Probability of hepatic decompensation in patients with cirrhosis resulting from recurrent hepatitis C following liver transplantation. (From Berenguer M, Prieto M, Rayon JM, et al. Natural history of clinically compensated hepatitis C virus-related graft cirrhosis after liver transplantation. Hepatology 2000; 32:852-8, with permission.)

shortly post–liver transplantation in an effort to prevent graft reinfection have resulted in low rates of sustained virologic response.57 On the basis of reports of rejection and graft loss in renal transplant recipients treated with interferon, as well as preliminary experience in liver transplant recipients, concern has been raised that therapy with interferon increases the risk of graft rejection (although as noted earlier the distinction between recurrent HCV infection and graft rejection is difficult even on histologic grounds).58 Increasingly, recurrent HCV infection is recognized as the cause of graft failure in transplant recipients, and the dilemma arises as to whether repeat liver transplantation in affected patients is justified.25 A subset of patients retransplanted for graft loss caused by recurrent hepatitis C have reasonable survival rates, if they do not have deep jaundice or renal failure at the time of retransplantation (see later).

ACUTE LIVER FAILURE

Acute liver failure (ALF) is an uncommon but important indication for liver transplantation because of its high mortality rate but excellent outcomes with prompt liver transplantation (unless major neurologic complications have occurred). ALF is defined as the onset of hepatic encephalopathy within 26 weeks of the initial recognition of acute liver disease and reflects a variety of causes (see Chapter 93). Despite an abrupt onset, antecedent chronic liver disease is absent, and hepatic recovery is possible. In the past, liver transplantation for ALF resulted in poorer patient survival rates than those for benchmark indications, such as PBC. Subsequent experience, however, has shown that excellent patient survival rates are possible if ALF is identified early in its course and the patient is listed for liver transplantation before irreversible complications, especially neurologic, have supervened.59 The absence of papilledema on funduscopy and of typical features of cerebral edema on computed tomography (CT) of the head do not preclude the presence of cerebral edema complicating worsening encephalopathy; therefore, direct

Criteria of King’s College, London Acetaminophen cases Arterial pH 6.5 and serum creatinine >3.4 mg/dL Non-acetaminophen cases INR >6.5, or Any three of the following: Age 40 years Duration of jaundice before encephalopathy >7 days Etiology: non-A, non-B hepatitis; halothane hepatitis; idiosyncratic drug reaction; indeterminate Serum bilirubin >17.6 mg/dL INR >3.5 (PT >50 seconds) Criteria of Hôpital Paul-Brousse, Villejuif Hepatic encephalopathy, and Factor V level 12 hr, serum bilirubin and creatinine levels Interval to retransplantation (better prognosis if within 30 days) Bilirubin and serum creatinine levels

UNOS database used; HCV positivity and donor age are significant predictors by univariate analysis83 King’s College84 University of Pittsburgh85 UCLA86 Mayo Clinic; limited to patients with PBC or PSC87 UNOS database used; limited to HCV-positive patients retransplanted for causes other than primary nonfunction88

HCV, hepatitis C virus; PBC, primary biliary cirrhosis; PSC, primary sclerosing cholangitis; UNOS, United Network for Organ Sharing. From Rosen HR. Disease recurrence following liver transplantation. Clin Liver Dis 2000; 4:675-89, with permission.

1611

1612

Section IX  Liver survival for an individual patient and do not take into account the patient’s quality of life, they can be used as adjuncts to clinical judgment. Application of retransplantation to low-risk patients is associated with survival com­ parable to that for primary liver transplantation; determining whether retransplantation is justified in patients with highrisk scores will require prospective studies. In the current era of extreme organ shortage, the utility of primary and repeat transplantation (i.e., duty to promote the best outcome in the aggregate) needs to be considered. An analysis by Burton and colleagues, using the Cobbs-Douglas equation, has demonstrated that for primary liver transplantation, maximal utility is achieved by allocating organs at the highest MELD score (i.e., “sickest first”). For retransplan­ tation, however, maximal utility for HCV and non-HCV diagnoses is achieved at MELD scores of 21 and 24, respectively. Utility starts to decline at MELD scores greater than 28.25 A multicenter U.S. study of 272 patients (73 retransplanted for causes not related to HCV) indicated that oneand three-year survival rates following retransplantation were equivalent in HCV-positive and HCV-negative patients. These data were likely shaped by selection bias, considering that nationally more than one third of patients with HCVrelated allograft failure are not even considered for retransplantation and only half of those reevaluated are even relisted. This study also showed that MELD scores greater than 30 are associated with particularly poor survival following retransplantation.99 Major challenges remain in liver transplantation, including the shortage of donor organs, threat of recurrent disease, and morbidity associated with lifelong therapeutic immunosuppression. Nevertheless, the availability of liver transplantation has transformed the lives of patients with advancing liver disease and their health care providers from an ultimately futile effort to manage the complications of cirrhosis into a life-prolonging and life-enhancing intervention.

KEY REFERENCES

Ahmed A, Keeffe EB. Current indications and contraindications for liver transplantation. Clin Liver Dis 2007; 11:227-47. (Ref 13.) Arjal RR, Burton JR Jr, Villamil F, et al. Review article: The treatment of hepatitis C virus recurrence after liver transplantation. Aliment Pharmacol Ther 2007; 26:127-40. (Ref 3.) Benten D, Sterneck M, Panse J, et al. Low recurrence of preexisting extrahepatic malignancies after liver transplantation. Liver Transpl 2008; 14:789-98. (Ref 29.) Cardenas A, Kelleher T, Chopra S. Review article: Hepatic hydrothorax. Aliment Pharmacol Ther 2004; 20:271-9. (Ref 40.) Fallon MB, Krowka MJ, Brown RS, et al. Impact of hepatopulmonary syndrome on quality of life and survival in liver transplant candidates. Gastroenterology 2008; 135:1168-75. (Ref 37.) Grewal P, Martin P. Pretransplant management of the cirrhotic patient. Clin Liver Dis 2007; 11:431-49. (Ref 14.) Ioannou GN, Perkins JD, Carithers RL Jr. Liver transplantation for hepatocellular carcinoma: Impact of the MELD allocation system and predictors of survival. Gastroenterology 2008; 134:1342-51. (Ref 61.) Kim WR, Biggins SW, Kremers WK, et al. Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med 2008; 359:1018-26. (Ref 45.) Lopez PM, Villanueva A, Roayaie S, et al. Neoadjuvant therapies for hepatocellular carcinoma before liver transplantation: A critical appraisal. Liver Transpl 2006; 12:1747-54. (Ref 17.) Lucey MR. Liver transplantation for alcoholic liver disease. Clin Liver Dis 2007; 11:283-9. (Ref 48.) McAvoy NC, Kochar N, McKillop G, et al. Prevalence of coronary artery calcification in patients undergoing assessment for orthotopic liver transplantation. Liver Transpl 2008; 14:1725-31. (Ref 32.) O’Leary JG, Lepe R, Davis GL. Indications for liver transplantation. Gastroenterology 2008; 134:1764-76. (Ref 1.) Oo YH, Neuberger J. Recurrence of nonviral diseases. Clin Liver Dis 2007; 11:377-95. (Ref 5.) Pham PT, Pham PC, Rastogi A, et al. Review article: Current management of renal dysfunction in the cirrhotic patient. Aliment Pharmacol Ther 2005; 21:949-61. (Ref 43.) Rosen HR. Transplantation immunology: What the clinician needs to know for immunotherapy. Gastroenterology 2008; 134:1789-801. (Ref 80.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

96 Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine Ellen Kahn and Fredric Daum

CHAPTER OUTLINE Anatomy  1615 Macroscopic Features  1615 Microscopic Features  1617 Embryology  1623 Molecular Regulation of Intestinal Morphogenesis  1623 Specific Structures and Systems  1624 Clinical Implications  1626 Abnormalities in Normal Embryologic Development  1626 Abdominal Wall  1626

ANATOMY MACROSCOPIC FEATURES Small Intestine

The small intestine is a specialized tubular structure within the abdominal cavity in continuity with the stomach proximally and the colon distally. The small bowel increases in length from about 250 cm in the full-term newborn to 600 to 800 cm in the adult. The duodenum, the most proximal portion of the small intestine, begins at the duodenal bulb, travels in the retroperitoneal space around the head of the pancreas, and ends on its return to the peritoneal cavity at the ligament of Treitz. The remainder of the small intestine is suspended within the peritoneal cavity by a thin broad-based mesentery that is attached to the posterior abdominal wall and allows relatively free movement of the small intestine within the abdominal cavity. The proximal 40% of the mobile small intestine is the jejunum, and the remaining 60% is the ileum. The jejunum occupies the left upper portion of the abdomen, and the ileum is positioned in the right abdomen and upper part of the pelvis. No distinct anatomic demarcation exists between jejunum and ileum. Visual examination of the luminal surface of the small intestine reveals mucosal folds, the plicae circulares. More numerous in the proximal jejunum, the plicae circulares

Meckel’s Diverticulum and Other Vitelline Duct Abnormalities  1628 Malrotations  1630 Proliferation  1631 Intestinal Atresia and Stenosis  1632 Anorectum  1633 Enteric Nervous System  1636 Miscellaneous and Genetic Defects  1640

decrease in number in the distal small bowel and are absent in the terminal ileum. Aggregates of lymphoid follicles are scattered throughout the small intestine but are found in highest concentration in the ileum, where they are designated Peyer’s patches. Peyer’s patches normally are more prominent during infancy and childhood than they are in adulthood. The small bowel transitions to the colon at the ileocecal valve, which consists of two semilunar lips that protrude into the cecum. The ileocecal valve provides a barrier to retrograde flow of colonic contents into the small intestine. This barrier appears to be a function of the angulation between the ileum and cecum and is maintained by the superior and inferior ileoceal ligaments; a true tonic sphincter-type pressure does not appear to be present in this region.

Colon and Rectum

The colon is a tubular structure approximately 30 to 40 cm in length at birth in the full-term infant. In the adult, the colon measures approximately 150 cm, about one quarter of the length of the small intestine. The diameter of the colon is greatest in the cecum (7.5 cm) and narrowest in the sigmoid (2.5 cm). The colon is continuous with the small intestine proximally at the ileocecal valve and ends distally at the anal verge (Fig. 96-1). The external appearance of the colon differs from that of the small intestine because the

1615

1616

Section X  Small and Large Intestine Hepatic flexure

Greater omentum (cut away) Epiploic taenia

Free taenia (taenia libera)

Transverse mesocolon Haustra

Hook exposing epiploic Appendices taenia epiploicae Semilunar folds (plicae semilunares)

Ileocecal valve Figure 96-1.  Macroscopic characteristics of the colon. Note the taeniae, haustra between the taeniae and appendices epiploicae on the outer surface, and the semilunar folds on the luminal side. (From Netter FH. The Netter Collection of Medical Illustration. vol 3. Teterboro, NJ: Icon Learning System; 2002.)

Greater omentum (cut away)

Free taenia (taenia libera) Sigmoid mesocolon

Splenic flexure

Hook exposing mesocolic taenia Haustra

Cecum Appendix

Rectosigmoid junction

longitudinal muscle fibers of the colon coalesce into three discrete bands called taeniae, located at 120-degree intervals about the colonic circumference: taenia liberis, taenia omentalis, and taenia mesocolica. The taeniae start at the base of the appendix and extend continuously to the proximal rectum. Outpouchings of the colon, the haustra, are found between the taeniae. Semilunar folds characterize the mucosa between haustra. Small sacs of peritoneum filled with adipose tissue, the appendices epiploicae, are found on the external surface of the colon. The most proximal portion of the colon, the cecum, lies in the right iliac fossa and projects downward as a blind pouch below the entrance of the ileum. The cecum is a sacculated structure 6 to 8 cm in length and breadth. Because of its large diameter, it is the part of the colon most apt to rupture with distal obstruction, and cecal tumors can grow to be quite large without producing symptoms of obstruction. The mobility of the cecum normally is fixed by a small mesocecum; an anomaly in fixation exists in 10% to 20% of people, especially women, predisposing them to cecal volvulus. The vermiform appendix is a blind outpouching of the ceum that begins inferior to the ileocecal valve. Appendiceal anatomy is discussed further in Chapter 116. The ascending colon extends from the cecum distally for 12 to 20 cm along the right side of the peritoneal cavity to the hepatic flexure. The ascending colon is covered with peritoneum and thus constitutes a retroperitoneal organ. At the hepatic flexure, the colon turns medially and anteriorly to emerge into the peritoneal cavity as the transverse colon. This is the longest portion (40 to 50 cm) and the most mobile segment of the colon and drapes itself across the anterior abdomen between the hepatic and splenic flexures. When a person assumes the upright posi-

tion, the transverse colon may actually dip down into the pelvis. The transverse colon may become fixed in this festooned position by adhesions, most commonly resulting from hysterectomy, potentially leading to a technically difficult colonoscopy. The descending colon, about 30 cm in length, travels posteriorly and then inferiorly in the retroperitoneal compartment to the pelvic brim. There, it emerges into the peritoneal cavity as the sigmoid colon. This is an S-shaped redundant segment of variable length, tortuosity, and mobility, which challenges the endoscopist and radiologist, and is susceptible to volvulus. Because the sigmoid is the narrowest part of the colon, tumors and strictures of this region typically cause obstructive symptoms early in the course of disease. The rectum, 10 to 12 cm in length in the adult, begins at the peritoneal reflexion and follows the curve of the sacrum, ending at the anal canal.

Anal Canal

The anal canal is approximately 5 cm in length in the adult and has discrete upper and lower demarcations. The anorectal ring is located proximally and is composed of the upper portion of the internal sphincter, the longitudinal muscle of the rectum, the deep portion of the external sphincter, and the puborectalis portion of the levator ani muscle; distally, the anal verge represents the transition of anoderm to true skin. The mucosa of the distal 3 cm of the rectum and the anal canal contains 6 to 12 redundant longitudinal folds called the columns of Morgagni, which terminate in the anal papillae. These columns are joined together by mucosal folds called the anal valves, which are situated at the dentate line. The muscularis mucosae disap-

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine pears in the anorectal canal, and the inner circular coat of muscularis propria thickens to form the internal anal sphincter. The external anal sphincter surrounds the anal canal, and its fibers blend with those of the levator ani muscle to attach posteriorly to the coccyx and anteriorly to the perineal body. The anatomy and function of these muscles are described in more detail in Chapter 125.

mp m

s

Vasculature

The superior mesenteric artery delivers oxygenated blood to the distal duodenum, the jejunum and ileum, the ascending colon, and the proximal two thirds of the transverse colon. The remainder of the colon is supplied by branches of the inferior mesenteric artery. The arterial supply of the anal area is from the superior, middle, and inferior hemorrhoidal arteries, which are branches of the inferior mesenteric, hypogastric, and internal pudendal arteries respectively. Venous drainage of the anus is by both the systemic and portal systems. The internal hemorrhoidal plexus drains into the superior rectal veins and then into the inferior mesenteric vein, which, with the superior mesenteric vein, joins the splenic vein to form the portal vein. The distal anus drains by the external hemorrhoidal plexus through the middle rectal and pudendal veins into the internal iliac vein. (See Chapter 114 for additional discussion of the intestinal blood supply.)

mm

sm

Figure 96-2.  Photomicrograph of the small intestine showing its general microscopic architecture. m, mucosa; mm, muscularis mucosae; mp, muscularis propria; s, serosa; sm, submucosa. (Hematoxylin and eosin, ×25.)

Lymphatic Drainage

The lymphatic drainage of both the small intestine and colon follows their respective blood supplies to lymph nodes in the celiac, superior preaortic, and inferior preaortic regions. Lymphatic drainage proceeds to the cisterna chyli and then via the thoracic duct into the left subclavian vein. Proximal to the dentate line, lymphatic drainage is to the inferior mesenteric and periaortic nodes, whereas distal to the dentate line, lymphatic drainage is to the inguinal lymph nodes. Therefore, inguinal lymphadenopathy can be seen with inflammatory and malignant disease of the lower anal canal.

ge

lp

Extrinsic Innervation

The autonomic nervous system—sympathetic, parasympathetic, and enteric—innervates the gastrointestinal tract. The sympathetic and parasympathetic nerves constitute the extrinsic nerve supply and connect with the intrinsic nerve supply, which is composed of ganglion cells and nerve fibers within the intestinal wall. Innervation of the small intestine and colon is discussed in detail in Chapters 97 and 98, respectively.

A

mv

MICROSCOPIC FEATURES General Considerations

The small and large intestine share certain histologic characteristics. The wall of the small intestine and colon is composed of four layers: mucosa (or mucous membrane), submucosa, muscularis (or muscularis propria), and adventitia (or serosa) (Fig. 96-2). Mucosa The mucosa is the innermost layer formed by glandular epithelium, lamina propria, and muscularis mucosae (Fig. 96-3A and B). The glandular epithelium forms cylindrical structures called crypts. The lamina propria, which supports the epithelium, is a layer of reticular connective tissue with elastin, reticulin, and collagen fibers, lymphocytes, plasma cells, and eosinophilic granulocytes, as well as lymphatics and capillaries. The muscularis mucosae consists of

B

mm

C

Figure 96-3.  Histologic and electron microscopic photographs of the small intestine. A, Components of the mucosa: ge, glandular epithelium; lp, lamina propria. Note the absorptive cells that appear as high columnar cells with eosinophilic cytoplasm (arrow). (Hematoxylin and eosin, ×250.) B, Goblet cells (arrow) and brush border are stained red. mm, muscularis mucosae. (Periodic acid–Schiff stain, ×150.) C, Microvilli (mv) are seen as delicate finger-like projections on electron microscopic examination, ×9000. (C, Courtesy of S. Teichberg, PhD.)

1617

1618

Section X  Small and Large Intestine a thin layer of smooth muscle at the boundary of the mucosa and submucosa. The glandular epithelium is composed of various cell types: stem cells, undifferentiated crypt cells, absorptive cells (also called columnar cells), secretory cells (goblet cells, Paneth cells, enteroendocrine cells), and M cells. Signaling pathways such as Wnt, bone marrow protein (Bmp), PTEN/PI3K, Notch, hedgehog, platelet-derived growth factor, and SOX9 play important roles in the development of the intestinal epithelium.1-4 Wnt signaling plays a role in promoting cell proliferation; maintains stem cells in an undifferentiated state; defines compartmentalization into Paneth cells, proliferative, and differentiation zones along the crypt-villus axis; and directs early secretory lineage development as well as terminal differentiation of Paneth cells through the transcription factor SOX9.1 Bmps belong to the transforming growth factor-β family. Bmp signaling is important in intestinal development and homeostasis. It antagonizes crypt formation and stem cell self-renewal and has a role in directing maturation of all three secretory cell types (goblet, enteroendocrine, and Paneth). Bmp signaling in the mesenchyme plays a signi­ ficant role in crypt morphogenesis; loss of Bmp leads to multiplication and elongation of crypts.2 PTEN/PI3K pathway plays a role in cell survival, proliferation, and growth.1 Notch proteins mediate cell fate decisions and pattern by regulating the helix-loop-helix factor that controls terminal differentiation. Notch directs development of absorptive cells and depletion of secretory lineage cells, and increases proliferation.1 The hedgehog (Hh) signaling pathway is important in crypt and villus morphogenesis and maintenance of stem cells.3 Both Sonic (Shh) and Indian (Ihh) play a role. Ihh is critical for the maintenance of intestinal stem cells, whereas Shh inhibits the growth of the villi. The contractile subepithelial pericryptal myofibroblasts represent a major target for Hh signaling. Hh signals sent to the epithelium-associated subepithelial myofibroblasts localize the precrypt structure and maintain the organization of the crypt-villus axis. Hh signaling also inhibits the proliferation or differen­tiation of smooth muscle and the proliferation compartment of the intestinal epithelium.3 Platelet-derived growth factor A stimulates mesenchymal condensation, proliferation, and evagination of overlying epithelium to form villi.3 Studies in animals also have contributed to the understanding of the molecular mechanism of the different pathways.1,2 Stem cells are pluripotential cells located at the base of the intestinal crypts. Stem cells give rise to all types of mature intestinal epithelial cells and at the same time replenish themselves through self-renewal. Undifferen­ tiated cells have fewer intracellular organelles and microvilli than do absorptive cells. The absorptive cells (see Fig. 96-3A) are high columnar cells with oval basal nuclei, eosinophilic cytoplasm, and a periodic acid–Schiff (PAS)– positive free surface, the brush border (see Fig. 96-3B). On electron microscopic examination, the brush border is seen to be composed of microvilli (see Fig. 96-3C), which are more numerous in the small intestinal than in the colonic epithelium. Small bowel enterocyte microvilli are estimated to increase the luminal surface area of the cell 14- to 40-fold. Goblet cells are oval or round, with flattened basal nuclei (Fig. 96-4A); their cytoplasm is basophilic, metachromatic (see Fig. 96-4B), and PAS positive (see Fig. 96-4C). Paneth cells are flask shaped and have an eosinophilic granular

A

B

C Figure 96-4.  Photomicrographs of the large and small intestine demonstrating goblet cells. A, Clear, empty-looking cytoplasm (arrow) and basal nuclei are seen with use of hematoxylin and eosin, ×250. B, Metachromatic staining of the cytoplasm results with use of the alcian blue stain, ×150. C, The cells demonstrate red staining with use of periodic acid–Schiff stain, ×150.

cytoplasm and a broad base positioned against the basement membrane (Fig. 96-5). Paneth cells contain zinc, antimicrobial peptides, and growth factors and secrete lysoenzymes. Enteric antimicrobial peptides produced by Paneth cells protect against intestinal infection and maintain enteric homeostasis.5 A cathelin-related antimicrobial peptide (CRAMP) identified in neonatal epithelium during the first weeks after birth, confers protection from Listeria monocytogenes.5 The mucosa also contains specialized cells that because of their specific endocrine function are called enteroendocrine or neuroendocrine cells. These neuroendocrine cells historically have been divided into argentaffin cells (granules able to reduce silver nitrate) and argyrophilic cells (granules that reduce silver nitrate only in the presence of a chemical reducer). Argentaffin cells stain positive with bichromate salts and also are called enterochromaffin cells. These cells are oval or triangular

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine

lc

Figure 96-5.  Photomicrograph of the small intestinal mucosa demonstrating the crypts of Lieberkühn (lc) and Paneth cells (arrow), which are characterized by granular eosinophilic cytoplasm. (Hematoxylin and eosin, ×250.)

(also called “halo cells”) and have a basal position in relation to the remaining epithelial cells (Fig. 96-6A) and a pale cytoplasm filled with dark-stained granules. Variation in shapes and cell types has been detected with immunohistochemical staining. The unifying APUD concept— amine precursor, uptake, and decarboxylation—ascribes common characteristics to these neuroendocrine cells. APUD cells are a group of cells with a common embryonic neural crest origin and with similar cytochemical and electron microscopic features; however, embryologic and morphologic data support an endodermal origin of these cells. Ultrastructurally, enteroendocrine cells contain membrane-bound granules with variably sized electrodense cores (see Fig. 96-6B), averaging 100 to 250 nm in diameter, and consisting of large dense-core vesicles and smaller, synaptic-type microvesicles. Neurosecretory granules can be demonstrated with the Grimelius stain by light microscopy as dark granules (see Fig. 96-6C), or, more specifically, by immunofluorescence, and with immunohistochemical stains such as neuron-specific enolase, chromogranin, and synaptophysin. Chromogranin enables visualization of the large dense-core vesicles, and synaptophysin targets the small synaptic-like microvesicles (see Fig. 96-6D).6 Vesicular monoamine transporter 1 (VMAT1) and 2 (VMAT2) are two isoforms of the adenosine triphosphate (ATP)– dependent vesicular monoamine transporters. These antigens, derived from both the large and small dense-core vesicles, are expressed differentially in small dense-core vesicles. Both are expressed in neuroendocrine cells, but VMAT1 is restricted to serotonin-producing enterochromaffin cells, and VMAT2 is expressed in histamine-producing cells, enterochromaffin-like cells, and pancreatic islet cells.7 Specific immunohistochemical stains allow for identification of individual protein products of the neuroendocrine cells. Besides releasing hormones in the blood, neuroendocrine cells also regulate secretion, absorption, motility, mucosal cell proliferation, and possibly immunobarrier control.6 Electron microscopy and immunohistochemistry have led to the identification of a variety of cell types (Table 96-1). Designation according to the nature of the stored peptide is preferable to characterization of neuroendocrine cells by letters. Serotonin-producing enterochro-

maffin cells, vasoactive intestinal polypeptide (VIP), and somatostatin D cells are distributed throughout the small and large intestine. Gastrin-, ghrelin-, gastric inhibitory peptide (GIP)-, secretin-, and cholecystokinin-producing cells are found predominantly in the stomach and proximal small intestine; peptide YY-, glucagon-like peptide (GLP)-1-, GLP-2-, and neurotensin-secreting cells are found in the ileum.8 Neuroendocrine cells originate from a common precursor cell in the intestinal crypt. The earliest cell fate is regulated by the Notch signaling pathway (see earlier). Math1 is the first factor involved in endocrine specification, followed by neurogenin3.8 Pax4 and Pax6, paired ox homeodomain transcription factors, and Nkx2.2 also are required for neuroendocrine differentiation.8,9 As mature neuroendocrine cells migrate to the tip of the villi, they undergo apoptosis and are extruded into the lumen. M cells are specialized epithelial cells overlying lymphoid follicles in the small intestine and colon. M cells selectively bind, process, and deliver pathogens directly to lymphocytes, macrophages, or other components of the mucosal lymphoid system. Interstitial cells of Cajal (ICC) are present in both the small intestine and the colon and are mesenchymal cells, located in the myenteric plexus, the muscularis propria, and the submucosa (Fig. 96-7). The distribution of the ICC is similar in children and in adults although a difference in their distribution is seen in fetuses of different gestational ages.10 Recognized as the pacemaker cells of the intestine, the ICC regulate intestinal motility by generating slow waves and determining frequency of smooth muscle contraction; they also amplify the neuronal signals, mediate neurotransmission from enteric motor neurons to smooth muscle cells, and set the smooth muscle membrane potential gradient. The ICC are spindle shaped or stellate, with long ramified processes, and have large, oval light-staining nuclei with sparse perinuclear cytoplasm. The ICC express the receptor for tyrosine kinase (c-Kit) or CD117 which is necessary for their maintenance. Serotonin regulates the number of the ICC by increasing their proliferation.11 Immunohis­ tochemical stains that use antibodies against c-Kit allow the ICC to be labeled. The distribution and onset of appearance of these cells in the gastrointestinal tract have been described.10 Submucosa The submucosa, between the muscularis mucosae and the muscularis propria, is a fibrous connective tissue layer that contains fibroblasts, mast cells, blood and lymphatic vessels, and a nerve fiber plexus—Meissner’s plexus—comprised of nonmyelinated, postganglionic sympathetic fibers, and parasympathetic ganglion cells. Muscularis or Muscularis Propria The muscularis propria, mainly responsible for contractility, consists of two layers of smooth muscle: an inner circular coat and an outer longitudinal coat arranged in a helicoidal pattern. A prominent nerve fiber plexus called the myenteric plexus, or Auerbach’s plexus, is found between these two muscle layers (Fig. 96-8). Parasympathetic and postganglionic sympathetic fibers terminate in parasympathetic ganglion cells, and postganglionic parasympathetic fibers terminate in smooth muscle. Adventitia or Serosa The adventitia is the outermost layer of connective tissue. When covered by a single layer of mesothelial cells, it is called the serosa.

1619

1620

Section X  Small and Large Intestine

A

B

C

D

Figure 96-6.  Microscopic characteristics of neuroendocrine cells of the small intestine. A, Features include clear cytoplasm and a round nucleus (arrow). (Hematoxylin and eosin, ×250.) B, Neurosecretory granules are seen as electron-dense, round black bodies (arrow) on electron microscopic examination, ×20,000. C, Granules in neuroendocrine cells are stained black with the Grimelius stain (arrow), ×150. D, Cells stained with synaptophysin have brown cytoplasm (arrow), ×250. (B, Courtesy of S. Teichberg, PhD.)

Table 96-1  Enteroendocrine Cells of the Intestinal Tract: Cell Types and Products, Vesicle Markers, and Distribution Vesicle Markers CELL TYPE

CELL product

LDCV

P/D1 EC D L PP G CCK S GIP M N

Ghrelin 5-HT Somatostatin GLI/PYY PP Gastrin Cholecystokinin Secretin, 5-HT GIP/Xenin Motilin Neurotensin

CgA, VMAT2 CgA, VMAT1 CgA SgII > CgA CgA, SgII, VMAT2 CgA CgA CgA CgA

SLMV Syn Syn Syn Syn Syn

DUOD

JEJ

ILEUM

APP

Colon

Rec

f + + f e + + + + + f

f + + +

f + f +

+ f +

+ f +

+ f +

+ + + + +

f f f +

App, appendix; CgA, chromogranin A; DUOD, duodenum; e, presence of cells in fetus and newborn; EC, enterochromaffin cell, 5-HT, 5-hydroxytryptamine (serotonin); f, presence of few cells; GIP, gastric inhibitory polypeptide; GLI, glucagon-like immunoreactants (glicentin, glucagon-37, glucagon-29, GLP-1, GLP-2); JEJ, jejunum; LDCV, large dense-core vesicles; NESP55, neuroendocrine secretory protein 55; PP, pancreatic polypeptide; PYY, PP-like peptide with N-terminal tyrosine amide; REC, rectum; SgII, secretogranin II (also known as chromogranin C); SLMV, synaptic-like microvesicles; Syn, synaptophysin; VMAT1, VMAT2, vesicular monoamine transporter 1, 2; +, presence of cells; >, heavier staining than. Adapted from Solcia E, Capela C, Fiocca R, et al. Disorders of the endocrine system. In: Ming SC, Goldman H, editors. Pathology of the Gastrointestinal Tract. Philadelphia: Williams & Wilkins; 1998. p 295.

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine v

A

Figure 96-7.  Photomicrograph showing the interstitial cells of Cajal in the small intestine. Brown-staining, elongated cells are evident around the myenteric plexus (arrow). (CD117 immunostain, ×250.)

il

bg

B Figure 96-9.  Photomicrograph of the duodenal mucosa. A, Villi are seen as finger-like projections. B, Brunner glands (bg) are found below the mucosa. (Hematoxylin and eosin. A, ×250; B, ×150.)

mp

ol

Figure 96-8.  Photomicrograph of the muscularis propria of the small intestine. The myenteric plexus (mp) is seen as a pale area with ganglion cells between the inner and outer layers (il, ol) of the muscularis propria (arrow). (Hematoxylin and eosin, ×250.)

Small Intestine

The mucosa of the small intestine is characterized by mucosal folds (plicae circulares, or valves of Kerckring) and villi. The mucosal folds are composed of mucosa and submucosa. Villi are mucosal folds that decrease in size from the proximal to distal small intestine and are of different shapes in the various segments of the small intestine: they may be broad, short, or leaf-like in the duodenum; tonguelike in the jejunum; and finger-like more distally (Fig. 96-9A). The villous pattern also may vary in different ethnic groups. Thus, for example, biopsy specimens from Africans, Indians, South Vietnamese, and Haitians have shorter and thicker villi, an increased number of leaf-shaped villi, and more mononuclear cells in comparison with specimens from North Americans. Various methods have been suggested to determine normal villus height. The height of the normal villus is 0.5 to 1.5 mm; villus height should be more than one half of the total thickness of the mucosa, and three to five times the length of the crypts. Villi are lined by enterocytes, goblet cells, and enteroendocrine cells.

Intestinal villus morphogenesis begins when mesen­ chymal aggregates impinge on the basal aspect of the epithelium to produce primitive folds. By nine to 10 weeks of gestation, the pseudostratified squamous epithelium converts to a single layer of columnar cells that lines mesenchymal stalks or the lamina propria.12 During mid- to late gestation, the basic tissue architecture of the intestine is established through epithelial-mesenchymal interaction. Induced by signals from mesoderm-derived mesenchyme, the endoderm-derived epithelium evaginates to form villi and intervillus regions. The intervillus region eventually invaginates into the mucosa to form crypts.1 Contractile subepithelial pericryptal myofibroblasts contribute mechanically to crypt formation and are the major source of instructive signals to the epithelium.3 Two types of glands are present in the small intestine: Brunner’s glands and crypts of Lieberkühn (intestinal crypts). Brunner’s glands are submucosal glands (see Fig. 96-9B) found primarily in the first portion of the duodenum and in decreased numbers in the distal duodenum; their function is to secrete a bicarbonate-rich alkaline secretion that helps neutralize gastric chyme. In children these glands also may be present in the proximal jejunum. Brunner’s glands open into the intestinal crypts and morphologically resemble pyloric glands. Crypts of Lieberkühn are tubular glands that extend to the muscularis mucosae (see Fig. 96-5). The crypts are occupied mainly by undifferentiated cells and Paneth cells. Cells are generated at the crypt base and proceed to migrate toward the villus. During this migration, these cells mature and differentiate into a secretory lineage (goblet cells, enteroen-

1621

1622

Section X  Small and Large Intestine docrine cells, Paneth cells) and enterocytes. The commitment of the stem cells to differentiate is acquired in the upper third of the crypt, where cells lose their ability to divide. The constant renewal of enterocytes is regulated by human acyl-coenzyme A synthetase.13 Paneth and columnar cells predominate in the base of the crypt. Above the base are absorptive cells and oligomucin cells; the latter originate from undifferentiated cells and differentiate into goblet cells. Goblet cells predominate in the upper half of the crypt. Enteroendocrine cells are admixed with goblet cells. A certain number of CD3+ intraepithelial T lymphocytes (up to 30 per 100 epithelial cells) normally are present in the villi. Smooth muscle is found in the lamina propria of the small intestinal villus, extending vertically up from the muscularis mucosae. Plasma cells containing primarily immunoglobulin A (IgA), and mast cells also are present. Lymphoid tissue is prominent in the lamina propria as solitary nodules and as confluent masses—Peyer’s patches—and is seen in the submucosa. Peyer’s patches are distributed along the antimesenteric border and are most numerous in the terminal ileum; their numbers decrease with age. Most types of enteroendocrine cells are present in the duodenum. Cells that produce ghrelin, gastrin, cholecystokinin, motilin, neurotensin, GIP, and secretin are restricted to the small intestine.6 The proportions of these cells differ in the villi and crypts, as well as in different segments of the intestine. Ninety percent of the villus epithelial cells are absorptive cells intermingled with goblet and enteroendocrine cells. The proportion of goblet to absorptive cells is increased in the ileum. The ICC are more abundant in the myenteric plexus of the small bowel than in the colon.10

Colon

Colonic epithelial cells are generated from stem cells at the base of the crypts and migrate toward the intestinal lumen after three to five days, on initiation of apoptosis. Most epithelial cells undergo apoptosis when they lose contact with the extracellular matrix and are shed into the lumen through caspase activation. Caspase activation is responsible for the cleavage of essential intracellular proteins leading to apoptosis and therefore loss of anchorage.14 The mucosa of the large intestine is characterized by the presence of crypts of Lieberkühn, associated predominantly with goblet cells intermixed with a few absorptive and enteroendocrine cells. Glucagon-like immunoreactant (GLI), pancreatic polypeptide-like peptide (PYY) with N-terminal tyrosine amide–producing L cells predominate in the large intestine. Enterochromaffin, enterochromaffin-like, and pancreatic polypeptide–producing cells also are found. Paneth cells are scarce and normally are noted only in the proximal colon. The lamina propria of the large intestine contains solitary lymphoid follicles extending into the submucosa. Lymphoid follicles are more developed in the rectum and decrease in number with age. Confluent lymphoid tissue is present in the appendix. Macrophages (muciphages) predominate in the subepithelial portion of the lamina propria. These cells are weakly PAS positive and are associated with stainable lipids.

Anal Canal

Microscopically the anal canal is divided into three zones: proximal, intermediate or pectinate, and distal or anal skin. The proximal zone is lined by stratified cuboidal epithelium, and the transition with the rectal mucosa, which is lined by high columnar mucus-producing cells, is called the

rg

ep

A ep as

B Figure 96-10.  Photomicrograph of the anal canal. A, Anorectal histologic junction. Transition from rectal glandular mucosa (rg) to proximal anal mucosa lined by stratified squamous epithelium (ep) is evident. B, The pectinate line is characterized by anal mucosa with stratified squamous epithelium (ep) and anal skin (as) containing adnexae (arrow). (A and B, Hematoxylin and eosin, ×150.)

anorectal histologic junction (Fig. 96-10A). The intermediate or pectinate zone is lined by stratified squamous epithelium but without adnexae (e.g., hair, sebaceous glands) and also is referred to as anoderm. Its proximal margin, in contact with the proximal zone, is called the dentate line; its distal margin, in contact with the anal skin, constitutes the pectinate line, also referred to as the mucocutaneous junction (see Fig. 96-10B). The anal skin is lined by squamous stratified epithelium and contains hair and sebaceous glands.

Vasculature

Large arterial branches enter the muscularis propria and pass to the submucosa, where they branch to form large plexuses. In the small intestine, two types of branches arise from the submucosal plexuses: some arteries branch on the inner surface of the muscularis mucosae and break into a capillary meshwork that surrounds the crypts of Lieberkühn. Other arteries are destined for villi, each receiving one or two arteries, and set up the anatomic arrangement that allows a

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine countercurrent mechanism during absorption. These vessels enter at the base of the villus and form a dense capillary network immediately underneath the epithelium of the entire villus structure. One or several veins originate at the tip of each villus from the superficial capillary plexus, anastomose with the glandular venous plexus, and then enter the submucosa joining the submucosal venous plexus. In the colon, branches from the submucosal plexus extend to the surface, giving rise to capillaries supplying the submucosa, and there branch to form a capillary meshwork around the crypts of Lieberkühn. From the periglandular capillary meshwork, veins form a venous plexus between the base of the crypts and the muscularis mucosae. From this plexus, branches extend into the submucosa and form another venous plexus, from which large veins follow the distribution of the arteries and pass through the muscularis propria into the serosa.

Lymph Vessels

The lymphatics of the small intestine are called lacteals and become filled with milky-white lymph called chyle after eating. Each villus contains one central lacteal, except in the duodenum, where two or more lacteals per villus may be present. The wall of the lacteal consists of endothelial cells, reticulum fibers, and smooth muscle cells. The central lacteals anastomose at the base of the villus with the lymphatic capillaries between the crypts of Lieberkühn. They also form a plexus on the inner surface of the muscularis mucosae. Branches of this plexus extend through the muscularis mucosae to form a submucosal plexus. Branches from the submucosal plexus penetrate the muscularis propria, where they receive branches from plexuses between the inner and outer layers. Lymphatic vessels are absent in the colonic mucosa, but the distribution of lymphatics in the remaining colonic layers is similar to that in the small intestine.

Nerves

The intrinsic nervous system (enteric nervous system) consists of subserosal, muscular, and submucosal plexuses. The subserosal plexus contains a network of thin nerve fibers, without ganglia, that connects the extrinsic nerves with the intrinsic plexus. The myenteric plexus, or Auerbach’s plexus, is situated between the outer and inner layers of the muscularis propria (see Fig. 96-8); it consists of ganglia and bundles of unmyelinated axons that connect with the ganglia forming a meshwork. These axons originate from processes of the ganglion cells and extrinsic vagus and sympathetic ganglia. The deep muscular plexus is situated on the mucosal aspect of the circular muscular layer of the muscularis propria. It does not contain ganglia; it innervates the muscularis propria and connects with the myenteric plexus. The submucosal plexus, or Meissner’s plexus, consists of ganglia and nerve bundles. The nerve fibers of this plexus innervate the muscularis mucosae and smooth muscle in the core of the villi. Fibers from this plexus also form a mucosal plexus that is situated in the lamina propria and provides branches to the intestinal crypts and villi. The ganglion cells of the submucosal plexus are distributed in two layers: one is adjacent to the circular muscular layer of the muscularis propria; the other is contiguous to the muscularis mucosae. Ganglion cells are large cells, isolated or grouped in small clusters called ganglia (Fig. 96-11). Ganglion cells have an abundant basophilic cytoplasm, a large vesicular round nucleus, and a prominent nucleolus. Ganglion cells are scarce in the physiologically hypogan­ glionic segment 1 cm above the anal verge.

g Figure 96-11.  Photomicrograph showing a normal submucosal plexus of the colon. The ganglia (g) are identified by their oval structure and the nerve trunks are thin (arrow). (Hematoxylin and eosin, ×150.)

EMBRYOLOGY The embryo begins the third week of development as a bilaminar germ disk. During week three, in a process called gastrulation, this disk becomes a trilaminar disk. The surface facing the yolk sac becomes the definitive endoderm; the surface facing the amniotic sac becomes the ectoderm. The middle layer is called mesoderm. The long axis and leftright axis of the embryo also are established at this time. The oral opening is marked by the buccopharyngeal membrane; the future openings of the urogenital and the digestive tracts become identifiable as the cloacal membrane. At four weeks of gestation, the alimentary tract is divided into three parts: foregut, midgut, and hindgut. The endoderm forms the intestinal tube, which communicates only with the yolk sac. Narrowing of the communication of the yolk sac with the endoderm forms the vitelline duct. With folding of the embryo during the fourth week of development, the mesodermal layer splits. The portion that adheres to endoderm forms the visceral peritoneum, whereas the part that adheres to ectoderm forms the parietal peritoneum. The space between the two layers becomes the peritoneal cavity.

MOLECULAR REGULATION OF INTESTINAL MORPHOGENESIS

The induction of endoderm appears to be governed by nodal or transforming growth factor-β signaling.15 Specification is initiated by transcription factors expressed in the different regions of the intestinal tube. Thus, PDX1 specifies the duodenum, CDXC the small intestine, and CDXA the large intestine and rectum.16 Differentiation of the gastrointestinal tract depends on the interaction between the endoderm and mesoderm through the Hox code. Signaling from the mesoderm to endoderm is regulated by the Hox genes that encode homeodomain-containing transcription factors. Induction of the Hox code in the mesoderm results from expression of Shh through the endoderm of the midgut and hindgut. Shh is a signaling molecule that acts as a morphogen or form-producing substance in a variety of organ systems. When prompted by this code, the mesoderm instructs the endoderm to form the various components of the midgut

1623

1624

Section X  Small and Large Intestine A Endoderm Ectoderm Angiogenic cell cluster Buccopharyngeal membrane

C Buccopharyngeal membrane

B

Amniotic cavity

Foregut Connecting stalk

Heart tube

Allantois

Pericardial cavity

Cloacal membrane

Cloacal membrane

Hindgut

D Lung bud

Liver bud Midgut

Heart tube

Remnant of the buccopharyngeal membrane Vitelline duct

Allantois

Yolk sac Figure 96-12.  Formation of the foregut, midgut, and hindgut (see text for details). (From Sun B, editor. Langman’s Medical Embryology. 9th ed. Philadelphia: Lippincott Williams & Wilkins; 2004.)

and hindgut regions, for example, the small bowel, cecum, colon, and cloaca.16 As indicated by animal studies, Hox genes contribute to the subdivision of the intestine, and formation of the ileocecal valve that separates the small and the large intestine. Shh also plays a crucial role in the development of the hindgut.17 The primitive gut results from incorporation of the endoderm-lined yolk sac cavity into the embryo, following embryonal cephalocaudal and lateral folding. The primitive gut is composed of a blind-ended tube in the cephalic and caudal portions of the embryo, which is the progenitor of the foregut and hindgut; the midgut (Fig. 96-12) is connected to the yolk sac by the vitelline duct. The endoderm gives rise to the epithelial lining of the gastrointestinal tract; muscle, connective tissue, and peritoneum originate from the splanchnic mesoderm. During the ninth week of development, the epithelium begins to differentiate from the endoderm with villus formation and differentiation of epithelial cell types. Organogenesis is complete by 12 weeks of gestation. Initially the foregut, midgut, and hindgut are in broad contact with the mesenchyma of the posterior abdominal wall. The intraembryonic cavity is in open communication with the extraembryonic cavity. Subsequently the intraembryonic cavity loses its wide connection with the extraembryonic cavity. By week five of embryonic development, splanchnic mesoderm layers are fused in the midline and form a double-layered membrane, the dorsal mesentery, between the right and left halves of the body cavity. The mesoderm surrounds the intestinal tube and suspends it from the posterior body wall, allowing it to hang into the body cavity. The caudal portions of the foregut, the midgut, and most of the hindgut thus are suspended from the abdominal wall by the dorsal mesentery extending from the duodenum to the cloaca. The dorsal mesentery forms the mesoduodenum in the duodenum, the dorsal mesocolon

in the region of the colon, and the mesentery proper in the region of the jejunum and ileum.16

SPECIFIC STRUCTURES AND SYSTEMS Duodenum

The duodenum originates from the terminal portion of the foregut and cephalic part of the midgut. With rotation of the stomach, the duodenum becomes C-shaped and rotates to the right; the fourth portion becomes fixed in the left upper abdominal cavity. The mesoduodenum fuses with the adjacent peritoneum; both layers disappear, and the duodenum becomes fixed in its retroperitoneal location. The lumen of the duodenum is obliterated during the second month of development by proliferation of its cells; this phenomenon is shortly followed by recanalization. Because the foregut is supplied by the celiac artery and the midgut by the superior mesenteric artery, the duodenum is supplied by both arteries and therefore is relatively protected from ischemic injury.16

Midgut

In a 5-week embryo, the midgut is suspended from the dorsal abdominal wall by a short mesentery and communicates with the yolk sac by way of the vitelline duct. The midgut gives rise to the duodenum distal to the ampulla, to the entire small bowel, and to the cecum, appendix, ascending colon, and the proximal two thirds of the transverse colon. The midgut rapidly elongates with formation of the primary intestinal loop. The cephalic portion of this loop, which communicates with the yolk sac by the narrow vitelline duct, gives rise to the distal portion of the duodenum, the jejunum, and a portion of the ileum; the distal ileum, cecum, appendix, ascending colon, and proximal two thirds of the transverse colon originate from the caudal limb. During week 6 of embryonic development, the primary

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine Diaphragm

Esophagus

Lesser omentum

Liver Falciform ligament Stomach

Gallbladder Duodenum Cecum Vitelline duct

Allantois

Descending colon

Jejunoileal loops

Cloacal membrane

Rectum

intestinal loop enters the umbilical cord (physiologic umbilical herniation) (Fig. 96-13), and by week 10 it re-enters the abdominal cavity. The proximal portion of the jejunum is the first portion of the intestine to re-enter the abdominal cavity and becomes located on the left side; the subsequent loop that re-enters the abdominal cavity locates to the right. The cecal bud is the last segment to re-enter the abdominal cavity. The cecum originates as a small dilatation of the caudal limb of the primary intestinal loop by approximately 6 weeks of development. Initially it lies in the right upper quadrant; then it descends to the right iliac fossa, placing the ascending colon and hepatic flexure in the right side of the abdominal cavity. The appendix originates from the distal end of the cecal bud. Because the appendix develops during descent of the colon, its final position frequently is retrocecal or retrocolonic. The primary intestinal loop rotates counterclockwise for approximately 270 degrees around an axis formed by the superior mesenteric artery. This rotation occurs in three stages (Fig. 96-14): the first stage occurs between six and eight weeks (90 degrees), the second stage is at nine weeks (180 degrees), and the third stage is at 12 weeks of gestation (270 degrees). Elongation of the bowel continues, and the jejunum and ileum form a number of coiled loops within the peritoneal cavity.16

Mesentery

When the caudal limb of the primitive intestine moves to the right side of the abdominal cavity, the dorsal mesentery twists around the origin of the superior mesenteric artery. After the ascending and the descending portions of the colon reach their final destinations, their mesenteries fuse with the peritoneum of the posterior abdominal wall, and they become retroperitoneal organs. The appendix, cecum, and descending colon retain their free mesentery. The transverse mesocolon fuses with the posterior wall of the greater omentum. The mesentery of the jejunum and ileum at first

Figure 96-13.  Physiologic umbilical herniation of the intestinal loop during normal development. Coiling of the small intestinal loops and formation of the cecum occur during the herniation. The first 90 degrees of rotation occur during herniation; the remaining 180 degrees occur during the return of the intestine to the abdominal cavity. (From Sun B, editor. Langman’s Medical Embryology. 9th ed. Philadelphia: Lippincott Williams & Wilkins; 2004.)

is in continuity with the ascending mesocolon; after the ascending colon becomes retroperitoneal, the mesentery only extends from the duodenum to the ileocecal junction.16

Hindgut

The distal third of the transverse colon, the descending colon and sigmoid, the rectum, and the upper part of the anal canal originate from the hindgut. Initially the urinary, genital, and rectal tracts empty into a common channel, the cloaca. They become separated by the caudal descent of the urorectal septum into an anterior urogenital sinus and a posterior intestinal canal. The lateral fold of the cloaca moves to the midline, and the caudal extension of the urorectal septum develops into the perineal body. In a man, the lateral genital ridges coalesce to form the urethra and scrotum; in a woman, no fusion occurs, and the labia minora and majora evolve. The cloaca is lined by endoderm and covered anteriorly by ectoderm. The most distal portion of the hindgut enters into the posterior region of the cloaca, the primitive anorectal canal. The boundary between the endoderm and the ectoderm forms the cloacal membrane. This membrane ruptures by the seventh week of embryonic development, creating the anal opening for the hindgut. This portion is obliterated by the ectoderm but recanalizes by week nine. Thus, the distal portion of the anal canal originates from the ectoderm and is supplied by the inferior rectal artery; the proximal portion of the anal canal originates from the endoderm and is supplied by the superior rectal artery. The pectinate line is situated at the junction of the endoderm and the ectoderm.

Arterial System

Vascular endothelial growth factor (VEGF)-A and its receptors, VEGFR-1 and VEGFR-2, are important for endothelial cell proliferation, migration, and sprouting. Angiopoietins

1625

1626

Section X  Small and Large Intestine 1st stage

3rd stage

2nd stage

Figure 96-14.  The three stages of normal intestinal rotation (see text for details). (From Gosche JR, Touloukian RJ. Congenital anomalies of the midgut. In: Wyllie R, Hyams JS, editors. Pediatric Gastrointestinal Disease. Pathophysiology, Diagnosis, Management. 2nd ed. Philadelphia: WB Saunders; 1999.)

and their receptors, Tie1 and Tie2, play a role in remodeling and maturation of the developing vasculature. Mutation in Tie2 has been reported in vascular dysmorphogenesis. Vascular malformation is briefly discussed in Chapter 36. Arteries of the dorsal mesentery, originating from fusion of the vitelline arteries, give rise to the celiac, superior mesenteric, and inferior mesenteric arteries. Their branches supply the foregut, midgut, and hindgut, respectively.

Venous System

Vitelline veins give rise to a periduodenal plexus that develops into a single vessel, the portal vein. The superior mesenteric vein originates from the right vitelline vein that receives blood from the primitive intestinal loop. The left vitelline vein disappears. The umbilical veins become connected to the hepatic sinusoids after which the right umbilical vein disappears and the left umbilical vein joins the inferior vena cava; ultimately the umbilical vein is obliterated and forms the ligamentum teres. The cardinal veins are involved with forming the inferior vena cava as is the proximal portion of the right vitelline vein.

Lymphatic System

Lymphatics originate from endothelial budding of veins, after which the peripheral lymphatic system spreads by endothelial sprouting into the surrounding tissues and organs. Flt4 (also known as VEGFR-3), a receptor for VEGF, plays a role in development of the vascular as well as the lymphatic systems. Overexpression of VEGF-C, a ligand of Flt4, results in hyperplasia of lymphatic vessels in transgenic mice. The homeobox gene Prox1 is essential for normal development of the lymphatic system based on animal studies. Homeobox genes contain a conserved sequence of 183 nucleotides. The proteins encoded by homeobox-containing genes act as regulatory molecules that control the expression of other genes. Several families of homeobox-containing genes are known, including the murine Hox family, which has been implicated in pattern formation during embryogenesis. Disruption of this gene in mice causes chyle-filled intestine. Abnormalities in the development of the lymphatic system can result in lymphangiectasia (see Chapter 28).

Enteric Nervous System

The enteric nervous system originates from vagal, truncal, and sacral neural crest cells. Most of the enteric nervous

system cells derive from the vagal and truncal neural crest, enter the foregut mesenchyma, and colonize the developing intestine in a cephalocaudal direction. The truncal neural crest gives rise to ganglia of the proximal stomach, whereas the vagal neural crest supplies ganglia to the entire intestine including the rectum; this colonization is complete by 13 weeks of embryonic development. A small component of the enteric nervous system originates from sacral neural crest cells. These cells form extraintestinal pelvic ganglia that colonize the hindgut mesenchyma before the arrival of the vagal-derived neural crest cells.18 The normal development of the enteric nervous system depends on the survival of cells developed from the neural crest, and their proliferation, movement, and differentiation into neurons and glial cells. Microenvironmental, genetic, or molecular mechanisms may intervene in these processes (see “Disturbance in the Enteric Nervous System”).

CLINICAL IMPLICATIONS

Table 96-2 summarizes the different congenital clinical entities that result from disturbances in embryologic development. Gastrointestinal malformations can be associated with extraintestinal defects when genes such as those that determine left-right asymmetry are involved. The CFC1 gene plays a role in establishing left-right axis. Mutations of this gene have been reported in extrahepatic biliary atresia, in the polysplenia syndrome (inferior vena cava abnormalities, preduodenal portal vein, intestinal malrotation, and situs inversus), and in right-sided stomach and congenital heart disease.19,20

ABNORMALITIES IN NORMAL EMBRYOLOGIC DEVELOPMENT ABDOMINAL WALL Omphalocele

Current theories suggest that a teratogenic event during the first three weeks of gestation prevents return of the bowel to the abdomen and causes failure of lateral embryonic fold development, which results in an omphalocele. Omphalocele occurs with a frequency of 2.5 in 10,000 births. Associated anomalies (e.g., sternal defects) result from failure of closure of the cephalic folds; failure of caudal fold develop-

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine Table 96-2  Causes of Abnormalities in Normal Embryologic Development Body wall   Omphalocele   Gastroschisis Mesentery   Mobile cecum   Volvulus Vitelline duct   Meckel’s diverticulum   Omphalomesenteric cyst   Patent omphalomesenteric duct Rotation   Malrotation   Nonrotation   Reverse rotation Proliferation   Duplication Intestinal atresia and stenosis   “Apple-peel” atresia   Duodenum   Small and large intestine Anorectum Enteric nervous system   Hirschsprung’s disease   Intestinal neuronal dysplasia   Pseudo-obstruction Miscellaneous   Intestinal epithelial dysplasia   Microvillus inclusion disease Other genetic defects   Congenital chloride diarrhea   Congenital glucose or galactose malabsorption   Congenital lactase deficiency   Congenital sodium diarrhea   Congenital sucrase/isomaltase deficiency   Cystic fibrosis

Failure of the intestine to return to the abdominal cavity after its physiologic herniation Weakening of the abdominal wall Persistence of mesocolon Failure of fusion of mesocolon with posterior abdominal wall Persistence of the vitelline duct (see Fig. 96-17) Focal failure of vitelline duct obliteration Total failure of vitelline duct obliteration Failure of rotation of the proximal midgut; distal midgut rotates 90 degrees clockwise Failure of stage 2 rotation (see Fig. 96-18) Rotation of 90 degrees instead of 270 degrees Abnormal proliferation of the intestinal parenchyma Coiling of proximal jejunum distal to the atresia around the mesenteric remnant Lack of recanalization Vascular “accident” Disturbance in hindgut development Failure of migration of ganglion cells; microenvironment changes Controversial Multifactorial (see Chapter 120) Abnormalities of the basement membrane Defective protein trafficking and abnormal cytoskeletal and microfilament function Abnormal Cl−-HCO3− exchange in the ileum and colon Absence of Na+-glucose cotransporter for glucose and galactose Decrease in lactase-phlorizin hydrolase Defective sodium-proton exchange Abnormal intracellular transport, aberrant processing, and defective function of sucrase or isomaltase Defective cystic fibrosis transmembrane conductance regulator

ment results in exstrophy of the bladder and, in extreme cases, exstrophy of the cloaca. Omphalocele is a congenital hernia involving the umbilicus. It is covered by an avascular sac composed of fused layers of amnion and peritoneum (Fig. 96-15). The umbilical cord usually is inserted into the apex of the sac, and the blood vessels radiate within the sac wall. Although a central defect is present in the skin and the linea alba, the remainder of the abdominal wall is intact, including the surrounding musculature. Because a small occult omphalocele of the umbilical cord may not be observed at birth, it is recommended that the umbilical cord be tied at least 5 cm from the abdominal wall at the time of delivery. Close inspection of the umbilical cord before clamping will avoid clamping an occult omphalocele. With a large omphalocele, the liver and spleen frequently are outside the abdominal cavity. Associated anomalies occur in about 75% of children with omphalocele and include chromosomal abnormalities such as trisomy 13 or 18, nonchromosomal syndromes such as BeckwithWiedemann syndrome (mental retardation, hepatomegaly, large body stature, hypoglycemia), fetal valproate syndrome, exstrophy of the bladder or cloaca, and OEIS (omphalocele, exstrophy of the bladder, imperforate anus, spinal defect). Malformations of the musculoskeletal, cardiovascular, and central nervous systems, also can occur.21,22 Prenatally, increased levels of maternal serum alpha fetoprotein suggest the possible presence of an omphalocele. Ultrasound examination during pregnancy allows the diagnosis of this abdominal wall defect in most infants.

Figure 96-15.  A newborn with an omphalocele. Note the translucent saclike structure with its attached umbilical cord.

1627

1628

Section X  Small and Large Intestine Increased maternal levels of alpha fetoprotein are suggestive of gastroschisis, as well as omphalocele. Most children with gastroschisis can undergo primary closure safely; however, for the child with significant intestinal atresia as a complication of gastroschisis, bowel exteriorization and secondary closure often are preferred treatment. It is crucial to conserve intestinal length in these children. Adhesive small bowel obstruction is a frequent and a serious complication, especially in the first year of life.23

MECKEL’S DIVERTICULUM AND OTHER VITELLINE DUCT ABNORMALITIES

Figure 96-16.  Gastroschisis. In this newborn, there are full-thickness disruption of the abdominal wall and protruding viscera without accom­ panying peritoneum. (From Feldman’s Online Gastro Atlas, Current Medicine.)

Fetal management, including possible termination of pregnancy, is determined by the physician in consultation with the family. If pregnancy is continued, mode of delivery and provision for care of a child with possibly coexisting anomalies should be considered before labor and delivery. Operative treatment is required in all patients with omphalocele. The size of the omphalocele determines whether a primary repair or delayed primary closure is selected. Escharification of the intact omphalocele sac has been used. Reoperation is necessary in up to 25% of cases of omphalocele, either for reclosure of stomas or for subsequent bowel obstruction.

Gastroschisis

Gastroschisis is an abdominal wall defect most commonly located to the right of an intact umbilical cord (Fig. 96-16). The incidence of gastroschisis is approximately 1 in 10,000 births overall, but approaches 7 in 10,000 among mothers younger than 20 years of age. Gastroschisis occurs more frequently in whites and in Hispanic infants than in other races or ethnicities. In gastroschisis, a sac is absent, and the extruded bowel is “padded” and thickened along its length from its extended exposure to the amniotic fluid. Histologically, the bowel usually is normal. Atresia occurs in 10% to 15% of children with gastroschisis. Almost all infants with gastroschisis also exhibit malrotation. Whereas prematurity is more common in children born with gastroschisis than it is in children with omphalocele, extraintestinal anomalies are much more common with omphalocele than they are with gastroschisis. The morbidity and mortality in patients with gastroschisis are largely related to intestinal atresia; other congenital anomalies also have been reported in a small number of patients.21,22 Gastroschisis may be complicated by necrotizing enterocolitis, with all its attendant short-term and long-term complications.

Persistence of the ductal communication between the intestine and the yolk sac beyond the embryonic stage may result in several anomalies of the omphalomesenteric (vitelline) duct (Fig. 96-17) including (1) a blind omphalomesenteric duct, or Meckel’s diverticulum; (2) a central cystic dilatation in which the duct is closed at both ends but patent in its center, an omphalomesenteric or vitelline cyst; (3) an umbilical-intestinal fistula (see Fig. 96-17A), resulting from the duct remaining patent throughout its length; and (4) complete obliteration of the duct, resulting in a fibrous cord or ligament extending from the ileum to the umbilicus, as an omphalomesenteric band.24 In approximately 1% to 4% of all infants, a remnant of the embryonic yolk sac is retained, making the omphalomesenteric or vitelline duct the most common site of congenital gastrointestinal anomaly. Between the fifth and seventh weeks of gestation, the omphalomesenteric duct, which has connected the embryo to the yolk sac, attenuates, involutes, and separates from the intestine. Before this separation, the epithelium of the yolk sac develops an appearance similar to that of the gastric mucosa. Under normal circumstances the omphalomesenteric duct becomes a thin fibrous band that fragments and is absorbed spontaneously during the fifth to tenth week of gestation. Partial or complete failure of involution of the duct results in the variety of retained structures described above. A Meckel’s diverticulum is an antimesenteric outpouching of the ileum that usually is found approximately 2 feet from the ileocecal junction (see Fig. 96-17B). It occurs in 1.2% to 2% of the population and has a male-to-female ratio of 3 : 1.25 Meckel’s diverticula account for 67% of all omphalomesenteric duct remnants.24 Length of the diverticulum varies, ranging from 1 to 10 cm. Ectopic gastrointestinal mucosa—duodenal, gastric, biliary or colonic, or aberrant pancreatic tissue—is present in about 50% of Meckel’s diverticula; most common is ectopic gastric mucosa, accounting for 80% to 85% of all Meckel’s diverticula– associated ectopic tissue (see Fig. 96-17C). Painless bleeding per rectum is the most common manifestation of a Meckel’s diverticulum. Blood in the stool usually is maroon, even in patients with massive bleeding and hypovolemic shock. Bright red blood per rectum, as might be seen with bleeding from the left colon, is almost never encountered, but melena may be seen in patients with intermittent or continual, less severe bleeding. The cause of bleeding is peptic ulceration secondary to acid production by the ectopic gastric mucosa within the Meckel’s diverticulum. A “marginal” ulcer often develops at the junction of the gastric and ileal mucosae. Although Helicobacter pylori has been observed in the gastric mucosa within a Meckel’s diverticulum, a relationship between bleeding from a Meckel’s diverticulum and presence of this organism is unlikely. Despite massive bleeding, death seldom, if ever, occurs in children from complications of a Meckel’s diver-

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine Meckel’s diverticulum

Vitelline cyst

Umbilicus

Vitelline ligament

Vitelline ligaments

Vitelline fistula

A

C

B 5 mins

10 mins

15 mins

20 mins

25 mins

30 mins

D

Figure 96-17.  Vitelline duct abnormalities and features of Meckel’s diverticulum. A, Schematic representations of a Meckel’s diverticulum, vitelline cyst, and vitelline fistula. B, Surgical specimen revealing an outpouching of the ileum (Meckel’s diverticulum). C, Photomicrograph showing replacement of the small intestinal mucosa by ectopic oxyntic mucosa lining a Meckel’s diverticulum. (Hematoxylin and eosin, ×150.) D, Meckel’s diverticulum scan demonstrating uptake of 99mtechnetium-pertechnetate (arrows) by the diverticulum by 10 minutes. (A from Sun B, editor. Langman’s Medical Embryology. 9th ed. Philadelphia: Lippincott Williams & Wilkins; 2004.) (See text for details.) (D, Courtesy of Dr. I. Zanzi.)

ticulum. Once hypovolemia occurs from blood loss, the splanchnic blood vessels contract, and bleeding tends to diminish or cease. Intestinal obstruction is the next most common manifestation of a Meckel’s diverticulum. This obstruction is caused either by intussusception with the diverticulum as a lead point or by herniation through or volvulus around a persistent fibrous cord remnant of the vestigial vitelline duct. In children older than 4 years of age, intussusception almost always is secondary to a Meckel’s diverticulum. However, diverticulum-related intestinal obstruction may occur at any age. Volvulus around a vitelline cord has been described in the neonatal period. Bilious vomiting and abdominal distention usually are the initial signs of obstruction. Intestinal obstruction in these patients, as with other causes of obstruction, can lead to intestinal ischemia and death. Diverticulitis of a Meckel’s diverticulum occurs as a result of acute inflammation. Most commonly, affected patients are diagnosed as having acute appendicitis, and the diagnosis of Meckel’s diverticulitis is made at exploratory laparotomy. Perforation occurs in approximately one third of patients with Meckel’s diverticulitis and may result from peptic ulceration.26 A chronic form of Meckel’s diverticuli-

tis (Meckel’s ileitis) may mimic the presentation of Crohn’s disease of the ileum. Meckel’s diverticulum may be an incidental finding.25 The presence of a Meckel’s diverticulum always should be considered in an infant or child with significant painless rectal bleeding. Standard abdominal plain films, barium contrast studies, and ultrasonographic imaging rarely are helpful in making the diagnosis. Because bleeding almost always is from ectopic gastric mucosa within the diverticulum, the Meckel’s scan, which allows imaging of the gastric mucosa, should be the initial diagnostic study (see Fig. 96-17D). Uptake of the 99mTc-pertechnetate is by the mucussecreting cells of the gastric mucosa, not the parietal cells. Unfortunately, this study has only 85% sensitivity and 95% specificity. When the diagnosis of a bleeding Meckel’s diverticulum is entertained and the Meckel’s scan is negative, splanchnic angiography and 99mTc-labeled red blood cell studies may be used; however, diagnosis is usually made at surgery. It is reasonable to perform esophagogastroduodenoscopy and colonoscopy to rule out other possible etiologic disorders. Although symptomatic Meckel’s diverticulum is far more common in pediatric patients, it may occur in adults.

1629

1630

Section X  Small and Large Intestine Omphalomesenteric (or Vitelline) Cyst

Omphalomesenteric (or vitelline) cyst is more common in men and is characterized by a mucosa-lined intestinal cystic mass within the center of a fibrous cord.24 Infection of the cyst or intestinal obstruction may result.

Patent Omphalomesenteric (Vitelline) Duct

Patent omphalomesenteric (vitelline) duct represents a persistent connection between the distal ileum and the umbilicus. This fistula has a male-to-female ratio of 5 : 1, and accounts for 6% to 15% of omphalomesenteric duct remnants. The diagnosis usually is made in the first few weeks of life after separation of the umbilical cord from the newborn umbilicus. Foul-smelling discharge from the umbilicus occurs.27 Examination of the umbilicus reveals either an opening or a polypoid mass resulting from limited prolapse of the patent omphalomesenteric duct. Definitive diagnosis can be made by fistulography. Complications of this type of fistula include prolapse of the patent duct, or of the duct and the attached ileum, through the umbilicus, which may lead to partial intestinal obstruction. Prolapse should not be mistaken for an umbilical polyp, because excision of involved tissue might result in perforation. Resection is warranted.27

Omphalomesenteric Band

In omphalomesenteric band, the solid cord connecting the ileum to the umbilicus remains intact. This cord may result in intestinal obstruction from an internal hernia or volvulus.

Vitelline Blood Vessel Remnants

Failure of involution of vitelline blood vessel remnants results in complications similar to those seen with a retained fibrous cord within the peritoneal cavity. Intestinal obstruction occurs when a portion of the small intestine wraps itself around the band. Treatment of all vitelline duct abnormalities is surgical.

MALROTATIONS

Rotation defects result from errors in the normal embryonic development of the midgut, which gives rise to the distal duodenum, jejunum, ileum, cecum, and appendix, as well as the ascending colon and proximal two thirds of the transverse colon. Aberrations in midgut development may result in a variety of anatomic anomalies, including (1) rotation and fixation, (2) atresias and stenoses, (3) duplications, and (4) persistence of embryonic structures. These congenital anomalies may cause symptoms not only in the newborn or neonatal period, but also later in childhood and adulthood. Therefore, congenital anomalies of the midgut are considerations in the differential diagnosis of intestinal obstruction and ischemia in patients of all ages. Because anomalies of intestinal rotation may remain asymptomatic throughout life, their true incidence is unknown; a prevalence of 1 in 500 live births has been reported.28 Symptoms usually manifest within the first month of life, with bilious emesis and abdominal distention, but presentation may be delayed in mild cases to the fourth decade of life. Patients may have cramping abdominal pain, vomiting, diarrhea, abdominal tenderness, and blood or even mucosal tissue in the stool from ischemia. If ischemia is allowed to progress, peritonitis and hypovolemic shock may develop, potentially culminating in death. Delay in surgery in patients with ischemic injury may result in a short bowel, necessitating chronic total parenteral nutrition therapy and eventually small bowel transplantation, with or without liver transplantation. Most adult

patients with anomalies of intestinal rotation have chronic symptoms for several months or years before diagnosis.

Classification

Anomalies of rotation usually are characterized by the stage in the rotational process at which normal embryonic development of the midgut has been interrupted. Most anomalies of midgut rotation occur during the second stage of rotation and have been characterized as nonrotation, reverse rotation, and malrotation (Fig. 96-18). Of these, nonrotation is most common and reflects a complete failure of the second stage of rotation. With this anomaly the intestinal tract occupies the same position in the abdomen as it does in an eight-week-old embryo; the small intestine is located to the right of the midline and the colon is positioned to the left. Defects in the first and third stages of rotation are uncommon. Abnormalities in the first stage are associated with extroversion of the cloaca; abnormalities of the third stage cause failure of cecal elongation, and the cecum remains in the right upper quadrant. In adults, reverse rotation of the midgut loop is the most commonly diagnosed defect of the midgut. Reverse rotation of the midgut loop is rare, however, and accounts for only 4% of all rotational anomalies. In reverse rotation, the midgut rotates 180 degrees clockwise during the second stage of rotation, resulting in a net 90 degrees of clockwise rotation. This may produce either the retroarterial colon type (the colon is located behind the superior mesenteric artery) or the liver and entire colon ipsilateral type of reverse rotation. Malrotation of the midgut loop, a developmental anomaly of intestinal fixation and rotation, occurs when the proximal midgut fails to rotate around the mesenteric vessels during the second stage of rotation. The distal midgut does rotate 90 degrees in a counterclockwise direction, however, with the result that the jejunum and ileum remain to the right of the superior mesenteric artery and the cecum is situated in the subpyloric region. With the potential for the small intestine and cecum to twist around the superior mesenteric artery and each other, this is the rotation anomaly in adults most frequently associated with ischemic damage, therefore mandating surgical correction.

Associated Abnormalities

Associated anomalies are seen in 30% to 60% of patients with defects in intestinal rotation. Nonrotation of the midgut is a significant finding in patients with omphalocele, gastroschisis, and diaphragmatic hernia. Rotation defects are seen in approximately 30% to 50% of infants with duodenal or jejunal atresia and in 10% to 15% of children with intestinal pseudo-obstruction; they also are associated with a variety of other conditions including Hirschsprung’s disease, esophageal atresia, biliary atresia, annular pancreas, meconium ileus, intestinal duplications, mesenteric cysts, Meckel’s diverticulum, urologic anomalies, and imperforate anus.29 Anomalies of rotation can cause acute or chronic intermittent obstruction from volvulus (see Fig. 96-18D,E). Venous and lymphatic obstruction, also from volvulus, can lead to malabsorption and abnormalities in intestinal motility. Patients may fail to thrive and present with chylous ascites and other symptoms and signs of lymphangiectasia resulting from chronic lymphatic obstruction. Duodenal obstruction can occur as the result of midgut volvulus, and as the result of peritoneal bands between a malpositioned cecum in the subpyloric region and the peritoneum. These bands, called Ladd’s bands, cross the second

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine

Duodenum

Duodenum

Ladd’s bands

Transverse colon

Ascending colon

Jejunoileal loops

Transverse colon Descending colon

A

B

Cecum

D

Jejunoileal loops

Descending colon

C

E

Figure 96-18.  Rotation defects. A and B, Two examples of nonrotation. A, Ladd’s bands are seen crossing the duodenum; some authors would refer to this as a “mixed rotation.” B, In nonrotation, the small intestine is located to the right of the midline, and the colon is to the left of the midline. C, Reverse rotation. The transverse colon passes behind the duodenum. D, Malrotation with volvulus characterized by a clockwise twist of the mesentery and strangulation. E, Radiologic appearance of malrotation depicting the duodenum to the right of the spine, with a volvulus. A and B, from Gosche JR, Touloukian J. Congenital anomalies of the midgut. In: Wyllie R, Hyams JS, editors. Pediatric Gastrointestinal Disease: Pathophysiology, Diagnosis, Management. 2nd ed. Philadelphia: WB Saunders; 1999. C, From Sun B, editor. Langman’s Medical Embryology. 9th ed. Philadelphia: Lippincott Williams & Wilkins; 2004. D, From Netter FH. The Netter Collection of Medical Illustration. vol 3. Teterboro, NJ: Icon Learning System; 2002. (D, Courtesy of Dr. J. Levenbrown.)

or third portion of the duodenum and cause obstruction by compression or kinking. Ladd’s bands are an anomaly of peritoneal embryogenesis and persist throughout life.

Diagnosis and Management

If time allows, diagnosis can be made by upper gastrointestinal contrast examination and delineation of the site of the duodenojejunal junction. Findings on ultrasonography may suggest malrotation if the superior mesenteric vein is seen located to the left of the superior mesenteric artery, in contradistinction to the normal anatomy. In the child with acute onset of bilious vomiting and peritoneal signs, no diagnostic studies should be performed if they delay surgical intervention. In the full-term infant with bilious emesis, anomalies of rotation should be considered first and foremost, to avoid the morbidity and mortality associated with these lesions. Ladd’s procedure, including dividing

Ladd’s bands, if present, widening of the mesentery, appendectomy, and fixation of the small intestine on the right and the colon on the left side of the abdomen, is the operation of choice.30

PROLIFERATION Enteric Duplication

Enteric duplications are rare with an incidence of 1 in 4500 births. Enteric duplications are either tubular or spherical; the tubular type communicates with the normal intestinal tract, whereas the spherical type does not. Tubular duplications may join the intestine at one or at both of its ends. Except for duodenal duplications, duplications occur on the mesenteric side of the bowel, and a common blood supply and muscular coat are shared by the duplicated segment and the adjacent bowel. Duplication cysts may be completely

1631

1632

Section X  Small and Large Intestine isolated and have their own exclusive blood supply. Small bowel duplications often contain pancreatic tissue or gastric mucosa; the latter can be diagnosed by 99mTc radioisotopic imaging.31 The etiology of duplications is unclear, but may involve a defect in intestinal recanalization. Enteric duplications occur throughout the gastrointestinal tract but are most common in the ileum.31 Gastric duplications occur least commonly. Depending on the site of the duplication, and whether ectopic gastric mucosa is present (seen in approximately 50% of the cases), complications include intestinal hemorrhage, ulceration, perforation, intestinal obstruction, volvulus, intussusception, infection, pancreatitis, jaundice, hematobilia, and cutaneous enteric fistulas. Duplication of the rectum is the most common of the large bowel duplications and may be associated with constipation or obstipation. Colonic duplications frequently involve the entire colon. Occasionally, large bowel duplications affect several segments of the colon, leaving “skip areas” of normal colon. A high percentage of children with duplications have associated malfor­mations. Adenocarcinoma, neuroendocrine carcinoma, and squamous carcinoma have been documented with gastric, small bowel, and colonic duplications,31,32 and carcinoid has been described in duplications of the rectum. Neuroenteric cysts attach posteriorly to the spinal cord, are associated with asymptomatic hemivertebrae, and may occur at any level of the gastrointestinal tract. An intra-abdominal mass may be appreciated in a child with intestinal duplication, either by abdominal palpation or on rectal examination. Stool may contain occult blood from ulcerated ectopic gastric mucosa or ischemic damage. Other symptoms and signs include abdominal distention, constipation, vomiting, and respiratory distress.33 Generalized peritonitis can be the first manifestation of a perforated duplication cyst. In adults, acute abdomen, intra-abdominal mass, symptoms of colonic diverticulitis and chronic abdominal pain have been observed.34 Preoperative diagnosis by radiographic evaluation is problematic, but radioisotope studies may prove diagnostic if ectopic mucosa is present in sufficient quantities.

INTESTINAL ATRESIA AND STENOSIS

Of all of the congenital anomalies of the midgut, atresias and stenoses occur most frequently. Intestinal atresia refers to a congenital complete obstruction of the intestinal lumen, whereas stenosis indicates a partial or incomplete obstruction. Atresias occur more commonly than do stenoses, and small bowel atresias have a reported incidence rate of 1 in 1500 live births.35 Small bowel atresias are more common in black infants, low birth weight infants, and twins. Jejunoileal atresias are distributed equally throughout the jejunum and ileum, and multiple atresias are found in up to 20% of children. Colonic atresia occurs infrequently and accounts for less than 10% of all atresias. In the duodenum, atresia results from failure of recanalization of the solid stage of duodenal development, whereas in the remaining small intestine and colon, atresia is the result of intestinal ischemia. Evidence of a vascular “accident” is noted in 30% to 40% of infants with atresia; proposed mechanisms include volvulus, constriction of the mesentery in a tight abdominal wall defect such as gastroschisis, internal hernia, intussusception, and obstruction with perforation. Jejunoileal atresia may follow maternal use of ergotamine (in Cafergot for headaches) or cocaine taken during pregnancy and also is associated with congenital rubella. Atresias also may result from low-flow states and placental insufficiency35; in such cases, evidence of a vas-

Figure 96-19.  Plain film of the abdomen showing a “double bubble,” typical of duodenal atresia. The larger bubble is the gastric bubble; the smaller bubble is the duodenal bubble. (Courtesy of Dr. J. Levenbrown.)

cular accident will be absent. Absence of fibroblastic growth factor 10 may result in intestinal atresia.36,37 In familial cases of jejunoileal atresia there is probably a disruption of a normal embryonic pathway, making this type of atresia a true embryologic malformation rather than an acquired lesion.38 Duodenal obstruction may result from atresia (40% to 60%), stenosis (35% to 40%), or an intestinal web (5% to 15%). Eighty percent of these atresias are contiguous with or distal to the ampulla of Vater, and virtually all webs are within a few millimeters of the ampulla. Atresias may be multiple. The incidence of duodenal obstruction varies, ranging from 1 in 10,000 to 20,000 live births. About 25% of patients with duodenal atresia are born preterm. Stenosis most often is due to extrinsic duodenal obstruction from an annular pancreas. Other anomalies that may cause duodenal obstruction in children with malrotation are Ladd’s bands, an anterior or preduodenal portal vein, or aberrant intramural pancreatic tissue. Clinically, the presentation is that of a proximal intestinal obstruction with bilious vomiting on the first day of life, usually without abdominal distention. With gastric dilatation, the epigastrium may appear to be full by inspection and palpation. Excessive retention of gastric bile–stained fluid is typical. Duodenal obstruction is diagnosed easily by abdominal films revealing a typical “double bubble” sign with a paucity of small intestinal air (Fig. 96-19). Mothers of infants with duodenal obstruction often have polyhydramnios, and uterine ultrasonography may even demonstrate a double bubble in the unborn fetus. Vomiting, abdominal distention, delayed meconium passage, and jaundice are more frequent with jejunoileal than duodenal atresia.39 The classification system of Grosfeld and colleagues comprises five different types of jejunoileal and colonic atresias

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine I

IIIa

II

IIIb

IV

Figure 96-20.  Classification of jejunoileal atresias. Type I: The mucosa and submucosa form a web or intraluminal diaphragm, resulting in obstruction. A defect in the mesentery is not present, and the intestine is not shortened. Type II: The dilated proximal intestine has a bulbous blind end connected by a short fibrous cord to the blind end of the distal intestine. The mesentery, however, is intact, and the overall length of the small bowel is not usually shortened. Type IIIa: The defect in type IIIa is similar to that in type II in that both types have blind proximal and distal ends. In type IIIa, however, complete disconnection exists. In addition, a V-shaped mesenteric defect is present. The proximal blind end is usually markedly dilated and not peristaltic. The compromised intestine undergoes intrauterine absorption, and, as a result, the intestine is shortened. Type IIIb: In addition to a large defect of the mesentery, the intestine is significantly shortened. This lesion is also known as Christmas tree deformity because the bowel wraps around a single perfusing vessel, like the tinsel coil wrapped around a Christmas tree; it also has been called an apple-peel deformity. The distal ileum receives its blood supply from a single ileocolic or right colic artery because most of the superior mesenteric artery is absent. Type IV: Multiple small intestinal atresias are present in any combination of types I to III. This defect often takes on the appearance of a string of sausages because of the multiple lesions. (From Grosfeld JL, Ballantine TVN, Shoemaker R. Operative management of intestinal atresia and stenosis based on pathologic findings. J Pediatr Surg 1979; 14:368.)

(Fig. 96-20).39a In the “apple-peel” atresia or “Christmas tree” deformity (type IIIb), proximal atresia with wide separation of the bowel loops is associated with absence of the distal superior mesenteric artery. The distal ileum receives its blood supply by retrograde perfusion through the ileocolic artery. Type IIIb atresias account for less than 5% of all atresias. Atresias are far more common than stenoses, with a frequency ratio of 15 : 1. With the exception of multiple atresias and perhaps the apple-peel atresia, heredity appears to be of little significance in most cases. Approximately 50% of children with duodenal atresia have associated malformations. Of this group, 30% have Down syndrome.39 Major anomalies occur less frequently with jejunoileal atresias and colonic atresias than with duodenal atresia. The most common anomalies are malrotation, volvulus, and gastroschisis, all of which can cause intestinal ischemia in utero.40 Extragastrointestinal anomalies associated with atresias include cardiovascular, pulmonary, and renal malformations, and skeletal deformities. Prematurity is common, ranging in incidence from 25% in ileal atresias to 40% in jejunal lesions; 50% percent of babies with multiple atresias are born prematurely. If the obstruction occurs beyond the ampulla of Vater, bilious or feculent vomiting with abdominal distention is seen. The presence of meconium in the colon is uncommon at surgery, but variable amounts may be noted. With distal obstruction, abdominal films may demonstrate multiple dilated air-filled bowel loops. If perforation has occurred in utero, extraluminal air and intraperitoneal calcifications or calcifications within the scrotal sac may be present, suggesting meconium peritonitis. A “soap bubble” appearance of the ileum may suggest meconium ileus (cystic fibrosis). Air-fluid levels

rarely are seen in meconium ileus. Prenatal ultrasonographic findings in jejunoileal atresia include dilated bowel and polyhydramnios.41 Considerations in the differential diagnosis of distal bowel obstruction include small intestinal and colonic atresias, meconium ileus, Hirschsprung’s disease, and meconium plug with or without small left colon syndrome. In the small left colon syndrome, the descending and sigmoid colon are narrowed, usually with a caliber transition at or near the splenic flexure. Typically, neonates are born to mothers with gestational diabetes and may experience resolution of obstruction without operation. Contrast studies of the colon are helpful in making a proper diagnosis. An upper gastrointestinal contrast study may provide additional important information. Surgery is required to relieve the intestinal obstruction in the atretic or narrowed segment. Postoperative complications include fluid and electrolyte disorders, nutritional and feeding problems from diarrhea due to short bowel and small bowel failure, and failure to thrive.

ANORECTUM

Anorectal malformations comprise a wide spectrum of diseases that can involve the male and female anus and rectum as well as the urinary and genital tracts.42 Anorectal malformations occur in 1 in 4000 to 5000 newborns and are more common among boys and in children with Down syndrome.43 During normal development, after appearance of the urorectal septum, migration of the primitive anus down the posterior wall of the cloaca may occur. Some experts postulate that a craniocaudal fusion of the lateral urorectal

1633

Section X  Small and Large Intestine

Type 2. Pouch ≤1.5 cm from the anal dimple

Type 3. A blind pouch >1.5 cm from the anal dimple

Type 4. Atresia of the rectum with a normal anus

In females

A

Type 1. A thin membrane over the anus

Rectovaginal

Rectofourchet

Rectoperineal

In males

1634

B

Rectovesical

Rectourethral

Rectoperineal

Figure 96-21.  Anorectal malformations. A, Types of imperforate anus. B, Types of associated fistulas. (From Netter FH. The Netter Collection of Medical Illustration. vol 3. Teterboro, NJ: Icon Learning System; 2002.)

ridges occurs from the walls of the cloaca. Migration of the anus is completed when the urorectal septum reaches the perineum. Anorectal malformations during the fourth to twelfth weeks of gestation are believed to result from failure of migration of the anus and excessive fusion. Vascular accidents, maternal diabetes, and maternal ingestion of thalidomide, phenytoin, and trimethadione all have been proposed causes. Defective development of the dorsal cloaca also has been implicated44 and distal 6q deletions have been reported in sacral or anorectal malformations.45 Alteration in Shh signaling also may play a role in producing abnormal notochord development and sacral or anorectal malformations.46,47 Anorectal malformations may occur with higher frequency in infants born after in vitro fertilization.48

Different types of anorectal malformations are illustrated in Figure 96-21. Anorectal malformations are divided into low (infra- or translevator), high (supralevator), and intermediate categories. A functional and practical classification of these malformations, the Wingspread classification, is summarized in Table 96-3A. The classification in Table 96-3B is designed, according to Pena,49 to increase the physician’s awareness of the possibility of the presence of these lesions, as well as to establish therapeutic priorities.

Anocutaneous Fistula

In anocutaneous fistula, the rectum traverses normally through most of the anal sphincter, but its lower portion

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine Table 96-3  Classifications of Anorectal Malformations Wingspread Classification MALE

FEMALE

Low* Anocutaneous fistula Anal stenosis

Anovestibular fistula Anal stenosis Anocutaneous fistula

†

Intermediate Anal agenesis without fistula Rectobulbar urethral fistula High‡ Anorectal agenesis   With rectoprostatic urethral fistula   Without fistula Rectal agenesis

Anal agenesis without fistula Rectovaginal fistula Rectovestibular fistula Anorectal agenesis With rectovaginal fistula Without fistula Cloaca

Classification Based on the Need for Colostomy49 MALE

FEMALE

Colostomy Not Required Perineal (cutaneous) fistula Colostomy Required Rectourethral fistula   Bulbar   Prostatic Rectovesical fistula Imperforate anus without fistula

Colostomy Not Required Perineal (cutaneous) fistula Colostomy Required Vestibular fistula

Rectal atresia

Persistent cloaca Imperforate anus without fistula Rectal atresia

*Low: infra-, or translevator. † Intermediate: between high and low. ‡ High: supralevator.

deviates anteriorly and ends as a perineal, cutaneous fistula anterior to the center of the external anal sphincter (anocutaneous or perineal fistula). These anomalies are similar in the male and the female child. Perineal fistula is the most benign of anorectal defects, and associated urologic defects are uncommon (10%). All patients achieve bowel control after proper surgical treatment. Examination of the perineum may demonstrate features indicative of a perineal fistula, including a prominent midline skin ridge (“bucket-handle” malformation) and subepithelial midline raphe fistula having the appearance of a black ribbon because of its meconium content. Surgery consists of a simple anoplasty, usually done without a protective colostomy.

Rectourethral Fistula

In rectourethral fistula, by far the most frequent anorectal malformation in male children, the rectum descends through a portion of the pelvic floor musculature but focally deviates anteriorly and communicates with the posterior urethra. This fistula may end in the lower posterior (bulbar) or in the upper posterior (prostatic) urethra.49 Prenatal echogenic calcifications within the bowel, due to a mixture of meconium and urine, should suggest an anorectal malformation with rectourinary fistula and bladder outlet obstruction.50 Children with prostatic urethral fistulas more commonly have sacral and urologic defects (60%) than do children with bulbar prostatic fistula (30%). Eighty-five percent of children with rectourethral bulbar fistula achieve fecal continence after repair, compared with 60% of children with rectoprostatic fistula.

Rectovesical Fistula

In rectovesical fistula, the most proximal anorectal defect in male children, the rectum opens into the bladder neck. Ninety percent of these malformations are associated with significant urologic defects, and only 15% of children achieve bowel control after surgical repair.

Vestibular Fistula

In vestibular fistula, the most common anorectal defect of female children, the rectum opens into the vestibular bulb of the clitoris. The vesticular bulbs are erectile structures situated on either side of the vulvovaginal orifice. The rectum and the vagina share a thin common wall. Thirty percent of affected children have associated urologic defects, and approximately 90% of these children achieve bowel control after surgery. In the vaginal fistula, the rectum opens in the lower or, less frequently, the upper half of the vagina.

Anorectal Agenesis (Imperforate Anus) Without Fistula

In anorectal agenesis the rectum ends blindly without a fistula approximately 1 to 2 cm above the perineum. Sphincter function usually is preserved, with 80% of these patients achieving bowel control after surgery. Approximately 50% of children with imperforate anus have Down syndrome. Conversely, 95% of children with Down syndrome who have anorectal malformations will have this specific type of defect.

Rectal Agenesis (Atresia)

Rectal agenesis occurs more frequently in female than in male children, and consists of complete (atresia) or partial (stenosis) interruption of the rectal lumen between the anal canal and the rectum. On visual inspection of the perineum, the anus appears normal; however, an obstruction can be found 1 to 2 cm above the mucocutaneous junction of the anus. Sphincter function is normal in these patients, and associated urologic defects are rare. In these children, prognosis is excellent, with 100% achieving full bowel control after anorectoplasty.

Anal Stenosis

Anal stenosis, a fibrous ring located at the anal verge, causes constipation and gives the stool a ribbon-like appearance. Response to dilation or surgical disruption is excellent.

Persistent Cloaca

In the complex defect of persistent cloaca, the rectum, vagina, and urethra are fused into a single common channel that opens into one perineal orifice situated at the site of what should be the opening of the normal urethra. Prognosis depends on the intactness of the sacrum and the length of the common channel. Prognosis is better in those children with a shorter common channel (less than 3 cm) than in those with a common channel longer than 3 cm; the latter have a higher incidence of urologic anomalies.51 Associated urologic problems are an important consideration with persistent cloaca. For example, urologic emergencies from obstructive uropathy are common, and hydrocolpos may compress the opening of the ureters, resulting in bilateral megaureters and massive vesicoureteral reflux.

Associated Abnormalities

Other associated abnormalities have been reported in 70% of children with anorectal malformation (Table 96-4).42,43 Anorectal malformations occur in malformation syndromes and with chromosomal anomalies.43,52 The higher and more complex the anorectal defect, the greater the chance of severe urologic anomalies (72%);

1635

1636

Section X  Small and Large Intestine Table 96-4  Common Abnormalities Associated with Anorectal Malformations Cardiovascular Atrial septal defect Dextrocardia Pulmonary stenosis Tetralogy of Fallot Ventricular septal defect Central nervous system Aqueductal stenosis Cerebral atrophy Microcephaly Myelomeningocele Teratoma Chromosomal abnormalities Trisomy 13 Trisomy 18 Trisomy 21 Craniofacial Cleft palate Epicanthal folds Low-set ears Potter facies Simian creases Gastrointestinal Duodenal atresia Esophageal atresia Malrotation Tracheoesophageal fistula Genitourinary Ambiguous genitalia Cryptorchidism Multicystic dysplastic kidney Renal agenesis Malformation sequences Caudal regression syndrome Malformation syndrome Cat’s-eye syndrome Opitz syndrome Potter syndrome type 1 Malformation associations VATER complex (vertebral defects, anal atresia, tracheoesophageal fistula with esophageal atresia, radial and renal anomalies) VATERL complex (vertebral, anal, cardiac, tracheal, esophageal, renal, and limb anomalies) Musculoskeletal Abnormal rib number Deformed or reduced number of sacral vertebrae Dislocated hip Hemisacrum Hemivertebra Micrognathia Omphalocele Polydactyly Respiratory Choanal atresia Diaphragmatic hernia Hypoplastic lungs Subglottic stenosis Data adapted from Cho S, Moore SP, Fangman T. One hundred three consecutive patients with anorectal malformations and their associated anomalies. Arch Pediatr Adolesc Med 2001; 155:587-91.

sacral abnormalities also are frequent. Children with a persistent cloaca or rectovesical fistula have a 99% chance of having an associated genitourinary anomaly, whereas less than 10% of children with low fistula have such abnormalities. Overall, patients with additional anomalies are more likely to have high lesions than are patients with isolated anorectal malformations.43 Boys with low and high anorectal malformation have a high incidence of genital and

gastrointestinal anomalies, whereas urologic anomalies are more frequent in girls with high anorectal malformations.53 Long-term bowel dysfunction occurs in one third of boys with perineal fistula. In the first 24 hours of life, a decision should be made whether a child needs a colostomy or simple anoplasty. The presence of an associated defect, either urologic or cardiac, that might be life threatening requires immediate evaluation. A cloaca with a common channel shorter than 3 cm can be repaired by posterior sagittal intervention, whereas a common channel longer than 3 cm requires a laparotomy.51

ENTERIC NERVOUS SYSTEM Hirschsprung’s Disease

Hirschsprung’s disease (HD) is due to a congenital absence of ganglion cells in both the submucosal (Meissner’s) and myenteric (Auerbach’s) plexuses. Aganglionosis extends continuously for a variable distance proximal to the internal sphincter. Short-segment HD is most common with a transition zone from aganglionic colon to ganglionic colon at the level of the sigmoid. In long-segment HD the entire colon and even the small intestine may lack ganglia. With an incidence of 1 in 5000 live births, approximately 700 new cases of HD occur each year in the United States. The incidence is lowest in Hispanic and highest in Asian individuals. Approximately 10% of babies with Down syndrome have HD. Deletion of 17q21 and other chromosomal anomalies also have been reported.54 Familial occurrence has been reported in about 7% of cases. Familial cases have a male predominance with an increased incidence of long-segment aganglionosis. Affected families carry a high risk of familial recurrence of long-segment HD.55 HD is seen most commonly in full-term infants but on occasion does occur in premature births. In the short-segment type, a 4 : 1 male preponderance is observed, and in the long-segment type, the ratio is reduced to about 2 : 1. Short-segment HD accounts for nearly 90% of cases in childhood, and long-segment HD accounts for the remainder. It is rare that ultrashort-segment HD manifests in the pediatric population, but it does explain certain cases of chronic constipation that come to attention in adulthood. Pathogenesis Two pathogenetic mechanisms have been proposed for HD: (1) failure of migration of neural cells and (2) alteration of the colonic microenvironment. Genetic, vascular, and infectious factors are invoked to explain these alterations. Failure of Migration.  Between the fifth and twelfth weeks of gestation, premature arrest of the craniocaudal migration of vagal neural cells will result in HD. Colonic Microenvironment Changes.  A basic defect in the microenvironment necessary for the migration, development, and survival of ganglion cells has been postulated. Levels of various substances such as laminin, nicotinamide adenine dinucleotide phosphate (NADPH)-diaphorase, and neural cell adhesion molecules, as well as other polypeptides, have been shown to be reduced in the aganglionic segment. Some investigators have postulated that an alteration in the extracellular matrix with decreased concentrations of laminin and collagen IV constitutes a barrier to neutrophin 3, thereby perhaps impairing the neuroblastic migration and colonization. Neutrophin 3 promotes survival of sympathetic and sensory neurons in vitro and supports the growth and survival of differing subsets of neurons.

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine Table 96-5  Genes Involved in Hirschsprung’s Disease GENE

chromosome LOCATION

INHERITANCE

PHENOTYPE

PENETRANCE OF HD TRAIT

RET GDNF NTN SOX10 EDNRB EDN3 ECE1 ZFHX1B (SIP1) PHOX2B TCF4

10q11.2 5p13 19p13 22q19 13q22 20q13 1p36 2q22 4p12 18q21

AD AD AD AD AR/AD AD AD AD AD AD

HD HD HD WS4 WS4/HD WS4/HD HD, CFD, CD MCA-MR CCHS Epileptic encephalopathy

70% (male), 50% (female) Low Low 80% Low 5% Low 60% 20% Low

AD, autosomal dominant; AR, autosomal recessive; CCHS, congenital central hypoventilation syndrome; CD, cardiac defect; CFD, craniofacial defect; HD, Hirschsprung’s disease; MCA-MR, multiple congenital anomalies–mental retardation syndrome; SIP1, Smad-interacting protein; WS4/HD, combination of Shah-Waardenburg syndrome with Hirschsprung’s disease (see Table 96-6). Data from Amiel J, Sproat-Emison E, Garcia-Barcelo M, et al. Hirschsprung disease, associated syndromes and genetics: A review. J Med Genet 2008; 45:1-14.

Nitric oxide synthase is reduced in the aganglionic segment in HD, explaining the failure of relaxation of the affected colonic segment. Isolated case reports have linked the destruction of ganglion cells in segmental HD to cytomegalovirus infection and muscular hyperplasia of pericolonic vessels. The genetics of HD have now been characterized.18 Inheritance of the disease can be autosomal dominant, autosomal recessive, or polygenic. Penetration of mutations generally is low and depends on the extent of aganglionosis in affected family members. RET (rearranged during transfection) mutation penetrance is incomplete and sex dependent. It appears that the mutation, although increasing a child’s odds of having HD, is not predictive of the specific abnormality. Alterations of several genes have been implicated (Table 96-5).56-59 RET, a proto-oncogene that codes for a receptor tyrosine kinase protein (c-Kit), is the major susceptibility gene in HD, and maps to chromosome 10q11.2. More than 100 mutations of this gene have been identified and reduced c-Kit levels in the colon of patients with HD have been observed.54 Identified gene mutations currently account for only approximately half of all cases of HD, but it is recommended that RET exon 10 mutation analysis be done in all children with HD18; germline RET mutations also can cause multiple endocrine neoplasia type IIA (MEN-IIA). Although the test results will be negative in the vast majority of cases, the significance of identifying MEN-IIA mutation carrier status for that individual and family appear to justify such testing.54 Mutation of the RET has been noted in familial and sporadic HD. Congenital birth defects are found 5% to 33% of patients with HD.54 Although HD usually occurs as an isolated event, in 30% of the patients it may be part of a syndrome (Table 96-6). Clinical Features Most children with HD should be diagnosed in the newborn nursery. Any full-term infant who does not pass meconium within the first 48 hours of life should be suspected of having this disorder. Frequently, such infants will have abdominal distention and feeding difficulties. They also may have bilious emesis from partial bowel obstruction. Dilation of the empty rectum by the first examiner usually results in the explosive expulsion of retained fecal material and decompression of the proximal normal bowel. HDassociated enterocolitis occurs more frequently in the first three months of life, in patients with delayed diagnosis, in

Table 96-6  Some Congenital Anomalies and Syndromes Associated with Hirschsprung’s Disease Congenital Anomalies Cardiac (5% of cases) Septal defects Central nervous system (4% of cases) Distal limbs Gastrointestinal (4% of cases) Meckel’s diverticulum Pyloric stenosis Small bowel atresia Genital (2%-3% of cases) Hypospadias Renal (4% of cases) Dysplasia Agenesis Sensorineural Skin Syndromes Congenital central hypoventilation MEN-II (medullary thyroid cancer, pheochromocytoma, parathyroid hyperplasia) Movat-Wilson (characteristic facies, microcephaly, mental retardation) Piebaldism (hypopigmentation of skin and hair) Shah-Waardenburg (regional hyperpigmentation, white forelock, bicolored irides, sensorineural deafness) Smith-Lemli-Opitz (anteverted nostrils, ptosis of eyelids, syndactyly of second and third toes, hypospadias and cryptorchidism in males) Syndromes with limb abnormalities (metaphyseal dysplasia, McKusick-type—mild bowing of legs, irregular metaphyses, fine sparse hair) MEN, multiple endocrine neoplasia. Data from Amiel J, Sproat-Emison E, Garcia-Barcelo M, et al. Hirschsprung disease, associated syndromes and genetics: A review. J Med Genet 2008; 45:1-14.

children with trisomy 21, and with long-segment involvement; girls and children with a positive familial history also are more frequently affected. Enterocolitis may develop due to ischemia from colonic distention proximal to the aganglionic segment, with secondary infection from colonic bacteria; cases also have been reported of HD-associated enterocolitis in the aganglionic segment; C. difficile has been isolated in children with this enterocolitis. Mortality rates of up to 30% have been reported for enterocolitis, which remains the major cause of death in HD. Colonic

1637

1638

Section X  Small and Large Intestine perforation, most frequently involving the cecum and rarely the appendix, may occur, even in utero. Most commonly, infants younger than six months of age with HD will continue to have variable but significant constipation, punctuated by recurrent obstructive crises or bouts of fecal impaction, often with failure to thrive. The abdomen may be distended with fecal masses, and peristaltic waves may be visible. Anemia and hypoalbuminemia are common. Blood-flecked diarrhea should suggest the presence of enterocolitis, and immediate evaluation should be undertaken. As the child with HD grows older, problems continue, and fecal soiling occasionally may occur. An infant with HD who is breast-fed may have fewer difficulties with defecation because the high concentration of lactose in breast milk causes watery stools that are passed more easily. Once breast milk is discontinued, symptoms of HD may worsen. Diagnosis The child with symptomatic HD usually demonstrates signs and symptoms of bowel obstruction. The diagnosis may be made by one or a combination of the following tests: contrast enema, rectal biopsy, and anal manometry. Flexible sigmoidoscopy plays a complementary role in diagnosis. A contrast enema performed on an unprepared colon will show the distal narrowed hypertonic segment of bowel (usually seen best in a lateral projection). The transition zone between the narrowed distal and dilated proximal

intestine will be seen in the most common form of HD—the rectosigmoid form (Fig. 96-22A)—but may not be seen with long- or ultrashort-segment intestinal involvement. In ultrashort-segment HD, a radiologic picture indistinguishable from that in functional constipation with dilated bowel extending to the anus usually is seen. The transition zone may not be evident in rectosigmoid HD if the patient has undergone cleansing enemas or colonic irrigation before the study. Although it has been suggested that the transition zone may not be evident in the first six weeks of life, it almost always is noted in the neonate with partial bowel obstruction. Flexible sigmoidoscopy typically reveals a normal but empty rectum. The dilated proximal bowel, if within reach of the scope, is traversed easily, unless there is abundant feces in the lumen; occasionally stercoral ulcers may be seen. Anal manometry is the most reliable method by which the gastroenterologist can make the diagnosis of ultrashortsegment HD. A normal physiologic response to distention of the rectum is relaxation of (smooth muscle) internal sphincter pressure. In HD, not only does rectal distention fail to induce internal sphincter relaxation, but a paradoxical rise in external sphincter pressure often is seen (see Fig. 96-22B). Sufficient volumes of air must be used to stimulate rectal distention for a reliable study. A false-positive result most commonly is caused by a capacious rectum in a constipated child or with megacolon, in which case balloon

External sphincter

External sphincter

Internal sphincter

Internal sphincter

dg

10 mm Hg 5 sec

na

A

B

Rectal balloon 50 mL air

Rectal balloon 50 mL air

nt

C

D

Figure 96-22.  Hirschsprung’s disease. A, Film from a barium enema examination showing the transition zone between the narrowed distal aganglionic segment (na) and the proximal dilated ganglionic segment (dg). B, Anal manometry. Left tracing illustrates normal function. In the right tracing note the lack of relaxation of the internal sphincter on rectal distention in a patient with Hirschsprung’s disease. C, Photomicrograph of a rectal suction biopsy specimen showing the absence of ganglion cells and the presence of thickened nerve trunks (nt) characteristic of Hirschsprung’s disease. (Hematoxylin and eosin, ×125.) D, Acetylcholinesterase-positive fibers stained brown (arrows) in the muscularis mucosae and lamina propria, ×250. (A, Courtesy of Dr. J. Levenbrown; B, from Markowitz J. Gastrointestinal motility. In: Silverberg M, Daum F, editors. Textbook of Pediatric Gastroenterology. 2nd ed. Chicago: Year Book Medical Publishers; 1988.)

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine distention may not stimulate the reflex. Up to 20% of normal children have a falsely absent reflex, especially if they are premature or of low birth weight. Nonetheless, a positive response such as internal sphincter relaxation is strong evidence against HD. Suction biopsy of the rectal mucosa is the most reliable method of diagnosis, except in patients with ultrashortsegment HD. The biopsy capsule should be placed at least 2 cm above the mucocutaneous junction in infants and 3 cm above the junction in older children to avoid the physiologic hypoganglionic zone. To be certain of the absence of ganglion cells in the submucosal plexus, an experienced pathologist may need to review many serial sections. Hyperplastic sympathetic nerve fibers and pro­ liferating Schwann cells are associated findings (see Fig. 96-22C), but can be absent in total aganglionosis. Controversy exists regarding the type of stains necessary to make a diagnosis of HD. Because acetylcholinesterase is increased in the muscularis mucosae and lamina propria in the aganglionic segment (see Fig. 96-22D), staining for this enzyme has been used for many years. This technique requires fresh, non–formalin-fixed tissue and technical expertise; at best, this stain is confirmatory. False-positive and false-negative reports have been documented in total colonic aganglionosis.60 A variety of histochemical staining methods have been proposed for the identification of ganglion cells, but all are expensive, time-consuming, and unnecessary. In the neonate, considerations in the differential diagnosis of HD include other causes of intestinal obstruction, such as meconium ileus, ileal atresia, meconium plug syndrome, and the microcolon seen in infants of diabetic mothers. When symptoms and signs of enterocolitis are present, diagnostic possibilities in the neonate also include primary necrotizing enterocolitis, HD-associated enterocolitis, milk protein–induced colitis (see Chapter 9), and sepsis with possible disseminated intravascular coagulation. In the toddler or older child, HD must be differentiated from functional constipation (stool withholding, fecal retention). In the latter condition, history indicates that the child did pass meconium in the newborn nursery and that clinical problems did not arise until the child usually was at least 18 months of age. Fecal impaction almost always is present in fecal retention, and fecal soiling is characteristic. Children with anterior displacement of the anus may be more prone to fecal retention. Idiopathic pseudo-obstruction and intestinal neuronal dysplasia generally can be distinguished from HD by rectal biopsy. Management Definitive treatment of HD is surgical. In all instances, biopsy of the muscularis propria of the bowel is indicated at the time of surgery to assess for the presence of ganglion cells in the myenteric plexus and to delineate the proximal extension of aganglionosis. All full-term babies with meconium plug in the newborn nursery should be evaluated for HD before discharge, because approximately 15% of children with HD have a history of meconium plug. Discharge of any newborn with undiagnosed HD and consequent delay in operative intervention may result in a greater frequency of enterocolitis, increased morbidity and even mortality. The specific method of surgery is operator dependent. Long-term prognosis varies and may depend on the length of the aganglionic segment. Even in the most common form of HD (short segment), it is usual to see older children continue to have defecatory issues with fecal retention and encopresis. The exact reasons for these continuing problems remain unclear, but the mechanism may involve an intrinsic

A

B Figure 96-23.  Photomicrographs of a rectal biopsy specimen from a patient with intestinal neuronal dysplasia. A, Increased number of enlarged ganglia (arrows). B, Active inflammation of the rectal mucosa with a crypt abscess (arrow). (A and B, Hematoxylin and eosin, ×250.)

abnormality in what is described as normal colon or in the pacemaker system of the colon. In the future, cell therapy using precursor cells from the developing human enteric system might prove to be a therapeutic option.61

Intestinal Neuronal Dysplasia

Intestinal neuronal dysplasia (IND) is a motility disorder that manifests with intestinal obstruction or severe chronic constipation; characteristic biopsy findings include an increased number of enlarged ganglia and neural hypertrophy (Fig. 96-23A).62 In addition, acetylcholinesterase activity is increased in the lamina propria and muscularis mucosae. A full-thickness surgical biopsy specimen is often necessary to diagnose IND. IND has been reported as an isolated lesion affecting especially premature infants, or infants with a history of formula protein intolerance, ileal stenosis, or small left colon–meconium plug syndrome. Three types of IND have been defined. IND type A usually manifests acutely in the neonatal period as severe constipation and enterocolitis. Biopsy features include mucosal inflammation (see Fig. 96-23B), ulceration with hyperplastic neural changes limited to the myenteric plexus, and increased acetylcholinesterase activity in the lamina propria and muscularis mucosae. The submucosal plexus in this type of intestinal neuronal dysplasia is histologically normal.

1639

1640

Section X  Small and Large Intestine IND type B usually is seen in children between six months and six years of age who have chronic constipation and megacolon. Histopathologic findings include hyperplastic submucosal ganglia with increased acetylcholinesterasepositive fibers in the muscularis mucosae and lamina propria. Ectopic ganglion cells in the muscularis mucosae and lamina propria also have been described. No changes are seen in the myenteric plexus. Significant interobserver variation has been documented for the pathologic diagnosis of IND type B by rectal suction biopsies. Some reports have speculated that some of the morphologic features described in type B are normal age-related phenomena. A third, mixed type of IND has an acute presentation and involves both the submucosal and the myenteric plexuses. The pathogenesis of IND is controversial. In some patients it is congenital malformation, whereas in others it is an acquired phenomenon. IND also can be seen in association with other syndromes such as neurofibromatosis or MENIIB, in proximal-segment HD, and with congenital anomalies predominantly of the gastrointestinal tract.63 Other associated conditions include cystic fibrosis, microvillus inclusion disease, congenital anomalies, lipoblastomatosis, and inflammatory bowel disease. Therefore, IND may not represent a well-defined entity but rather may constitute a secondary phenomenon related either to age or to obstruction or inflammation.64 IND can resolve with age.

Chronic Intestinal Pseudo-obstruction

Congenital forms of neuropathic and myopathic pseudoobstruction are rare and sporadic, perhaps representing new mutations (see Chapter 120). In these situations, a family history of pseudo-obstruction is lacking, as are any asso­ ciated syndromes and evidence of other predisposing factors such as toxins, infections, ischemia, or autoimmune disease. Children with chromosomal abnormalities such as Down syndrome, as well as those with MEN-III or with Duchenne’s muscular dystrophy, may suffer from pseudo-obstruction.

MISCELLANEOUS AND GENETIC DEFECTS Microvillus Inclusion Disease

Congenital microvillus atrophy, also known as microvillus inclusion disease, is an autosomal recessive disorder that may manifest with severe diarrhea shortly after birth and is characterized by atrophy of the intestinal villi and characteristic electron microscopic findings.65 Although the prevalence of microvillus inclusion disease is not known, it is reported to be the most common cause of familial intract­ able diarrhea.66 A female gender predominance has been observed, and consanguinity is reported in 20% of cases. The incidence of microvillus inclusion disease may be higher among Navajo Indians and persons from the Middle East. Defective protein trafficking and abnormal cytoskeletal and microfilament function have been proposed as possible etiologies.67 A blockage in the transport pathway from the Golgi apparatus leads to fusion of the small vesicles into microvillus inclusions.68 Secretory diarrhea is severe, with intolerance to oral feeding and unresponsiveness to most therapeutic modalities. Three variants of microvillus inclusion disease are recognized: congenital, the most frequent and severe, manifesting within the first week of life; late-onset, starting at six to eight weeks; and atypical, with either early or late onset. The wall of the small intestine is paper-thin in micro­ villus inclusion disease. The mucosa of the duodenum and small bowel is characterized by villus atrophy, hypoplastic

lp

A

i

B Figure 96-24.  Photomicrographs of the duodenum from a patient with microvillus inclusion disease. A, Villus atrophy with crypt hyperplasia (arrow) and decreased cellularity of the lamina propria (lp). (Hematoxylin and eosin, ×250.) B, On electron microscopy, lack of or shortened microvilli (arrow) and a cytoplasmic inclusion (i), composed of a vesicle lined by microvilli, can be seen. (×15,000.) (Courtesy of S. Teichberg, PhD.)

or normal crypts, and normal or decreased cellularity of the lamina propria (Fig. 96-24A). The absence of the brush border membrane is demonstrated by lack of linear staining with PAS, carcinoembryonic antigen (CEA), and CD10.69 These stains also visualize the microvillus inclusions on light microscopy. Evaluation by electron microscopy reveals ultrastructural abnormalities of the microvillus membrane, including disruption or absence of the brush border membrane, shortening and absence of the microvilli, and microvillus inclusions (see Fig. 96-24B). Although these lesions are most commonly noted in biopsies from the small intestine, micro­ villus inclusions also may be seen in specimens from the rectum and colon. Total parenteral nutrition must be used to prolong survival. The secretory diarrhea persists but becomes less voluminous. Small bowel transplantation should be considered because without it, microvillus inclusion disease is fatal.70 Intestinal Epithelial Dysplasia Intestinal epithelial dysplasia (IED), also known as tufting enteropathy, is a congenital enteropathy with early onset, severe intractable diarrhea, and characteristic microscopic findings.71 In IED, there is a variable degree of villus atrophy. Surface epithelial cells are arranged in tufts with a round apex. Tufts can also been seen in the colonic mucosa. These epithelial cells have an abnormal expression of E-cadherin and do not contain inclusions on electron microscopic examination. In the basement membrane, heparin sulfate proteoglycan is increased, and laminin is faint and irregular.71

Chapter 96  Anatomy, Histology, Embryology, and Developmental Anomalies of the Small and Large Intestine The diarrhea is secretory, malabsorption intractable, and growth is impaired. Several cases of IED have been associated with congenital anomalies.71 Nonspecific punctate keratitis is observed in more than 60% of patients with IED. Most patients with IED have consanguineous parents or affected siblings. In the Middle East, IED is even more common than microvillus inclusion disease. IED is characterized by a basement membrane with abnormal distribution of 2 β1 integrin adhesion molecules along the crypt-villus axis.71 Tufts result from nonapoptotic epithelial cells that are no longer in contact with the basement membrane. Small bowel transplantation is required.

Congenital Glucose and Galactose Malabsorption

Familial glucose and galactose malabsorption, transmitted as an autosomal recessive trait, due to mutation in the SGlLT1 gene, is characterized by an absence of the active transport carrier protein (Na+-glucose cotransporter) for glucose and galactose.72 Ingestion of any formula containing glucose or galactose results in severe life-threatening watery diarrhea in the newborn period. Stools are strongly positive for reducing substances. Neither blood nor white blood cells are present in the stool. Findings on biopsy of the small bowel and colon are normal. Discontinuation of formula containing glucose, galactose, or lactose (lactose is metabolized to glucose and galactose) and institution of a fructosecontaining formula with resultant therapeutic benefit usually are sufficient to make a clinical diagnosis of glucose or galactose malabsorption. Diarrhea abruptly ceases and the newborn begins to thrive when fructose-containing formula feedings are instituted. Some reports indicate that the severity of the diarrhea from glucose or galactose malabsorption diminishes with age because of the increased capacity of the intestinal flora to metabolize glucose.

Congenital Sucrase and Isomaltase Deficiency

Because sucrose is not a common dietary carbohydrate during the first six months of life, watery stools generally do not develop in children with this disorder until sucrose is administered in baby food. An exception to this rule is in the newborn receiving a formula (usually with soy protein or casein hydrolysate) with sucrose as the carbohydrate. Because sucrose itself is not a reducing substance, to make the diagnosis, the stool must be hydrolyzed by boiling it with 1N hydrochloric acid for 20 minutes, thereby changing sucrose to glucose and fructose. Congenital sucrase or isomaltase deficiency, although extremely rare, is the most common congenital disaccharidase deficiency.

Congenital Lactase Deficiency

Congenital absence of lactase is extremely rare. Affected babies receiving a lactose-containing formula develop severe watery diarrhea, which resolves with the institution of a non–lactose-containing formula. Biopsy specimens of the small intestine are normal histologically, but assay for disaccharidases reveals diminished or absent lactase.

Congenital Chloride Diarrhea

Congenital chloride diarrhea is an autosomal recessive disorder of intestinal Cl-HCO3 exchange caused by mutations of the SLC26A3 gene.73 The chloride-bicarbonate exchange mechanism in the ileum and colon is reversed, and chloride is actively secreted, resulting in a chloride-rich diarrhea. The baby with congenital chloride diarrhea often is premature and may present with an ileus or absence of passage of meconium. Watery diarrhea with a high stool chloride content and low stool pH is lifelong. Increased absorption

of bicarbonate may result in dehydration, a hypochloremic metabolic alkalemia, hyponatremia, and marked hypokalemia. The stool contains no blood, no white blood cells, and no reducing substances. Urinary chloride is low. Biopsy specimens of the small intestine and colon are normal. Treatment is fluid and electrolyte replacement. Acid reduction with proton pump inhibitors has been tried with variable results.

Congenital Sodium Diarrhea

Congenital sodium diarrhea is caused by defective sodium or proton exchange.74 Patients have acidemia and hyponatremia. The stool concentration of HCO3 and sodium are increased.

Cystic Fibrosis

Cystic fibrosis is an autosomal recessive disorder of cyclic adenosine monophosphate chloride transport that results from a defect in the cystic fibrosis transmembrane regulator (CFTR) (see Chapter 57). Approximately 10% to 15% of newborns with cystic fibrosis present with neonatal meconium ileus or its complications. Meconium plug syndrome also may occur, resulting in colonic obstruction, rather than small bowel obstruction, as is seen with meconium ileus. Antenatally, small intestinal ischemia and perforation may occur, resulting in meconium cyst, intestinal atresia, or meconium peritonitis with intra-abdominal or scrotal calcifications.

KEY REFERENCES

Abdullah F, Arnold MA, Nabaweesi R, et al. Gastroschisis in the United States 1988-2003: Analysis and risk categorization of 4344 patients. J Perinatol 2007; 27:50-5. (Ref 22.) Amiel J, Sproat-Emison E, Garcia-Barcelo M, et al. Hirschsprung disease, associated syndromes and genetics: A review. J Med Genet 2008; 45:1-14. (Ref 54.) Auclair BA, Benoit YD, Rivard N, et al. Bone morphogenetic protein signaling is essential for terminal differentiation of the intestinal secretory lineage. Gastroenterology 2007; 133:887-96. (Ref 2.) Cho S, Moore SP, Fangman T. One hundred three consecutive patients with anorectal malformations and their associated anomalies. Arch Pediatr Adolesc Med 2001; 155:587-91. (Ref 43.) Fairbank TJ, Sala FG, Kanard R, et al. The fibroblast growth factor pathway serves a regulatory role in proliferation and apoptosis in the pathogenesis of intestinal atresia. J Pediatr Surg 2006; 41:132-6. (Ref 36.) Goulet O, Salomon J, Ruemmele F, et al. Intestinal epithelial dysplasia (tufting enteropathy). Orphanet J Rare Dis 2007; 2:20-29. (Ref 71.) Karnak I, Ocal T, Senocak ME, et al. Alimentary tract duplication in children: Report of 26 years experience. Turk J Pediatr 2000; 42:11825. (Ref 33.) Madison BB, Braunstein K, Kuizon E, et al. Epithelial hedgehog signals pattern the intestinal crypt-villus axis. Development 2005; 132:27989. (Ref 3.) Mastroiacovo P, Lisi A, Castilla EE, et al. Gastroschisis and associated defects: An international study. Am J Genet A 2007, 143:660-71. (Ref 21.) Meier-Ruge WA, Ammann K, Bruder E, et al. Updated results on intestinal neuronal dysplasia (INDB). Eur J Pediatr Surg 2004; 14:384-91. (Ref 62.) Pena A. Imperforate anus. In: Wyllie R, Hyams JS, editors. Pediatric Gastrointestinal Disease. Pathophysiology, Diagnosis, Management. 2nd ed. Philadelphia: WB Saunders; 1999. p 499. (Ref 49.) Penco JM, Murillo JC, Hernandez A, et al. Anomalies of intestinal rotation and fixation: Consequences of late diagnosis beyond two years of age. Pediatr Surg Int 2007; 23:723-30. (Ref 29.) Schonhoff SE, Giel-Moloney M, Leiter AB. Minireview: Development and differentiation of gut endocrine cells. Endocrinology 2004; 145:2639-44. (Ref 8.) Scoville DH, Sato T, He XC, Li L. Current view: Intestinal stem cells and signaling. Gastroenterology 2008; 134:849-64. (Ref 1.) Full references for this chapter can be found on www.expertconsult.com.

1641

CHAPTE R

97 Small Intestinal Motor and Sensory Function and Dysfunction Jane M. Andrews and L. Ashley Blackshaw

CHAPTER OUTLINE Anatomy  1644 Key Elements in Normal Small Intestinal Motor and Sensory Function  1644 Smooth Muscle  1644 Interstitial Cells of Cajal  1644 Neurons  1645 Gastrointestinal Hormones  1648 Integrated Control of Motility  1648 Peristalsis  1648 Interdigestive Motor Cycle  1648 Mechanisms Underlying Abnormal Motor and Sensory Function  1649 Smooth Muscle Dysfunction  1649

The two most important goals of small intestinal motor and sensory function are the efficient absorption of nutrients and the maintenance of orderly aboral movement of chyme and indigestible residues along the small intestine. Small intestinal motility is also critically important in preventing bacterial overgrowth within the intestinal tract. This is achieved by the net aboral flow of luminal contents during both the fed and the fasting states, probably with the assis­ tance of the gatekeeper function of the ileocecal junction, which prevents backflow of cecal contents and keeps small intestinal bacterial concentrations at their usual relatively low levels. Net movement of contents along the small intestine is antegrade, but retrograde flow also occurs normally over short distances. Optimal progression of luminal contents allows the optimal mixing of digested food with intestinal secretions and contact of contents with the epithelium; such contact is important for absorption and sensing of nutrients within the lumen. Both absorption and mucosal sensing of nutrients exert significant feedback control on gastric and small intestinal motor function, an interplay thought to optimize the rate at which additional nutrients are pre­ sented to the absorptive epithelium, and to minimize the amount of nutrients lost to the colon. Preceding emesis, and in association with nausea, gross retrograde movement of small intestinal contents occurs over long distances, when a unique pattern of a strong zone of phasic small intestinal contractions travels in an orad direction over a large portion of the small intestine. These contractions deliver luminal contents back to the stomach for ejection into the esophagus

Intrinsic Neural Dysfunction  1649 Extrinsic Afferent Dysfunction  1650 Measurement of Small Intestinal Motility  1651 Basic Principles  1651 Clinical Approach  1652 Normal in Vivo Small Intestinal Motility Patterns  1654 Contractions at a Fixed Point  1654 Contractions that Travel along the Small Intestine  1654 Patterned Motility  1654 Clinical Consequences of Disordered Small Intestinal Motility  1656 Approach to Patients with Possible Small Intestinal Motor Dysfunction  1657

during emesis. This coordinated motor pattern underscores the versatile modulation of small intestinal motility accord­ ing to physiologic need. The motor function of the small intestine depends directly on smooth muscle in the intestinal wall, which contains the basic control mechanisms that initiate contractions and control their frequency. Overlying these basic control mech­ anisms are the enteric nervous system (ENS) and the auto­ nomic nervous system (ANS). In addition, a number of hormones modulate the frequency and patterning of small intestinal contractions. Each of these factors plays a role in the motility of the small intestine in health; specific damage to each component in some diseases has helped to define their discrete roles. This chapter concentrates on the physiology of normal small intestinal motility. Anatomy is considered first, with discussion of the structural and functional elements that control sensory and motor activity. Neurophysiology, in­ tegrative control, and patterning of small intestinal motility are reviewed next, along with some insights into possible mechanisms underlying motor and sensory dysfunction. Basic, and then clinical, measurement techniques and limi­ tations in the evaluation of motility then are discussed, followed by descriptions of commonly recognized motor patterns. Finally, we present a more clinically directed commentary on specialized tests used to assess small intestinal motility, disease states in which small intestinal motor and/or sensory function is disturbed, and a general approach to the patient with suspected small intestinal motor dysfunction.

1643

1644

Section X  Small and Large Intestine ANATOMY The small intestine is approximately 3 to 7 meters long and extends from the duodenal side of the pylorus to the ileocecal valve. It is divided into three regions—duodenum, jejunum, and ileum—based on structural and functional considerations. Although some structural and functional differences exist among these three regions, they exhibit similar motor characteristics. At each end of the small intes­ tine, however, physiologic sphincters—the pylorus and the ileocecal valve—have distinctly different motor patterns that give them the ability to act as controllers of flow between the antrum and duodenum and between the ileum and colon, respectively. The motor function of the pylorus is discussed in Chapter 48, the ileocecal region is discussed in Chapter 98, and general anatomy is discussed in Chapter 96. The duodenum is a fixed, largely retroperitoneal struc­ ture located in the upper abdomen, and the distal ileum generally is anchored in the right iliac fossa by its attach­ ment to the cecum. Except for these regions, the small intestine is mobile within the peritoneal cavity.

KEY ELEMENTS IN NORMAL SMALL INTESTINAL MOTOR AND SENSORY FUNCTION SMOOTH MUSCLE The wall of the small intestine comprises the mucosa, con­ sisting of the epithelium and lamina propria; submucosa; muscular layer (muscularis); and serosa (Fig. 97-1). The muscularis is composed of inner circular and outer longi­ tudinal layers of smooth muscle, which are present in con­ tinuity along the length of the small intestine. Contractions within these layers are responsible for gross small intestinal motility. A much smaller additional muscular layer, the muscularis mucosae, is present between the mucosa and the submucosa and plays a role in mucosal or villus motility1 but does not contribute to gross motility and is not con­ sidered further in this chapter. The smooth muscle cells within each muscle layer form a syncytium. Myocytes communicate electrically with each other through physically specialized areas of cell-to-cell contact, called gap junctions, which are visible by electron microscopy. This intimate contact between adjacent myo­ cytes gives low-resistance electrical contact or coupling among them, thereby enabling them to be excited as a unit. Mechanical connections among myocytes in each layer enable them to function as a contractile unit. At a cellular level, the mechanical connections are provided by inter­ mediate junctions, and at a tissue level, mechanical connec­

tions are provided by the dense extracellular stroma of collagen filaments between bundles of smooth muscle cells. Within each layer, the smooth muscle cell bodies are arranged in parallel, so that the circular muscle layer encircles the lumen, and the longitudinal layer extends axially along the small intestine; each may be controlled independently. Hence, small intestinal muscle contractions reduce luminal diameter and/or shorten small intestinal length. The myocytes themselves are spindle-shaped cells that derive their contractile properties from specialized cyto­ plasmic filaments and from the attachment of these fila­ ments to cytoskeletal elements. On electron microscopy, condensations of electron-dense, amorphous material are noted around the inner aspect of the cell membrane (dense bands) and throughout the cytoplasm (dense bodies). The contractile filaments—actin and myosin—are arranged in a fashion similar to that in skeletal muscle and insert onto the dense bands and bodies approximately in parallel with the long axis of the cell. Thus, when the contractile filaments are activated to slide over each other, cell shortening results. Most of the Ca2+ required for activating the contractile appa­ ratus enters the cells via L-type Ca2+ channels (Fig. 97-2). Ca2+ entry also can be supplemented to a varying extent by release of Ca2+ from the sarcoplasmic reticulum membrane via IP3 receptor–operated Ca2+ channels. IP3 is generated by phospholipase C, which in turn is activated by G-proteins, coupled to receptors for excitatory transmitters (G-protein– coupled receptors). The increased cytoplasmic Ca2+ binds to the Ca2+ binding protein calmodulin, enabling it to activate myosin light chain kinase, which phosphorylates the 20 kD light chain of myosin (MLC20). Phosphorylation of MLC20 facilitates actin binding to myosin and initiates cross-bridge cycling and development of mechanical force. Phosphorylation of MLC20 is reduced by MLC phosphatase. Dephosphory­ lation of MLC20 reduces cross-bridge cycling and leads to muscle relaxation. The dephosphorylation process is under a complex system of hierarchical control, which is impor­ tant in setting the gain of smooth muscle contractility.2

INTERSTITIAL CELLS OF CAJAL

Interstitial cells of Cajal (ICC) are specialized cells within the smooth muscle layer that are vital for normal small intestinal motor function. ICC are pleomorphic mesenchy­ mal cells that form an interconnecting network via long, tapering cytoplasmic processes. ICC lie in close proximity to both nerve axons and myocytes, with which they form electrical gap junctions.3 ICC serve two roles in control of small intestinal motility: first, they act as pacemakers gen­ erating the electrical slow wave that determines the basic rhythmicity of small intestinal contractions4; second, they

Epithelium Mucosa Lamina propria Muscularis mucosae Submucosa Figure 97-1.  Diagram showing the layers and components of the small intestinal wall. DMP, deep muscular plexus; ICCIM, intramuscular interstitial cells of Cajal; ICCMY, myenteric interstitial cells of Cajal. (Advice from Dr. Elizabeth Beckett is acknowledged.)

Circular muscle Muscularis propria

Submucous plexus ICCIM (DMP) ICCMY Myenteric plexus

Longitudinal muscle Serosa

ICCIM

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction G-protein coupled receptors L-type

Ca2+ channel IP3

PLCβ

Ca2+ Calmodulin

Ca2+

Relaxation

MLC 20 MLCK MLCP MLC 20 (P)

Multiple intracellular pathways

Contraction

Figure 97-2.  Diagram of a smooth muscle cell showing pathways that lead to contraction and relaxation. See text for details. MLC 20; 20-kD myosin light chain; (P), phosphorylated; PLC, phospholipase C; MLCK, myosin light chain kinase; MLCP, myosin light chain phosphatase. (Modified from Sanders KM. Regulation of smooth muscle excitation and contraction. Neurogastroenterol Motil 2008;20 Suppl 1:39-53.)

transduce both inhibitory and excitatory neural signals to the myocytes5 and thus can vary the myocyte membrane potential and, in turn, contractile activity. This transduction occurs because ICC are interposed functionally between nerve terminals and the smooth muscle that the nerves supply. The neuroeffector junctions of the small intestine are not just simple contacts between nerve terminals and smooth muscle cells; they are contacts between enteric nerve terminals and ICC, and from there with myocytes by means of electrical gap junctions. Thus, effective neuro­ transmission results from the activation of specific sets of receptors on ICC, rather than by direct action on smooth muscle cells. At least three separate functional groups of ICC exist. They are the myenteric ICC (ICCMY), intramuscular ICC (ICCIM), and ICC in the deep muscular plexus (ICCDMP). Cells of the ICCMY population form a dense, electrically coupled network within the intermuscular space at the level of the myenteric plexus between the circular and longitudi­ nal muscle layers. ICCMY are the pacemaker cells in the small intestine that trigger generation of slow waves in the smooth muscle. These cells possess a specialized mecha­ nism that uses their oxidative metabolism to generate an inward (pacemaker) current, resulting from the flow of cations through nonselective cation channels in the plasma membrane. A primary pacemaker initiates slow waves. This depolarization from the primary event then entrains the spontaneous activity of other ICC within the network. This sequence results in a propagation-like phenomenon by which slow waves spread, without decrement, through the ICC network by means of gap junctions. A specialized type of ICCMY line the septa (ICCSEP) between circular muscle bundles; these cells form a crucial conduction pathway for spreading excitation deep into muscle bundles of the human jejunum, which is necessary for the motor patterns underly­ ing mixing.6 The second main population of ICC, ICCIM, is distributed within the muscle layers. ICCIM are innervated preferentially by intrinsic enteric motor neurons. In the small intestine, a third population, ICCDMP, which may be a specialized type of ICCIM in the small intestine, is concentrated at the

inner surface of the circular muscle layer at the region of the deep muscular plexus; it also receives preferential innervation. Both inhibitory and excitatory enteric nerve terminals selectively target intramuscular ICC. Their responses are transduced, in turn, to smooth muscle cells through gap junctions. Inputs from enteric excitatory motor neurons are mediated by muscarinic acetylcholine receptors (M2 and M3) and NK1 substance P receptors that result in increased inward currents, thereby causing depolarization. When depolarization reaches smooth muscle, it increases the opening of L-type Ca2+ channels during slow waves. These conditions result in greater Ca2+ entry and more forceful phasic contractions. Inputs from inhibitory enteric motor neurons are mediated by neurotransmitters including nitric oxide and vasoactive intestinal polypeptide, which activate both receptor and nonreceptor mechanisms in ICCIM. The result of these inputs is increased opening of K+ channels and, in turn, a stabilizing effect on membrane potential, reduced Ca2+ channel opening, and less forceful contrac­ tions of smooth muscle. Therefore, the mechanical response of small intestinal muscle to the ongoing slow wave activity depends strongly upon regulation of its excitability by the enteric nervous system via ICCIM. ICC, in general, play broadly similar roles in the small intestine and colon, and the reader is referred to Chapter 98 for a discussion of their roles in the large bowel (see also Fig. 98-2), as well as recent reviews by Sanders, Ward, and their colleagues.4,5 Absence or inactivity of ICC has been implicated in a number of clinical disorders that manifest as disturbed intestinal motility (see Chapter 20).

NEURONS

The small intestine is richly innervated with both extrinsic and intrinsic neurons. Intrinsic neurons have their cell bodies within the wall of the small intestine and constitute the ENS. These intrinsic neurons greatly outnumber the neurons of the extrinsic supply, which have their cell bodies outside the gut wall, but they have projections that end within the intestinal wall. Extrinsic neurons can be classi­ fied anatomically according to the location of their cell bodies and the route along which their projections travel. Extrinsic motor neurons belong to the ANS and connect the central nervous system (CNS) with the ENS and, from there, the small intestinal smooth muscle through the ICC. Some extrinsic motor neurons terminate directly in the muscle layers. Extrinsic sensory neurons from the small intestine do not belong to the ANS and are classified as spinal or vagal, depending on the route they follow to the CNS (Fig. 97-3). Neurons supplying the intestine are designated either afferent or efferent, depending on the direction in which they conduct information. By convention, information is conducted centrally by afferent neurons and peripherally by efferent neurons. Thus, the term afferent in regard to neural supply is used to describe pathways conducting informa­ tion that is detected in the intestine; in most texts “afferent” is interchangeable with the “sensory,” although most sensory information from the small intestine is not per­ ceived at a conscious level. The terms efferent and motor in regard to neural supply are used to describe pathways con­ ducting signals toward the effector small intestinal smooth muscle. Although the importance of motor innervation for motility is self-evident, the pivotal role of afferent function in determining motor responses has been less well appreci­ ated. The importance of the extrinsic afferent innervation is emphasized by the observation that at least 80% of vagal fibers are afferent.7

1645

1646

Section X  Small and Large Intestine CNS

Brain Nodose ganglion

Autonomic nervous system and sensory connections

Spinal cord Parasympathetic motor supply

Preganglionic sympathetic motor supply

Spinal sensory supply

Prevertebral ganglia Intestinofugal neuron Vagal sensory supply

Axon collaterals

Postganglionic sympathetic motor supply

Muscularis ICCs

Intestinal wall ENS

Figure 97-3.  Schematic representation of the relationships among components of the small intestinal motor control system. For further details, see text. CNS, central nervous system; ENS, enteric nervous system; ICC, interstitial cells of Cajal.

Intrinsic Neurons

ENS elements of the small intestine can be subdivided into three major functional groups: primary sensory (afferent) neurons, motor (efferent) neurons, and interneurons. Other categories of neurons, including secretomotor and vaso­ motor neurons and motor neurons to endocrine cells, are recognized, but they are not considered further in this chapter. Many distinct groups of enteric neurons are now well characterized both structurally and functionally and are reviewed in detail elsewhere.8,9 The cell bodies of ENS neurons are grouped together in the ganglia (clusters of cell bodies) of two main intramural plexuses. These plexuses lie in the submucosa (submucosal plexus) and between the two muscle layers (myenteric plexus). A deep plexus exists within the circular muscle but does not contain ganglia. The ganglia in the submucosal and myenteric plexuses are connected by interganglionic fasci­ cles. These fascicles are composed predominantly of the axons of motor neurons and interneurons, because sensory nerve processes do not often extend for any distance outside the ganglia. The myenteric plexus consists of ganglia spaced at regular intervals connected by a network of interganglionic fasci­ cles; this major network is known as the primary plexus. Within this main structure, smaller branches of nerve bundles arise from the primary plexus and form the second­ ary plexus, and still smaller branches form the tertiary plexus. The submucosal plexus has two layers, one close to the mucosa and another nearer to the circular muscle layer. These two layers are connected by interganglionic fascicles. The submucosal plexus does not have a hierarchy of subor­ dinate plexuses. Afferent Supply The primary afferent neurons of the ENS morphologically are Dogiel type II neurons (neurons with numerous pro­

cesses).10 Intrinsic primary afferent neurons that respond to mucosal chemical stimuli have their cell bodies in the myenteric plexus, and they project axons toward the mucosa. The myenteric plexus also contains the cell bodies of intrinsic afferent neurons that discharge in response to mechanical stimulation of the muscle layer induced by muscle activity or stretch. Intrinsic afferent neurons that respond to mechanical stimulation of the mucosa also are believed to exist, based on enteric reflexes seen in extrinsically denervated preparations. The cell bodies and processes of these neurons have not yet been identified definitively, although available evidence is consistent with the presence of their cell bodies in the submucosal ganglia.10 Intrinsic sensory neurons synapse in the intramural plexuses with intrinsic motor neurons and interneurons, which they excite mainly by release of acetylcholine and substance P. A more detailed account of the function and role of intrinsic afferent neurons can be found in a review by Furness and coworkers.10 Observations indicate that other classes of enteric neurons also respond to mechanosensory stimuli, suggesting that the ENS behaves as a sensorimotor network rather than as sepa­ rate components.11 Efferent Supply The axons of the intrinsic motor neurons that supply small intestinal smooth muscle exit the intramural ganglia and enter either the circular or the longitudinal muscle layer, where they pass in close proximity to both the myocytes and ICC. No specific neuromuscular junctions are present in small intestinal smooth muscle as in skeletal muscle, although the multiple varicosities along the motor axons probably represent specialized areas of neurotransmission. The motor axons discharge along their length, potentially activating large numbers of myocytes through ICC but pos­ sibly also directly activating them. The lack of exclusive, specific neuromuscular junctions, the electrical gap junc­ tions among myocytes, and the overlap of innervation of myocytes from more than one motor axon mean that func­ tionally discrete motor units in the intestinal smooth muscle do not appear to exist, in contrast with skeletal muscle. The ENS motor supply itself is both inhibitory and excitatory, and intrinsic motor neurons generally contain both a fast and a slow neurotransmitter. The pre­ dominant excitatory transmitters are acetylcholine (fast) and substance P (slow), and the predominant inhibitory trans­mitters are nitric oxide (fast), vasoactive intestinal polypeptide (VIP) (slow), adenosine triphosphate (ATP) (fast), and the nucleotide β-nicotinamide adenine dinucleo­ tide (β-NAD).12 Interneurons Interneurons connect ENS neurons of the same class or of different classes with one another. They permit local com­ munication within limited lengths of intestinal wall (mea­ sured in millimeters or centimeters) and are implicated in simple local responses by means of release of acetylcholine or nitric oxide, depending on their oral or aboral direction of projection. Some evidence also suggests the presence of connections within the intestinal wall along greater dis­ tances, but these neural pathways are not well defined. These connections may be provided anatomically by the ENS or by connections between the ENS and ANS. Inter­ neurons that play an additional sensory role have been identified.11 A special type of interneuron, the intestinofugal neuron, may be important for controlling local reflexes. Intestinofu­

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction gal neurons have cell bodies within the myenteric plexus. These cell bodies receive input from several local enteric neurons and project to the prevertebral ganglia, where they synapse with sympathetic motor neurons (see Fig. 97-3).13

Extrinsic Neurons

Afferent Supply The small intestine is innervated by vagal and spinal extrin­ sic afferents. The pathway of small intestinal vagal afferent innervation is relatively straightforward. The vagal afferent neurons have endings in the intestinal wall and cell bodies within the nodose and jugular ganglia, which deliver input directly to the brainstem. Spinal afferent fibers travel along perivascular nerves to the prevertebral ganglia, where neurons do not end but might give off axon collaterals that synapse on postganglionic sympathetic motor neurons; these fibers then pass into the thoracic spinal cord along the splanchnic nerves. Spinal afferent neurons have their cell bodies throughout the thoracic dorsal root ganglia and enter the spinal cord through the dorsal roots; they synapse mainly on neurons of the superficial laminae of the spinal gray matter. These neurons, in turn, can send projections to numerous areas of the brain involved in sensation and pain control. Spinal afferent neurons also can give off axon collaterals closer to the intestinal wall, which synapse on components of the ENS, blood vessels, smooth muscle, or secretory elements (see Fig. 97-3). The different stimulus response profiles of vagal and splanchnic mechanoreceptors are generally interpreted as evidence that vagal afferents subserve physiologic regulation, and splanchnic afferents mediate pain.14-16 Functionally, three distinct and characteristic patterns of terminal distribution can be identified within the intestinal wall. Extraluminal afferent fibers have responsive endings on blood vessels in the outer, serosal layer and in the mes­ enteric connections. Muscular afferents form endings either in the muscle layers or in the myenteric plexus.17 Mucosal afferents form endings in the lamina propria, where they are positioned to detect substances absorbed across the mucosal epithelium or released from epithelial and subepithelial cells, including enterochromaffin and immunocompetent cells.17 These three different populations of afferent endings have different sensory modalities, responding to both mechanical and chemical stimulation.14,18 Serosal and mesenteric affer­ ents are found mainly in the splanchnic innervation and are activated by distortion of the intestine and its attachments; they do not normally signal distention or contraction of the bowel wall unless it is strong enough to cause distortion of the outer layers. Serosal and mesenteric receptors also com­ monly show evidence of chemosensitivity. This observation hints at potential responsiveness to circulating or locally released factors, especially in view of the localization of these receptors on or near blood vessels.19 Muscular afferents respond to distention and contraction with lower thresholds for activation, and they reach maximal responses within levels of distention that are encountered normally during digestion. Muscular afferents show main­ tained responses to distention of the small intestine and signal each contractile event, giving rise to the term in-series tension receptors. Nerve tracing studies have identified vagal afferent terminals in the longitudinal and circular muscle layers described as intramuscular arrays (IMAs), consisting of several long (up to a few millimeters) and rather straight axons running parallel to the respective muscle layer and connected by oblique or right-angled short connecting branches.17,20 IMAs were proposed to be the inseries tension receptor endings, possibly responding to both

passive stretch and active contraction of the muscle, although direct evidence for this proposal is currently lacking. Vagal afferent terminals surrounding the myenteric plexus throughout the gastrointestinal tract have been described as intraganglionic laminar endings (IGLEs). These endings are in intimate contact with the connective tissue capsule and enteric glial cells surrounding the myenteric ganglia, and they have been hypothesized to detect mechan­ ical shearing forces between the orthogonal muscle layers. Evidence for such a mechanosensory function of IGLEs has been elaborated by mapping the receptive field of vagal afferent endings in the esophagus, stomach, and large intes­ tine, showing morphologically that individual hot spots of mechanosensitivity correspond with single IGLEs.21 Func­ tional evidence exists for muscular afferents in both the vagal and the spinal innervation, but the appearance of spinal distention-sensitive afferents in the small intestine is yet to be determined. It is likely to be distinct from that of vagal afferents due to their higher thresholds for distention.22 Small intestinal mucosal afferents have been found in the vagal supply, but their existence in the spinal supply can be inferred only from the fact that they exist in the colon.14,19 Mucosal afferents do not respond to distention or contrac­ tion but are exquisitely sensitive to mechanical deformation of the mucosa, as might occur with particulate material within the lumen.16,23 In the rat duodenum and jejunum, vagal afferent fibers penetrate the circular muscle layer and submucosa to form networks of multiply branching axons within the lamina propria of both crypts and villi.17 Termi­ nal axons are in close contact with, but do not seem to penetrate, the basal lamina and thus are in an ideal position to detect substances including absorbed nutrients and mediators that are released from epithelial cells and other structures within the lamina propria. Efferent Supply The extrinsic efferent pathways to the small intestine are supplied by the parasympathetic and sympathetic divisions of the ANS. The small intestinal parasympathetic supply is cranial and cholinergic, whereas the sympathetic supply is spinal (thoracic) and adrenergic. These two motor pathways are not entirely separate, however, because postganglionic sympathetic fibers arising from cervical ganglia sometimes are found within the vagus nerve. The parasympathetic motor neurons of the small intestine have cell bodies within the dorsal motor nuclei of the vagi in the medulla oblongata. Their axons extend through the vagi to the intestinal intramural plexuses, where they synapse with motor neurons of the ENS. The sympathetic motor supply is more complex: Primary motor neurons within the intermediolateral horn of the thoracic spinal cord synapse with second-order neurons in the prevertebral ganglia, which then synapse with ENS motor neurons within the intestinal intramural plexuses, directly with smooth muscle, or possibly with ICC. Both excitatory and inhibitory extrinsic motor outputs to the small intestine are recognized. Excitatory outputs depo­ larize, and inhibitory outputs hyperpolarize the smooth muscle, thereby facilitating and impeding the development of contractions, respectively. In general, the sympathetic motor supply is inhibitory to the ENS, and this ENS inhibi­ tion leads to decreased smooth muscle activity, with the opposite effect seen in sphincter regions. Direct sympathetic inhibitory and excitatory outputs to smooth muscle also exist. The parasympathetic motor output to the ENS is more diffuse, each primary motor neuron supplying a large area.

1647

1648

Section X  Small and Large Intestine Excitatory parasympathetic motor output occurs to either inhibitory or excitatory ENS motor neurons, through which parasympathetic efferents can selectively inhibit or excite smooth muscle.

Central Connections of Neural Control Elements

Centrally, the sensory and motor supplies to the small intes­ tine are closely interrelated; the vagal sensory input and the parasympathetic motor output are closely located, as are the spinal sensory input and the sympathetic motor output. Both the vagal parasympathetic and the spinal sympathetic supplies have widespread connections to many other areas throughout the CNS that are implicated in feeding, arousal, mood, and other reflex behavior. The proximity of these CNS areas involved in small intestinal regulation, and their interconnections, makes it likely that the vagal parasympa­ thetic and the spinal sympathetic control mechanisms are interconnected and might function less independently than has been previously thought. The parasympathetic primary motor neurons are located bilaterally in the dorsal motor nuclei of the vagus in the medulla, which lie close to and receive substantial input from neurons of the nuclei tracti solitarii (NTS). The NTS is the site of terminals of vagal afferent fibers, which enter through the tracti solitarii and have cell bodies in the nodose ganglia. Each NTS also has extensive connections to other CNS regions, and several of these regions have input to the dorsal motor nuclei of the vagus, thereby influencing vagal motor output to the intestive. The central connections of the spinal and sympathetic supply to the gut are less well described. The spinal sensory neurons enter the spinal cord, where they synapse ipsilaterally on second-order sensory neurons and also provide direct feedback to sympathetic preganglionic motor neurons through axon collaterals. The second-order sensory neurons then ascend the spinal cord either contra­ laterally or ipsilaterally, after which they terminate in numerous areas,15 including the raphe nuclei and periaq­ ueductal gray matter in the brainstem and the thalamus. The thalamus has extensive ramifications throughout the CNS. The central influence on sympathetic motor output to the small intestine is complex and not well understood, but stress and arousal level play a role. These influences have their output through the brainstem and descending tracts to the sympathetic preganglionic motor neurons in the intermediolateral horn of the spinal cord, which send their axons to the prevertebral ganglia, whereupon they synapse with sympathetic postganglionic adrenergic nerves.13

GASTROINTESTINAL HORMONES

Gastrointestinal hormones are dealt with in detail in Chapter 1, but it is important to emphasize here their vital role in modulation of small intestinal motor and sensory function. Gastrointestinal hormones relevant to small intestinal function can act in either a humoral or paracrine fashion on both enteric neurons and myocytes, and generally they are released in response to the presence (or anticipation) of enteral nutrition. The best known of these hormones include CCK, somatostatin, VIP, glucagon-like peptide-1 (GLP-1), gastric inhibitory peptide (GIP), ghrelin, and motilin. Most of the hormones released in response to the presence of food in the lumen lead to slowing of small intestinal transit, signals of satiety, and increased mixing or segment­ ing contractions (see later). For a detailed description of these hormones and their effects, the reader is referred to Chapter 1.

INTEGRATED CONTROL OF MOTILITY So far we have considered the structure and function of individual components of the neuromuscular apparatus of the small intestine. When we consider how these com­ ponents operate together to produce known motility pat­ terns, several gaps are revealed in our knowledge, because the evidence for contribution of specific mechanisms is often circumstantial. Two important examples of motility patterns—peristalsis and the interdigestive motor cycle (IDMC)—are described next. These motor patterns illustrate the involvement of integrated hierarchical levels of control and our current level of understanding of the control systems.

PERISTALSIS

Peristalsis is the fundamental integrated motility pattern of the small intestine and can be coordinated entirely within the ENS and muscular layers. It may be initiated in response to a number of mechanical and chemical stimuli in the lumen and consists of progression of contractile activity usually, but not always, in an aboral direction. Therefore, both sensory and motor aspects to peristalsis are recognized. The populations of intrinsic primary afferent neurons described earlier probably are responsible for detection of luminal stimuli, either directly or following release of medi­ ators from mucosal enteroendocrine cells. Their activation results in transmitter release onto neighboring interneurons and motor neurons whose activity is coordinated subse­ quently as a network to provide synchronous activation of circular and longitudinal muscles on one side of the bolus (usually the oral side) and synchronous inhibition of muscle on the other side. This networked activity normally travels aborally, but the mechanism of propagation is not yet under­ stood. It might result from patterns of activity in interneu­ rons that can project over distances of several millimeters and thus mediate a general descending excitation. The mechanism by which peristalsis is reversed—for example, in conditions of luminal toxicity—is not known, but the fact that reverse peristalsis does occur in the small intestine illustrates that the pattern is not a totally polarized phenomenon. Debate is ongoing about the precise interactions of trans­ mitters and mediators in the normal function of peristalsis, but peristalsis is known to be affected by exogenous activa­ tion of several pre- and postsynaptic mechanisms, some of which also may be active endogenously. Of particular inter­ est are serotoninergic mechanisms, which have been shown to have involvement in initiation of peristalsis and modula­ tion of transmission between subclasses of enteric neurons.

INTERDIGESTIVE MOTOR CYCLE

The IDMC is discussed here because it serves to demon­ strate the extraordinary integrative capacity of the ENS; other aspects of the IDMC are described later in this chapter. The IDMC is a complex series of periods of variable con­ tractile activity with distinct phases showing different con­ tractile amplitudes, propagation, and regularity. The pattern as a whole sweeps slowly down the small intestine in the fasting state and recurs at regular intervals. Although a number of candidate hormones are proposed to be involved in its initiation and recurrence, the switch between quies­ cent and active phases and their orderly migration along the bowel are functions of the ENS; this ENS autonomy is dem­ onstrated by occurrence of the IDMC in extrinsically dener­

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction vated or autotransplanted intestine. The ENS therefore is capable of controlling large segments of the small intestine independent of extrinsic input, probably by virtue of its extensive interneuronal connections and constant sensory feedback. Although the ENS has this regulatory capacity, normal function is modulated by ANS efferent output, which in turn may be influenced by locally or centrally processed information gathered from primary spinal or vagal afferents. In particular, synapses outside the CNS in the prevertebral ganglia are capable of subserving inhibitory intestinointestinal reflexes that are potentially important in the minute-to-minute regulatory control of motility.13 Small intestinal neuromuscular function is also influenced by a number of hormones acting in either endocrine or paracrine fashion. Little direct information is available on the precise con­ tribution of each extrinsic pathway to motor function of the small intestine in humans. Vagal reflexes generally are thought to make an important contribution in the integra­ tion of major homeostatic functions, such as motility, secre­ tion, blood flow, and the control of food and water intake.14-16 The role of sympathetic reflexes is thought to be concerned primarily with inhibition of motility and other functions in response to noxious stimuli, rather than in digestive small intestinal functions.

MECHANISMS UNDERLYING ABNORMAL MOTOR AND SENSORY FUNCTION Much of the evidence for the mechanisms involved in dys­ function of the small intestine is derived from animal models in which mucosal inflammation or infection has been induced, after which alterations in physiology, pharma­ cology, and anatomy of motor and sensory elements are assessed. These models provide some clues to the underly­ ing mechanisms involved in motor abnormalities seen clini­ cally; however, because many clinical manifestations are of unknown etiology, this approach is limited in the extent to which basic findings can be translated directly. Infection and inflammation of the intestine can result in long-term changes in all elements, including myocytes, ICC, and intrinsic and extrinsic neurons. Symptoms in func­ tional gastrointestinal diseases such as functional dyspepsia and irritable bowel syndrome (IBS) may be attributable partly to specific sensorimotor abnormalities occurring locally in the intestine, but they also are attributable to alterations in the extrinsic neural control system of the intestine and possibly to alterations in central perception, processing of afferent information, or both (see Chapters 13 and 118). Abnormalities in pain control systems in the brain and disordered processing of affective components of vis­ ceral sensations also have been described in these condi­ tions24 and can produce symptoms through the central connections described in the preceding sections. Some clinical scenarios in which discrete abnormalities have been identified or hypothesized in small intestinal motility are outlined in Table 97-1.

SMOOTH MUSCLE DYSFUNCTION

It is often difficult to separate pathologic changes in the function of smooth muscle from those in neural control mechanisms; however, a number of changes can be attrib­ uted directly to alterations in smooth muscle. Cytokines

play an important role in the abnormal smooth muscle function associated with gastrointestinal inflammation and infection. Different insults induce different patterns of cyto­ kines, which in turn determine the type of infiltrating immune or inflammatory cells, which in turn release spe­ cific mediators. Thus, the resultant effect on smooth muscle function depends on the origin of disease. For example, nematode infection induces mastocytosis and eosinophilia, which lead to activation of intracellular signaling pathways in smooth muscle by IL-4 and IL-13, ultimately resulting in hypercontractility of smooth muscle.25 By contrast, chemically induced inflammation is characterized by the presence of neutrophils and macrophages among other cells. Inflammation and infection can lead to changes at sites in the small intestine distant from the affected site, and the functional effects of inflammation in smooth muscle can persist following recovery from the acute insult as is seen with post-infection IBS. Smooth muscle hyperresponsiveness may be characterized by enhanced responses to cholinergic and noncholinergic excitation and are observable in human inflammatory bowel disease.26

INTRINSIC NEURAL DYSFUNCTION

Several abnormalities of small intestinal intrinsic control are attributable to developmental dysfunction and are dealt with separately in Chapter 96. Changes in the ENS also can occur after a bout of intestinal infection or inflammation. Many of these changes are centered on the intrinsic primary afferent neurons. These neurons become more excitable because of changes in the expression of ion channels that initiate generation of action potentials and those that deter­ mine recovery of membrane potential after an action poten­ tial. Thus, the long after-hyperpolarization that characterizes intrinsic primary afferent neurons from other classes is shortened, and they are able to fire in longer trains. This ability directly affects the responses of other interneurons and motor neurons that receive inputs from these afferent neurons and that therefore are involved in intrinsic (ENS) reflexes. Changes in excitability may be observed during an acute phase of infection or inflammation,26 or for several weeks afterward,27 at least in the large intestine. These longer-term changes are referred to as plasticity and might partly explain the occurrence of exaggerated motor responses to a given stimulus in the acute phase and after recovery of mucosal lesions. Changes can result from alterations in gene expression in enteric neurons that persist beyond the initial insult, from persisting increases in locally released media­ tors following alterations in mucosal cell types, or from both types of responses.28 In animal models of insulin-dependent diabetes mellitus, altered levels of neuropeptides may be seen, which might explain the disordered motility noted clinically in diabetes mellitus. The only reported neuroanatomic human study in a patient with type 1 diabetes mellitus showed that ICC were markedly decreased throughout the entire thickness of the jejunum. A decrease in neuronal nitric oxide synthase, VIP, pituitary adenyl cyclase–activating peptide (PACAP), and tyrosine hydroxylase–immunopositive nerve fibers was observed in the circular muscle layer, and substance P immunoreactivity was increased.29 Although patients with type 1 diabetes mellitus and sympathetic denervation have abnormally slow gastric emptying (see Chapter 48), their transit of a liquid meal through the distal small intestine is more rapid, which might play a part in the production of diarrhea. Diabetic patients also show abnormal duodenal motility patterns such as early recurrence of phase III after

1649

1650

Section X  Small and Large Intestine Table 97-1  Disorders Associated with Abnormal Small Intestinal Motility DISORDER

SMOOTH MUSCLE ABNORMALITIES

Irritable bowel syndrome

None identified

Post-infection and inflammatory changes (subset of patients)

Acute severe illness

Decreased strength of contractions

Altered neurotransmission (? related to metabolic or electrolyte disturbances)

Pregnancy

Decreased strength of contractions None identified

None identified

Metabolic disturbances*

Possible decreased strength of contractions

Altered neurotransmission

Drugs†

Possible decreased strength of contractions

Intestinal obstruction

NEURAL ABNORMALITIES

SENSORY ABNORMALITIES

POTENTIAL OUTCOMES

Altered afferent function Increased visceral sensitivity Heightened sensitivity to neurohumoral feedback loops None identified

Heightened sensation Disordered motility

Altered neurotransmission

Enhanced perception of gastrointestinal stimuli Nausea Altered sensory perception None identified

Hypertrophy, if chronic

None identified

None identified

Pseudo-obstruction syndromes

Myopathy of hollow viscera

None identified

Scleroderma and other connective tissue disorders

Ischemia and fibrosis

Neurologic syndromes‡

NA

Multiple neural abnormalities: neuron loss, plexus abnormalities, altered distribution of neurotransmitters Nerve loss in intestinal wall Extrinsic neural supply may also be damaged by vasculitis Neural absence or loss

Abnormal patterning of contractions Slow or rapid transit Abnormal patterning of contractions Slow or rapid transit Ileus Slow or rapid transit Disordered contractions High-amplitude forceful contractions Feeble contractions Absent phase III of the IDMC Slow or failed transit

Rare myopathies

Myocyte and mitochondrial abnormalities; inadequate contractile force Fibrosis (? related to ischemia)

NA

None identified

Diabetes mellitus

Radiation enteritis

Acromegaly

Altered neurotransmission

Delayed transit Ileus Decreased absorption Slowed transit

None identified

Feeble contractions Thickening of bowel wall Slow transit

Loss of afferent neurons with consequent loss of sensory information for reflux control

Disorganized IDMC Failure to convert to fed pattern Transit failure

Increased muscarimic (m) 3 receptors

None identified

Postulated autonomic nerve dysfunction

None identified

Less mixing, disordered transit Stasis, bacterial overgrowth, diarrhea Delayed orocecal transit time Bacterial overgrowth

Insufficient force for transit and mixing

IDMC, interdigestive motor cycle; NA, not appropriate. *Examples include disturbances of potassium, calcium and magnesium homeostasis, and renal and hepatic failure. † Examples include antidepressants, calcium channel blockers, and beta-blockers. ‡ Examples include dysautonomia and Parkinson’s disease.

a meal (see later). No consistent correlation, however, has been found between changes in manometric parameters and the degree of cardiac autonomic neuropathy, nor has any correlation yet been established between changes in enteric neurotransmitters and ICC and manometric and transit observations.

EXTRINSIC AFFERENT DYSFUNCTION

Mechanisms leading to extrinsic afferent dysfunction after infection or inflammation probably are similar to those involved in intrinsic primary afferent and smooth muscle dysfunction. It is well established that a wide range of chemical mediators can influence mechanosensitivity of extrinsic primary afferents, in addition to evoking direct responses as detailed earlier. These chemical mediators can be released in conditions of inflammation, injury, or

ischemia from a variety of cell types including platelets, neutrophils, lymphocytes, macrophages, mast cells, glial cells, fibroblasts, blood vessels, muscles, and neurons. Each of these specific cells can release several modulating agents, some of which act directly on the sensory nerve terminal; others act indirectly, causing release of other agents from other cells in a series of cascades. The end result of these actions is that the response properties of extrinsic afferents, like their intrinsic counterparts, are subject to plasticity, usually resulting in an increased sensitivity of the afferent endings; this process is described as peripheral sensitization. Some evidence supports the involvement of algesic medi­ ators, including prostaglandins and purines, in changes leading to peripheral sensitization.30 Other endogenous chemical mediators, including somatostatin, can down-

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction regulate small intestinal afferent sensitivity such that an imbalance in prosensitizing and antisensitizing mecha­ nisms leads to a disordered sensory signal. Such mecha­ nisms are likely clinically relevant to functional bowel disorders, such as IBS, in which increased perception of mechanical and chemical stimulation is apparent. More­ over, because these afferents also serve to trigger reflex mechanisms that control and coordinate intestinal motor function, their sensitization can contribute to chronic dys­ motility, resulting in a cycle of disordered sensory and motor function.

1.5 cm intervals

Antrum

MEASUREMENT OF SMALL INTESTINAL MOTILITY BASIC PRINCIPLES Spatiotemporal Measurements

The outcomes of small intestinal motor activity depend on the patterning of small intestinal contractions in both space and time: Where and when do the contractions occur with respect to each other? Measurement methods must therefore gather functionally relevant information on the temporospa­ tial organization of small intestinal motility. This presents substantial challenges, especially in humans, because of the length of the small intestine, the spatiotemporal complexity of motor events, and the long time frame (several hours) over which small intestinal motility determines the success­ ful absorption and movement of each meal. In health, the occurrence and patterning of a large number of individual motor events determine the outcomes of absorption and transit, so that whole-animal measures of small intestinal transit and absorption yield a gross, or summary, report. More-detailed descriptions of small intes­ tinal motility report great variability in the patterning of individual contractile events, depending in part on the tech­ nique used, the time frame over which motility is observed, and the temporospatial resolution of the measurement tech­ nique itself. To understand the relationship between individual motor events and transport in the small intestine, the temporal resolution of the measurement technique must be greater than the duration of each discrete motor event. Based on similar principles, the spatial resolution of measurements is also an important parameter to consider if relationships between motor events and intraluminal flow(s) are to be defined. The importance of spatiotemporal resolution can be appreciated by considering Figure 97-4. Direct evalua­ tion of small intestinal motility requires methods of mea­ surement with a time resolution of at least two seconds, because in humans, the intrinsic frequency of duodenal contractions is up to 12 per minute. The optimal spatial resolution for studies of small intestinal motor function has not been determined, but the spatial patterning of pressures is known to vary over relatively small distances,31 with most propagating pressure wave sequences traveling less than 6 cm. Because of practical limitations of data handling and the number of sensors one can place in the small intestine, measurement techniques usually either achieve high tem­ porospatial resolution over a short distance or low tempo­ rospatial resolution over a far greater distance. Realistically this means that data gained from different studies are usually interpreted alongside one another to provide more complete information.

Evaluation of Single Cell Functions

At the cellular level, a number of techniques can be used to yield insights into small intestinal motor physiology.

Distal duodenum

Antrum

4.5 cm intervals

Distal duodenum Antrum

Distal duodenum

6 cm intervals

50 mm Hg 0 mm Hg

30 seconds

Figure 97-4.  Multichannel manometric recordings of the human antrum and duodenum, with recording points placed at varied intervals: 1.5 cm (top panel), 4.5 cm (middle panel), and 6 cm (bottom panel). These data demonstrate some of the limitations of varying the interval between recording points: As a phasic contraction travels along a section of intestine, the associated rise in pressure is detected only at each measurement point. If the interval between recording points is too wide, unrelated pressures may be judged to be related to the propagated pressure wave, or a propagated pressure wave sequence may be judged to be a limited phasic event. Spatial detail is lost as the recording interval is widened. (Courtesy Dr. J.M. Andrews.)

Intracellular recordings of electrical potential can be obtained from a number of cell types within the small intes­ tine and its extrinsic neural control system. These record­ ings give detailed information about the signals received and transmitted by individual cells, with excellent temporal resolution, but generally they cannot be applied concur­ rently over a significant length of intestine and therefore have limited real-time spatial resolution with regard to motor events. A combined functional and neuroanatomic approach whereby imaging of specific neurons with intracellular or extracellular recordings and chemical coding using immu­ nohistochemistry are performed concurrently has allowed important correlations to be made between structure

1651

1652

Section X  Small and Large Intestine and function. In particular, this approach has led to under­ standing of the function of IGLEs and IMAs (see “Extrinsic Afferent Supply” earlier). Although electrophysiologic and anatomic methods provide information on structure, neurotransmitters used, and proximity to other elements, they cannot describe pre­ cisely how these relate to the actual resulting motility and its temporospatial organization. Although these single-cell techniques generally have been applied to animal tissues, the results also probably apply to humans, because a similar structural organization of control elements is seen in human tissue.

Recording of Muscle Contractions

Increased muscle tension generally is directly recorded with strain gauges; these can be used in muscle strips, isolated loops of intestine, and whole-organ preparations or even chronically implanted in animals. Strain gauges are capable of excellent temporal resolution of motor events, but spatial resolution is limited by the size and number of strain gauges that are used concurrently. Over short lengths of intestine, a spatial resolution of approximately 1 cm is possible. Unfortunately, strain gauges are not suitable for use in human subjects, although they have provided much valu­ able information on the organization of motor events in animals.32 Muscle contractions also can be measured by surrogate measurement techniques that record associated phenom­ ena. One such approach is fluorescence measurement of calcium transients (rapid increases in free intracellular calcium) in smooth muscle.33 Over short sections of intes­ tine (1-2 mm), such measurements provide excellent tem­ porospatial resolution and are helpful in elucidating neurophysiologic control rather than describing wholeorgan function. Other measurement techniques that record phenomena resulting from contractions of smooth muscle include luminal manometry (reflecting intraluminal pres­ sure increases), fluoroscopy (showing wall movement and movement of intraluminal contrast), and transit studies per­ formed by a number of approaches. Luminal manometry measures the change in intraluminal pressure that results mainly from lumen-occlusive or near– lumen-occlusive contractions. Fortunately, because the small intestine is tubular, with a relatively small diameter, a large portion of motor events are recognized as pressure rises. Researchers have hypothesized that contractions not resulting in a detectable change in intraluminal pressure are less important in determining flow, and therefore little mechanical information is lost by failure to detect them, but small changes in intraluminal pressure can be pivotal in producing flows in some regions of the small intestine.34 Manometry can be applied in several settings, ranging from short isolated intestinal segments in the laboratory to clinical use in humans. Modern computer-based recording systems allow excellent temporal resolution (~10 Hz), and spatial resolution can be tailored to give either close spatial resolution (intervals of 1-2 cm) over 20 to 40 cm, or wider resolution, while still covering a longer segment of small intestine. Manometric assemblies are either of the perfused side-hole or the solid-state sensor design and are capable of routinely recording at up to 22 sites.

Wall Motion and Transit Studies

Contrast fluoroscopy is the most widely available wall motion study. It yields detailed information on the time and space patterning of motor events in vivo and useful insights into associated movements of luminal contents. When this technique is used in combination with other techniques,

such as manometry, intraluminal impedance, or strain gauges, useful correlations can be made between contrac­ tions or luminal pressures and transit of contents. These insights are likely to lead to improved understanding of pressure patterns, which in turn might enable us to better interpret less-intrusive techniques such as manometry and impedance in humans. Improving the interpretation of these other techniques is important, because risks associated with radiation exposure restrict the use of fluoroscopy in humans. Other in vivo imaging methods for assessing small intes­ tinal wall motion and movement of intraluminal contents include magnetic resonance imaging (MRI), ultrasono­ graphy, and intraluminal impedance recording. These approaches are suitable for human use with good temporal resolution, although they have significant practical limitations. These limitations previously restricted their applications outside of research centers as an alternative to contrast fluoroscopy, but they are now gaining more widespread use. MRI allows prolonged observation but, because of the anatomic complexity of the small intestine, difficulties with spatial resolution can limit views. Additionally, MRI is expensive, and not all centers have sufficient MRI capacity for it to become a routine clinical tool for this indication. Ultrasonography also allows prolonged observation and repeat measurements, but only of short segments and with relatively poor spatial resolution. Ultrasonography is limited in many instances by patient factors, such as body habitus and intestinal gas, and it is operator dependent. Multichannel intraluminal impedance (MII) is a tech­ nique for assessing intraluminal bolus transit rather than motility. The technique is based on the different conductivi­ ties of intraluminal air and liquids compared with those of opposed sections of bowel wall. Voltage is applied to a recording assembly along which several electrodes are sited. The current recorded between electrode pairs depends on the conductivity and thickness of any air or fluid bolus straddling the electrode pair. In this fashion, MII sequen­ tially measures the transit of a conducting bolus between electrode pairs. Recordings in the small intestine can, there­ fore, depend on the state of its filling,35 and motility in an empty bowel might not be assessed accurately. Other transit and absorption measurements demonstrate whether mass transit occurs but give no information on the mechanical pattern by which the transport of contents is achieved. Methodology for transit studies includes breath tests and scintigraphy. Breath tests are based on the exhalation of gases such as H2 or CO2 (labeled with 13C or 14C), which are generated when a test meal reaches the colon and undergoes bacterial degradation. Scintigraphic tests of small intestinal transit visually assess the arrival of a labeled meal at the cecum. These two transit techniques yield the lowest temporospa­ tial resolution in assessing small intestinal motility but are clinically useful and are discussed later in this chapter. In vitro techniques for detailed assessment of small intes­ tinal wall movements reveal subtle motility patterns that cannot be detected with manometry or in vivo wall motion studies. For example, one technique using digitized video recording can measure changes in diameter and length of an immobilized segment of intestine36 and has the unique capacity to appreciate discrete changes in the longitudinal and circular muscle layers.

CLINICAL APPROACH

The broader issues of measurement of small intestinal motor function were considered earlier, and the discussion that

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction follows is limited to the clinically relevant techniques used to assess small intestinal motor function.

Small Intestinal Transit Studies

Small intestinal transit time can be measured with breath tests or scintigraphic observation of the movement of intra­ luminal contents. Unless the test substance is delivered past the pylorus by tube, these techniques also include gastric emptying (and thus gastric function) in their measurements; they are, therefore, imprecise about actual small intestinal transit time and are more accurately termed tests of orocecal transit time. Because each technique measures a different aspect of motility, the results obtained from different tech­ niques are not directly comparable. The lactulose breath test is perhaps the best known and most widely used of these techniques. Lactulose is a non­ absorbable disaccharide that is fermented on reaching the bacteria-laden environment of the colon. The H2 gas that is formed is rapidly absorbed and exhaled from the lungs. Samples of exhaled gases are taken at baseline and at regular intervals after the ingestion of lactulose. The orocecal transit time is taken as the time at which a sustained rise in exhaled H2 is seen. An early rise, or a high baseline level, may be evidence of small intestinal bacterial overgrowth, but this measure is relatively insensitive for bacterial overgrowth. The administration of lactulose itself is known to hasten small intestinal transit and so the result is not directly com­ parable to other transit time measures. Similar principles are used in 13C or 14C breath tests, which measure gastric emptying combined with the evalu­ ation of small intestinal absorption of specific nutrients. Acetate, octanoic acid, and triolein have been used in this regard. Acetate appears to be a good liquid marker, octanoic acid is better suited for solids, and triolein is useful in sus­ pected cases of malabsorption. This nutrient-focused assess­ ment of small intestinal function can be combined with the H2 lactulose breath test to measure orocecal transit time as well. The more familiar visual and anatomic scintigraphic measurement of small intestinal transit is also widely avail­ able. The major difficulty with these studies is the lack of a reliable anatomic landmark for the cecum. Either the cecum is defined arbitrarily as the right iliac fossa and a skin marker is used or it is considered retrospectively as the area in which radioisotope accumulates. Two approaches are used to report the scintigraphic orocecal transit time. In the simpler approach, the time of first appearance of isotope in the cecum is given; in the other, the initial activity of the radiolabeled meal is quantified in the stomach, and the orocecal transit time is reported as the time taken for 50% of this initial gastric activity to reach the cecum. Values obtained vary depending on which of these methods is used, and each laboratory should set its own normal range.

Fluoroscopy

Contrast fluoroscopy is useful for detecting mural disease and fixed narrowings of the intestinal lumen that can induce secondary changes in motility, transit, and absorption. Fluoroscopy is insensitive for detecting abnormal nutrient absorption and measuring transit time. Clinical fluoroscopy is limited by the necessarily short observation times because of concern with radiation exposure; therefore, only gross disturbances of motor activity may be detected. Once a substantial amount of contrast has entered the small intes­ tine, the usefulness of fluoroscopy is reduced further, because overlying loops of bowel hinder the interpretation of the movement of contrast.

Manometry

Manometry of the small intestine gives direct measurement of the forces that are applied to luminal contents as a result of motor function. Manometry can be performed over hours or even days and over long or short segments; it is capable of excellent spatial resolution, although it has major practi­ cal limitations. Manometric assemblies can be placed in any part of the human small intestine and are moderately well tolerated, although placement of such an assembly along the small intestine can be demanding even in healthy persons, and it is especially challenging in patients who have major abnormalities of motor function. Manometry allows recognition of some abnormal patterns of pressure over time at individual recording points, but no studies have yet performed a critical evaluation of the best spacing of pressure recording points and of diagnostic cri­ teria for abnormal pressure patterns to distinguish between health and disease. This lack of specific criteria reflects the limited understanding of the relationship between small intestinal intraluminal time-space pressure patterning and the achievement of mixing and propulsion within the small intestine. Because of practical limitations, one must choose between high spatial resolution over a short segment and lower spatial resolution over a longer segment of intestine. Both approaches are likely to be necessary in achieving an accu­ rate understanding of small intestinal motor physiology, perhaps in conjunction with a technique to assess wall motion or intraluminal flow.

Multiple Intraluminal Impedance

Recording assemblies can be used to measure impedance in humans in much the same fashion as manometry. Multiple intraluminal impedance (MII) can measure episodes of bolus transit in any tubular section of the upper intestine. MII gives good spatiotemporal resolution, but owing to tech­ nical limitations of how far apart the sensors can be spaced, it cannot give continuous cover in measuring transit over long lengths of the intestine. MII is increasingly used for the clinical evaluation of esophageal motility, and in recent times it has been applied to the proximal small intestine with success.35 In combination with manometry, MII has the potential to show real-time pressure-flow relationships.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) is capable of excellent spatiotemporal resolution of small intestinal wall motion and movement of intraluminal contents. Because the small intestine is not all within one plane, however, it cannot, at present, be viewed routinely all at once; it does not involve a radiation dose and thus is not time limited on this basis. Because it is also an anatomic imaging technique, it has a substantial advantage over other techniques of being able to offer additional information in the assessment of patients with suspected small intestinal motility problems. Wall thickening, fibrosis, inflammatory changes, and stenoses all can be revealed, and this information can help with direct­ ing diagnosis and even therapy.37 MRI has several disadvantages: it is expensive; some claustrophobic subjects find it too confining; patients who have pacemakers or other metallic prostheses cannot be assessed with MRI. At present, its use in assessing small intestinal motility is restricted to units with a research interest in functional gastrointestinal MRI.38 Its use is increasing in clinical gastroenterology, however, especially for small intestinal Crohn’s disease,37 and it also has the potential to encompass clinical motility assessments.39

1653

1654

Section X  Small and Large Intestine NORMAL IN VIVO SMALL INTESTINAL MOTILITY PATTERNS CONTRACTIONS AT A FIXED POINT

The increased smooth muscle tension arising from muscular contractions can result in increased intraluminal pressure, decreased intraluminal diameter, small intestinal shorten­ ing, or a combination of these effects. Smooth muscle con­ tractions can be tonic or phasic, but common usage has labeled tonic contractions as tone and phasic motor events as contractions. Human phasic small intestinal contractions generally last from 0.8 to 6.0 seconds. Small intestinal electrical recordings reveal continuous cyclical oscillations in electrical potential, referred to as the slow wave, basic electrical rhythm, or pacesetter potential. This slow wave is generated by the ICC (see earlier). In humans, the slow-wave frequency decreases from a peak of 12 per minute in the duodenum to approximately 7 per minute in the distal ileum. A small intestinal contraction arises when an electrical action potential, or spike burst, is superimposed on the slow wave (Fig. 97-5). Spike bursts may be caused by the intrinsic motor output from the ENS to the ICC and are likely also to be modulated by the extrinsic motor supply. Except during phase III of the IDMC (interdigestive migrating motor complex), not every slow wave leads to a phasic contraction. The region-specific frequency of the slow wave thus controls small intestinal rhythmicity by determining the timing and maximal fre­ quency of contractions. The rapid increases in free intracellular calcium, or calcium transients, that underlie smooth muscle contraction can be visualized with fluorescence techniques and appear to spread in a coordinated fashion over an area of smooth muscle and to extend over variable distances of the bowel wall. These calcium transients are extinguished by collision with each other or by encountering locally refractory regions.33

Intracellular recording

Extracellular recording

Muscle tension

Time Figure 97-5.  Schematic representation of the relationship among slow waves, spike bursts, and muscle contraction. The top tracing is from an intracellular electrode in the muscle; the middle tracing is from an extracellular electrode; and the bottom tracing shows muscle tension. The cyclical fluctuation in membrane potential in the top tracing is the slow wave. When spike bursts are superimposed on the peak of the slow wave, the muscle depolarizes, and contraction occurs. (From Christensen J. Gastrointestinal motility. In West JB, editor. Best and Taylor’s Physiologic Basis for Medical Practice. Baltimore: Williams & Wilkins; 1990. p 614.)

CONTRACTIONS THAT TRAVEL ALONG THE SMALL INTESTINE

The electrical slow wave migrates along the small intestine in an aboral direction so that each subsequent site along the intestine is depolarized sequentially. When a slow wave results in contraction, the propagation of the slow wave along the small intestine also leads to the contraction pro­pagating along the small intestine. The propagation velocity of the slow wave thus determines the maximal rate at which contractions can travel along the small intes­ tine. Because not every slow wave leads to a contraction, however, contractions do not always travel at this maximal rate. The distance over which muscular excitation or inhibition spreads appears to be determined by ENS influ­ ences mediated through local inhibitory and excitatory circuits.33 Contraction sequences travel aborad (in an antegrade direction) or orad (in a retrograde direction). From animal data and some human studies at high spatial resolution, it is known that a large portion of contractions travel along the small intestine, rather than remaining static, although most contractions are limited to only a few centimeters in extent.40,41 Further data are needed to determine the con­ tribution that these short contraction sequences make to overall transit compared with the less-frequent longer sequences.

PATTERNED MOTILITY

From isolated small intestinal segments, ascending excita­ tion and descending inhibition are the simplest wellrecognized patterns of motility. Ascending excitation refers to the contraction that occurs proximal (oral) to a stimulus, and descending inhibition refers to the inhibition of motor activity that occurs distal to a stimulus. These simple reflexes can be demonstrated in the absence of any extrinsic innervation and are thus entirely attributable to the ENS, although extrinsic influences can modulate their occur­ rence. These two patterns are thought to be responsible for peristalsis and retroperistalsis when they travel in a coordi­ nated fashion along the intestine. Recordings of human small intestinal motility show iso­ lated (stationary) phasic contractions, but often, spatial pat­ terns are more complex. The limited spatial resolution of many recording techniques can lead to over-reporting of the fraction of stationary contractions. Commonly, phasic motor activity consists of a recognizable group of contractions associated along the small intestine in space and time; phase III activity of the IDMC (see later) is a good example of this association. Several other types of grouped small intestinal contractions have been described and include contractions associated with emesis42 and discrete clustered contractions, which are said to be common in IBS (see Chapter 118).43 The most commonly observed motor pat­ terns in the healthy small intestine, however, are described simply as the fed or postprandial pattern and the fasting (interdigestive) pattern, or IDMC (Fig. 97-6). The motor pattern is determined by the presence or absence of a significant amount of nutrient within the small intestine. Despite a large number of studies on fasting motil­ ity, few studies have been performed on human postpran­ dial small intestinal motility; this paucity probably exists because of the difficulty in knowing which aspects of post­ prandial motility to study, in contrast to fasting motility, which has an easily recognized cyclic pattern and thus easily studied parameters. The fed motor pattern ensures transit of small intestinal contents at a rate consistent with normal digestion and absorption. The fasting motor pattern is less involved with orderly luminal transport and

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction FASTING STATE Phase II (IDMC)

Phase III Phase I 50 mm Hg

5 min FED STATE

Liquid meal

variable amounts of the same nutrient, generate recogniz­ ably different motor responses.31,32,46,47 In general, the pres­ ence of unabsorbed small intestinal nutrients slows small intestine transit by decreasing the frequency and length of travel of phasic contractions, so that the rate at which a substance is absorbed limits its transit rate. In the absence of sufficient proximal small intestinal nutrient stimulation, the fasting motor pattern re-emerges four to six hours after a meal. In the absence of its interruption by intraluminal nutrients, the IDMC repeats continuously. Distention, intraluminal pH changes, and hyperosmolar contents are capable of stimulating small intestinal motor activity. Hyperosmolar contents and pH changes probably are sensed by receptors in the mucosa, whereas distention is signaled by receptors in the muscle. In the normal course of events, these stimuli occur concurrently with the pres­ ence of nutrients, and the significance of their isolated effects in healthy subjects is unclear. The small intestine also exerts negative feedback control on the rate of gastric emptying through neural and humoral means. This negative feedback is achieved by the release of neural signals and intestinal hormones that suppress phasic gastric motor activity, relax the gastric fundus, and increase tonic and phasic pyloric pressures subsequent to mucosal sensing of small intestinal nutrients.48 This process indi­ rectly also prolongs whole-meal small intestine transit time by slowing the input of small intestinal chyme. The small intestine, in particular the duodenum, also is thought to offer direct mechanical resistance to gastric emptying by acting as a capacitance resistor49 and by reaugmenting gastric contents as a result of duodenogastric reflux.50

Fed Motor Pattern

80 mm Hg 10 min Figure 97-6.  Manometric tracings demonstrating small intestinal motility in the fasting (top) and fed (bottom) state. The three phases of the interdigestive motor cycle (IDMC) (courtesy Dr. J.M. Andrews) and the conversion to a fed motor pattern by infusing a liquid nutrient meal into the small intestine are shown. In the top set of tracings, at a given time point (dashed vertical line), all three phases of the IDMC can be encountered at different points along the small intestine. The similarity of phase II and the fed motor pattern can be appreciated by comparing the top (fasting) and bottom (fed) sets of tracings. (Professor R.J. Fraser provided data for this figure.)

is thought to serve important roles in clearing the upper intestine of solid residues, which otherwise can accumulate and form bezoars; in maintaining relative sterility of the small intestine by keeping it empty; and in preventing net oral migration of colonic bacteria. Within 10 to 20 minutes of consumption of a meal, the IDMC that is in progress at the time of eating is inter­ rupted.44 The presence of intraluminal nutrients is sensed by mucosal nutrient contact, as evidenced by the fact that portally administered or intravenous nutrients do not have the same effects as those consumed orally.45 Several neural and humoral signals result from mucosal nutrient contact and are implicated in the induction of the fed motor pattern, including vagal afferent signals, cholecystokinin, and GLP-1. Moreover, the sensing of intraluminal nutrients is relatively complex, because different types of nutrients, or

Radiologic Observations Early radiologic observations of the small intestine in animals described several different patterns of wall motion and transit of intestinal contents. Walter Cannon42,51 observed both localized contractions over short segments of intestine in association with to-and-fro movement of con­ tents and intermittent episodes of propulsion of contents over greater distances caused by aborally traveling waves of peristalsis. In the fed state, the most common pattern of wall motion consisted of localized circular contractions that recurrently divided and formed short columns of chyme into new aliquots by temporary local occlusion of the lumen over distances of less than 1 to 2 cm, this pattern being labeled rhythmic segmentation.42,51 These contractions did not travel along the small intestine and did not result in much, if any, net oral movement of contents.42,51 Peristalsis also was commonly observed, often in combi­ nation with segmentation. During small intestinal nutrient loading, peristalsis was noted to have two forms: One was a slow advance of chyme over short distances in association with segmentation and the other was a rapid transit of chyme over longer distances, sometimes several loops, of the small intestine. The “fast peristalsis” was often seen in the cat duodenum.51 Similar observations have been made in other animal species32,42 and correlate with some of the motor patterns seen during clinical radiologic studies in humans (although these studies usually are performed when the subject is fasting and show the rapid peristaltic pattern more than the segmenting postprandial activity). Transit Time Observations The small intestinal transit time for a meal varies greatly according to the amount and nature of what is consumed, because both caloric content and physical form of a meal determine the gastric emptying rate and the rate of transport

1655

1656

Section X  Small and Large Intestine along the intestine.41,52-54 Depending on the test and param­ eter used, postprandial orocecal transit time usually is less than six hours. As assessed by lactulose breath testing, however, orocecal transit time can be as rapid as about 70 minutes with low nutrient loads. A systematic evaluation of the optimal conditions for nutrient loading is much needed to reveal abnormal small intestinal motor function, using transit studies. Manometric Observations Postprandial small intestinal motility is characterized by irregular phasic pressure waves without a discernible cycli­ cal pattern. Most small intestinal motility data are quite limited in spatial resolution because of the length of the small intestine. Nevertheless, most phasic pressures (pres­ sure wave sequences) are thought to travel only a short distance31,32 and probably represent the mixing and seg­ menting contractions noted in earlier radiologic studies.42,51 In animal studies, postprandial small intestinal motility is more segmenting than is fasting phase II activity, and phasic pressures occur less frequently and travel shorter distances along the bowel, resulting in slower transit of the contents.32 A similar suppression in the frequency of pressure wave sequences now has been found in the human duodenum.31 This segmenting motor pattern is thought to assist in mixing food with digestive enzymes and in maximizing the expo­ sure of food to the mucosa to optimize absorption.

Fasting Motor Pattern

During fasting, small intestinal motor activity adopts a repetitive cyclic motor pattern, the IDMC. The IDMC is absent in a number of disease states, presumably because of a primary neuropathic process. This absence is associated clinically with stasis of small intestinal contents, malab­ sorption, and small intestinal bacterial overgrowth. For detailed reviews, see the articles by Husebye43 and Sarna.44 Radiologic Observations Contrast agents can stimulate small intestinal mucosal receptors sensitive to pH, caloric content, and osmolarity changes. It is possible, therefore, that radiologic studies of “fasting” motility do not truly represent the fasting state. In general, however, contrast agents appear to move more swiftly through the small intestine during fasting than during the postprandial state and to be associated with more episodes of peristalsis over one or more loops and fewer segmenting contractions. When the phase of the IDMC is assessed concurrently (see later), little net movement of small intestinal contents is seen during phase I, but residual luminal contents are swept through the small intestine and into the terminal ileum during late phase II and phase III of the IDMC. This finding is not surprising because, by definition, phase I is the absence of measurable phasic pres­ sure waves, which are likely to be necessary to generate a sufficient intraluminal pressure gradient to cause intralumi­ nal flow. Transit Time Observations Studies of transit time through the small intestine also prob­ ably do not represent a true assessment of fasting motor function, because most of the substrates used to measure transit also interact with small intestinal mucosal receptors. The lower the caloric content, the more closely fasting motility will be assessed (see earlier). Manometric Observations The IDMC is defined manometrically and comprises three main phases. Phase I is defined as motor quiescence (less

than three pressure waves per 10 minutes at any one site); phase II is characterized by random pressure waves at less than the maximal rate; and phase III is characterized by pressure waves at the maximal rate (for the region) for longer than two minutes and, ideally, extending over a segment longer than 40 cm. Some authors also include a fourth phase (phase IV) as a transitional period between phases III and I, although this approach is not universal. Phases I and III are quite distinctive and easily recognized, whereas phase II can be recognized reliably only when sandwiched between phases I and III, because it superfi­ cially resembles the fed pattern. The phases of the IDMC start proximally and migrate distally over variable dis­ tances; few phase IIIs reaching the ileum.40 Phase III of each IDMC can start at any of a variety of locations; approxi­ mately one third of IDMCs have a gastroduodenal compo­ nent, and most onsets of phase III occur near the proximal jejunum.40 Because of the length of the small intestine and the velocity of travel of the IDMC, one part of the small intestine can be in phase I while other parts are in phase II or III (see Fig. 97-6). The normal periodicity of the IDMC varies greatly both within and between subjects; however, its median duration is 90 to 120 minutes.

CLINICAL CONSEQUENCES OF DISORDERED SMALL INTESTINAL MOTILITY Most of the time, the overall outcome of small intestinal motility is achieved without conscious awareness; a range of symptoms can arise, however, when an optimal outcome is not attained. Fortunately, like other organs, the small intestine has a substantial reserve capacity and copes with many insults, including infection, resection, inflammation, and denervation, before clinical problems become manifest. In IBS, the most common clinical syndrome in which altered motility is implicated, the sufferer’s physical wellbeing rarely is threatened even when symptoms are con­ siderable. Infrequently, the motor disturbances are severe enough to disrupt a person’s ability to maintain oral nutrition. The most important diseases and clinical settings asso­ ciated with abnormal small intestinal motility are listed in Table 97-1. Because these disorders are covered elsewhere in this book, they are mentioned here only with regard to the associated small intestinal motor disturbances. In IBS, a number of abnormalities of visceral sensation have been documented. These sensory abnormalities prob­ ably also lead to disordered motility; however, whereas motor abnormalities have been documented in some patients with IBS, they are absent in others (see Chapter 118). Because it appears likely that IBS is an as-yet-undefined generalized enteric neuropathy or low-grade neuroinflam­ matory disorder,55,56 failure to detect specific motor abnor­ malities might simply reflect our current poor understanding of normal small intestinal motor physiology and the rela­ tively gross measures by which motility has been assessed in patients with IBS. Small intestinal motility is severely disrupted in acutely ill persons and is increasingly recognized as an important factor to consider in postoperative and intensive care unit patients. Such motility disturbances likely result from several factors, including sepsis and drugs, which disrupt the slow wave rhythm; abdominal trauma and surgery, which stimulate reflex motor responses; and inflammatory mediators and cytokines, which affect neurotransmission within the CNS, ANS, and ENS. For a more detailed review,

Chapter 97  Small Intestinal Motor and Sensory Function and Dysfunction see the articles by Ritz and colleagues,57 and Chapman and colleagues.58 Pregnancy is known to alter the function of the lower esophageal sphincter, delay gastric emptying and disturb the frequency of gastric slow waves, and it is often asso­ ciated with constipation. In view of these widespread findings related to altered intestinal motility, it is likely that small intestinal motor function also is altered. In guinea pigs, the strength of the contraction of intestinal circular smooth muscle has been shown to be impaired during pregnancy by down-regulation of Gαq/11 proteins (which mediate contraction) and up-regulation of Gs alpha protein (which mediates relaxation).59 It is intriguing that G protein associations now are also being reported in functional gas­ trointestinal disorders, suggesting a final common pathway for sensorimotor intestinal disturbances.60 Diabetes has widespread effects on gastrointestinal motil­ ity. Acute effects result from changes in blood glucose levels, but they also can result from the autonomic neuropa­ thy that develops in patients with long-standing disease. As indicated predominantly by studies of the stomach, hyperglycemia can alter the rhythm of the slow wave, modulate sensory signaling, lead to changes in the tempo­ rospatial pattern of phasic contractions, and even stimulate inappropriate phase III–like IDMC activity in the small intestine. Metabolic disturbances of potassium, magnesium, and calcium homeostasis are likely to impair small intestinal motor function because these chemicals are vital for normal neuromuscular function. The effects of abnormal levels of these electrolytes on normal human small intestinal func­ tion have not been studied specifically, but in organ bath experiments, their alterations have caused gross distur­ bances in neural and muscular function. In addition, renal and hepatic failure are likely to alter small intestinal motil­ ity because of the multiple homeostatic inputs of the affected organs; however, altered motility usually is not a prominent clinical feature in these conditions. Many drugs affect small intestinal motility, especially those that alter ion transport, such as antidepressants, calcium channel blockers, and beta blockers. Sedatives and narcotic analgesics also alter motility but usually do not cause clinically important small intestinal motor dysfunc­ tion, except in critically ill patients or those with acute severe pain. Pseudo-obstruction, scleroderma and other connective tissue diseases, dysautonomia, visceral myopathies, and other rare diseases in which abnormal small intestinal motor function occur are discussed in detail in other chapters. These diseases may be uncommon causes of dis­ ordered small intestinal motility, but they have increased our understanding of normal motility, because in some cases, the neural and myopathic processes are impaired separately.

APPROACH TO PATIENTS WITH POSSIBLE SMALL INTESTINAL MOTOR DYSFUNCTION Taking a thorough history is a vital first step in approaching a patient who may have abnormal small intestinal motility. A review of exposures to drugs and toxins, family history, and, in the younger patient, milestones of growth and devel­ opment are especially important to consider. Findings on physical examination in this setting often are unremarkable. First-line investigations generally are suggested by the history, physical examination, and age of the patient and

may include a plain abdominal film (to look for dilated small intestinal loops, thickened bowel wall, or air-fluid levels), complete blood count with determination of red blood cell indices (to look for evidence of malabsorption), measurement of serum albumin and electrolyte levels, and random testing of blood glucose or glycosylated hemoglobin level. How much further to proceed with investigation depends on these results and on the severity of the patient’s condition. Special investigations may be indicated to answer par­ ticular questions. No standard approach has been recog­ nized, however, and local interest and expertise often determine which investigations are available. Fluoroscopy is widely available and can help exclude medically or surgi­ cally treatable problems. Endoscopy with small intestinal biopsy or aspiration is useful if celiac sprue, small intestinal bacterial overgrowth, or intestinal infection is considered likely. Analysis of stool may be necessary to exclude mal­ absorptive or secretory causes of small intestinal diarrhea. Small intestinal manometry, if available, can help distin­ guish neuropathic from myopathic forms of disordered motility, although in many settings, the abnormalities asso­ ciated with these two forms overlap (see Table 97-1). Manometry can show features typical of intestinal obstruc­ tion, although abdominal imaging by a variety of radiologic techniques is a better tool to identify an obstruction. In selected cases, full-thickness biopsy of the small intestine is necessary, but such biopsy should be performed only in centers with expertise in immunohistochemistry of intesti­ nal neurons, because standard histologic approaches often yield little useful information. Unfortunately, there are few therapies to date, beyond supportive measures, that can be offered to patients with disordered small intestinal motility. Nutritional status is of prime importance, and where patients can manage this independently, no further specific treatment may be needed. Symptomatic treatment approaches include modifications in diet (small frequent meals, lower fat intake), exercise (which is shown to improve bloating symptoms and expul­ sion of intestinal gas), antinausea agents, antispasmodics, and drugs to modulate sensory function. Thus far, there are no clinically available agents that spe­ cifically modify visceral hypersensitivity, and simple anal­ gesics, opiates, and antidepressants are all used. Apart from the tricyclic antidepressants and selective serotonin reup­ take inhibitors, there is little proof that these offer sig­ nificant benefit, and opiates can even worsen symptoms, leading to the narcotic bowel syndrome. Treatment of psy­ chological comorbidities also is important because anxiety and depression can heighten the perception of, and distress caused by, intestinal symptoms. Prokinetic agents have been limited in their therapeutic benefit, and because of safety concerns, availability of several, such as cisapride and tegaserod, has been restricted. There is hope, however, that prokinetics and visceral-specific analgesics might offer a better balance between safety and efficacy in the future.

KEY REFERENCES

Berthoud HR, Blackshaw LA, Brookes SJ, Grundy D. Neuroanatomy of extrinsic afferents supplying the gastrointestinal tract. Neurogastro­ enterol Motil 2004; 16 Suppl 1:28-33. (Ref 17.) Blackshaw LA, Gebhart GF. The pharmacology of gastrointestinal nociceptive pathways. Curr Opin Pharmacol 2002; 2:642-649. (Ref 14.) Cannon WB. The Mechanical Factors of Digestion. London: Edward Arnold, 1911. (Ref 42.) Furness JB. Types of neurons in the enteric nervous system. J Auton Nerv Syst 2000; 81:87-96. (Ref 8.)

1657

1658

Section X  Small and Large Intestine Hennig GW, Costa M, Chen BN, Brookes SJ. Quantitative analysis of peristalsis in the guinea-pig small intestine using spatio-temporal maps. J Physiol 1999; 517(Pt 2):575-90. (Ref 36.) Hunt JN, Smith JL, Jiang CL. Effect of meal volume and energy density on the gastric emptying of carbohydrates. Gastroenterology 1985; 89:1326-30. (Ref 53.) Husebye E. The patterns of small bowel motility: physiology and impli­ cations in organic disease and functional disorders. Neurogastroen­ terol Motil 1999; 11:141-61. (Ref 43.) Krauter EM, Strong DS, Brooks EM, et al. Changes in colonic motility and the electrophysiological properties of myenteric neurons persist following recovery from trinitrobenzene sulfonic acid colitis in the guinea pig. Neurogastroenterol Motil 2007; 19:990-1000. (Ref 27.) Sanders KM. Regulation of smooth muscle excitation and contraction. Neurogastroenterol Motil 2008; 20 Suppl 1:39-53. (Ref 2.) Sanders KM, Koh SD, Ward SM. Interstitial cells of Cajal as pacemakers in the gastrointestinal tract. Annu Rev Physiol 2006; 68:307-43. (Ref 4.)

Sarna SK. Cyclic motor activity; migrating motor complex: 1985. Gastroenterology 1985; 89:894-913. (Ref 44.) Schwizer W, Steingoetter A, Fox M. Magnetic resonance imaging for the assessment of gastrointestinal function. Scand J Gastroenterol 2006; 41:1245-60. (Ref 38.) Vermillion DL, Huizinga JD, Riddell RH, Collins SM. Altered small intestinal smooth muscle function in Crohn’s disease. Gastroenter­ ology 1993; 104:1692-9. (Ref 26.) Ward SM, Sanders KM. Involvement of intramuscular interstitial cells of Cajal in neuroeffector transmission in the gastrointestinal tract. J Physiol 2006; 576:675-82. (Ref 5.) Wood JD. Enteric nervous system: reflexes, pattern generators and motility. Curr Opin Gastroenterol 2008; 24:149-58. (Ref 9.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

98 Colonic Motor and Sensory Function and Dysfunction Ian J. Cook, Simon J. Brookes, and Philip G. Dinning

CHAPTER OUTLINE Anatomy and Basic Control Mechanisms of the Colon and Anorectum  1660 Macroscopic Structure of the Colon  1660 Structure and Activity of Colonic Smooth Muscle  1660 Ion Channels in Colonic Smooth Muscle  1661 Interstitial Cells of Cajal  1661 Innervation of the Colon  1662 The Enteric Nervous System  1662 Sympathetic Innervation  1664 Parasympathetic Innervation  1664 Extrinsic Afferent Pathways  1664 Anorectal Anatomy and Innervation  1664 Relationships among Cellular Events, Pressure, and Flow  1665 Recognizable Colonic and Anorectal Motor Patterns and Putative Functions  1665 Nonpropagating Motor Patterns  1665 Propagating Motor Patterns  1665 Rectal Motor Complexes  1665 Regional Variation of Propagating Sequences  1665 Regional Linkage among Propagating Sequences  1665

Each day, 1200 to 1500 mL of ileal effluent enter the colon, 200 to 400 mL of which are finally excreted as stool. The colon mixes its contents to facilitate the transmural exchange of water, electrolytes, and short-chain fatty acids and stores stool for extended periods. The mixing process involves rhythmic to-and-fro motions, together with short stepwise movements of contents, resulting in an overall net aboral flow rate that averages approximately 1 cm per hour. When dehydration threatens survival, such as with water deprivation or severe diarrhea, the ability of the colon to reabsorb fluids is of major physiologic significance; appropriate motility patterns are important in achieving this function. For example, the colon has the capacity to increase its fluid absorption five-fold when required, but this ability is greatly impaired when transit is accelerated. Under normal circumstances, viscous contents occasionally are propelled aborally at a rapid rate, and, if circumstances are appropriate, stool is evacuated under voluntary control. Thus, the colon is capable of showing a diverse range of motor patterns that are suited for particular physiologic functions. The generic term motility describes the range of motor patterns and the mechanisms that control them. Common sensorimotor symptoms, such as constipation, diarrhea, bloating, abdominal pain, or rectal urgency, can arise from disturbances of ileocolonic delivery, colonic propulsion, or stool expulsion. Clearly, these symptoms and dysmotility must be linked, although our current under-

Regulation of Colonic Filling and Transit  1666 Role of the Ileocecal Junction  1666 The Colon as a Storage Organ  1666 Relationships between Colonic Motor Patterns and Flow  1667 Defecation  1668 Rectal Filling, Capacitance, and Accommodation and Motility of the Anal Sphincters  1669 Anorectal Motility during Defecation  1670 Modulators of Colonic Motility  1670 Physiologic  1670 Pharmacologic  1671 Nonpharmacologic  1672 Disorders of Colonic Motility  1672 Constipation  1672 Diarrhea  1673 Irritable Bowel Syndrome  1673 Colonic Motility Disturbances Secondary to Nonmotor Intestinal Disorders  1673

standing of such linkages is limited, largely because of technical difficulties involved in studying the human colon. Because of differences among species, care is required in extrapolating data from animal studies to humans. For many years, intraluminal motility recordings in humans were obtained mainly from the rectum and sigmoid, but it is now clear that the motor activity of these distal regions is not representative of the colon as a whole. The contents of the colon become increasingly viscous distally, and this alteration complicates the relationship between propulsion and the contractile activity of the smooth muscle. Colonic movements are much less frequent and transit is considerably slower than in other regions of the gastrointestinal tract. The highly propulsive, stereotypical motor patterns that are associated with stool expulsion generally occur only once or twice daily. Hence, study of the motor patterns in the human cannot be achieved using contrast radiography. Prolonged recording techniques must be used to capture such infrequent motor patterns. Recording of intraluminal pressure, by means of manometric catheters inserted per rectum, requires prior bowel cleansing, which can modify colonic motility. Furthermore, interpretation of intraluminal pressure measurements is complicated, because many contractions of the colonic wall do not occlude the lumen and therefore are detectable manometrically only if they cause significant pressure changes. Measurement of colonic wall tone using a barostat

1659

1660

Section X  Small and Large Intestine provides information on these nonocclusive colonic wall movements, but it tells us nothing of the spatiotemporal patterning of motility. Smooth muscle electromyography provides insight into the patterning of muscle activity but generally requires access to the muscular wall of the colon, which ethically is problematic in humans. Scintigraphy with suitably high frame rates can resolve discrete movements of the contents but is suboptimal for measuring actual wall motion. In vitro study of the cellular basis of motility using isolated specimens of colon faces fewer technical and ethical limitations; however, data obtained at the cellular level, often under highly nonphysiologic conditions, can be difficult to extrapolate to the more complex integrated responses of the entire organ in vivo. Nonetheless, while recognizing the intrinsic limitations of all these measurement techniques, in combination they have allowed us to piece together a number of concepts that have provided important insights into the relationships among muscle activity, wall motion, intraluminal pressure, and flow.

ANATOMY AND BASIC CONTROL MECHANISMS OF THE COLON AND ANORECTUM MACROSCOPIC STRUCTURE OF THE COLON

The human colon is just over one meter long and is divided anatomically into the cecum; the ascending, transverse, descending, and sigmoid colon; and the rectum, which lies between the rectosigmoid junction and the anal canal. The outer longitudinal smooth muscle layer forms three thick, cord-like structures called the teniae coli, which are spaced evenly around the circumference of the colon. Between the teniae, the longitudinal smooth muscle is much thinner, allowing the wall to bulge noticeably. Irregularly spaced circumferential constrictions pinch the colon into a series of pockets, called haustra, which give the colon a sacculated appearance for much of its length. Haustra are not fixed structures and appear to be caused by sustained contractions of the circular muscle. Myogenic activity alone, however, does not seem sufficient to explain haustration, and neural input is likely to contribute to their formation; haustra move, disappear, and re-form during the propulsion of colonic contents. The teniae fuse to form a continuous outer longitudinal smooth muscle layer at the rectosigmoid junction, which then continues down to the distal margin of the anal canal, insinuating itself between the internal and the external anal sphincters. Throughout the length of the colon, the circular smooth muscle layer consists of thick bundles of cells, which are separated by connective tissue septa. The internal anal sphincter consists of a thickening of the circular muscle layer over the last 2 to 4 cm of the anal canal. Macroanatomy of the colon is also discussed in Chapter 96.

STRUCTURE AND ACTIVITY OF COLONIC SMOOTH MUSCLE Structure

Smooth muscle cells in the human colon, as in other muscular organs, are spindle-shaped, nucleolated cells with tapered ends. The surface area of the smooth muscle cell membrane is increased greatly by numerous caveolae, or small pits. Individual smooth muscle cells are connected mechanically to neighboring cells by intermediate junctions and electrically by gap junctions, which allow ions and

small molecules—those with molecular weights up to approximately 1000 kD—to diffuse between the cells, thereby ensuring that the cells are functionally coupled to one another. Therefore, the smooth muscle cells do not contract as individual cells; rather, they contract together in a large, coordinated assembly, a syncytium.

Activity

Like smooth muscle throughout the gastrointestinal tract, colonic smooth muscle shows spontaneous, oscillatory electrical activity, even when all neural activity is blocked. Two types of rhythmic myoelectrical activity occur1: myenteric potential oscillations (MPOs) and slow waves. MPOs are small-amplitude, rapid oscillations, with a frequency of 12 to 20 per minute, that originate from the plane of the myenteric plexus. These small oscillations spread, by means of gap junctions, into both the longitudinal and the circular smooth muscle layers and often reach the threshold potential for generating smooth muscle action potentials in both muscle layers. In the circular muscle layer, MPOs, with superimposed action potentials, generate small phasic contractions of the circular muscle layer. When the muscle is strongly excited by neurotransmitters released by enteric excitatory motor neurons, each MPO evokes an action potential, and the phasic contractions summate into powerful contractions that last several seconds. Although the functions of the colon circular smooth muscle are well understood, the role that the longitudinal muscles play in colonic motility, mixing, and propulsion is a matter of some controversy. The longitudinal muscle probably acts in an antagonistic role to the circular muscle, contracting largely in concert with it and thus preventing excessive lengthening when the circular muscle contracts, which would be mechanically disadvantageous. It might also contribute to propulsion by pulling the colon over its contents, so that circular muscle contractions gain more purchase on them. Some evidence from modeling suggests that it also might play a role in mixing of liquid contents, at least in the small intestine. A second pacemaker region is located at the submucosal border of the circular muscle. This region produces largeramplitude, slower myogenic oscillations in membrane potential called slow waves, which also spread through the thickness of the circular smooth muscle by means of gap junctions. Slow waves also often reach the threshold for triggering smooth muscle action potentials and can evoke strong contractions. Slow waves occur throughout the human colon at a frequency of approximately 2 to 4 per minute. In the small intestine, a gradient of slow wave intrinsic frequencies causes slow waves to propagate predominantly aborad. This is not the case in the colon: Slow waves propagate over short distances up or down the colon, and complex interactions occur as waves coming from different initiation sites collide, leading to mixing of contents with little propulsion. The currents produced by pacemaker cells at the myenteric and submucosal borders decay as they spread through the thickness of the circular muscle layer. Thus, operating in the middle of the circular smooth muscle layer is complex spontaneous electrical activity consisting of a mixture of MPOs and slow waves, with superimposed smooth muscle action potentials. Most of the time, slow waves determine the contractile activity of the smooth muscle and cause nonpropulsive mixing movements. During times of strong enteric neuronal activity, however, MPO-related contractions summate, giving rise to powerful patterned contractions of much longer duration than those produced by slow waves. These contractions can propagate for long distances

Chapter 98  Colonic Motor and Sensory Function and Dysfunction along the colon and are known as propagating sequences. Action potentials in the smooth muscle can be recorded in vivo with electrodes attached to the serosal surface, thereby giving a high-resolution measurement of myoelectric activity or spike bursts.

ION CHANNELS IN COLONIC SMOOTH MUSCLE

The membrane of colonic smooth muscle cells contains a variety of ion channels, including several types of potassium channels, calcium channels, chloride channels, and nonselective cation channels.2 Although the exact physiologic roles of many of these ion channels are unknown, the high-threshold, voltage-operated calcium channels (L-type calcium channels) do play a crucial role in colonic muscle contractility. These channels open when the membrane potential of smooth muscle cells is depolarized beyond a voltage threshold, and they are responsible for the rapid upstroke of smooth muscle action potentials. The influx of calcium through L-type calcium channels during action potentials is a major trigger for activation of the contractile apparatus. It is not surprising that pharmacologic blockade of L-type calcium channels by dihydropyridine drugs such as nifedipine can reduce the contractility of colonic smooth muscle substantially. Release of calcium from intracellular stores, which is triggered by excitatory neurotransmitters, also may play a role in muscle contraction.

and ICCSM form extensive networks along the colon and are electrically coupled to one another and to the smooth muscle layers by gap junctions (Figs. 98-1 and 98-2). ICCMY probably are the pacemakers for the small, rapid (12-20/ min) oscillations in membrane potential (MPOs) of longitudinal and circular smooth muscle layers. ICCSM are the pacemakers for the large-amplitude slow waves (2-4/min) originating in the plane of the submucosal plexus; these slow waves have a powerful influence on the patterning of circular muscle contraction. The exact ionic basis of rhythmicity in ICCMY and ICCSM that gives rise to MPOs and slow waves is not entirely clear; however, oscillations in membrane potential are an intrinsic

Mucosa ICCSM CM Nerve ICCIM CM ICCMP

INTERSTITIAL CELLS OF CAJAL

LM

Since 1991, the interstitial cells of Cajal (ICC) have been shown to play at least two important roles in the control of gastrointestinal motility: control of myogenic activity and mediating or amplifying the effects of motor neurons on the smooth muscle apparatus. ICC are non-neuronal in origin and are derived from common progenitors of smooth muscle cells. Mutant mice and rats that are deficient in ICC have profoundly disturbed intestinal motility, an observation that provides insight into the roles of ICC in the human gastrointestinal tract. In the human colon, three types of ICC are recognized and are named according to their locations: ICC in the plane of the myenteric plexus (ICCMY), ICC near the submucosal plexus (ICCSM), and intramuscular ICC located between the circular and the longitudinal muscle layers (ICCIM). ICCMY

A

B

Figure 98-1.  Schematic cross section of the muscularis externa of the human colon. The outer longitudinal smooth muscle layer (LM) is thickened at the teniae. In the plane of the myenteric plexus (not shown) is a network of interstitial cells of Cajal (ICC), which generate a rapid myenteric potential oscillation (ICCMY; see lower waveform on the right). The circular muscle layer (CM) is innervated by axons of enteric motor neurons with transmitter release sites (clusters of clear vesicles) that are associated with specialized intramuscular ICC (ICCIM). At the outer border of the circular muscle is another network of submucosal ICC, which generate slow waves (ICCSM; see upper waveform on the right). Also present are axons of motor neurons in the longitudinal muscles and ICCIM (not shown in this cross section). The tiny white squares represent gap junctions, which electrically couple cells.

100 µm

Figure 98-2.  Micrographs of interstitial cells of Cajal (ICC) in the human colon, labeled by c-Kit im­munohistochemistry. A, ICC in the plane of the myenteric plexus (ICCMY) have an irregular shape, form a dense network of cells, and probably function as pacemakers. B, A different plane of focus of the same region shows spindle-shaped intramuscular ICC (ICCIM) in the overlying circular muscle layer. These cells probably are involved in neuromuscular transmission to the smooth muscle. (Courtesy of Liz Murphy and David Wattchow.)

1661

1662

Section X  Small and Large Intestine property of both ICCMY and ICCSM. Intramuscular ICC (ICCIM) are a major target of neurotransmitters released from the axons of excitatory and inhibitory enteric motor neurons. Acetylcholine and nitric oxide (and probably several other motor neuron transmitters) evoke changes in the membrane potential of ICCIM, which then spread through the smooth muscle by means of gap junctions. ICCIM also may be involved in amplifying the slow waves as they spread through the muscle layers. Thus, these cells appear to be key players in integrating non-neuronal pacemaker activity and neuronal inputs to the smooth muscle. The discovery that cellular mechanisms long considered to be the properties of smooth muscle cells actually are mediated by ICC may have important clinical implications. For example, in the distal bowel, reduced numbers of ICC, or a reduction in the total volume of ICC, has been associated with anorectal malformations, colonic manifestations of Chagas’ disease, and possibly some cases of slow-transit constipation.3 Some reports have suggested that the density of ICC may be reduced in aganglionic segments of colon in Hirschsprung’s disease and that this deficit might contribute to diminished propulsive activity; this finding, however, has not been consistent between studies.3

INNERVATION OF THE COLON 1 THE ENTERIC NERVOUS SYSTEM

Direct neuronal control of colonic motility is mediated mostly by the enteric nervous system (ENS). Although the ENS is capable of expressing a diverse repertoire of motor patterns, its functions are modulated by sympathetic, parasympathetic, and extrinsic afferent pathways (Fig. 98-3). In terms of numbers of nerve cells, the ENS is by far the largest component of the autonomic nervous system, with considerably more neurons than those of the parasympathetic and sympathetic divisions combined. The nerve cell bodies of the ENS are located in plexuses of myenteric ganglia (Auerbach’s plexus), which lie between the longitudinal and the circular muscle layers of the muscularis externa, or in the submucosal ganglia, which lie between the circular muscle and the mucosa (Fig. 98-4). The submucosal plexus is divisible into at least two networks: Meissner’s plexus, which lies closer to the mucosa, and Schabadasch’s plexus, which lies adjacent to the circular muscle. Some authors have identified an additional intermediate plexus. Internodal strands that contain hundreds of axons run within and between the different plexuses. Finer nerve trunks innervate the various target tissues of the intestinal wall, including the longitudinal muscle layer, circular muscle, muscularis mucosae, mucosal crypts, and mucosal epithelium. Within the ganglia of each plexus, different functional classes of enteric nerve cell bodies are intermingled, although differences in the proportions of cell types among the plexuses have been observed. It has become clear that an exquisite degree of organization is characteristic of the ENS, each class of nerve cell making highly specific and precise projections to its particular target. The ENS uses many transmitters in addition to the major transmitters acetylcholine and nitric oxide, including tachykinins, purines, numerous other modulatory peptides, and some amines. Many other substances, released from neural and non-neural cells, also modulate neuronal and muscular excitability, including gaseous mediators (carbon monoxide and hydrogen sulfide) and, in inflammation, prostanoids, cytokines, purines, bradykinin, H+ ions, and neurotrophins.

Vagus nerve Lumbar colonic nerves

Nodose ganglion DRG

Ascending colon

Transverse colon

Descending colon

Inferior mesenteric plexus Pelvic plexus ganglia

Sigmoid colon Rectum External anal sphincter

Lesser splanchnic nerve

DRG

Pelvic nerves

Pudendal nerve Figure 98-3.  The extrinsic innervation of the human colon. Parasympathetic efferent pathways ( filled cell bodies) arise from the dorsal motor nucleus of the vagus in the brainstem and pass through the vagus nerve and prevertebral sympathetic ganglia, through the lumbar colonic nerves to the proximal colon. Parasympathetic pathways also extend from nuclei in the sacral spinal cord that run through the pelvic nerves and either synapse in the pelvic plexus ganglia or run directly into the bowel wall. Sympathetic pathways (open cell bodies) consist of preganglionic neurons in the thoracic spinal cord that synapse with sympathetic postganglionic neurons either in the inferior mesenteric plexus or in the pelvic plexus. Enteric nerve cell bodies in the colon receive input from both parasympathetic and sympathetic pathways. Viscerofugal enteric neurons project out of the bowel to the prevertebral ganglia. Afferent pathways consist of vagal afferent neurons from the proximal colon with cell bodies in the nodose ganglia. In addition, spinal afferent neurons with cell bodies in lumbar dorsal root ganglia (DRG) run through the lesser splanchnic and colonic nerves to the colon and mediate nociception. Another population of spinal afferents, with cell bodies in the sacral DRG, runs through the pelvic nerves and pelvic ganglia to the rectum; these include sensory neurons that transmit non-nociceptive information about the distention of the rectum. The striated muscles of the pelvic floor (including the external anal sphincter) are supplied by motor neurons with cell bodies in the spinal cord and axons that run in the pudendal nerves. Triangles represent transmitter release sites; combs represent sensory transduction sites. Anus

Primary Afferent Neurons

Much of the motor and secretory activity of the intestine can be conceptualized as a series of reflexes evoked by mechanical or chemical stimuli. These reflexes involve activation of enteric primary afferent neurons, integration by interneurons, and execution of appropriate responses by motor neurons. The first neurons in these reflex circuits are primary afferent neurons (sometimes called “sensory” neurons, although they do not give rise to conscious sensation). These neurons are located in both myenteric and submucosal plexuses and characteristically have several long axonal processes. Some primary afferents fire action potentials in response to stretch or tension in the bowel

Chapter 98  Colonic Motor and Sensory Function and Dysfunction

Mucosa Meissner’s submucosal plexus

Submucosa

Schabadasch’s submucosal plexus

Circular muscle

Auerbach’s myenteric plexus

Longitudinal muscle

Oral Sensory neuron ending

Aboral Motor (output) neuron ending

Interneuron axon projection

Figure 98-4.  Diagram showing the layers and components of the intestinal wall. The lumen is at the top and the longitudinal muscle layer is at the bottom. Auerbach’s myenteric plexus and the submucosal plexuses (Meissner’s and Schabadasch’s plexuses) are shown, along with some of their major classes of enteric neurons. The networks of interstitial cells are shown in Figure 98-1.

wall; others are activated by chemical or mechanical stimuli of the mucosa. These mucosal stimuli probably work, at least in part, by activating specialized enteroendocrine cells in the mucosal epithelium, such as the serotonin-containing enterochromaffin cells. The primary afferent neurons then release synaptic transmitters, such as acetylcholine or tachykinins or other peptides, to excite other classes of enteric neurons in nearby ganglia. Enteric primary afferent neurons also make excitatory synaptic contacts onto other neurons of their own class, so that they fire in coordinated assemblies.

cyclase-activating peptide (PACAP). The varicose transmitter release sites of inhibitory motor neurons also are associated with ICCIM, just as are the release sites of excitatory motor neurons. Interstitial cells probably mediate a large component of the electrical effects on smooth muscle of neurotransmitters released by enteric motor neurons. Inhibitory motor neurons usually are tonically active, modulating the ongoing contractile activity of the colonic circular smooth muscle. Inhibitory motor neurons are particularly important in relaxing sphincteric muscles in the ileocecal junction and the internal anal sphincter.

Motor Neurons

Interneurons

Enteric motor neurons typically have smaller cell bodies than afferent neurons, with a few short dendrites and a single long axon. Separate populations of motor neurons innervate the circular and longitudinal muscle layers. Excitatory motor neurons synthesize acetylcholine, which they release from their varicose endings in the smooth muscle layers; some also release the tachykinin peptides, substance P and neurokinin A, which excite smooth muscle. Typically, axons of excitatory motor neurons project either directly to the smooth muscle close to their cell bodies or orad for up to 10 mm.4 Once in the smooth muscle layers, the axons turn and run parallel to the smooth muscle fibers for several millimeters; they branch extensively and form many small varicosities, or transmitter release sites, closely associated with intramuscular ICC (ICCIM). Inhibitory motor neurons typically are slightly larger than excitatory motor neurons, and there are fewer of them. They also have short dendrites and a single axon but, unlike excitatory motor neurons, they project aborally to the smooth muscle layer for distances of 1 to 15 mm in the human colon.4 Once the axon reaches the smooth muscle, it branches extensively to form multiple varicose release sites. Inhibitory motor neurons release a cocktail of transmitters that inhibit smooth muscle cells, including nitric oxide, adenosine triphosphate (ATP), and peptides, such as vasoactive intestinal polypeptide (VIP) and pituitary adenyl

When a region of colon is stimulated, such as by a bolus that distends it, intrinsic primary afferent neurons (IPANs) are activated. These neurons then activate excitatory and inhibitory motor neurons, which, because of their polarized projections, cause contraction of the muscle orad to the bolus and relaxation aborally. These effects tend to propel the contents aborally. From the new position of the bolus, another set of polarized reflexes is triggered, and peristaltic propulsion results. The ascending excitatory reflex and the descending inhibitory reflex sometimes are called “the law of the intestine.” These reflexes spread farther than is predicted by the projections of the excitatory and inhibitory motor neurons, because interneurons also are involved in these reflex pathways. Ascending cholinergic interneurons in the human colon have axons that project up to 40 mm orad and extend the spread of ascending excitatory reflex pathways. In addition, several classes of descending interneurons are present in the human colon, with axons that project up to 70 mm aborally. Some of these interneurons are involved in spreading descending inhibition along the colon, but others are likely to be involved in the pro­ pagation of migratory contractions. It also is likely that some interneurons are themselves stretch sensitive, thereby functioning as primary afferent neurons. In addition to the sensory neurons, interneurons, and motor neurons, viscerofugal nerve cells project to the sympathetic prevertebral

1663

1664

Section X  Small and Large Intestine ganglia, vasomotor neurons innervate blood vessels, and secretomotor neurons stimulate secretion from the colonic epithelium.

SYMPATHETIC INNERVATION

The major sympathetic innervation of the proximal colon arises from the inferior mesenteric ganglion and projects through the lumbar colonic nerves to the ascending and transverse colon (see Fig. 98-3). A small number of sympathetic neurons in the celiac and superior mesenteric ganglia, in the paravertebral chain ganglia, and in the pelvic plexus ganglia also project to the colon (see Fig. 98-3). These neurons receive a powerful cholinergic drive from preganglionic nerve cell bodies in the intermediolateral column of the spinal cord (segments L2-L5). This is a major pathway by which the central nervous system modifies bowel activity, such as during exercise. Sympathetic efferent neurons also receive input from the enteric viscerofugal neurons and from extrinsic, spinal sensory neurons with cell bodies in the dorsal root ganglia, forming several reflex loops with the distal bowel. Sympathetic nerve fibers from prevertebral ganglia cause vasoconstriction of the mucosal and submucosal blood vessels. Other cells project to the enteric ganglia, where they cause presynaptic inhibition of synaptic activity in the ENS and thus depress reflex motor activity. Another target for sympathetic axons is the circuitry of the submucosal plexus (largely Meissner’s plexus) involved in controlling epithelial secretion. Hence, these pathways inhibit colonic motor activity, reduce blood flow, and inhibit secretion to limit water loss from the body during times of sympathetic activation. In addition, some sympathetic axons innervate the smooth muscle directly, particularly the ileocecal junction and internal anal sphincter, where they cause contraction; these effects also are consistent with closing down enteric motor activity during sympathetic arousal.

PARASYMPATHETIC INNERVATION

The colon receives parasympathetic innervation from both the vagus nerve and pathways in the sacral spinal cord. Branches of the vagus nerve reach the prevertebral ganglia (superior hypogastric plexus) and then run with sympathetic axons to the cecum and the ascending and transverse colon. The distal colon is supplied largely by sacral parasympathetic axons via the pelvic nerves (pelvic splanchnic nerves). Some of these cholinergic spinal efferent neurons synapse first onto nerve cell bodies in the pelvic plexus (inferior hypogastric plexus), and others project directly to the colon. From their point of entry into the colon, many of the axons run in an oral direction and form thick trunks called shunt fascicles. Parasympathetic axons project to the enteric ganglia in the colon, where they make excitatory cholinergic synapses onto enteric nerve cell bodies. Sacral parasympathetic pathways play an important role in increasing the propulsive activity of the distal colon before defecation and also may be involved in triggering the pro­ pagating complexes that start more proximally before defecation.

EXTRINSIC AFFERENT PATHWAYS

Sensation from the colon is mediated by primary afferent neurons with cell bodies outside the bowel wall. Vagal afferent neurons, with nerve cell bodies located in the nodose and jugular ganglia, project to the proximal colon and run with the vagal efferent parasympathetic pathways. Currently, their exact role in reflex control and sensation is not clear, but they are unlikely to be involved in the transmission of pain sensation from the colon.

The entire colon also is innervated by spinal primary afferent neurons with nerve cell bodies in the lumbar dorsal root ganglia. Lumbar spinal afferents project along the lumbar splanchnic nerves, through the prevertebral inferior mesenteric ganglion, and through the lumbar colonic nerves to the colon, where they terminate in sensory endings in the mesentery, serosa, muscular layers, and mucosa throughout the entire colon and rectum. In addition, a population of spinal afferents, with cell bodies in the sacral dorsal root ganglia, projects along the pelvic nerves to the colon and traverses the pelvic plexus en route. Evidence indicates that some of these sacral spinal afferent neurons form a functionally different population from the lumbar spinal afferents, encoding different types of information, particularly from the rectum. Sacral afferents include many mechanoreceptors with a low threshold and wide dynamic range; these mechanoreceptors probably are responsible for graded sensations of rectal filling and for activating defecatory reflexes.5 By contrast, lumbar spinal afferents and some higherthreshold sacral afferents are responsible for generating pain sensations from all regions of the colon and rectum. They respond to gross distention of the bowel wall, traction on the mesenteric membranes, powerful colonic contractions, or chemical stimulation of the mucosa by bile acids, high osmolarity, and other stimuli. It is well established that the sensitivity of many spinal afferents is increased greatly by inflammation in the colon wall. In addition to their role in sensation, spinal afferents also have axon branches (colla­t­ erals) in enteric ganglia and prevertebral sympathetic ganglia and on mucosal blood vessels, where they might play a role in generating peripheral reflex responses to noxious stimuli. In summary, sacral afferent and efferent (parasympathetic) pathways run in parallel and connect the distal bowel with neural circuitry in the sacral spinal cord via pelvic and rectal nerves. The important role of these pathways in both rectal sensation and in generating the enhanced motility required for defecation is clearly demonstrated by the effects of nerve lesions at several levels. Thus, severing of peripheral nerves and distal spinal cord injury can lead to loss of rectal sensation and to severely impaired defecatory ability.

ANORECTAL ANATOMY AND INNERVATION Although the rectum is in direct continuity with the colon, the longitudinal muscle layer within this region is not organized into teniae; rather, it forms a continuous outer layer, uniformly encircling the rectum, and insinuating between the internal and external anal sphincters to the distal end of the anal canal. The narrowed distal rectum, or anorectal junction, is formed by the longitudinal muscle coat of the rectum, which is joined by the sling fibers of the pubo­ rectalis muscle, attachments of the levator ani muscles, and proximal margins of the internal and external anal sphincters. The puborectalis and levator ani muscles have important roles in maintaining continence and in defecation. These striated muscles form part of the pelvic floor and are in a state of constant tone that serves to pull the rectum anteriorly and elevate it, thereby reducing the anorectal angle; this mechanical effect tends to prevent entry of stool into the upper anal canal. The internal anal sphincter is a thickened band of smooth muscle, with relatively high spontaneous tone, that is in

Chapter 98  Colonic Motor and Sensory Function and Dysfunction continuity with the circular smooth muscle of the rectum. By contrast, the external anal sphincter is a striated muscle and is located distal to, but partly overlying, the internal sphincter. The external sphincter also has a high resting tone, but unlike that of its internal counterpart, its tone can be influenced by voluntary efforts, to help maintain continence. As expected, the sources of innervation of the internal and external anal sphincters are different. The internal sphincter directly receives a powerful inhibitory innervation from intrinsic, enteric inhibitory motor neurons and also extrinsic input from lumbar sympathetic and sacral parasympathetic nerves that project via the pelvic plexus ganglia. The external anal sphincter and other pelvic floor muscles are innervated, through the pudendal nerve (S3S4), by motor neurons with cell bodies in the spinal cord. The rectum and proximal anal canal are richly supplied with sensory receptors that respond to rectal stretch and the composition of the intraluminal contents. These receptors are important for detecting rectal filling, triggering sensations of urgency, facilitating rectal accommodation, and differentiating the composition (stool or gas) of rectal content (see Chapters 96 and 125).

ICCSM at the submucosal border of the circular muscle.1 The timing of these nonpropagating contractions is affected relatively little by enteric motor neural activity but is very dependent on the degree of wall distention. This nonpro­ pagating activity also displays a circadian rhythm, being significantly reduced during sleep.

PROPAGATING MOTOR PATTERNS

When excitatory motor neurons are active, contractions evoked by MPOs summate, giving rise to powerful lumenocclusive contractions that can last longer than slow waves and that can propagate substantial distances along the colon. These identifiable colonic motor patterns are commonly referred to as propagating sequences or high-amplitude propagating sequences (also termed propagating contractions or high-amplitude propagating contractions [HAPCs]). Depending upon their direction of propagation, these patterns can be further qualified by the terms antegrade (aboral) or retrograde (orad). In the healthy colon, antegrade pro­ pagating sequences are recorded with a three-fold higher frequency than retrograde propagating sequences.6 As with nonpropagating activity, propagating sequences display nocturnal suppression and can be stimulated by a meal.

RECTAL MOTOR COMPLEXES

RELATIONSHIPS AMONG CELLULAR EVENTS, PRESSURE, AND FLOW Smooth muscle activation often is divided into two com­ ponents. The first component is the tonic, ongoing activation that gives smooth muscle its basal resistance to stretch, its tone. The second component comprises the dynamic, phasic contractions that mix and propel contents. Compliance is a term used to describe the extent to which the bowel wall can stretch to accommodate contents. For example, a muscle that is very distensible—for example, because of powerful inhibitory motor neuron activity—is said to have a high compliance. During phasic contractions, a transient increase occurs in the resistance of the bowel wall to stretch, namely, a decrease in its compliance. If bowel contents are fluid and no downstream resistance is present to impede flow, the smooth muscle rapidly shortens. The contents are then propelled, with a minimal increase in intraluminal pressure. By contrast, if resistance to forward flow of contents is encountered, as by a lumenoccluding contraction occurring distally, the smooth muscle does not shorten significantly, although its tension increases. This increase in tension increases intraluminal pressure, but it does not cause propulsion. In most situations in vivo, smooth muscle contraction causes a mixture of shortening, increased tension, increased pressure, and propulsion. The process of propagation is controlled by pathways intrinsic to the enteric neural circuitry and by triggering sequences of polarized reflexes that cause peristaltic propulsion.

RECOGNIZABLE COLONIC AND ANORECTAL MOTOR PATTERNS AND PUTATIVE FUNCTIONS NONPROPAGATING MOTOR PATTERNS

Nonpropagating, apparently random activity makes up the majority of the recorded colonic motor activity and is presumed to serve segmenting or mixing functions. The frequency of nonpropagating colonic contractions in vivo is generally 2 to 4 cycles per minute, similar to the frequency of the spontaneous myogenic slow waves generated by

Periodic contractile activity predominates in the sigmoid colon and rectum. This activity is commonly termed the rectal motor complex (RMC) or periodic rectal motor activity (PRMA). The mean RMC amplitude ranges from 15 to 60 mm Hg with a duration of 3 to 30 minutes.6 In contrast to all other colonic contractile patterns, the circadian trend for RMCs is reversed, i.e., the RMC is more prevalent during sleep, suggesting the relevance of the extrinsic neural control of this pattern. The relationship between the RMC and flow is still incompletely understood. RMCs can be triggered by propagating pressure waves from the proximal colon and by the arrival of stool or gas from the sigmoid colon,7 suggesting the RMC provides a braking mechanism to keep the rectum empty.

REGIONAL VARIATION OF PROPAGATING SEQUENCES

Contractile activity in the human colon demonstrates marked regional variation. For example, propagating pressure waves originate nearly four times as frequently in the proximal colon than in the distal colon (Fig. 98-5). The mean distance covered by antegrade pressure waves arising from the cecum is 50 cm, compared with only 20 cm for sequences originating in the descending colon. Still, pressure waves arising proximally generally do not propagate beyond the mid-colon (see Fig. 98-5). It is now clear that slower propagation rates favor the effective propulsion of contents. The conduction velocity of pressure waves increases as the waves migrate caudally. Indeed, such events often accelerate to the point of synchronicity, which arrests the progress of content moving ahead of the contractile front. In addition, nonpropagating (segmenting) pressure waves make up a higher proportion of activity in the distal colon than in more proximal regions.6 Thus, most motor activity in the distal colon functions to retard forward flow, thereby minimizing challenges to continence.

REGIONAL LINKAGE AMONG PROPAGATING SEQUENCES

In addition to the regional variation in propagating sequence frequency, these sequential motor patterns are linked in an

1665

Section X  Small and Large Intestine organized spatiotemporal pattern.8 Many of these regionally linked propagating sequences also form series in which three or more consecutive propagating sequences demonstrate a regional shift in the same direction. Each propagating sequence in a linked series originates in either a more proximal or more distal colonic location. Although most single propagating sequences do not span the length the colon, collectively a series of linked propagating sequences can do so (Fig. 98-6), and it is likely that such linkage is

Cecum

Rectum

20

50

10

25

0 Cecum Hepatic flexure

Splenic flexure

Sigmoid colon

0 Rectum

Propagating pressure waves per colonic region (per 24 hr)

Antegrade propagating sequence frequency (per 24 hr)

1666

Colonic regions Figure 98-5.  Regional variation in the frequency of initiation and extent of propagation of antegrade propagating sequences. The blue histogram (bottom) shows the distribution of antegrade propagating sequences grouped according to the site of origin. The green horizontal bars at the top show the mean extent of propagation by sequences originating at the same site. Note that propagating sequences originate significantly more frequently in the proximal than in the distal colon, and the extent of propagation is much greater for sequences originating in the proximal colon. The solid purple line (bottom) is proportional to the propagating sequence frequency shown at the bottom and indicates that the density of component pressure waves is highest in the mid-colon and lowest at the ends of the colon.

important for the transport of content over great lengths of the colon. The mechanisms underlying regional linkage are yet to be determined.

REGULATION OF COLONIC FILLING AND TRANSIT ROLE OF THE ILEOCECAL JUNCTION

In humans, the ileocecal junction regulates colonic filling and prevents coloileal reflux, thereby preventing contamination of the small bowel with colonic bacteria.9 In the fasting state, cecal filling is slow and erratic, and chyme is retained in the distal ileum for prolonged periods.10 The close physical link between the terminal ileum and the cecum by the ileocecal ligaments behaves functionally as a valve and is responsible in part for continence of the ileocecal junction. A specialized band of muscle forms a low-pressure tonic sphincter11 and prominent 6 cyclesper-minute (cpm) phasic contractions contribute to the regulatory function of the ileocecal junction. Phasic and tonic activity are inhibited concurrently with episodic terminal ileal flow or distention of the ileum, and the tone of the ileocecal junction increases in response to cecal distention.11 Phase III of the interdigestive motor cycle (IDMC) (or migrating myoelectric-motor complex [MMC]), a motor pattern that occurs every 90 to 120 minutes in the upper intestine during fasting (see Chapter 97), does not contribute to ileocecal transit, because it rarely reaches the terminal ileum in the human. Most ileal chyme, driven by ileal propagating contractions in synchrony with inhibition of phasic contractions of the ileocecal junction, enters the cecum in a pulsatile fashion within 90 minutes of a meal. Prolonged studies, over several hours, correlating ileocecal movement of isotope with intraluminal pressures show that 72% of episodes of ileocecal transport result from monophasic, ileal propagating pressure waves.9 Furthermore, 93% of cecal propagating pressure waves were temporally associated with episodes of cecal filling, a finding that suggests episodic cecal filling is one of the triggers for proximal colonic propagating contractions (Fig. 98-7).9

THE COLON AS A STORAGE ORGAN

The region of preferential storage of colonic content is not entirely settled. In 1902, Cannon proposed on the basis of radiologic observations that the proximal colon is the site

mm Hg >150 Cecum

100

Splenic flexure

50

Rectum

0

4

2

6

0

Time (hr)

Figure 98-6.  Six-hour spatiotemporal map of colonic propagating sequences in a female healthy control subject. In this map each individual ridge represents an antegrade propagating sequence. The start of each ridge indicates the site of origin and the time the propagating sequence occurred. The length of the ridge indicates the extent of propagation. The shading within the ridge indicates the amplitude of the component pressure waves. The hatched white arrows link the site of origin of sequential propagating sequences. Although no single propagating sequence spans the entire colon, a linked series of propagating sequences can do so.

Chapter 98  Colonic Motor and Sensory Function and Dysfunction 1

7.5 cm

2

Terminal 3 ileum 4 1

2

5 Cecum 6

3

7 8 Hepatic flexure 20 mm Hg

T = 0 sec 20 sec

T = 10 sec

8 7 3 2 1 6 5 4

T = 20 sec T = 40 sec

Ascending colon

Terminal ileum

Recording sites

Figure 98-7.  Propagating pressure wave sequences identified in the terminal ileum and proximal colon during prolonged combined scintigraphic and manometric recordings. The bottom left corner of the far right box shows a scintigraphic image of technetium sulfur colloid in the terminal ileum and ascending colon of a healthy control subject. The pink circles indicate the location of the recording side holes, each spaced at 7.5 cm. The green hatched lines indicate the regions from which the luminal flow was recorded. Four scintigraphic images have been selected to indicate flow across the ileocolonic junction (solid bars 1 and 2) and mid-ascending colon (solid bar 3). The black arrows correspond to the time (horizontal axis) of acquisition of each 10-second scintigraphic frame. Small blue arrowheads on the scintiscans indicate the location of the manometric side hole from which the corresponding pressure tracing was recorded. Corresponding with the scintigraphic frame at T = 0, a cecal pressure wave is recorded. This cecal pressure wave initiates an ascending colonic propagating sequence that was temporally associated with coloileal reflux (solid bar 1) and flow across the mid-ascending colon (solid bar 2). During the coloileal reflux, an ileal propagating sequence is initiated (hatched black arrow), and this ileal propagating sequence is temporally associated with antegrade flow across the ileocolonic junction (solid bar 3). The red circle on the scintiscan images T = 0 to T = 40 follows the direction of the retrograde flow from cecum to ileum (T = 0 and T =10 sec) and then antegrade flow from the ileum to cecum (T = 20 sec and T = 40 sec).

of storage and mixing, whereas the distal colon acts as a conduit for expulsion. Subsequent studies, however, found no difference in the dwell time for radiopaque markers in the proximal, middle, and distal colon: roughly 11 hours in each. Composition of the diet influences regional transit and probably accounts for some of the discrepancies among studies. With a liquid diet, the ascending colon empties rapidly, within one to two hours, whereas the transverse colon retains isotope for 20 to 40 hours.12 Solid diet retards transit through the cecum and ascending colon. With a mixed diet, particulate matter and liquids are stored in both the ascending and transverse colon.13

RELATIONSHIPS BETWEEN COLONIC MOTOR PATTERNS AND FLOW

Emptying of the proximal colon occurs more rapidly when wall tone is increased (e.g., by intraluminal fatty acids) than when the tone is low; the volume and consistency of the contents also affect the rate of emptying. Isotonic fluid infused into the proximal colon stimulates proximal colonic emptying, suggesting that distention, per se, can activate propulsive motor patterns. Irritant laxatives (which act by stimulating mucosal receptors) in the proximal colon, however, trigger propagating contractions much more reliably than distention alone.14 Hence, proximal colonic emptying is influenced by a combination of increased wall tone and the initiation of propagating contractions, probably under the influence of both chemical and mechanical factors.

Mass movements, first detected radiologically, are infrequent movements of stool over long distances. More often, movement of colonic content occurs in a stepwise manner over short distances and in both antegrade and retrograde directions.15 Studies combining manometry with radiography in animals and with high frame rate scinti­graphy in humans have shown that 93% of all propagating sequences in the proximal colon, regardless of amplitude or polarity, are temporally associated with discrete movements of isotope-labeled colonic contents within the unprepared colon (Fig. 98-8).15 The strength of this pressure-flow relationship is region dependent, being stronger in the transverse colon than in the cecum and ascending colon. Most episodic antegrade movements of colonic content, which are not associated with propagating sequences, can be attributed to repetitive, nonpropagating pressure waves (over short distances in either direction), but there remain a number of movements of content that currently cannot be attributed to identifiable changes in intraluminal pressure. This observation might reflect the occurrence of contractions at points remote from the recording sites. Alternatively, propulsion is sometimes caused by motor events that do not significantly affect intraluminal pressure, such as longitudinal muscle shortening, non–lumen-occluding circular muscle contractions, or alterations in regional wall tone. Retrograde movements occur frequently. About half of retrograde contractions follow immediately after an antegrade movement, indicating frequent reflux of content back into the region from which it had just moved. This subtle to-and-fro motion is likely to help maintain maximal absorp-

1667

1668

Section X  Small and Large Intestine Cecum 1

Ascending colon 2

T = 0 sec Hepatic flexure 3

T = 30 sec Transverse colon 4

T = 45 sec Splenic flexure 5

T = 60 sec Proximal descending colon 6 50 mm Hg 15 sec T = 75 sec Figure 98-8.  Intracolonic pressure measurements and corresponding scintiscans that show a clear correlation between a propagating pressure wave sequence and discrete movement of colonic contents from the cecum to the sigmoid colon. This particular movement of contents was not associated with defecation or sensation. The oblique lines ending in vertical arrows (scintiscan images) correspond to the time of acquisition of each 15-second scintigraphic frame. Small arrowheads on the scintiscans indicate the location of the manometric side hole from which the corresponding pressure tracing was recorded. In the proximal colon and mid-colon (channels 2, 3, and 4 from the top), a close temporal relationship exists between movement of the isotope and the onset of the propagating pressure wave upstroke. When the pressure wave reaches the splenic flexure, however, the proximal descending colon is seen to expand to accommodate the isotope, consistent with loss of lumen occlusion at this region. The pressure waves in channels 5 and 6 do not appear to correspond to lumen-occluding contractions. Note also that propagating pressure-wave amplitudes in channels 3 and 4 are only 30 and 39 mm Hg, respectively, yet the motor pattern is clearly propulsive. (From Cook IJ, Furukawa Y, Panagopoulos V, et al. Relationships between spatial patterns of colonic pressure and individual movements of content. Am J Physiol 2000;278:G329.)

tion, retard colonic transit, and, therefore, reduce stool frequency. In summary, outside of the immediate predefecatory phase (see later), the distal colon displays a combination of fewer propagating sequences, shorter extent of propagation, higher conduction velocity, and lower probability of content propulsion than is observed in the proximal colon. In addition, proportionally more nonpropagating (segmenting) pressure waves occur in the distal colon than proximally. Considered together, these features would be expected to result in retardation of flow into the distal sigmoid and rectum, thus minimizing challenges to continence while maximizing the mixing of content more proximally.

DEFECATION Variations in propagating motor activity along the colon, as just described, would limit or might even prevent colonic contents from ever reaching the rectum and being expelled.

Clearly, additional mechanisms must occur from time to time that lead to defecation. Traditionally, defecation was conceptualized as an exclusively anorectal function; however, evidence for the integration of colonic motor activity with defecation has come from several sources. Radio-opaque markers and scintigraphic recordings confirm that the greater proportion of the entire colonic contents is evacuated in some cases. Furthermore, pancolonic manometric studies have demonstrated that the preparatory phase of defecation not only involves the greater part of the colon but also commences up to one hour before stool expulsion.16 In this predefecatory phase, a characteristic progressive increase occurs in the frequency of propagating pressure wave sequences. These sequences start first in the proximal colon, with each successive sequence originating slightly more distal to the preceding one; these priming sequences do not evoke conscious sensation. By contrast, in the 15 minutes leading up to defecation, a dramatic increase occurs in the frequency of these propagating sequences, which leads to a strong defecatory urge. In the last 15 minutes of the predefecatory

Chapter 98  Colonic Motor and Sensory Function and Dysfunction A

Stool expulsion

Terminal ileum

Cecum

Splenic flexure

Sigmoid colon

1

2

3

4 Attempted stool expulsion

B Cecum

Splenic flexure

Rectum

1

2 80 mm Hg

1 min Figure 98-9.  Intracolonic pressures leading to spontaneous defecation by the healthy human colon. Recordings were made with a perfused Silastic catheter passed transnasally to produce 15 recording sites at 7.5-cm intervals. A, Stool expulsion is preceded by four propagating sequences (1-4), the last of which actually was associated with stool expulsion. Each propagating sequence originates from a site more proximal than the preceding sequence. Note also the increase in amplitude and slowing of propagation velocity with successive sequences leading to stool expulsion. B, Two propagating sequences (1-2) precede defecation; however, the attempted stool expulsion is associated with straining only. (From Bampton PA, Dinning PG, Kennedy ML, et al. Spatial and temporal organization of pressure patterns throughout the unprepared colon during spontaneous defecation. Am J Gastroenterol 2000;98:1027.)

phase, propagating pressure waves begin to originate in the distal colon; however, in this late phase, each successive propagating sequence originates from a site proximal to the preceding one. Each sequence also tends to run for a slightly longer distance and has a higher amplitude compared with the preceding propagating sequence (Fig. 98-9). These final sequences provide potent forces to fill and distend the rectum, activating specialized low-threshold sacral spinal afferent mechanoreceptors. These mechanoreceptors then give rise to the defecatory urge, prompting the expulsive phase in which the anorectum comes into play.

RECTAL FILLING, CAPACITANCE, AND ACCOMMODATION AND MOTILITY OF THE ANAL SPHINCTERS

When stool or gas enters the rectum, the rectal wall is stretched, thereby simultaneously activating an enteric descending inhibitory reflex that causes transient relaxation of the internal anal sphincter and an extrinsic reflex pathway that leads to a brief contraction of the external anal sphinc-

ter. The anorectal inhibitory reflex can be demonstrated and tested by balloon distention of the rectum, and its presence reflects the integrity of enteric neural pathways. For example, the rectoanal inhibitory reflex is absent in Hirschsprung’s disease, which is characterized by loss of enteric ganglia in the rectal myenteric plexus. In health, this reflex permits entry of a small amount of content into the upper anal canal, and continence is maintained by the reflexive contraction of the external anal sphincter. This sampling of content by sensory receptors in the proximal anal canal permits the distinction between solid or liquid stool and gas. Sampling reflexes of this kind occur many times each day in response to low-volume rectal distentions, are not registered consciously, and do not cause an urge to defecate. A large-volume rectal distention causes an internal sphincter relaxation of longer duration, which is registered consciously and which necessitates extra voluntary contraction of the external anal sphincter to maintain continence while the person decides how best to deal with the intraluminal content (stool or gas). Suppression of the defecation urge at this time, together with receptive accommodation of

1669

1670

Section X  Small and Large Intestine

A

B

C

D

E

F

Figure 98-10.  Some of the mechanical processes that facilitate stool expulsion, as illustrated by sequential films of a simulated defecation of thickened barium during defecation proctography. A, The rectum at rest with a normal resting angle of approximately 90 degrees; the anal canal is closed. B, On straining, as the anterior rectal wall begins to flatten, the proximal anal canal begins to funnel as barium contrast is forced into it. C, As more pressure is exerted, the anterior rectal wall flattens further, contrast fills the anal canal, and evacuation begins. At this time, the puborectalis muscle and external anal sphincter are relaxing, resulting in the onset of descent of the rectoanal junction. At the same time, the levator ani muscles are activated and help control the descent of the rectoanal junction (note the posterior indentation resulting from contraction of the pubococcygeus muscle). D, The puborectalis is fully relaxed; this, in combination with vigorous straining, has resulted in nearly complete descent of the rectoanal junction. Note the position of the rectoanal junction, which in this frame is well below the horizontal pale artifact (due to the water-filled toilet seat), compared with that in the previous frame, in which the junction is level with this artifact. This descent has now opened up the anorectal angle, thereby further reducing the resistance to outflow through the anal canal. E, Rectal emptying continues, and anterior rectal compression is more obvious. F, After evacuation, the anorectal junction has ascended to its original position, and the anorectal angle has returned to its more acute resting angle. (Courtesy of Prof. D. Z. Lubowski.)

the rectum (see later), results in temporary storage of stool or gas in the rectum or retrograde transport of the stool or gas back to the sigmoid colon. Although the rectum is usually empty, it has the capacity to temporarily store feces until convenient evacuation can be arranged. Moreprolonged rectal storage is made possible by the ability of the rectum to accommodate an increasing volume without a corresponding increase in intrarectal pressure, in a manner similar to gastric fundic relaxation.17 This adaptive increase in rectal compliance, mediated by inhibitory nerves, is important for maintaining continence by permitting prolonged fecal storage without a constant urge to defecate. Such rectal distention also has negative feedback effects on the proximal bowel and inhibits gastric emptying, slows small bowel transit, reduces the frequency of proximal colonic propagating pressure waves, and delays colonic transit.18 Typically, rectal tone is increased following a meal. A pathologic reduction of rectal compliance, such as after pelvic radiotherapy, causes rectal urgency. Conversely, excessive compliance, as in megarectum, attenuates the urge to defecate.

ANORECTAL MOTILITY DURING DEFECATION

If the processes just described give rise to the urge to defecate and the social circumstances are appropriate, the full defecation process is activated. This process involves a combination of pelvic reflexes coordinated in the medulla and pons. Rectal distention by stool stimulates reflexinduced complete relaxation of the internal anal sphincter,

and the stool moves into the upper anal canal, heightening the sense of urge. Postural changes and straining facilitate this process in several ways: Sitting or squatting causes descent of the anorectal junction, and straining produces further rectal descent. Both activities serve to increase the anorectal angle, thereby reducing resistance to outflow. At this point, if the person wishes to proceed to expel stool, the external anal sphincter is relaxed voluntarily. At the same time, the puborectalis muscle is relaxed (further increasing the anorectal angle); the levator ani muscles contract; the perineum descends further; and stool is funneled into the anal canal and expelled by increasing strain-induced, intrarectal pressure (Fig. 98-10). Once the expulsion phase has commenced, evacuation of stool can proceed in some cases without further straining, as a consequence of colonic contractions propagating toward the anus (see Fig. 98-9).16 Expulsion of stool is possible in response to strain alone without rectosigmoid contractions, although a contribution from increased rectal wall tone cannot be excluded.

MODULATORS OF COLONIC MOTILITY PHYSIOLOGIC

Twenty-four hour recordings of myoelectric activity or intraluminal pressure show that colonic phasic and tonic activity predictably are increased one to two hours after a

Chapter 98  Colonic Motor and Sensory Function and Dysfunction meal (the gastrocolonic response) and are markedly suppressed at night.19 The entire colon responds to the meal, with an increase in colonic wall tone, migratory long spikebursts, and propagating and segmenting contractile patterns. A minimum caloric load of approximately 300 kcal is required to generate the colonic response to a meal, and a meal of only 200 kcal increases rectal muscle tone.20 The meal response also is highly dependent on the fat content of the caloric load. For example, 600 kcal of fat induces the response, whereas an equicaloric load of protein or carbohydrate does not. The mechanism of the colonic meal-response remains unclear, although it is known that neither the stomach nor the spinal cord needs to be intact to display the response. Non-nutrient gastric distention, by balloon or water, also can stimulate rectosigmoid motility, yielding a similar response to that following intraduodenal lipid infusion. Both of these responses are markedly attenuated by prior intravenous administration of the 5-hydroxytryptamine-3 (5-HT3) receptor antagonist granisetron, which suggests that 5-HT3 receptors on vagal afferents may be involved in the gastrocolic response.21 Cholecystokinin (CCK), which is released by fats and fatty acids in the duodenum, can replicate the gastrocolic response, but only at doses exceeding those occurring postprandially. The CCK-A antagonist loxiglumide blocks the effects of CCK on the colon but does not abolish the gastrocolic response, thus making CCK an unlikely mediator of the response. Colonic myoelectric and pressure activities are profoundly suppressed at night.19 During stable sleep, colonic motility virtually ceases (except for the antipropulsive rectal motor complexes, which increase), thereby reducing the challenges to continence at a time when anal sphincter

Sleep stage mm Hg 80 Ascending 0 80 Hepatic flexure 0 80 Transverse 0 Splenic 80 flexure 0 80 Descending 0 80 Proximal sigmoid 0 80 Midsigmoid 0 80 Distal sigmoid 0

Stage 2

Stage 1

tone and awareness of colorectal sensations are minimal. If the subject shifts to a lighter level of sleep, even without actually awakening, an immediate increase occurs in propagating and nonpropagating pressure waves (Fig. 98-11). Forced awakening at night and spontaneous early-morning awakening both stimulate an immediate increase in colonic propagating pressure waves. This phenomenon clearly is linked with the readily identifiable habit of defecation soon after awakening in the morning and demonstrates the potential for profound modulation of colonic motor activity by the central nervous system. Stress and emotional factors long have been believed to influence colonic motility, but experimental evidence for this is conflicting, possibly because of a reliance on measurements from the distal colon, which might not be representative. In light of the profound waking-response, it is likely, but unproved, that stress does induce propagating pressure waves. Due to technical difficulties associated with trying to record physical activity and colonic motility simultaneously, data on the colonic response to physical activity are sparse; however, physical exercise, perhaps through increased sympathetic tone, decreases colonic motility.22 The colonic response to stress and exercise highlight the importance of the autonomic nervous system in modulating colonic function. Similarly, autonomic dysfunction, resulting from pelvic surgery, childbirth, or neural degradation, has been implicated in several colonic disorders including slow-transit constipation and irritable bowel syndrome (IBS).23

PHARMACOLOGIC

Laxatives exert their diarrheal actions by increasing mucosal secretion or by stimulating colonic propulsive activity.

Awake

Arousal Arousal

Stage 1

Stage 2

Arousal

1 min Figure 98-11.  Relationship of propagating pressure wave sequences to nocturnal arousals. An arousal represents a lightening of the level of sleep and need not necessarily culminate in awakening. This example demonstrates an arousal-induced event propagating from the proximal to the distal sigmoid colon and followed by another arousal-induced event propagating from the ascending colon to the sigmoid. Only the second arousal culminated in a brief period of wakefulness. Repetitive propagating sequences of this type also are seen on early-morning awakening. (From Furukawa Y, Cook IJ, Panagopoulos V, et al: Relationship between sleep patterns and human colonic motor patterns. Gastroenterology 1994;107:1372.)

1671

1672

Section X  Small and Large Intestine For example, the irritant laxative bisacodyl and the bile acid chenodeoxycholic acid both stimulate high-amplitude colonic propagating pressure wave sequences, thereby leading to mass movements. Bisacodyl exerts its motor effect through mucosal afferent nerve fibers, because the response can be blocked by topical mucosal application of lidocaine. In addition to the local response, these agents, when administered rectally, can stimulate motor activity in the proximal colon, thereby indicating the existence of long reflex pathways between the rectum and proximal colon. Colchicine, a natural alkaloid, is well known to cause diarrhea. Colchicine increases the frequency of spontaneous bowel movements and accelerates colonic transit in patients with chronic constipation24; the mode of action is not yet clear but it has been shown to increase prostaglandin synthesis and to promote intestinal secretion, the latter mediated through cyclic AMP. In the rat, small intestinal colchicine stimulates myoelectric activity. Lubiprostone, a type 2 chloride channel (ClC2) activator, is a member of a new class of compounds known as prostones. Activation of ClC2 increases intestinal chloride secretion resulting in increased intraluminal fluid accumulation, which accelerates intestinal transit, softens stools, and increases spontaneous stool frequency in patients with constipation.25 Serotonin (5-HT) is an important mediator of bowel physiology, and both 5-HT3 and 5-HT4 receptors play a role in colonic peristalsis and transit. For example, the 5-HT3 receptor antagonists granisetron and ondansetron blunt the gastrocolic response and delay colonic transit, respectively.21 Alosetron, another antagonist of the 5-HT3 receptor, exerts a significant constipating affect by slowing colonic transit.26 In contrast, 5-HT4 agonists (e.g., tegaserod, pru­ calopride, renzapride), act on presynaptic receptors and facilitate release of acetylcholine and CGRP (calcitonin gene-related peptide), thereby inducing colonic propagating contractions and accelerating colonic transit. Although this class of drug shows promise for the treatment of constipation,27 tegaserod, a 5-HT4 agonist, was withdrawn from the market because of concern about associated adverse cardiovascular events (see Chapter 118). Other highly selective 5-HT4 agonists, such as prucalopride, might be attractive options because they do not interact with 5-HT3 or 5-HT1B receptors, and prucalopride does improve stool frequency and symptoms in severe constipation.28 Further trials with these agents are awaited. Opiates are well known to have an antidiarrheal effect, but their mechanism of action is less clear. In the human colon, morphine increases phasic segmenting activity, reduces colonic tone, and attenuates the bowel’s response to a meal.29 Opiates are known to inhibit presynaptic and postsynaptic enteric neural circuitry. The reduction in neurally dependent propagating contractions and the enhancement of myogenic mixing movements and fluid absorption contribute to the constipating effect of the drug. Specific constipation syndromes, such as opiate-induced constipation or postsurgical ileus, might respond to opiate antagonists such as methylnaltrexone and alvimopan (see Chapter 120).30 Nitric oxide is a potent endogenous inhibitor of colonic propulsive activity and the human colon appears to be under a state of tonic nitrergic inhibition. For example, infusion of the nitric oxide synthase inhibitor, l-NMMA (NGmonomethyl l-arginine), is a potent stimulator of colonic propagating contractions.31 Alternatively, segmental lengthening of the colon induced by the entry of content triggers nitric oxide release from descending pathways, which in turn inhibits colonic propulsive activity.32

NONPHARMACOLOGIC

Probiotics are living organisms that, when ingested in adequate amounts, are claimed to exert a health benefit to the host. Relatively few rigorously designed studies have been conducted with probiotics but some strains have been shown to have a beneficial effect in IBS (see Chapter 118), ulcerative colitis (see Chapter 112), and diarrhea.33 In the colon, probiotics are likely to modulate the inflammatory response through activation of signals with the epithelium and immune system.33 Probiotics may well inflence colonic motility, but this has not been systematically evaluated. Sacral nerve stimulation modulates the extrinsic nerves innervating the pelvic floor and colon. Electrical stimulation of the S3 sacral root induces a modest increase in external anal sphincter tone and has been used successfully in the management of fecal incontinence. Stimulation of the S3 root also induces propulsive activity throughout the entire colon and has been shown to increase stool frequency in patients with slow-transit constipation.34 Randomized trials of this promising technique for treating slow-transit constipation are in progress; the precise mode of this action remains unknown. The substantial latency between stimulus and pelvic floor or colonic contractile responses is longer than would be expected via a polysynaptic efferent pathway, which suggests possible involvement of extrinsic neural pathways.35 Magnetic stimulation of the sacral nerve S3 also shows promise in modulating colonic and anorectal function.36 Because this approach is less invasive than electrical stimulation of sacral nerves, it may be a reasonable treatment option in children with colonic or anorectal dysfunction. Acupuncture has been shown to have significant effects upon upper gastrointestinal tract disorders such as nausea and vomiting. Only two studies have evaluated its potential in constipation, one in children and one in adults.37 Acupuncture improved stool frequency in children, but these results weren’t replicated in adults; this warrants further study. Acupuncture is known to activate neural, opioid, humoral, and serotoninergic pathways and potentially has a clinical role in treating disorders such as constipation and IBS.37 Biofeedback has been shown to improve stool frequency and rectal evacuation in patients with pelvic floor dyssynergia, and the technique has been shown to accelerate colonic transit in this subset of patients with constipation (see Chapter 18).38 The mode of action of biofeedback is not fully understood, but evidence suggests that extrinsic autonomic efferent pathways mediate the response.39

DISORDERS OF COLONIC MOTILITY Disorders attributable to disturbed colonic motor function are discussed elsewhere in this book (Chapter 120). It is useful, however, to consider how disturbances in the mechanisms of colonic motility described in this chapter might relate to symptoms or pathophysiologic phenomena.

CONSTIPATION

Intuitively, one would expect that constipation and diarrhea should be manifestations of hypomotility and hypermotility, respectively. Sometimes this is true, but in the distal colon, at least, the converse may be true. A paradoxical increase in nonpropagating (segmenting) contractions and myoelectrical short spike-bursts has been reported in the rectosigmoid region in constipated patients. Conversely, patients with diarrhea have hypomotility in this region.

Chapter 98  Colonic Motor and Sensory Function and Dysfunction A

Morning awakening

Healthy female control subject

Lunch (1000 kCal)

mm Hg >150

Cecum Splenic flexure Rectum

B

100 14:00

22:00 Nocturnal 06:00 suppression Defecation

Patient with slow-transit constipation

Morning awakening

Lunch (1000 kCal)

14:00

50

0 Retrograde Antegrade

Cecum Splenic flexure Rectum

14:00

22:00

06:00

14:00

Time of day (24 hr) Figure 98-12.  Twenty-four-hour spatiotemporal maps of colonic propagating sequences in a female healthy control subject (A) and a female patient with slow-transit constipation (B). Within these maps, each individual ridge represents a propagating sequence. Antegrade propagating sequences (green) originate at the orad end of the ridge, and retrograde propagated sequences (red) originate at the anal end of the retrograde ridge. The start of each antegrade and retrograde ridge indicates the site of origin and the time of day the propagating sequence occurred. The length of the ridge indicates the extent of propagation. The shading within the ridge indicates the amplitude of the component pressure waves. In health (top), the maps indicate several physiologic characteristics of colonic propagating sequences. These include increased frequency and amplitude before defecation, immediately after morning awakening, and in response to a high-calorie meal. The map also indicates nocturnal suppression of these motor patterns. In contrast, the patient with severe constipation (B) demonstrates a clear ability to generate propagating sequences, but there are notable differences in the characteristics of these motor events compared with those of healthy control subjects. For example, the patient demonstrates an increased frequency of low-amplitude antegrade and retrograde propagating sequences in the proximal and distal colon and few propagating sequences spanning the mid-colon. This patient also demonstrates a lack of the normal colonic nocturnal suppression of propagating sequences and a lack of the normal colonic response to a highcalorie meal. (Courtesy of Dr. P.G. Dinning, Kogarah, New South Wales.)

It is likely that segmenting activity retards forward flow, whereas suppression of such activity permits unrestricted access of stool to the rectum, where a defecatory urge is initiated. Thus, constipation can result from either infrequent or ineffective propagating pressure waves, or from an increase in segmenting distal colonic pressure waves, or perhaps both. In severe slow-transit constipation, prolonged manometric studies have confirmed a reduction in the overall number of high-amplitude propagating pressure waves19; however, the overall number of propagating pressure waves of any magnitude is often normal or increased. Studies examining the spatiotemporal patterning of propagating sequences have revealed colonic regions in which activity is diminished or absent, particularly within the vicinity of the splenic flexure (Fig. 98-12).40 Furthermore, there appears to be a loss of the normal linkage in patients who have sequential progressive systemic sclerosis and constipation.8 The underlying pathogenesis of severe slow-transit constipation is unclear, but changes in enteric excitatory motor inner­ vation of the smooth muscle in patients with severe slow-transit constipation are likely to contribute to this disorder.41 Constipation is fully discussed in Chapter 18.

entire colon in patients with diarrhea would help explain these observations, but are lacking. A relative lack of distal colonic segmenting activity, perhaps in combination with increased proximal colonic propagating pressure waves, might explain this preferential acceleration of proximal colonic transit, but proof of this hypothesis is awaited. Diarrhea is fully discussed in Chapter 15.

DIARRHEA

COLONIC MOTILITY DISTURBANCES SECONDARY TO NONMOTOR INTESTINAL DISORDERS

Detailed scintigraphic studies in patients with diarrhea have shown the dominant feature to be early and rapid transit through the ascending and transverse colon.42 Normally, propagating sequences are more frequent in these proximal regions than elsewhere. Manometric data from the

IRRITABLE BOWEL SYNDROME

Although colonic transit generally is slower in constipationpredominant IBS and faster in diarrhea-predominant IBS, no colonic motor pattern is specific for IBS.43 Exaggerated responses to stimuli such as meals, CCK, and mechanical stimuli have been reported, but a consistent disturbance has not emerged, probably because of the heterogeneity of the disease and the methodologies used for characterization. In addition, remarkably little study of the proximal colon in IBS has been conducted to date. At present, compelling evidence regarding the pathophysiology of IBS suggests a major contribution by afferent hypersensitivity, in addition to a variable alteration in colonic motor function. IBS is fully discussed in Chapter 118.

Altered motility secondary to underlying inflammation or a hormonal disturbance can contribute to the colonic symptoms of a nonmotor disease. The diarrhea of idiopathic inflammatory bowel disease, for example, results from a

1673

1674

Section X  Small and Large Intestine combination of enhanced secretion, reduced absorption, and altered colonic motor function. In ulcerative colitis, rectosigmoid-segmenting, nonpropagating pressure waves are diminished, whereas postprandial propagating pressure waves are increased.13 Rectal compliance also is reduced, and together, these effects can exacerbate diarrhea, as suggested by studies demonstrating rapid rectosigmoid transit in ulcerative colitis.13 The motility of the healthy colon also can be perturbed by ileal diseases. For example, exposure of the healthy proximal colon to supranormal concentrations of bile salts, such as from terminal ileal disease or resection, not only stimulates net colonic secretion but also initiates high-amplitude propagating pressure waves, thereby accelerating colonic transit.13

KEY REFERENCES

Bampton PA, Dinning PG, Kennedy ML, et al. Spatial and temporal organization of pressure patterns throughout the unprepared colon during spontaneous defecation. Am J Gastroenterol 2000; 95:1027-35. (Ref 16.) Dickson EJ, Spencer NJ, Hennig GW, et al. An enteric occult reflex underlies accommodation and slow transit in the distal large bowel. Gastroenterology 2007; 132:1912-24. (Ref 32.) Dinning PG, Szczesniak MM, Cook IJ. Removal of tonic nitrergic inhi­ bition is a potent stimulus for human proximal colonic propagating sequences. Neurogastroenterol Mot 2006; 18:37-44. (Ref 31.) Dinning PG, Szczesniak MM, Cook IJ. Determinants of postprandial flow across the human ileocecal junction: A combined manometric and scintigraphic study. Neurogastroenterol Mot 2008; 10:1119-26. (Ref 9.)

Di Stefano M, Miceli E, Missanelli A, et al. Meal induced rectosigmoid tone modification: A low caloric meal accurately separates functional and organic gastrointestinal disease patients. Gut 2006; 55:1409-14. (Ref 20.) Farrugia G. Ionic conductances in gastrointestinal smooth muscles and interstitial cells of Cajal. Ann Rev Physiol 1999; 61:45-84. (Ref 2.) Hagger R, Kumar D, Benson M, Grunday A. Periodic colonic motor activity identified by 24-h pancolonic ambulatory manometry in humans. Neurogastroenterol Mot 2002; 14:271-8. (Ref 7.) Hardcastle JD, Mann CV. Physical factors in the stimulation of colonic peristalsis. Gut 1970; 11:41-6. (Ref 14.) Kamath PS, Phillips SF, O’Connor MK, et al. Colonic capacitance and transit in man: modulation by luminal contents and drugs. Gut 1990; 31:443-9. (Ref 29.) O’Brien MD, Phillips SF. Colonic motility in health and disease. Gastroenterol Clin North Am 1996; 25:147-62. (Ref 13.) Porter AJ, Wattchow DA, Brookes SJ, Costa M. The neurochemical coding and projections of circular muscle motor neurons in the human colon. Gastroenterology 1997; 113:1916-23. (Ref 4.) Rae MG, Fleming N, McGregor DB, et al. Control of motility patterns in the human colonic circular muscle layer by pacemaker activity. J Physiol 1998; 510:309-20. (Ref 1.) Rao SS, Sadeghi P, Beaty J, Kavlock R. Ambulatory 24-hour colonic manometry in slow-transit constipation. Am J Gastroenterol 2004; 99:2405-16. (Ref 19.) Sanders KM, Koh SD, Ward SM. Organisation and electrophysiology of intersitial Cajal and smooth muscle cells in the gastrointestinal tract. In: Johnson LR, editor. Physiology of the Gastrointestinal Tract. Amsterdam: Elsevier; 2006. p 533-76. (Ref 3.) Scott M. Manometric techniques for the evaluation of colonic motor activity: current status. Neurogastroenterol Mot 2003; 15:483-513. (Ref 6.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

99 Intestinal Electrolyte Absorption and Secretion Jayashree Venkatasubramanian, Mrinalini C. Rao, and Joseph H. Sellin

CHAPTER OUTLINE Intestinal Architecture and Transport  1675 Basic Epithelial Cell Model  1677 Segmental Heterogeneity of Transport  1677 Movement across the Intestinal Epithelium  1678 Tight and Leaky Epithelia  1678 Transepithelial Transport  1679 Transcellular Transport  1679 Water Movement  1679 Channels, Carriers, and Pumps  1679 Ion Transporters  1681 Apical Sodium Channel  1681 Nutrient-Coupled Sodium Transport  1681 Sodium-Hydrogen Exchangers  1681 Electroneutral Sodium Chloride Absorption  1682 Chloride (Anion) Absorption  1683 Chloride Secretion  1683 Chloride Channels  1683

The gastrointestinal (GI) tract processes 8 to 9 L of fluid daily that is derived from oral intake and endogenous exocrine secretions. Intestinal fluid absorption is a process that functions with 98% efficiency, allowing only 100 to 200 mL to be excreted each day. The intestine also extracts nutrients, vitamins, and minerals; excludes destructive antigens and microbes; and excretes waste (Fig. 99-1). This multitasking is achieved by the unique tissue, cellular, and molecular architecture of the small and large intestine in combination with a complex array of intricate regulatory mechanisms (Fig. 99-2). Regulation is accomplished by cross-talk between endocrine and paracrine hormones, neurotransmitters, immunomodulators, and luminal factors. Remarkably, this orchestration proceeds smoothly on a daily basis; however, when the balance is perturbed, as occurs with an enteric infection, diarrhea ensues. Over the past four decades, our understanding of intestinal ion transport processes has been revolutionized by the elucidation of the molecular basis of two devastating diseases, cholera and cystic fibrosis. Although the two diseases effect opposite ends of the physiologic spectrum—too much versus insufficient fluid secretion, respectively—examination of their underlying physiologic, regulatory, and genetic parameters have vastly advanced our knowledge. This increased insight of how the intestine transports fluid and electrolytes has had significant clinical impact, most notably in the development of oral rehydration therapy (ORT) for

Calcium-Activated Chloride Channels  1684 Potassium Transport  1684 Bicarbonate Transport  1685 Short-Chain Fatty Acid Transport  1685 Paracrine Immunoneuroendocrine System  1686 Extracellular Regulation  1687 Endocrine, Paracrine, Juxtacrine, and Autocrine Regulation  1687 Neural  1687 Immunologic  1688 Systemic Effects  1689 Osmotic Effects  1689 Specific Regulatory Factors  1690 Absorptive  1690 Secretory  1690 Intracellular Mediators  1692 Homocellular Regulation  1693

diarrheal diseases, one of the major health advances of the 20th century. In this chapter, we review the current understanding of the cellular and molecular underpinnings of the trafficking of ions and solutes in different regions of the small and large intestine and their regulation in health and disease states. The functional activities of intestinal transporters have long been recognized; however, only recently has it become apparent that there are a plethora of transport proteins that carry out these specific functions. This understanding is critical for appreciating normal intestinal function, the pathophysiology of intestinal absorptive abnormalities, and the development of therapeutic strategies for specific diseases.

INTESTINAL ARCHITECTURE AND TRANSPORT The structural and functional design of the intestine is optimally geared to its functions of absorbing nutrients and transporting fluids. In the small intestine, a 600-fold amplification of the absorptive surface is achieved by structural features, such as the circular folds of Kerckring (plicae circulares), villus-crypt architecture, and microvilli. Using a cylinder as the model, it has been estimated that the surface area of the small intestine is about 3300 cm2; the plicae

1675

Section X  Small and Large Intestine Duodenum

Jejunum

Ileum

Cecum

Proximal colon

Distal colon

Absorption 7000 mL Diet 1500 mL 1400 mL Salivary 1500 mL Gastric 2500 mL Biliary 500 mL Pancreatic 1500 mL Intestinal 1000 mL

100 mL

1500 mL Secretions 7000 mL

Tissue permeability

Leaky

Mod. leaky

Mod. tight

Tight

3 mV

6 mV

12 mV

20 mV

Na-nutrient Na+-H+

Na-Cl Na-nutrient Na-bile acid

Na+ SCFA

PD Absorptive mechanisms

Na-Cl SCFA

Na+

[Na+] 20–140 mM

–40

3–20 mV

10

[Na+] 140 mM 3Na+

2K+

Zona occludens (Tight junction)

Gap junction

0 mV

Desmosomes

Basolateral

Apical

la u JA dins M O cc lu di ns Zona adherens

Figure 99-1.  Overview of intestinal fluid balance. Approximately 8.5 L of fluid flow into the intestine daily. Salivary, gastric, biliary, pancreatic, and intestinal secretions make up most of this amount. The bulk of this fluid is absorbed in the small intestine, and approximately 1500 mL cross the ileocecal valve. The colon efficiently reabsorbs most of this fluid, with only 100 to 200 mL lost in stool. Permeability can be viewed as a surrogate to conductance, which is a reciprocal of resistance. Based on Ohm’s law, current = potential difference divided by resistance (I = PD/R). If I is a constant, PD increases as R increases. Permeability of the intestinal epithelium decreases down the length of the cephalocaudal axis, the distal colon having a relatively tight epithelium. Thus, the spontaneous PD demonstrates a corresponding rise with resistance along the cephalocaudal axis. Absorptive mechanisms in each segment of the intestine differ; chloride secretion is found throughout the intestine. SCFA, short-chain fatty acid.

C

1676

Spectrin/actin Paracellular Transcellular

Passive Active Passive

Myosin/ terminal web

ZO-1 ZO-2 ZO-3

Rab3b MUPP1 aPKC PAR-3

Figure 99-2.  Architecture of intestinal epithelia. Intestinal epithelial cells are structurally and functionally geared for vectorial transport: The cell membrane is divided into distinct apical and basolateral domains by the tight junctions with an asymmetrical distribution of transporters; the Na+ pump on the basolateral membrane is integral to maintaining an electrochemical profile; this profile permits downhill entry of sodium from either the apical or basolateral side; water and solutes can cross the epithelium either between the cell (paracellular) or through the cell (transcellular). Transcellular transport across the membrane can be passive or active. The paracellular pathway is characterized by a series of structures that are defined by specific molecular distributions. The tight junction, or zona occludens (ZO), is made up of a network of strands and grooves that consist of membrane proteins (e.g., occludins, claudins, and junctional adhesion molecules [JAMs]) that attach to a group of scaffolding proteins (zonula occludens proteins [ZO-1, ZO-2, ZO-3], multi-PDZ domain protein-1 [MUPP1]). These scaffolding proteins are then linked to the cytoskeleton, participate in vesicular transport (via monomeric guanosine triphosphatase [GTPase] of the Ras superfamily (Rab3b) and in the activation of signaling molecules that regulate junction assembly (partition-defective protein, PAR-3 and -6, and atypical protein kinase C, [aPKC]). Cadherins span the paracellular pathway across the zona adherens and are responsible for cell-to-cell attachment and maintenance of cell polarity. Cadherins bind to catenins, which are linked to the actin cytoskeleton through an additional family of molecules, including radixin, vinculin, and α-actinin. Molecules associated with the zona adherens, including rab, src, and yes, are involved in intracellular signaling through second messengers. Desmosomes are cadherin-like molecules that are linked to intermediate filaments. Gap junctions, made by an assembly of membrane spanning proteins called connexins, allows exchange of small molecules between neighboring cells. PD, potential difference.

Chapter 99  Intestinal Electrolyte Absorption and Secretion circulare, villi, and microvilli amplify the surface area by factors of 3, 10, and 20, respectively, ultimately giving a surface area of about 2,000,000 cm2. In the large intestine, the spatial separation of crypts and surface cells allows efficient reabsorption of fluid. The overall architecture of the intestinal musculature can influence bulk fluid flow and transit time via changes in motility patterns (see Chapters 97 and 98), but the work of fluid transport occurs in the epithelia. Most epithelia serve as semipermeable barriers: They act as the first line of defense between the mucosal (luminal) and serosal (blood-side) compartments and are capable of bulk transport of fluid from one compartment to the other. These epithelia, including those of the intestine, share common characteristics. One fundamental property of epithelia is cellular polarity, with molecularly distinct apical (luminal) and basolateral (serosal) membranes demarcated by intercellular tight junctions. The permeability of the tight junctions vary from being relatively leaky in the small intestine to fairly tight in the large intestine, and these differences determine an individual epithelium’s effectiveness as a barrier. A loss of tight junction integrity disrupts the barrier function and the vectorial transport capabilities of the tissue.

BASIC EPITHELIAL CELL MODEL All GI epithelial cells have two fundamental similarities: discrete apical and basolateral membranes, with distinct biochemical and biophysical properties, separated by tight junctions; and a basolateral Na+ pump (ouabain-inhibitable Na+,K+-ATPase [adenosine triphosphatase]) that establishes a specific intracellular electrochemical environment with a low intracellular Na+ concentration ([Na]i) and a negative intracellular voltage. This basic cell model is modified by insertion of transporters into either the apical or basolateral membrane or by the characteristics of tight junctions that determine the unique qualities of a specific epithelial segment. A complex interaction of protein-sorting signals, cytoskeletal elements, and intracellular trafficking processes determines whether a newly synthesized protein is targeted to either the apical or basolateral membrane. For example, proteins with a glycosyl phosphatidyl inositol (GPI) anchor (e.g., alkaline phosphatase or carcinoembryonic antigen) are often associated with lipid rafts, and the GPI anchor serves to direct them toward the apical membrane.1 Membrane proteins destined to be delivered to the basolateral membrane carry specific membrane-sorting signals (amino acid sequences) in their cytoplasmic tails. In contrast, other proteins can insert randomly into either apical or basolateral domain, but they may be retained in the basolateral pole by specific components such as ankyrin.2 Regulation of intracellular trafficking ensures delivery of the right protein to the right membrane and is critical for establishing epithelial polarization and vectorial transport. When tight junctions are disrupted in vitro, diffusion and intermingling of apical and basolateral proteins in the fluid phase of the membrane result in a loss of epithelial cell polarity. There is some evidence that the distribution of Na+ pumps is altered during postischemic injury.3 The most prominent feature of epithelial cell polarity is targeting of the Na+,K+-ATPase pump to the basolateral membrane, for which expression of the beta subunit of Na+,K+-ATPase is critical. The Na+ pump is electrogenic, extruding three Na+ ions in exchange for two K+ ions, and

thereby maintaining relatively low intracellular Na+ and high intracellular K+ concentrations compared with con­ centrations of these electrolytes in the extracellular environment (see Fig. 99-2). There also is greater membrane permeability for K+ over Na+, which favors diffusional exit of K+ from the cell over diffusional cellular entry of Na+. These features, in combination with the large number of intracellular proteins with fixed negative charges, lead to the characteristic negative intracellular potential difference compared with either the mucosal or serosal compartments.* Low [Na+] and electronegativity establish a favorable electrochemical gradient for passive Na+ entry into the cell. Functionally, the epithelial cell uses the energy of the favorable Na+ gradient to transport not only Na+ ions but also a variety of nutrients, vitamins, and electrolytes. These properties provide the basic mechanisms of ion and water transport that apply to all epithelia. In the intestine, differences in transport can be seen along its cephalocaudal length as well as along the surface-crypt axis within a particular segment of intestine. Tissue- and segment-specific nuances arise from structural-functional and regulatory differences of both intracellular and inter­ cellular proteins.4

SEGMENTAL HETEROGENEITY OF TRANSPORT All intestinal segments from the duodenum to the distal colon have mechanisms for absorbing and secreting water and electrolytes. The diverse physiologic functions along the length of the GI tract are supported by the varied array of transporters encountered in its different segments. For example, the glucose- and amino acid-coupled transporters in the jejunum are well suited for absorption of large volumes of nutrients and water. The cecum, proximal colon, and distal colon, however, exhibit distinctly different transporters, with electrogenic Na+ absorption in the distal colon accomplishing the necessary final fluid extraction in preparation of feces.5-8 Different transporter molecules have been identified in specific segments of the GI tract. What is not clear, however, is why an individual transporter is located only in a specific segment of the intestine. For example, the DRA (down-regulated in adenoma) protein is an anion exchanger, and although anion exchange function is recognized in different segments of the intestine, DRA is predominantly expressed in the colon9-10 (see the later discussion of bicarbonate transport). There also is segmental heterogeneity along the cryptvillus axis. Stem cells near the base of the crypt differentiate and migrate upward to form villus enterocytes in the small intestine or surface colonocytes in the large intestine while undergoing important changes in their transport and barrier properties (Fig. 99-3).11,12 As epithelial cells migrate away from the proliferative zone, the complexity of their tight junctions increases, the microvillus architecture of their apical membrane becomes more pronounced, and underlying cytoskeleton and signaling molecules undergo change; there also is increased expression of Na+ nutrient-coupled transporters, apical Na+-H+ exchangers, and brush border membrane hydrolases. In contrast, the levels of the Na+ pumps remain relatively constant and others, such as *There are several potential differences across the epithelium: across the apical membrane into the cell, from the cell interior across the basolateral membrane, across the epithelium, across the mucosa, and across the entire GI tract. By convention, the potential differences across the epithelium, the mucosa, and the entire GI tract generally are considered the same.

1677

1678

Section X  Small and Large Intestine Lumen

Cells at tip undergo apoptosis and slough into lumen

Villus/surface

Na-channel Na-nutrient NHE-3 PAT

CFTR

Na pump DRA NHE-I

Enterocytes

Endocrine cell

Villus region

Turnover 3-5 days

Mucin-secreting goblet cell

Proliferative zone

Crypt region

Cells mature as they migrate up to the villus region

Paneth Submucosa cells Figure 99-3.  Types of epithelial cells of the intestinal mucosa: enterocytes, endocrine cells, goblet cells, and Paneth cells. All of these cell types originate from the proliferative zone near the base of the intestinal crypt. With the exception of Paneth cells, these cells migrate up the villus axis, mature during this process, and eventually undergo apoptosis and slough after three to five days at the tip of the villus. Paneth cells remain at the base of the crypt and make defensins, which are important in host defense.

the signaling molecule adenylate cyclase and the cyclic adenosine monophosphate (cAMP)-associated Cl− channel CFTR (cystic fibrosis transmembrane conductance regulator), decrease in more mature villus cells. This spatial distribution of transporters (Fig. 99-4) is consistent with a model in which secretory function resides primarily in the crypts and absorption occurs in villus or surface cells. This dichotomy between absorptive surface cells and secretory crypt cells, however, is far from absolute; for example, in the colon, crypts absorb Na+ and fluid, and surface cells secrete Cl−.13,14 Thus, depending on their relative position along the crypt-villus axis, the crosstalk between transporters and their signaling molecules can vary and fine tune intestinal function. This segregation of absorptive and secretory functions might explain why, in diseases that selectively damage villi or surface epithelia—such as enteric infection, inflammatory bowel disease (IBD), and celiac disease—secretion predominates.

MOVEMENT ACROSS THE INTESTINAL EPITHELIUM Movement of ions and solutes across the epithelium is bidirectional and occurs via the transcellular and paracellular routes. Paracellular movement is largely passive, in response

Crypt Figure 99-4.  Spatial location of transport protein gradients. There is a significant spatial geometry of transport proteins along the crypt-villus (crypt-surface) axis. Some transport proteins are found at relatively constant concentrations along this axis, whereas some proteins exhibit a greater density in the base of the crypt and others are denser toward the villus or surface. CFTR, cystic fibrosis transmembrane conductance regulator; DRA, down-regulated in adenoma; NHE, sodium-hydrogen exchanger; PAT, putative anion transporter.

to a variety of gradients, including concentration, electrical, osmotic, and hydrostatic; transcellular movement of ions and solutes occurs by active and passive transport mechanisms. Net transport is termed absorptive if the mucosal-toserosal flux (Jms) is greater than the serosal-to-mucosal (Jsm) flux and, and it is termed secretory if Jsm > Jms. Changes in either or both can alter the direction of the net movement; for example, the ileum, which normally exhibits an absorptive flux, responds to cholera toxin with a decrease in Jms and an increase in Jsm for Cl−, resulting in massive fluid secretion. Characteristics of the tight junctions—for example, tight versus leaky—vary along the length of the intestine and dictate the contribution of paracellular fluxes to overall transport. The effectiveness of a transepithelial gradient may be modified by series of physical barriers, including an unstirred layer created by the glycocalyx above the apical membrane, the lipid composition of the apical and basolateral membrane, the tight junctions, the geometry of the basolateral space between cells, and the basement membrane. Generally, movement of an uncharged particle is dictated solely by concentration gradients. In contrast, the transport of an ion is governed by the electrical potential and concentration differences (the electroche­ mical gradient) across the transported surface. Solvent drag, a nonspecific entraining of solutes along with the movement of water across paracellular pathways, is an absorptive mechanism that may be especially important in the small intestine, for example, for Na+-coupled solute absorption.

TIGHT AND LEAKY EPITHELIA The paracellular space and junctional complexes between cells define the barrier function of epithelia. Epithelia with a low transepithelial voltage and low resistance are con­ sidered leaky, and those that exhibit a high transepithelial voltage and high resistance are considered tight. The tight junctions in villi have higher resistance than do those in crypts. Transepithelial resistance increases in a cephalocaudal direction (see Fig. 99-1).13

Chapter 99  Intestinal Electrolyte Absorption and Secretion Since the 1990s, the model of paracellular transport and tight junctions has rapidly evolved from a static rigid barrier to a dynamic complex structure that is finely regulated (see Fig. 99-2). Movement through the space is exclusively passive, but it is influenced by electrical conductivity, charge selectivity, and its ability to be regulated. Cell-to-cell communications along the paracellular pathway occur in several discrete structures: zona occludens (ZO; tight junction), zona adherens (ZA), desmosomes, and gap junctions. The ZO is composed of several families of proteins that determine its physical and biological properties. For example, claudins belong to a family of 24 membranespanning proteins (24-27 kd) that form pores by interactions of the extracellular domains of claudins of adjoining cells; homotypic adhesion claudins are important in determining the charge selectivity of the tight junction.15,16 Additional proteins in the tight junction include occludins, junctional adhesion molecules (JAMs), and scaffolding proteins such as the zona occludens proteins (ZO-1, ZO-2) and multi-PDZ domain protein 1 (MUPP-1). The scaffolding proteins serve to link membrane proteins to an array of protein kinases, phosphatases and, via filamentous actin, to myosin in the terminal web, thereby influencing paracellular permeability.17,18 For example, disruption of tight junctions by enteropathogenic Escherichia coli is specifically associated with protein kinase Cζ activation.19 Another junctional complex that allows cell-to-cell interaction is the zona adherens. In epithelia, the zona adherens primarily is made up of E-cadherins, 120-kd transmembrane glyco­ proteins, with extracellular motifs that engage in calciumdependent homotypic interaction with cadherins of adjoining cells. Intracellularly, cadherins bind to a family of adhesion molecules, the catenins, which in turn anchor to a dense actin-filament network. Alterations in cadherincatenin distribution or function have been implicated in carcinogenesis.20 Desmosomes are junctional complexes that are structurally similar to zona adherens junctions, although instead of actin, they link to intermediate filaments through a dense plaque of intracellular anchor proteins. Gap junctions have a unique function: They bridge gaps between cells, thus allowing neighboring cells to exchange small molecules. They are made up of an assembly of connexins, a four-pass membrane-spanning protein, six of which join to form a hemichannel. When these hemichannels in two adjoining cells are aligned, they form a continuous pore that connects the interior of the two cells.1

TRANSEPITHELIAL TRANSPORT Our current understanding of the movement of ions, solutes, and fluid across epithelia is gleaned from a combination of in vitro studies using reductionist models of cell lines or isolated epithelial sheets, and from complex in vivo metho­ dologies such as the triple-lumen perfusion technique. All these models underscore that transepithelial ion (largely Na+) movement from the mucosa to the serosa drives fluid absorption, whereas net ion (largely Cl−) movement in the reverse direction drives fluid secretion. Although different approaches help elucidate a complex mechanism, at times they give confounding results. For example, some in vitro studies report decreased Cl− secretion and increased Na+ absorption in the jejunum of cystic fibrosis patients, implying that the intestinal manifestations of the disease are due to hyperabsorption of water. In contrast, in vivo studies show decreases in both Cl− secretion and passive Cl− absorp-

tion, suggesting that rather than a hyperabsorption of fluid, the severity of the disease is reflected by decreased fluid absorption.21 The reductionist models allow us to focus on transport processes at the cellular and paracellular level. In the intact intestine, however, things are more complicated. The geometry of the intestinal wall and the unstirred layer influence the distance that an individual molecule must traverse to reach the apical membrane. The extracellular glycosylated domains of apical membrane proteins make up the glyco­ calyx, which contributes to the thickness and permeability of the unstirred layer; this layer can be a diffusive barrier to the movement of large lipophilic molecules in a chiefly aqueous milieu. Physical parameters such as the mixing of luminal contents by peristalsis, villus motility, and the finer movement of the microvilli influence this rate.

TRANSCELLULAR TRANSPORT Transcellular transport of ions and solutes can be passive or active. Because of the semipermeable nature of the lipid membrane, movement through the cell requires the deployment of specialized membrane proteins, such as channels, carriers, and pumps. The negative intracellular potential favors cation entry into, and anion exit from, the cell. This leads to the curious situation in which ions can move passively against their concentration gradient. For example, although the chemical concentration of Cl− in the cell is relatively low (~35 mmol) compared with the outside concentration (~110 mmol), the intracellular electronegativity creates a driving force for Cl− exit out of the cell.

WATER MOVEMENT Although water movement is a major property of the intestine, the mechanism(s) of intestinal water transport have not been clearly delineated. The movement of water is inextricably linked to the movement of solutes, in response to osmotic gradients. The standing-gradient hypothesis of water absorption suggests that even a small increase (2-3 mOsm) in the osmolarity of the intercellular and subepithelial spaces can cause movement of water across the epithelium, both through and around the cells.22 In the early 1950s, “water pores” were postulated to explain transepithelial water movement, but it was not until the remarkable discovery of the aquaporin (AQP) family of water transporters that a role of specific membrane proteins was implicated in erythrocyte and renal water transport.23,24 Although AQP2, AQP3, and AQP7 have been localized to the small intestine and AQP1, AQP3, AQP4, and AQP8 have been localized to the large intestine, the specific intestinal apical water channel, if any, has eluded discovery.25 Wright and associates proposed that the apical Na+-glucose transporter (SGLT) also may be able to transport water, perhaps as much as 5 L/day,26,27 but whether this can compensate for the puzzling lack of functional apical AQPs remains to be determined.

CHANNELS, CARRIERS, AND PUMPS Small hydrophobic and uncharged molecules move across the lipid bilayer of the cell by diffusion, the rate of transport

1679

1680

Section X  Small and Large Intestine A

Lipid-soluble solutes

B

Lipid-insoluble solutes*

C

D

Lipid-insoluble solutes

Water molecules Extracellular space

Membrane

Cytoplasm Simple diffusion across membrane

Channel-mediated facilitated diffusion via channel proteins

Na+

E

Na+

Carrier-mediated facilitated diffusion via carrier proteins

Osmosis: diffusion via a channel protein or across the membrane

Glucose

Na+

Na+ Extracellular space

Na+

Membrane Active transport K+

K+

1. Primary active transport: creates an ionic gradient

Cytoplasm

Na+

Na+ 2. Secondary active + transport: utilizes the Na gradient to transport uphill

Glucose

Figure 99-5.  Channels, carriers, and pumps. Because only nonpolar solutes freely cross a lipid domain by simple diffusion (A), transfer of ions and charged molecules necessitates specific transmembrane proteins to modulate entry and exit. Ion-specific channels mediate membrane transport by facilitated diffusion (B). Carriers permit facilitated diffusion and transfer specific solutes across the membrane by undergoing a conformational change (C). Trans­ cellular transport of water molecules occurs via channel proteins or carrier proteins (D). Active transport occurs against an electrochemical gradient and can be driven by adenosine triphosphate (ATP) (primary active transport; E1) or an ionic gradient (secondary active transport; E2)

determined by the concentration gradients and diffusion coefficients (Fig. 99-5). Oxygen, carbon dioxide, fat-soluble vitamins, and unconjugated bile acids are examples of substances transported by diffusion. Because the majority of ions and solutes cannot cross the phospholipid membrane by diffusion, the cell employs an array of distinct integral membrane proteins, including channels, carriers, and pumps to cross cell membranes (Fig. 99-5).1 Channels are pores that allow the swift (>106 ions/sec) and controlled (by rapid opening and closing) transit of ions across the membrane, driven by the electrochemical gradient. The advent of molecular cloning techniques, patch clamp methodology (which allows the measurement of function of single channels), and membrane protein crystallography has greatly advanced our knowledge of how these proteins function. Channels tend to be ion selective. For example, Na+ channels exclude K+ despite its same charge and smaller size. Selectivity is determined by the hydration radius of the ion and the physiochemical nature of the pore. The overall transport of a particular ion is determined by the electrochemical gradient, the density of channels, and the gating (open-close time) of the channel; gating may be modulated by voltage, ion, concentration, or intracellular regulation. Mutations of critical residues in the channel protein can have dire functional consequences; for example,

in cystic fibrosis, specific mutations of the CFTR affect the ability to transport chloride and bicarbonate.28,29 Carriers are another class of integral membrane proteins responsible for transport of ions and solutes at rates several orders of magnitude lower than channels. Carrier-mediated transport exhibits substrate specificity, saturation, and inhibitory kinetics. Carriers undergo a series of sequential conformational changes to facilitate the transport of substrates across a membrane. When concentration or electrochemical gradients drive carrier-mediated transport, the process is downhill and is termed facilitated diffusion. For example, the entry of fructose into the enterocyte via Glut-5 is by facilitated diffusion. The entering fructose is rapidly isomerized to glucose, maintaining the downhill gradient for fructose. In contrast, other carriers harness the electrochemical energy established by the downhill movement of a second ion, usually Na+, to move a solute or another ion uphill. This process is termed secondary active transport because the specific gradient is indirectly created by a distinct energy-using process. For example, glucose uptake via the apical membrane Na+-dependent glucose transporter, SGLT, is driven by the Na+ gradient generated by the basolateral Na+,K+-ATPase. Carriers exhibit substrate specificity; thus, SGLT transports d-glucose but not l-glucose. Equally impor-

Chapter 99  Intestinal Electrolyte Absorption and Secretion tant, carriers can transport single or multiple substrates and perform the transport in different directions. Uniporters, such as Glut-2 in the basolateral membrane, transport one type of substrate, hexoses. Symporters, such as the Na+/ K+/2Cl− cotransporter, move Na+, K+, and Cl− in the same direction, whereas antiporters, such as the Na+-H+ exchangers, move the two ions in opposite directions. Pumps are the third class of integral membrane proteins and directly use energy, generally adenosine triphosphate (ATP) hydrolysis, to move ions against an electrochemical gradient. This process is termed primary active transport. Na+,K+-ATPase is the quintessential pump in the intestine. Other pumps important in GI epithelial transport include the luminal gastric and colonic H+,K+-ATPases and the basolateral Ca2+-ATPases.

ION TRANSPORTERS APICAL SODIUM CHANNEL

In the GI tract, the surface epithelial cells of the distal colon and rectum exhibit electrogenic Na+ absorption against a fairly steep concentration gradient. The downhill electrochemical gradient created by the Na+ pump drives Na+ entry via an apical membrane Na+-specific ion channel (Fig. 99-6), which belongs to the family of epithelial Na+ channels (ENaCs). Members of the ENaC family are found in many epithelia.30-32 They are multimeric proteins composed of α, β, and γ subunits; they exhibit a high sensitivity to the diuretic amiloride; and they are stimulated by mineralocorticoids and cAMP. Colonic ENaCs are inhibited by increases in intracellular Ca2+. Unlike many other channels that regulate transport by gating, ENaCs modulate transport by varying the channel density in the cell membrane; this variation may be accomplished through changes (increases or decreases) in synthesis (e.g., aldosterone) or exocytosis (e.g., cAMP, vasopressin) of the channels. Additionally, both aldosterone and cAMP block the association of ENaC with Nedd4-2, a ubiquitin protein ligase, which normally flags the protein for degradation; this block of degradation increases ENaC. Mutations in this pathway can result in the Cl−

HCO3− H+

6 Cl−

1 3 Na+

Na+

+ H+ K

Cl− 3

Na+

5

~

Na+ 2

K+

4 Na+

H2O

H2O

~ GLUT2

Na+

–Glucose Figure 99-6.  Apical sodium transporters. Sodium crosses the apical membrane of the epithelial cell down an electrochemical gradient. The mechanisms may be (1) an ion-specific channel that can be blocked by amiloride; (2) a carrier (e.g., SGLT1) that couples the movement of sodium and a nutrient such as glucose; or (3) a carrier (antiport carrier; e.g., NHE3) that allows electroneutral entry of sodium in exchange for intracellular hydrogen. The common exit pathway across the basolateral membrane is the Na+ pump (4). K+ channels help maintain the electrochemical gradient (5). Cl− moves passively through the paracellular pathway or via cellular transporters (6) . Glucose exits the basolateral membrane via the facilitated diffusion hexose transporter, GLUT2. NHE3, sodium-hydrogen exchanger 3; SGLT1, sodium-glucose cotransporter 1.

increased Na+ absorption and hypertension characteristic of Liddle’s syndrome, an autosomal dominant disorder with features of hyperaldosteronism that is a cause of infantile hypertension.

NUTRIENT-COUPLED SODIUM TRANSPORT

By and large, nutrient transporters are the purview of the small intestine. Transport of many hydrophilic nutrients including glucose, amino acids, and some vitamins occurs against their concentration gradients via secondary active transport at the apical membrane and facilitated diffusion across the basolateral membrane. Glucose transport processes, elegantly elucidated by Wright and coworkers, provide a good example of nutrient transport.33 Transport across SGLT1 is electrogenic (2 Na+ to 1 glucose), is stereospecific (d-isomer), and transports galactose but not fructose.33 Glucose exit across the basolateral membrane occurs via a separate family of facilitated diffusion carriers, the glucose transporters (GLUT-2) (see Fig. 99-6). Fructose enters the cell via another member of this family, GLUT-5, and exits via GLUT-2. Although it is clear that Na+ and glucose absorption stimulates water absorption, the mechanism is not fully delineated. The classic explanation is that basolateral exit of glucose creates a hypertonic compartment in the paracellular space, thereby generating an osmotic gradient for the entry of fluid from the lumen. Some enticing evidence suggests that secondary to transcellular transport through SGLT, passive processes triggers the contraction of the actomyosin ring in the terminal web, resulting in increased paracellular permeability to glucose and to water. Additionally, SGLT activation results in a protein kinase-dependent recruitment of GLUT2 to the apical membrane, which then serves as a high-capacity, low-affinity route for sugar entry during feeding.34,35 Evidence suggests that SGLT can serve as a water channel (210 water to 2 Na+ to 1 glucose) and might account for up to 5 L of fluid reabsorption in the fed state.36 How molecular regulation of transporters is translated into net nutrient absorption during feeding in vivo is a critical area of inquiry. For a description of similar advances made in our understanding of amino acid and vitamin transport, see Chapters 100 and 101.

SODIUM-HYDROGEN EXCHANGERS

The exchange of extracellular Na+ for intracellular H+ is a process that is driven by the electrochemical gradient for Na+ and by a pH gradient resulting from a moderately acidic intracellular environment; this process occurs in almost every cell. In mammalian intestine, members of the Na+-H+ exchange (NHE) gene family play an important role in electroneutral sodium absorption. Electroneutral Na+ absorption may be down-regulated during eating and increases upon postprandial nutrient absorption. Ten mammalian isoforms of NHE have been cloned. NHE1-4 and 6-9 exhibit species- and segment-specific distribution in the GI tract.37-39 NHE1 is a plasma membrane protein found in epithelial and nonepithelial cells. It is expressed on epithelial basolateral membranes and functions as the housekeeper regulator of intracellular pH, cell volume, and growth. NHE2 and NHE3 are apical membrane proteins restricted to epithelia and are the major conduits for electroneutral Na+ absorption in the intestine (see Fig. 99-6). NHE2 is expressed throughout the GI tract, but maximal expression is in the proximal colon. NHE3 is considered a marker for the absorptive cells of the small intestine and colon; it is expressed only in the villus or surface cells, and not in the crypts. NHE4 is located in the basolateral membrane and is primarily expressed in gastric parietal and chief cells, where it might have a role in acid

1681

1682

Section X  Small and Large Intestine and NHERF2, which anchors cGMP kinase protein (GKAP). Activation of guanylate cyclase C by guanylin or E. coli– stable toxin A increases cGMP content near the brush border to locally activate cGKII (Fig. 99-7), which then inhibits NHE3 activity.40

secretion. NHE5 and NHE10 are not expressed in the GI tract, and the roles of NHE6-9 remain to be determined. NHE activity is differentially modulated by neural, paracrine, or endocrine stimuli through intricate scaffolding complexes that include the exchanger itself, a family of NHE regulatory factors (NHERFs) that act as a bridge between the exchanger, and a variety of kinases and phosphatases.40,41 Different stimuli use differing scaffolding complexes to exert their effect. For example, glucocorticoids stimulate Na+ absorption and up-regulate NHE3 but not NHE1, 2, or 4, consistent with their respective roles in vectorial transport and housekeeping. Glucocorticoids act via a serum and glucocorticoid inducible kinase, SGK1; SGK1 stimulates the activity of NHE3 by interacting directly with NHERF2. Alternatively, cAMP inhibits NHE3 by activating protein kinase A, which is recruited to the C-terminus of NHE3 by NHERF1, NHERF2, and an additional cytoskeletal protein, ezrin. In this location, PKA induces its inhibitory effect by phosphorylating NHE3. Cyclic guanosine monophosphate (cGMP) can inhibit NHE3 by triggering the formation of a complex between cGMP-dependent protein kinase II (cGKII) Guanylin STa 1

Extracellular space

ELECTRONEUTRAL SODIUM CHLORIDE ABSORPTION

Sodium absorption is coupled to the movement of Cl− through a Cl−-HCO3− anion exchanger, located specifically in the ileum and proximal colon. The rates of the transporters are similar and coordinated by cell pH and HCO3−. Alkalinization of the cell by NHE drives the exit of HCO3− in exchange for Cl−, resulting in electroneutral Na+ and Cl− absorption, maintenance of cell pH, and luminal release of H+ + HCO3− (water and CO2). The coupling of the two exchangers exhibits species and segmental variations. Thus, it is fairly tight in the ileum and proximal colon, whereas in other intestinal segments, although NHE is dependent on Cl− or HCO3−, the linkage between Na+ and Cl− transport is VIP Prostaglandins 1

Somatostatin

Membrane

R

2

4

Active PKG II

GTP Cyclic GMP

Cytoplasm

Gs 2

3

Adenylate cyclase

ATP

GKAP

Cyclic AMP

3

Active PKA

4

Gi

AKAP

Activity of membrane channels/transporters 5 Figure 99-7.  Second messengers: cAMP and cGMP. Five steps are involved in the transduction of an external signal into a change in cellular function: (1) Binding of either a stimulatory or an inhibitory agonist to an appropriate receptor of the membrane-bound adenylate cyclase or guanylate cyclase system. (2) Binding of ligand to receptor, modulates the cyclase activity either within the same molecule in the case of guanylate cyclase, or by activating the corresponding membrane-bound heterotrimeric guanine nucleotide regulatory proteins (G proteins) in the case of adenylate cyclase. (3) An intracellular signal results from production of cAMP from ATP and cGMP from GTP. (4) Increase in [cAMP]i (intracellular cAMP concentration) activates protein kinases such as PKA and increase in [cGMP]i activates protein kinases such as protein kinase G II, which is fixed to the membrane by myristoylation. Involvement of kinase-anchoring proteins such as A kinase-anchoring proteins (AKAPs) and G kinase-anchoring proteins (GKAPs) has been demonstrated in the signaling. (5) Protein kinase phosphorylation of specific target proteins results in change in the activity of channels or transporters such as chloride channel or the Na+-H+ exchanger. In cAMP signaling, binding of stimulatory regulators, such as VIP and prostaglandins, to specific receptors causes activation. Activated receptors couple via Gs to signal adenylate cyclases to catalyze the conversion of ATP to cAMP, which then activates specific cAMP kinases. An inherent GTPase returns Gs to its nascent state; in cholera, the toxin prevents this occurrence by covalently modifying Gs, leaving enterocyte turnover as the only recourse to returning the tissue to its basal state. Other hormones, such as somatostatin, trigger the activation of inhibitory G proteins (Gi) to decrease cAMP. The adenylate cyclase cascade is localized to the basolateral membrane of epithelial cells. In cGMP signaling, cGMP is generated by the activation of membrane or soluble guanylate cyclases (GCs). In contrast to the adenylate cyclases, membrane GCs are single-pass transmembrane proteins for which the extracellular domain serves as the receptor-binding domain and the intracellular domain catalyzes conversion of GTP to cGMP. Thus, the GCs are specific for their ligands, which include the endogenous atrial natriuretic peptides, guanylin and uroguanylin, as well as enterotoxins such as the heat-stable enterotoxin of E. coli. The intestinal cGMP protein kinase (PGII) is tethered to the membrane via a myristoylated N-terminal region. The soluble GCs are the target of nitric oxide (NO) activation; they are minimally expressed in the small intestinal epithelium, but they are present in colonic epithelia, subepithelial elements, and smooth muscle, where they cause muscular relaxation. ATP, adenosine triphosphate; cAMP, cyclic adenosine monophosphate; cGMP, cyclic guanosine monophosphate; Gi, inhi­bitory G protein; Gs, stimulatory G protein; GTP, guanosine triphosphate; PKA, protein kinase A; PKG, protein kinase G; STa, heat-stable toxin; VIP, vasoactive intestinal peptide.

Chapter 99  Intestinal Electrolyte Absorption and Secretion H2O

Na+

Cl− HCO3−

Lumen

cAMP, cGMP Ca2+ K+ ~ K+

Na+

Na+

2Cl− Figure 99-8.  Intestinal chloride secretion. Discrete basolateral entry steps and apical exit steps are integral to chloride secretion. The Na+-K+-2Cl− carrier couples the movement of Na+, K+, and Cl− in a 1 : 1 : 2 stoichiometric relationship and permits Cl− to accumulate in the cell in a concentration greater than its electrochemical equilibrium. Cl− then exits the cell across the apical membrane by means of a chloride channel; Na+ and water follow passively. The Na+ and K+ that entered with the Cl− are recycled by, respectively, the Na+ pump and a basolateral K+ channel, both critical to maintaining the driving force. These transporters can be regulated by second messengers such as Ca2+, cAMP (cyclic adenosine monophosphate), and cGMP (cyclic guanosine monophosphate).

variable. The molecular nature of the anion exchangers are described under HCO3− transport later.

CHLORIDE (ANION) ABSORPTION

The transepithelial, lumen negative potential difference contributes to the passive movement of Cl− and other anions via the paracellular pathway in the jejunum.7 While coupled Cl−-HCO3− and Na-H exchangers govern electroneutral transport in the ileum and proximal colon, sodium-independent Cl−-HCO3− exchange occurs in the distal colon (see Fig. 99-6) (see the discussion of HCO3− transport, later).

CHLORIDE SECRETION

The principal driving force for the secretion of fluid is the transcellular movement of Cl− from the serosal to the luminal compartment. Na+ and water follow passively in response to the ensuing electrical and osmotic gradients (Fig. 99-8). The small and large intestine exhibit a basal rate of Cl− secretion that is maintained by the interplay of cell volume, [Cl]i, and paracrine, autocrine, neuronal, endocrine, luminal, and immune modulators. Disruptions in the balance of these regulatory processes can lead to secretory diarrhea. Several epithelia in the GI tract exhibit electrogenic Cl− secretion. Although there are some tissue-specific regulatory differences, the mechanisms underlying this secretion are remarkably similar. The Na+ pump provides the driving force, Cl− enters the cell across the basolateral membrane via an electroneutral cotransporter (NKCC1) that couples the movement of 1Na+:1K+:2Cl−, and Cl− leaves the cell via specific channels on the apical membrane. The Na+ entering the cell via NKCC1 exits via the Na+ pump, and the K+ leaves via K+ channels either on the apical or the basolateral membrane. This complex interplay of transporters is an elegant demonstration of cellular economy. The NKCC1 cotransporter effectively moves 2 Cl− and 1 K+ uphill for the expenditure of a single Na+ ion. The pump-to-leak relation between K+ channels and the Na+ pump helps to maintain the interior of the cell as electronegative, thereby providing the driving force for Cl− exit. Basolateral K+ exit electrically balances the large Cl− flux across the apical membrane. NKCC cotransporters belong to a superfamily of cation transporters and are characterized by their inhibition

by the loop diuretics bumetanide and furosemide.42-44 NKCC1 is regulated by many kinases, including a unique PASK (proline-alanine–rich STE20-related kinase), phosphatases, actin-myosin interactions, cell volume, and intracellular Cl−. At least three classes of Cl− channels, belonging to distinct protein families and with distinct electrophysiologic characteristics, have been identified in secretory epithelia. The most important of these is the cystic fibrosis transmembrane conductance regulator (CFTR), which is coded for by the gene that is defective in cystic fibrosis. The crucial role of this channel is underscored by the exocrine pathologies that are the hallmark of cystic fibrosis. In addition to abnormalities of the lungs, sweat glands, and pancreas, infants with cystic fibrosis often present with meconium ileus, and 15% of adults with cystic fibrosis exhibit distal intestinal obstructive syndrome. Interestingly, the major pathology of the CFTR-deficient mouse is meconium ileus, which results in early mortality unless treated with an osmotic laxative. CFTR has been localized to the apical membrane of various segments of the small and large intestine, with greater expression in the crypts than the villus or surface cells.45 CFTR is a 250-kd membrane protein belonging to the superfamily of ATP binding cassette proteins. CFTR is a small-conductance (8-10 pS) linear channel, with an ion selectivity of Br− > Cl− > I− > F− that also can transport HCO3−.46,47 It has two membrane-spanning domains, two ATP-binding domains, and a regulatory (R) domain that has many consensus sequences for phosphorylation by kinases, specifically for PKA. Gating of the channel is regulated by sequential binding of ATP to the two domains and phosphorylation and dephosphorylation of the R domain. In secretory diarrhea such as cholera, PKA activates the R domain to increase channel activity while simultaneously stimulating the recruitment of CFTR-bearing endosomes to the apical membrane, increasing the number of channels (Fig. 99-9). Although more than 1000 mutations of CFTR have been identified, approximately 70% of cystic fibrosis patients carry the ΔF508 mutation, a single amino acid deletion that results in improper folding and diversion of the protein to the endoplasmic reticulum-associated degradative pathway rather than to the apical membrane. The protein is pleiotropic and interacts with a variety of other proteins, influencing their expression and regulation, and being modulated, in turn, by mechanisms that are not fully deciphered. Some of this interaction can occur through crosstalk of CFTR with scaffolding proteins it shares with other transporters; for example, the C-terminal of CFTR interacts with NHERF, a protein that also binds to NHE3. Diminished intestinal fluid secretion in CF−/− mice is associated with goblet cell hyperplasia, increased crypt cell proliferation, Paneth cell abnormalities, and increased inflammation.48 Whether all these changes are secondary to decreased fluid secretion or result from more direct effects of CFTR on other protein function remains to be determined (see later). By combinatorial chemistry, a series of compounds have been designed to specifically inhibit or stimulate CFTR, thereby serving as therapeutic strategies for secretory diarrhea or cystic fibrosis.49

CHLORIDE CHANNELS CIC Family Channels

Chloride channels of the distinct ClC family,50,51 in parti­ cular the widely distributed plasma membrane protein ClC2, are of clinical interest as a potential salvage pathway for Cl− transport in cystic fibrosis patients and as the specific

1683

1684

Section X  Small and Large Intestine A

B

Vibrio cholerae

Cholera toxin

A B ZOT Cl−

HCO−3

Na+

H+

GM1 No effect

Cl− −

Na+ −

Gs cAMP

+

2

Na+ H2 O

3 AC

~

1

Retrograde endocytosis

cAMP

4

EC cells

Na+ K+ VIP 5-HT

+

PG

+

Mesenchymal cells Myenteric plexus

Secretory reflex via interneurons

To colon

Figure 99-9.  Mechanism of action of Vibrio cholerae enterotoxin. V. cholerae produces an enterotoxin, and a zonula occludens toxin (ZOT) that disrupts tight junction permeability, and other toxins (not shown here) that are not fully identified. CT induces diarrhea by inhibiting the absorptive processes in the villus and surface epithelial cells (A) and by stimulating the secretory processes in the crypt epithelial cells (B). As shown in A, a CT-mediated increase in cAMP leads to an inhibition of salt absorption (Na+-H+ and Cl−-HCO3− transporters), but does not affect Na-glucose transport. The events that follow enterotoxin binding leading to cAMP generation are similar in villus and crypt cells. As shown in B, the enterotoxin binds to ubiquitous GM1 gangliosides via its B subunits on the intestinal brush border membrane (1) and, by capturing elaborate cellular processes, including lipid rafts, retrograde endocytosis, and endoplasmic reticular proteolysis (2), transmits the A1 peptide of its A subunit to the basolateral membrane. At the basolateral membrane, A1 catalyzes the adenosine diphosphate ribosylation of Gαs (Gs). This permanently activates Gs by covalently inhibiting the nascent GTPase and preventing activated Gs from returning to its nascent state (3). Activated Gs then stimulates adenylate cyclase to produce an unregulated increase in cAMP (4). Although the major pathophysiologic effects are attributed to the enterotoxin, V. cholerae also recruits multiple components of the PINES, including enteric neurons, enterochromaffin cells, prostaglandins, and serotonin (as shown at the bottom of the figure) that together contribute to the copious fluid output in the small intestine. In addition, interneurons between the small and large intestine underlie the ability of cholera toxin in the small intestine to trigger a reflex secretory response in the colon. AC, adenylate cyclase; cAMP, cyclic adenosine monophosphate; CT, cholera toxin; EC, enterochromaffin; GM1, monosialotetrahexosylganglioside; Gs, stimulatory G protein; 5-HT, serotonin; PG, prostaglandin; PINES, paracrine, immunologic, neural, and endocrine systems; VIP, vasoactive intestinal polypeptide.

target of lubiprostone, a novel laxative.52 ClC2 is involved in regulation of epithelial transport, intracellular pH, intracellular chloride and cell volume. Because of significant differences in its localization in a diverse set of human and animal models, however, its exact role in chloride secretion is uncertain. Thus, although lubiprostone activates ClC2 in the apical membrane of the human colon carcinoma cell line T84, it cannot rescue Cl− secretion in the CFTRdeficient mouse. This dichotomy suggests that the channel is located on the basolateral membrane of the CF mouse. Further research will be needed to clarify how these channels and their pharmacologic agonist work in regulating chloride secretion in humans.

infection. The rotavirus toxin NSP4 activates calciumdependent Cl− secretion, especially in the colons of young mammals. Despite an understanding of the regulation of calcium-dependent Cl− secretion in many epithelia, however, the molecular nature of these channels has eluded precise identification; recent evidence points to a group of proteins, the bestrophins, as potential candidates. It remains to be determined if bestrophins are regulated channels per se or regulators of channels or whether they use auxiliary proteins to function as channels.53,54 The CLCA proteins also might play a role in goblet cell function: Mouse CLCA3 is severely reduced in the CF−/− mouse, and its restoration ameliorates the severity of intestinal cystic fibrosis disease.55

CALCIUM-ACTIVATED CHLORIDE CHANNELS

POTASSIUM TRANSPORT

A third class of Cl− channels, the calcium-activated Cl− (CLCA) channels, are involved in the diarrhea of rotavirus

A plethora of K+ transport processes help the intestine cope with its need to balance fluid and electrolyte movement.56,57

Chapter 99  Intestinal Electrolyte Absorption and Secretion Potassium secretion and absorption occur along the length of the intestinal tract, although the specific pathways are segment specific. In the small intestine, the luminal negative potential difference drives the passive absorption of K+. In contrast, K+ absorption in the distal colon occurs by primary active transport via H+,K+-ATPase pumps located on the luminal membrane. These H+,K+-ATPase pumps are P-type ATPases, related to the gastric H+,K+-ATPase, and least two colonic isoforms have been identified: a ouabain-sensitive isoform in the crypt cells and a ouabain-insensitive isoform in the surface cells. Depletion of aldosterone and K+ upregulates the ouabain-insensitive H+,K+-ATPase and stimulates K+ absorption. Potassium channels are the largest group of ion channels in the human genome. A number of K+ channels have been localized to intestinal cells, but only a few are mentioned here (using the nomenclature of the Human Genome Organization). Found primarily on the basolateral membrane of intestinal epithelial cells, K+ channels contribute significantly to intestinal electrolyte homeostasis through several mechanisms. K+ channels modulate the hyperpolarization of the cell interior that is needed for vectorial, voltagedriven transport processes. In the basolateral membranes of the small and large intestine, the cAMP-activated KCNE3/KCNQ1 channels and the Ca2+-calmodulin– activated KCNN4 channels hyperpolarize the membrane and promote Cl− secretion. Basolateral K+ channels contribute to the transepithelial potential difference, which influences paracellular movement; K+ secretion in the colon occurs via apical KCNMA1 (MaxiK) channels, which are regulated by mineralocorticoids. In response to cell swelling, K+ channels are activated and cause a regulatory volume decrease, a critical function for intestinal cells faced with constant fluctuations in osmolarity. K+ channels also play roles in differentiation, apoptosis, and carcinogenesis, functions that involve different channel proteins and that can be regulated by a number of processes including protein modifications (phosphorylation, sumoylation), membrane voltage, cytosolic calcium, pH, cell swelling, and cell metabolism.

BICARBONATE TRANSPORT

Bicarbonate is a metabolic product and a critical anion in fluid homeostasis in the intestine. In clinically significant diarrhea, bicarbonate is the major anion in the stool. It is secreted by electrogenic and electroneutral processes in the duodenum, ileum, and colon. Being a metabolic product, intracellular HCO3− can arise from intracellular metabolism, diffusion of CO2 or the action of transporters such as the basolateral Na+-HCO3− cotransporter. Electrogenic HCO3− secretion can occur via apical anion channels, including CFTR; however, the major mechanism for HCO3− secretion in the small and large intestine is inexorably linked to the inward movement of Cl− and occurs through apical Cl−HCO3− exchangers.58 It is postulated that electrogenic Cl− secretion via CFTR provides luminal Cl−, which is then recycled across the apical membrane in exchange for intracellular HCO3−. SCFA-dependent bicarbonate secretion also has been observed in surface cells of the colon.59 Although the prototype of anion exchangers, the red cell Cl−-HCO3− exchanger (AE1), has been extensively studied, identification of intestinal exchangers is relatively recent. In a reorganization of nomenclature, the more than 360 identified solute carriers (SLC) are now classified in one of 46 families; of these, the SLC4 bicarbonate transporter family encompasses the red cell anion exchanger (AE1) and its epithelial isoforms (e.g., AE2, AE3, and AE4).60-62 The precise role of the AEs in overall intestinal Cl−-HCO3− trans-

port remains to be clarified, and apical and basolateral localizations have been suggested. Structurally distinct from the SLC4 bicarbonate transporter family is the SLC26 multifunctional anion exchange family. SLC26 exchangers can transport Cl−, HCO3−, sulfate, formate, oxalate, hydroxyl ions, and other anions with differing affinities. Their varied distribution along the GI tract provides them with the flexibility to handle a variety of luminal anions. Two members of this family, SLC26A3 and SLC26A6, are of special interest.10,63-65 The former, first identified as DRA (down-regulated adenoma), is expressed abundantly on the apical membranes of colonocytes, but not enterocytes, and transports more than 2Cl−:1HCO3− ion. Mutations in DRA cause congenital chloride diarrhea, which manifests with severe diarrhea, volume depletion, and metabolic alkalosis. In contrast, SLC26A6, also known as the putative anion transporter 1 (PAT1), is expressed abundantly in the apical membrane of villus enterocytes and less so in the colon. PAT1 transports more than 2HCO3−:Cl− ion; a separate HCO3− conductive pathway might mediate bicarbonate secretion into the duodenum. In contrast to the apical NHE isoforms, anion exchangers are present on the apical membranes of crypt and surface cells. A number of regulatory pathways, including cGMPmediated pathways, strongly influence HCO3− secretion. The physiologic implications of the spatial distribution and varied anion transporters in net HCO3− secretion remain to be elucidated. It is, however, intriguing that HCO3− transport is dependent on CFTR and that anion exchange processes are down-regulated in cystic fibrosis.

SHORT-CHAIN FATTY ACID TRANSPORT

The 2- to 4-carbon short-chain fatty acids (SCFAs), such as acetate, propionate, and butyrate, are generated by the fermentative action of the bacterial microflora on poorly absorbed carbohydrates. Unlike the small intestine, in which Cl− and HCO3− predominate, the major anions (60 to 150 mmol/kg) in the colon are SCFAs; thus, there is a large SCFA concentration gradient across the colonic epithelium. The magnitude of the daily colonic load and absorption of SCFAs is comparable to that of colonic Na+. SCFAs are a major source of metabolic fuel for the colonocytes, modulate epithelial growth and differentiation, and have been implicated in the pathogenesis of and therapy for several inflammatory diseases of the colon, such as bypass colitis. SCFAs are rapidly absorbed in the colon and also greatly enhance Na+ and fluid reabsorption through linked transport mechanisms and by upregulating the expression of NHE3 on the apical membrane of colonocytes.66 SCFAs are weak electrolytes and can be ionized or protonated. Ionized SCFAs need specific carriers, whereas nonionized protonated species can diffuse across the colonocyte membrane; at luminal colonic pH, SCFAs are 95% to 99% ionized. A picture of the molecular basis of SCFA transport is beginning to emerge.32,67 First, apical NHEs can create an acidic pH microclimate and enhance the diffusion of protonated SCFAs into the cell. Monocarboxylate transporters (MCTs), members of two different SLC families, are involved in electroneutral carrier-mediated SCFA transport. Members of the SLC16 family, specifically SLC16A1 (MCT1), transport 1H:1SCFA and require an ancillary protein for their function.68 High-affinity (SLC5A8) and low-affinity (SLC5A12) Na+-dependent MCTs have been identified in the colon and intestine. Although their molecular identity is unclear, Cl−-butyrate exchangers and SCFA-HCO3− exchangers, functionally coupled to Na+-H+ exchange, might

1685

1686

Section X  Small and Large Intestine account for SCFA promotion of electroneutral Na+ and Cl− absorption.59,69

PARACRINE IMMUNONEUROENDOCRINE SYSTEM Extracellular factors from classic paracrine, immunologic, neural, and endocrine systems (PINES) as well as autocrine factors and luminal content regulate intestinal ion transport. The borders separating members of PINES are arbitrary at best, because there is considerable overlap and crosstalk of the underlying factors, and patterns can be altered in disease states (Fig. 99-10). Luminal mechanical (stroking and stretch) or chemical (toxins) stimuli can activate mechanoreceptors and chemoreceptors, respectively, to activate one or more arms of PINES. All these interactions are compounded by many of the factors acting through cell-specific multiple receptors and signaling pathways. Within the subepithelium, structural elements of PINES, including blood vessels, are in close proximity (see Fig. 99-10), so release of mast cell mediators can easily target neurons and vice versa; this proximity and interplay contributes to the minute-by-minute local regulation necessary in the intestine. Although it is possible to separate the specific effects of an individual component in vitro, clinically they are inextricably intertwined. For example, VernerMorrison syndrome (pancreatic cholera) is classified as an endocrine-mediated diarrhea, because pancreatic islet cell tumors produce large amounts of vasoactive intestinal

Food

peptide (VIP). In the healthy adult, however, VIP is not found in the pancreas but is a peptidergic neurotransmitter in the enteric nervous system that stimulates epithelial cell secretion and smooth muscle relaxation. Alternatively, a single agonist such as cholera toxin directly acts on epithelial cells while simultaneously stimulating neural, paracrine, and immune responses. In a third example, serotonin (5-hydroxytryptamine [5-HT]) released by mucosal enterochromaffin cells acts via distinct receptors on epithelial cells to directly stimulate secretion; it acts on myenteric neurons to release acetylcholine (Ach) and elicit migratory contractions; or it acts on submucosal neurons to release Ach and calcitonin gene-related peptide to stimulate peristalsis and secretory reflexes. Fluid secretion is the major component in the production of diarrhea, and the organism’s defensive response to intestinal challenge; motility, mucus secretion, and blood flow, all regulated by PINES, are important adjuncts to the process. The involvement of PINES in motility helps to explain diarrhea associated with rapid intestinal transit (e.g., following gastrectomy), altered anorectal motility (e.g., small-volume diarrhea), or decreased motility (e.g., bacterial overgrowth). Decreased motility leads to a increase in the bacteria in the small intestine, which causes diarrhea by a variety of mechanisms (see Chapters 97 and 102). Alternatively, inflammatory mediators such as prostaglandins or bacterial enterotoxins target both the epithelial and muscle layers to elicit a coordinated secretory response, while promoters of absorption such as opiates and enkephalins suppress motility and promote electrolyte absorption. Thus, PINES allow for a coordinated and integrated response to multiple extracellular signals.

Bacterial toxins/ rotavirus Bile acids

Villus

Immune

Submucosal plexus

Neurocrine 2 Myenteric plexus 1

Autocrine

Ach

PG Mesenchymal cells

3

Paracrine

Endocrine 4 Capillary Circular muscle

Crypt Figure 99-10.  Model depicting the integral components of the enteric nervous system and the immune system that regulate intestinal ion transpart. The components include (1) neurons responsive to intraluminal mechanical and chemical stimuli (e.g., food, bile acids, bacterial toxins, rotavirus); (2) interneurons in either the myenteric or submucosal plexuses; (3) secretory neurons that release acetylcholine (Ach) which acts on epithelial cells; and (4) interactions among secretory neurons and blood vessels, immune cells, and paracrine cells. Intestinal cells can release an array of secretory factors, which can act either directly on the epithelium or indirectly by stimulating the mesenchymal cells or enteric neurons to release prostaglandins (PGs) or acetylcholine (Ach). See text for details.

Chapter 99  Intestinal Electrolyte Absorption and Secretion EXTRACELLULAR REGULATION Tables 99-1, 99-2 and 99-3 list the major neurohumoral substances and toxins that modulate intestinal fluid transport. Agents that promote net fluid secretion generally inhibit Na+ absorption and stimulate Cl− secretion, whereas agents that promote net fluid absorption increase Na+ uptake and attenuate Cl− secretion. In a healthy person, net absorption prevails, and when this balance is disrupted, diarrhea

Table 99-1  Agents That Stimulate Intestinal Absorption of Fluid and Electrolytes Endogenous Absorbagogues α-Adrenergic agonists Aldosterone Angiotensin Enkephalins Glucocorticoids Growth hormone Neuropeptide Y Peptide YY Prolactin Short-chain fatty acids Somatostatin Pharmacologic Agents Berberine Clonidine (α2-agonist) Cyclooxygenase inhibitors Glucocorticoids Lithium Mineralocorticoids Octreotide Opiates Propranolol

can ensue. It is unclear if there is a corollary for a pre­ dominant absorptive pattern in a subset of patients with constipation.

ENDOCRINE, PARACRINE, JUXTACRINE, AND AUTOCRINE REGULATION Intestinal transport is modulated by classic endocrine, paracrine, juxtacrine, and autocrine processes. Intestinal endocrine cells are interspersed between epithelial cells and function as sensors that rapidly respond to changes in the luminal environment by releasing secretory granules containing biogenic amines and hormones; these mediators cross the basolateral membrane. These hormones can act either in a classic endocrine manner on distant target cells (via the circulation) or in a local (paracrine) manner by affecting neighboring cells in the intestinal wall. Juxtacrine mediators are those released from nonendocrine cells, such as neural and inflammatory cells, and they affect neigh­ boring cells. Intestinal mesenchymal cells, in particular myofibroblasts, are a rich source of cytokines, chemokines, eicosanoids, and growth factors that can alter intestinal transport. Epithelial cells can self-regulate (autocrine) by secreting factors such as eicosanoids, which act on epithelial cell receptors to alter function.70

NEURAL

Neural input is critical in the regulation of fluid and electrolyte transport (see Fig. 99-10) and involves interactions of the parasympathetic and sympathetic divisions of the autonomic nervous system with the labyrinthine enteric nervous system (ENS) (see Chapters 1, 97, 98 and 120). Cholinergic stimulation of secretion, predominantly through parasympathetic vagal input, and adrenergic stimulation of

Table 99-2  Endogenous Agonists of Intestinal Secretion AGONIST Acetylcholine Adenosine Arachidonic acid Atrial natriuretic peptide Bombesin Bradykinin Calcitonin, calcitonin gene-related peptide Galanin Gastric inhibitory polypeptide Gastrin Guanylin Histamine Leukotrienes Motilin Neurotensin Nitric oxide Peptide histidine isoleucine Platelet activating factor Prostaglandins Reactive oxygen metabolites Secretin Serotonin Substance P Vasoactive intestinal polypeptide

INTRACELLULAR MEDIATOR 2+

Ca cAMP cAMP cGMP Ca2+ cAMP ? Ca2+ ? Ca2+(PKC/MaPK?) cGMP Ca2+ ? Ca2+ Ca2+ cGMP cAMP cAMP cAMP cAMP cAMP Ca2+ Ca2+ cAMP

cAMP, cyclic adenosine monophosphate; cGMP, cyclic guanosine monophosphate; ENS, enteric nervous system.

SOURCE ENS Immune cells Immune cells, cell membranes Heart ?? Immune cells ENS ?? Endocrine cells Goblet, epithelial cells Immune cells Immune cells Endocrine M cells ENS Immune cells, mesenchymal cells ENS Immune cells Immune cells, mesenchymal cells Immune cells Endocrine cells ENS ENS ENS

1687

1688

Section X  Small and Large Intestine Table 99-3  Luminal Agents That Stimulate Intestinal Secretion AGENT Bacterial Enterotoxins Aeromonas sp. Campylobacter jejuni Clostridium difficile (toxin A) Clostridium perfringens Escherichia coli (heat labile toxin) E. coli (heat stable toxin) Rotavirus NSP4 Salmonella sp. Vibrio cholerae   accessory cholera enterotoxin   enterotoxin   zona occludens toxin Vibrio parahaemolyticus Yersinia enterocolitica Miscellaneous Agents Bile salts Laxatives Long-chain fatty acids

INTRACELLULAR MEDIATOR cAMP cAMP Ca2+ ?? cAMP cGMP Ca2+ cAMP ?? cAMP ?? Ca2+ cGMP cAMP/Ca2+ ?? cAMP/Ca2+

cAMP, cyclic adenosine monophosphate; cGMP, cyclic guanosine monophosphate; NSP, nonstructural protein.

absorption through prevertebral and sympathetic ganglia have long been recognized as fundamental neural pathways affecting the intestinal epithelium. The ENS is the end controller of neural activity in the intestinal wall, however, independently integrating the regulation of the epithelia, muscles, and blood vessels, with input and modification from the central nervous system. Like other neural networks, the ENS has reflexes that can have important clinical implications. Target cells for neurons include components of PINES, blood vessels, and, of course, epithelial cells. Sensory input into the ENS comes from changes in the luminal content (e.g., acidity, dietary content, pathogens) or volume (e.g., stretch). Thus, acid or distension can activate TRVP1 vanilloid receptor on capsaicin-sensitive afferent nerves, which in turn evokes secretion by stimulating submucosal neurons and causes vasodilation by directly activating submucosal arterioles.71,72 Endocrine, dendritic, and/or paracrine cells releasing serotonin, adenosine (see later), and other signals are implicated as auxiliary sensors. Primarily cholinergic, interneurons are believed to underlie the ENS-mediated regulation of colonic epithelial responses to distant small intestinal challenges. Motor neurons that innervate epithelial and submucosal cells can be cholinergic or VIPergic, each releasing additional neuroactive substances. A basal cholinergic secretory drive is tempered by the sympathetic tone; loss of adrenergic sympathetic innervation in diabetic neuropathy is associated with the development of “diabetic diarrhea” and may be corrected by α2-adrenergic agonists.73 As in the brain, the number and variety of compounds that act as neuroactive agents has ballooned, and this has added to the complexity of our elucidating the underlying mechanisms. Thus, neurons can release specific combinations of mediators, such as VIP, cholecystokinin, gastrin-releasing peptide, and ATP, rather than a single substance. Individual neurotransmitters can have biphasic effects varying with concentrations. Furthermore, agents can act as classic neurotransmitters; alternatively, they can act as neuromodulators, fine-tuning the neuronal circuits of presynaptic sites of neurons, or, as in the case of serotonin, they can function as paracrine substances.

IMMUNOLOGIC The clinical correlation between intestinal inflammation and diarrhea is obvious (see Chapter 2), with ulceration, exudation of protein, changes in motility, and loss of absorptive surface area implicated as causing the fluid losses of (inflammatory) diarrheas. Immunocompetent cells of the intestine reside chiefly in the lamina propria, form the gutassociated lymphoid tissue (GALT), and often are distinct from their systemic counterparts; in noninflamed intestine, T lymphocytes account for 60%, with smaller numbers of B lymphocytes and plasma cells (25% to 30%), macrophages (8% to 10%), mast cells, and polymorphonuclear cells (2% to 5%) (usually eosinophils).74 These cells secrete a vast array of soluble products (chemokines, cytokines, eicosanoids, nucleotides, biogenic amines) whose actions are intimately intertwined with other aspects of PINES. Intestinal inflammation increases the number of immunocytes, the cause of the inflammation determining the type of inflammatory cells; for example, acute bacterial infections increase polymorphonuclear leukocytes, whereas parasitic infections dramatically enlarge the mast cell population and celiac disease is characterized by intraepithelial lymphocytes. In IBD, there is activation of all components of GALT with an increase in immunoglobulin (Ig)Gsecreting cells.75,76 Thus, the cause of the inflammatory reaction can determine the type of immunocytes recruited, the range of cytokines released, and the specific effects on transport and motility. Many inflammatory mediators are potent secretagogues, including peptides (e.g., cytokines, platelet-activating factor, substance P, interferon gamma, kallikreins, and bradykinin), eicosanoids (e.g., arachidonic acid, leukotrienes, and prostaglandins), and oxidants (e.g., superoxides). These mediators either interact directly with epithelia to alter ion transport and barrier function or elicit these effects indirectly by activating other PINES elements (see Fig. 99-10). Prostaglandins are central to the secretory response associated with inflammation, affecting various PINES components, such as enteric neurons, and with mediators such as bradykinin, which liberates arachidonic acid to stimulate prostaglandin production. In interpreting the effects of the inflammatory mediators in normal model systems, it is

Chapter 99  Intestinal Electrolyte Absorption and Secretion important to recognize that in vivo, cells damaged by the inflammatory process might not be able to function normally. With this caveat, a few examples are provided. Mucosal mast cells are strategically located in close proximity to enteric neurons, blood vessels, and epithelial cells, and they are central to several inflammatory reactions. Mast cell mediators including histamine, eicosanoids, and cytokines elicit secretion by direct effects on the epithelial cells and by indirect neural stimulation and prostaglandin release. The mechanisms of secretion resulting from polymorphonuclear infiltration of the mucosa recently have become better understood. Polymorphonuclear leukocytes, responding to chemoattractants (e.g., fMLP [formyl-Met-Leu-Phe]), uniquely present in the inflamed colon, leave the vasculature and interact with epithelial cells.77 The white cells burrow through the intercellular space of the colonic cells in a complex integrin-dependent process. The migrating leukocytes release 5-AMP, which is converted to adenosine by apical membrane enzymes. Adenosine is a potent secretagogue, and this adenosine-stimulated secretion might serve as a mechanism to cleanse the crypt lumen.78 Thus, complex specific immunocyte-epithelial cell interactions are important in alterations of electrolyte secretion associated with mucosal inflammation. Anti-integrin targeted therapy may be effective in treating Crohn’s disease (see Chapter 111). During inflammation, oxidants such as superoxides, hydrogen peroxide, and hydroxyl radicals released from neutrophils stimulate Cl− secretion; cytokines such as interleukin-1 and interleukin-3 also stimulate secretion.79 In contrast, interferon-γ and TNF-α can cause diarrhea more through an antiabsorptive effect, by down-regulating particular transporters 80-82 or by altering permeability of tight junctions. Similarly, studies on colonic specimens from patients with IBD or ulcerative colitis indicate that diarrhea in these entities results not from stimulated secretion but from increased tissue permeability combined with decreased apical Na+ channel, basolateral K+ channel, and Na+,K+ATPase expression, which decreases NaCl reabsorptive capacity.76 The multiplicity of transport malfunctions seen in IBD might reflect a sick cell syndrome rather than specific alterations modulated by one or two unique cytokines.

SYSTEMIC EFFECTS Acid-base balance modulates intestinal electrolyte transport in vivo and in vitro. Changes consistent with metabolic acidosis are potent stimulators of electroneutral NaCl absorption, whereas metabolic alkalosis inhibits this process.83-84 Intracellular bicarbonate concentrations can modulate basal Cl− secretion, and intracellular pH and Pco2 can alter Na+:H+ exchange. Volume status and intestinal blood flow also alter ion transport. Any decrease in intravascular volume, such as with hemorrhage, elicits a series of responses that increase fluid absorption. Cardiopulmonary mechanoreceptors and carotid baroreceptors increase sympathetic input into the ENS, resulting in decreased secretion. Angiotensin II, antidiuretic hormone, and atrial natriuretic peptide also can contribute to regulation of intestinal fluid transport in these conditions.85-86 The metabolic status of the intestine has an impact on its transport capability; a well-fed intestine transports more effectively.87 There also are segmental preferences for metabolic fuels. Although the entire intestinal tract uses glucose, the small bowel

effectively uses glutamine, and the colon preferentially uses SCFAs, particularly butyrate.88,89

OSMOTIC EFFECTS Unlike the kidney, the intestinal epithelium cannot maintain an osmotic gradient. Under normal physiologic conditions, the duodenum and upper jejunum are subject to major fluid shifts as they adjust to dietary intake of hypertonic foods and liquids. Rapid equilibration usually is accomplished by movement of water into the intestinal lumen, and absorptive processes along the remainder of the intestine steadily decrease the luminal volume. The continued presence of a nonabsorbable solute within the intestinal lumen, however, can negate functioning absorptive pathways in the distal intestine. This is the basis for osmotic diarrhea (see Chapter 15). Carbohydrates, usually disaccharides, are a common source of nonabsorbable solute. Disaccharides must be hydrolyzed to monosaccharides before they can cross the apical membrane of the small intestine, (see Chapters 100 and 101). The most common clinical example of maldigestion resulting from a deficiency of a specific disaccharidase is lactose intolerance, in which the glucose-galactose disaccharide cannot be broken down because of lactase deficiency. Because the human intestine does not naturally possess a lactulase, the disaccharide lactulose reliably increases small intestinal fluid because of luminal hyperosmolarity and bacterial fermentation of the disaccharide. The limited intestinal absorptive capacity for several sugars found in processed foods and drinks, such fructose and sorbitol, can play important, but often overlooked, roles in osmotic diarrhea, bloating, abdominal pain, and irritable bowel syndrome. The physiology of carbohydrate-induced osmotic diarrhea is complicated by the fact that nonabsorbable solutes in the small bowel can be converted into absorbable solutes by colonic bacteria. Almost all classes of carbohydrates not absorbed by the small intestine are rapidly converted to SCFAs once they cross the ileocecal valve and encounter the colonic bacterial flora; these SCFAs can be absorbed and serve as metabolic fuel for the colon. Thus, depending on the rate of carbohydrate conversion to SCFAs and the colonic capacity for SCFA absorption, small intestinal fluid loss may be compensated by colonic fluid absorption. However, if the capacity of the intestinal flora to convert carbohydrates to SCFAs is maximized, or if the SCFA absorptive capacity of the colon is exceeded, additional unmetabolized carbohydrates could pass through the colon and exacerbate the osmotic effects of nonabsorbable solute.90 Cations (e.g., magnesium) or anions (e.g., sulfate and phosphate) are absorbed poorly by the normal intestine. Increased ingestion of these ions easily leads to osmotic diarrhea, which may be intended, such as with laxatives, or unintended, such as with ingestion of magnesiumcontaining antacids or supplements. In clinical situations in which there is malabsorption or a generalized destruction of the epithelium, solutes normally absorbed readily can remain in the intestinal lumen and thereby contribute an osmotic component to an inflammatory diarrhea or a malabsorptive state. Osmolality is an important factor in patients receiving enteral nutrition (see Chapters 4 and 5). For example, complex carbohydrates provide a significant amount of calories with minimal osmolality compared with simple sugars. Absorption of dipep-

1689

1690

Section X  Small and Large Intestine tides and tripeptides instead of amino acids reduces intestinal osmolality. This balance between calories and osmolality becomes clinically relevant in effectively designing appropriate tube-feeding regimens. Osmolality also is an important factor in designing second-generation oral rehydration therapy (ORT) formulations; by replacing glucose with complex carbohydrates such as rice, intestinal absorption is stimulated further by creating a hypotonic luminal environment, thereby enhancing water absorption. In addition to providing numerous sugar molecules per milliosmole, complex carbohydrates such as rice and amylase-resistant starches have another advantage in ORT: They are metabolized by commensal bacteria in the colon to release SCFAs, which in turn promote fluid absorption (see the earlier discussion of SCFAs).91

SPECIFIC REGULATORY FACTORS ABSORPTIVE

Intestinal agents that stimulate absorption are listed in Table 99-1. Mineralocorticoids (e.g., aldosterone) primarily affect electrogenic Na+ absorption in the distal colon and have little effect on the small intestine, which exhibits electroneutral Na+ absorption. Aldosterone increases the activity and numbers of the apical membrane ENaC (see the earlier discussion of Na channels) and stimulates activity of the Na+/K+ pump and SGK1, resulting in an increase in Na+ absorption. Aldosterone increases K+ absorption and K+ secretion.92 Neonates exhibit a correlation between high circulating levels of aldosterone and enhanced colonic Na+ absorption. Clinically, the physiologic role of aldosterone can be seen in the increased colonic absorption after Na+ depletion (aldosterone stimulation) or in the diarrhea associated with Addison’s disease (aldosterone deficiency). Glucocorticoids are also potent stimulators of Na+ absorption in the small intestine and the colon, in addition to having well-documented anti-inflammatory effects. At low concentrations, glucocorticoids stimulate electroneutral Na+ absorption and suppress electrogenic Na+ absorption, whereas at high concentrations they stimulate both processes. The actions of glucocorticoids are complex and are species and segment specific, and they may be directed at the level of apical Na+ transporters and at the Na+ pump. For example, in the rabbit jejunum, ileum, and colon, glucocorticoid-stimulated increases in Na+ absorption are associated with selective increases in NHE3 but not NHE2 or NHE1 mRNA and protein. Both glucocorticoids and aldosterone evoke rapid cellular responses involving the SGK1 pathway as well as genomic transcriptional effects. These effects might account, in part, for the potent antidiarrheal action of glucocorticoids in a wide variety of clinical settings.38 Catecholamines, enkephalins, and somatostatin all stimulate electroneutral Na+ absorption and often decrease HCO3− secretion with similar patterns of action. They bind to specific heptahelical membrane receptors and activate the Gαi cascade, which in turn suppresses the prosecretory cAMP signaling cascade. Catecholamines (e.g., dopamine and epinephrine) acting as α-adrenergic agonists, all have similar absorptive properties. The theoretical basis for the use of clonidine as an antidiarrheal agent, particularly in diabetic diarrhea, is rooted in this adrenergic absorptive pathway.93 The use of plant opiates as antidiarrheal agents dates back two millennia to the early Egyptians and underscores

their effectiveness. Elucidating their therapeutic effect led to the characterization of the mammalian opioid peptides— enkephalins, endorphins, and dynorphins—a classic example of molecular mimicry.94 Acting via one of three main opioid receptor subtypes—mu, delta, and kappa—the opiates and opioid peptides decrease secretion and promote nonpropulsive motility patterns, thereby increasing transit time. They can act directly on the epithelial and smooth muscle cells or can modify the electrical and synaptic behavior of ENS neurons. The constipation associated with morphine intake can be due to hyperpolarization of secretomotor neurons and suppression of secretion or to a centrally mediated stimulation of sympathetic noradrenergic discharge, or both. Direct activation of K+ channels and inhibition of Ca2+ channels via a G protein-mediated mechanism underlie these effects. Chronic treatment with opiates leads to tolerance, and diarrhea ensues upon abrupt withdrawal. Clinically, management of constipation in patients receiving opiates as analgesics can be a clinical challenge. The development of long-acting analogs of somatostatin has transformed this ubiquitously distributed hormone from a physiologically fascinating regulator to a clinically relevant pharmacologic agent (see Chapter 1). In the intestine, enterochromaffin D cells produce somatostatin, which stimulates salt and water absorption in the ileum and colon and blocks the effects of several secretagogues.95,96 Somatostatin analogs such as octreotide are effective in treating several types of diarrheal diseases, particularly endocrine-related secretory diarrhea. Their therapeutic effect is due to a combination of actions, including inhibition of hormone release by tumors, slowing of intestinal transit, and a direct effect on epithelial cells. Paradoxically, elevated somatostatin levels, as encountered in somatostatinomas or with large pharmacologic doses of octreotide, can precipitate diarrhea secondary to steatorrhea.93 Other peptide hormones, including peptide YY, angiotensin II, and insulin, have been implicated as proabsorptive agents, but their physiologic significance is as yet unknown.

SECRETORY

Endogenous agents that stimulate secretion are listed in Table 99-2. Although there are subtle differences in their biologic actions, in general, eicosanoids, including arachidonic acid, prostaglandins, and leukotrienes, inhibit electroneutral NaCl absorption and stimulate electrogenic Cl− secretion. Most intestinal prostaglandins arise from submucosal immunocytes. They have a major autocrine effect on epithelial cells, but they also modulate enteric nerves and affect intestinal motility and blood flow. Depending on the type of prostaglandin and receptor subtype, prostaglandins primarily act via cAMP and, to a lesser extent, via intracellular Ca2+. Prostaglandins can contribute to the basal secretory tone of the epithelium, because cyclooxygenase inhibitors, such as indomethacin or aspirin, increase basal rates of absorption. Increased intestinal production of eicosanoids contribute to the diarrhea of IBD (see Chapters 111 and 112). Glucocorticoids can decrease prostaglandin synthesis. Although the 5-ASA (acetylsalicylic acid) class of medications, which are a mainstay of the treatment of IBD, target cyclooxygenase and decrease prostaglandin production, their clinical efficacy probably depends on additional mechanisms of action, including effects on other inflammatory pathways, including nuclear factor κB (NFκB) and reactive oxygen species (ROS).97,98 Leukotrienes are considered also to have a role; the mechanism of their action is less well understood

Chapter 99  Intestinal Electrolyte Absorption and Secretion but might involve activation of secretomotor neurons in the subepithelium. Acetycholine, serotonin, guanylin, VIP, and other hormones and neurotransmitters have been implicated as stimulators of intestinal secretion, generally acting to inhibit electroneutral NaCl absorption and stimulate Cl− secretion. They are classified by their mechanisms of action. Acetylcholine acts via the muscarinic receptor M3. Serotonin and neurotensin increase intracellular Ca2+, and VIP and related peptide hormones (e.g., secretin, peptide histidine leucine, and peptide histidine methionine) increase intracellular cAMP. Guanylin acts by increasing cell cGMP.

Serotonin

Serotonin plays a critical role in modulating intestinal motility, sensation, and secretion and is responsible for the diarrhea associated with carcinoid tumors. About 95% of the body’s serotonin is produced by enterochromaffin cells, and the remainder is produced by serotoninergic neurons of the myenteric plexus. Sensory receptors on enterochromaffin cells are activated by mechanical stimuli, acidity, invading pathogens, and dietary contents; for example, SGLT-like protein activates enterochromaffin cells, which serve as a glucose sensor to secrete serotonin (see Fig. 99-9).99 Although not completely elucidated, this signaling involves a complex sequence of autocrine and paracrine actions: ATP is released and converted extracellularly to ADP, which in turn activates a purinergic (P2Y) receptormediated calcium signaling cascade in the enterochromaffin cell to release 5-HT.99 5-HT then acts in a paracrine manner to stimulate epithelial cells, intrinsic primary afferent neurons (IPANs), and extrinsic primary afferent neurons (EPANs). Specific 5-HT receptor subtypes on different IPANs modulate the secretory reflex. Thus, 5-HT1PR on submucosal IPANs and amplifying presynaptic 5-HT4 receptors cause the release of acetylcholine and calcitonin generelated peptide, which stimulate peristaltic and secretory reflexes. In contrast, 5-HT3Rs on myentric IPANs trigger the release of acetylcholine to stimulate giant migrating contractions.99-101 Serotonin stimulation of EPANs results in CNSmediated responses of nausea and discomfort. The major mechanism of serotonin inactivation is by a serotonin reuptake transporter (SERT) on enterocytes and neurons. Interestingly, SERT may be decreased in patients with diarrhea-predominant irritable bowel syndrome (IBS-D) and ulcerative colitis and might account for their increased colonic motility and diarrhea. There has been a concerted effort to clinically alter intestinal function by pharmacologic manipulation of specific serotonin receptors. Thus, 5-HT3 receptor antagonists, such as alosetron, are used to treat IBS-D, and tegaserod, a partial 5-HT4 agonist, can alleviate constipation associated with IBS. Unfortunately, significant side effects have limited their clinical use (see Chapter 118).93,101,102 Adenosine and related purine nucleotides play unique and complex roles in modulating secretion in vivo. As described earlier, they can stimulate secretion directly as it occurs in response to polymorphonuclear leukocyte infiltration of the mucosa or indirectly via release of 5-HT.99 Adenosine acting via P1 purinoreceptors, however, has been shown to attenuate secretion evoked by mechanical stimuli. Activation or inhibition of different populations of channels can underlie these seemingly opposite effects.

Guanylin, Nitric Oxide, and Reactive Oxygen Metabolites

The search for an endogenous activator of the E. coli heatstable enterotoxin receptor led to the discovery of guanylin and uroguanylin, another example of molecular mimicry.

These small peptides, synthesized in goblet and columnar cells activate membrane guanylate cyclase to increase intracellular cGMP and elicit fluid secretion. The guanylin family of peptides, in conjunction with the atrial natriuretic peptides, calibrate the intestinal-renal acid-base response axis. In contrast, nitric oxide, a neuroimmune regulator, stimulates soluble guanylate cyclase to increase cGMP.103 This enzyme is far more prevalent in the subepithelium of the small intestine, but it is expressed in colonic epithelia. Reactive oxygen metabolites, including oxygen free radicals and hydrogen peroxide, generally are produced by immune cells, but they can be released by epithelial cells under certain conditions. They stimulate fluid secretion (see the discussion of immunologic regulation). Additional agonists stimulate secretion (see Table 99-2). Interestingly, most of these agents also affect intestinal motility.

Microbial Pathogens

Microbial pathogens including bacteria, viruses, and fungi can alter electrolyte transport, increase intestinal permeability, and trigger inflammation to elicit diarrhea. They do this by a variety of mechanisms, including attaching to epithelial cells to insert their own products and alter host cell machinery, and by elaborating enterotoxins, which may be cytotoxic or can capture cell-signaling mechanisms to elicit secretion or disrupt tight junctions (see Table 99-3; see also Chapter 107).104-108 A few examples are provided here. Shigella causes dysentery by release of Shiga cytotoxins, which enter the epithelial cell, inhibit protein synthesis, impair absorption, and damage the mucosa. In contrast, secretory diarrheas, such as those associated with cholera and traveler’s diarrhea, result from noninvasive pathogens, which elaborate enterotoxins that capture and turn on the secretory machinery of the epithelium. The archetypal enterotoxin-mediated diarrhea is cholera. Vibrio cholerae carry a virulence cassette that produces at least three different molecules: an enterotoxin; a zonula occludens toxin (ZOT), which disrupts tight junction permeability; and a channel-like protein.109 The enterotoxin causes an unregulated increase in cAMP, activates CFTR, and inhibits NHE3, thus resulting in copious fluid secretion (see Fig. 99-9 legend for details). Mice lacking CFTR do not respond to cholera toxin. Despite voluminous secretion, specific intestinal Na+-coupled nutrient absorptive (Na+-glucose, Na+amino acids) pathways are unaltered by the toxin, forming the physiologic basis for ORT. Bacteria such as Salmonella species, Campylobacter jejuni, and E. coli elaborate enterotoxins similar to cholera toxin, thereby employing the cAMP machinery to elicit fluid secretion. Strains of E. coli and Yersinia enterocolitica associated with traveler’s diarrhea elaborate small-molecular-weight, heat-stable enterotoxins, which increase cGMP to stimulate fluid secretion; signaling cascades distinct from cAMP and cGMP also activate CFTR and inhibit NHE3 (see the discussion of intracellular mediators). Vibrio parahaemolyticus elaborates a thermostable direct hemolysin (TDH) and is a major cause of gastroenteritis; its associated intestinal fluid secretion is attributed to an increase in intracellular calcium and activation of the Ca2+-calmodulin and protein kinase C signaling pathways.93,108 Rotavirus, the major cause of infantile gastroenteritis, induces watery diarrhea. The virus predominantly infects the mature enterocyte of the villus and elaborates an enterotoxin NSP4.110,111 NSP4 can inhibit brush border membrane disaccharidases and SGLT1 activity, thereby limiting Na+glucose and fluid absorption, with resultant diarrhea. NSP4 in vitro elicits Cl− secretion via a calcium-phospholipase C pathway that stimulates Ca2+-activated Cl− channels; this

1691

1692

Section X  Small and Large Intestine effects a transient secretion similar to the actions of carbachol. Unlike other secretagogues, NSP4 has no effect on crypt cell secretion, but under favorable electrochemical conditions it stimulates secretion from villus cells. Paradoxically, NSP4 also can stimulate Cl− absorption from villus cells. The role of NSP4 in vivo has yet to be defined. Many bacterial pathogens use different signaling molecules (e.g., kinases and phosphatases) to perturb the delicate balance of tight junctional proteins and cytoskeletal elements, thus disrupting intestinal permeability. Enteropathogenic (EPEC) and enterohemorrhagic (EHEC) E. coli decrease transepithelial resistance using different signaling cascades. The EPEC strains use a fascinating arsenal to alter host cell responses: They adhere to intestinal cells and in the process recruit a complex network of host cytoskeletal elements. Infectious bacteria use a variety of mechanisms, classified as types I to IV, to infect host cells. EPEC strains use a type III secretion apparatus to insert effector molecules into the host cell and use these molecules to co-opt the cell machinery to cause changes in the actin-myosin network, alter tight-junction proteins, and modulate ion transport processes to decrease absorption and increase fluid accumulation.105,107 The anaerobic bacterium Clostridium difficile, which causes antibiotic-associated pseudomembranous colitis, and Clostridium perfringens, which is associated with food-borne illnesses, alter intestinal permeability by using two distinct processes: The C. difficile toxins A and B interact with the Rho family of cellular proteins to disrupt the perijunctional actin-myosin ring, whereas the tight-junction claudin proteins serve as receptors for C. perfringens enterotoxin (CPE), and binding results in a disruption of tight-junctional fibrils. Other bacteria, including Bacteroides fragilis and V. cholerae, elaborate proteases that attack junctional proteins, including occludins, claudins, and cadherin, to disrupt the integrity of tight junctions.93,108 Many bacterial strains, including EPEC, EHEC, ETEC, Salmonella, and Shigella, trigger a highly specialized cascade to stimulate ion secretion. These pathogens induce the expression of receptors for the peptide neurohormone galanin, whereas uninfected cells do not possess galanin receptors. Galanin in turn activates Cl− secretion via Ca2+dependent signaling processes.112-114,117 The ever-expanding spectrum of toxin-induced mechanisms underscores the importance of delineating the intrinsic regulatory processes and the molecular pathophysiology of infectious diarrhea.

Bile Acids

An increase in colonic bile acids secondary to ileal malabsorption or oral supplementation can cause diarrhea (see Chapters 15 and 64).13,14 Only 7α dihydroxy bile acids such as chenodeoxycholic acid (3α; 7α), but not 7β, are associated with diarrhea. At high concentrations, bile acids act as a detergent and increase intestinal permeability. At more physiologic concentrations, bile salts indirectly increase cAMP and activate mast cells and, more importantly, stimulate epithelial cell Cl− secretion via the Ca2+ and PKCδ cascade.115,116 The ability of bile salts to stimulate Cl− secretion occurs only in the adult and is absent in the neonatal animal. Long-chain fatty acids are seen in increased concentration in the colonic lumen in conditions such as sprue, when long-chain triglycerides are digested by lipase but the fatty acids are malabsorbed within the small intestine (see Chapters 101 and 104). Hydroxylated fatty acids are more potent secretagogues than the corresponding long-chain fatty acids and arise from colonic bacterial metabolism; ricinoleic acid is the long-chain fatty acid that is derived from oral castor

oil, which is a nontoxic oil before it is hydroxylated. Specific fatty acid transporters have been identified in the intestine, and their mechanisms of action in electrolyte secretion are similar to those of bile acids.118,119

INTRACELLULAR MEDIATORS The barrage of extracellular stimuli need to be translated into an intracellular language so the cell can regulate its transport machinery. The second messenger cascades of the cell include the cyclic nucleotides cAMP and cGMP, intracellular Ca2+, and the inositol phosphate-diacyl glycerol and tyrosine kinases. These messenger systems are common to several organ systems, and many cell-specific and tissuespecific structural and functional nuances contribute to the net biological response. More detailed descriptions of second messenger systems can be found elsewhere, but an overview is provided here.120 Epithelial cells require rapid response cascades for turning on and turning off ion transport systems. For example, cyclases are poised to synthesize cyclic nucleotides, and phosphodiesterases are set to degrade them. The net biologic response is governed by the relative contributions of the accentuating and attenuating processes. In addition to the burgeoning increase in new signaling molecules awaiting definition of their roles, there are other variables compounding the resulting net biologic response. The molecules at almost every step in the signal transduction cascade, from the activating hormone to the receptors, cyclases, kinases, and phosphatases, and, finally, the transporters themselves, exist as multiple isoforms and variants (see Fig. 99-7). These isoforms exhibit differences in species, tissue, cell type, and subcellular distribution and are subject to regulation during their development as well as in response to routine physiologic demands. The cannonical cascade of stimulus → second messenger → kinase → response is an oversimplification; extensive crosstalk exists between different signaling pathways. Other post-translational modifications such as glycosylation, myristoylation, nitration, and sumoylation are increasingly recognized as important modulators. Within a signaling cascade there are critical feed-forward and feedback regulatory steps. Thus, protein kinases catalyze the transfer of the terminal phosphate from ATP to the hydroxyl group of a serine, threonine, or tyrosine of a target protein, leading to both conformational and functional changes, such as altered affinity for substrate. They exhibit specificity in their activators and substrates, and their action is essentially irreversible in living cells. Phosphoproteins can only be dephosphorylated by protein phosphatases, which are a separate class of enzymes that are also subject to regulation. Phosphoproteins may be the transporters themselves or may be modulator proteins in the membrane or cytosol, or they may be both. Equally important is the recognition that protein phosphorylation is not synonymous with activation; the dephosphorylated protein may be the active form. In general, Ca2+-specific and cyclic nucleotide-specific protein kinases are serine-threonine kinases, whereas tyrosinespecific protein kinases are associated with receptors of cytokines and hormones involved in growth, such as EGF. In addition to specific serine-threonine and tyrosine phosphatases, dual-specificity kinases and phosphatases add to the complexity of crosstalk. Compartmentalization of components of the signaling cascade via cytoskeletal runners, anchoring domains, or

Chapter 99  Intestinal Electrolyte Absorption and Secretion sequestration in vesicles as a means of regulation is especially germane to the polarized enterocyte. Localizing transporters and their signaling systems into specific subcellular domains is the norm rather than the exception. This is a dynamic and highly regulatable process. Scaffolding proteins can promote docking of various proteins, kinases, and phosphatases by protein-protein interactions—to each other, to the cytoskeleton, and/or to the membrane (see Fig. 99-7); for example, guanylate cyclase C and the intestinal protein kinase GII have cytoskeletal and membrane interacting domains that bring them in close proximity to CFTR in the brush border membrane (Fig. 99-7). Cholesterol-rich membrane domains such as lipid rafts influence membrane fluidity and anchor specific transporters and their regulators. It should come as no surprise that many anchoring domains serve as multienzyme signaling complexes. Finally, trafficking of transporters into and out of the membrane via endosomal vesicles is an effective way of rapidly altering the Vmax of the transporter. For example, cAMP increases CFTR translocation to the membrane and NHE3 retrieval from the membrane, resulting in an increase in Cl− secretion and decrease in Na+ absorption (see Fig. 99-9). Thus, the cAMP-stimulated increases in Cl− secretion and decreases in Na+ absorption (see Fig. 99-9), respectively, result from an increase in CFTR translocation to, and NHE-3 retrieval from, the membrane. Although it is not associated with rapid responses, neurohumoral stimulation can cause changes at the transcriptional level, leading to the synthesis of new proteins, such as aldosterone, which increases ENaC synthesis in the distal colon. Diseases such as cystic fibrosis underscore the importance of intracellular quality control in the cell machinery; thus ΔF508 CFTR, the most common mutation, is a misfolded protein and is tagged for degradation. All signaltransduction mechanisms need to be assessed with respect to their physiologic relevance in the intact intestine because reductionist models, although a necessity, do not provide the complete picture. With these caveats, some common themes have emerged. Generally, agents that elevate intracellular cAMP, cGMP, or Ca2+ increase fluid secretion (see Tables 99-2 and 99-3). They can activate one or more transporters associated with electrogenic Cl− secretion: apical Cl− and K+ channels, basolateral K+ conductances, and NKCC1; they also inhibit the apical NHEs, NHE2 and NHE3. Conversely, fluid absorption is associated with a decrease in these messengers or with an activation of some tyrosine kinase pathways. cAMP plays an additional role in promoting trafficking of transporter-bearing vesicles (CFTR in crypts and distal colonic Na+ channels) to the apical membrane.91,111 The cAMP cascade is triggered by a hormone, , such as VIP, binding to a specific member of the superfamily of heptahelical membrane-spanning receptors (7TM-VPAC1 and VPAC2) (see Fig. 99-7).121 Cyclic GMP is generated by the activation of membrane or soluble guanylate cyclases (GCs) by the natriuretic peptides guanylin and uroguanylin (see Fig. 99-7). Guanylins share their receptor GC-C with the heat-stable enterotoxins (see the earlier discussion of guanylin). Hormones and neurotransmitters such as substance P and acetylcholine activate secretion by increasing intracellular Ca2+ (Fig. 99-11; see figure legend for details). A number of growth factors, cytokines, and inflammatory mediators use entirely different signaling pathways, which involve a combination of extracellular regulated kinases (ERKs), dual-specificity kinases, receptor kinases, and receptor-associated tyrosine kinases.

Acetylcholine, neurotensin 1

Bile acids

Substance P Ca2+ 2 G-protein IP3

PLC 3

DAG

PIP2

G-protein

4 PKC

IP3R 5

Ca2+ Ca/CAM kinase

Ca2+ Calcium stores

Activity of membrane channels/transporters

Figure 99-11.  Calcium signaling in intestinal epithelial cells. Certain hormones and neurotransmitters (e.g., substance P, acetylcholine) activate secretion by increasing intracellular Ca2+. Substance P can stimulate Ca2+ channel activity, and acetylcholine binds to M3 muscarinic heptahelical membrane spanning (HHMS) receptors coupled to the Gaq class of G proteins (1). Activated Gaq stimulates phospholipase C-β (PLC) (2) to hydrolyze PIP2 to release DAG and inositol IP3 (3). DAG can also be produced from phosphatidic acid by the activation of phospholipase D by tyrosine kinase receptors. DAG is rapidly metabolized and does not increase intracellular Ca2+; its major action is to stimulate PKC, a family of phosphatidyl serine-dependent enzymes that have far-reaching biological actions (4). In contrast to DAG, IP3 binds to specific receptors to release Ca2+ from intracellular compartments (5). Intracellular free Ca2+ is tightly regulated and maintained at less than micromolar concentrations, in contrast to the 1-2 mmol/L in the plasma. Transient elevations in intracellular Ca2+ are sufficient to elicit a host of biological responses including ion transport. Calcium directly activates target proteins, such as Ca2+ channels, or binds to the ubiquitous Ca2+-binding protein calmodulin to activate specific calcium-calmodulin protein kinases. Ca2+-dependent secretagogues may be responsible for the minute-by-minute regulation needed in the intestine. This is underscored by the transient nature of Ca2+ signaling and its desensitization to Ca2+-dependent secretagogues. Ca2+ is rapidly resequestered by Ca2+-dependent adenosine triphosphatases on the endoplasmic reticulum or effluxed by Na+-Ca2+ exchange on the plasma membrane. The transient receptor potential channels allow the replenishment of intracellular Ca2+ from the extracellular compartment. PLC activation can concomitantly release polyinositol phosphates, such as inositol 3,4,5,6tetrakisphosphate, which function as ileal brakes and dampen Ca2+-induced Cl− secretion. Phenothiazines and loperamide can interfere with Ca2+ metabolism. CAM, calmodulin; DAG, diacylglycerol; IP3, inositol triphosphate; IP3R, inositol triphosphate receptor; PIP2, phosphatidyl inositol bisphosphate; PKC, protein kinase C; PLC, phospholipase C.

HOMOCELLULAR REGULATION Epithelia need to exhibit homocellular regulation. Given the vicissitudes in luminal content and osmolality, intestinal epithelial cells must be prepared for large and rapid changes in the rates of ion and nutrient transport. What enters the cell on one side must exit the cell at the other end at a similar rate. If not, the cell will either shrink or explode, owing to a rapid change in ionic content and osmolality.

1693

1694

Section X  Small and Large Intestine The concept of homocellular regulation is best exemplified by Na+ absorption. Thus, when Na+-glucose cotransport is maximally stimulated, the amount of Na+ flowing through the cell each minute is several orders of magnitude greater than the basal cell Na+ content and needs to be regulated. The rate-limiting step for transepithelial Na+ transport is entry across the apical membrane, for both channel transport and carrier-mediated transport. Changes in rates of Na+ entry initiate a series of coordinated changes in the Na+ pump, basolateral K+ conductances, and the apical entry mechanism itself such that the intracellular environment remains constant. Potential regulators of this dialog between membranes include cell volume, ATP, stretch-sensitive ion channels, and specialized intracellular pools of either Na+ or Ca2+. The ability of the cell to fine-tune discrete events at its opposite borders allows it to function effectively.

KEY REFERENCES

Alberts B, Johnson S, Lewis J, et al. Molecular Biology of the Cell. New York: Garland Science, Taylor & Francis Group; 2007. (Ref 1.) Alper SL. Molecular physiology of SLC4 anion exchangers. Exp Physiol 2006; 91:153-61. (Ref 61.) Barrett KE, Seely SJ. Integrative Physiology and Pathophysiology of Intestinal Electrolyte Transport. San Diego: Academic Press; 2006. pp 1931-51. (Ref 4.) Field M. Intestinal ion transport and the pathophysiology of diarrhea. J Clin Invest 2003; 111:931-43. (Ref 13.)

Gershon MD, Tack J. The serotonin signaling system: from basic understanding to drug development for functional GI disorders. Gastroenterology 2007; 132:397-414. (Ref 101.) Jentsch TJ, Neagoe I, Scheel O. ClC chloride channels and transporters. Curr Opin Neurobiol 2005; 15:319-25. (Ref 51.) Kunzelmann K, Mall M. Electrolyte transport in the mammalian colon: mechanisms and implications for disease. Physiol Rev 2002; 82:24589. (Ref 32.) Kunzelmann K, McMorran B. First encounter: how pathogens com­ promise epithelial transport. Physiology (Bethesda) 2004; 19:240-4. (Ref 108.) Schneeberger EE, Lynch RD. The tight junction: a multifunctional complex. Am J Physiol 2004; 286:C1213-28. (Ref 16.) Thiagarajah JR, Verkman AS. New drug targets for cholera therapy. Trends Pharmacol Sci 2005; 26:172-5. (Ref 109.) Rao, M. Absorption and secretion of water and electrolytes. In: Ratnaike R, editor. Small Bowel Disorders. London: Hodder Headline Group; 2000. pp 116-34. (Ref 120.) Rao MC. Oral rehydration therapy: new explanations for an old remedy. Ann Rev Physiol 2004; 66:385-417. (Ref 91.) Weber CR, Turner JR. Inflammatory bowel disease: is it really just another break in the wall? Gut 2007; 56:6-8. (Ref 82.) Wright EM, Loo DD, Hirayama BA, Turk E. Surprising versatility of Na+-glucose cotransporters: SLC5. Physiology (Bethesda) 2004; 19:370-76. (Ref 33.) Zachos NC, Tse M, Donowitz M. Molecular physiology of intestinal Na+/ H+ exchange. Annu Rev Physiol 2005; 67:411-43. (Ref 38.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

100 Digestion and Absorption of Nutrients and Vitamins James J. Farrell

CHAPTER OUTLINE Digestion and Absorption of Nutrients  1695 An Overview of the Digestive Process  1695 Fat  1698 Dietary Intake  1698 Digestion and Absorption  1699 Carbohydrate  1706 Dietary Intake  1706 Digestion and Absorption  1707 Protein  1712 Dietary Intake  1712 Proteins from Endogenous Sources  1712 Digestion and Absorption  1712 Vitamins  1717 Water-Soluble Vitamins  1717 Fat-Soluble Vitamins  1720 Minerals and Trace Elements  1722 Calcium  1722 Magnesium  1723

Most nutrients are absorbed with remarkable efficiency: Less than 5% of ingested carbohydrate, fat, and protein is excreted in the stool of adults consuming a normal diet.1 Even much of the indigestible dietary fiber is absorbed from the colon as short-chain fatty acids that are liberated by bacterial breakdown of fiber.2 The intestinal tract of neonates is less efficient: infants fail to absorb 10% to 15% of their dietary fat, and in premature infants as much as 25% to 35% of fat may be lost in the stool.3,4 In old age, nutrient absorption remains highly efficient unless the intestine becomes diseased. Despite considerable variations in types of food ingested and nutritional intake across national and racial groups, absorption remains efficient. Absorptive mechanisms adapt to the nature and amount of various nutrients presented to the intestinal tract. Such changes occur not only during early development5 but also throughout life and at times of specific need, such as during pregnancy.6 In achieving the overall objective of nutrient absorption, the different parts of the gastrointestinal tract act in a closely integrated and coordinated manner under the control of neural and humoral regulatory mechanisms. The understanding of intestinal digestion and absorption at a molecular level has improved our knowledge of the integration and coordination of these functions within the gastrointestinal tract. The pharmacokinetics and pharmaco-

Iron  1723 Trace Elements  1725 Adaptation to Changes in Need or Load  1726 Mucosal Hypertrophy  1726 Specific Reversible Adaptation  1727 Vitamins and Trace Elements  1728 Signaling for Intestinal Adaptation and Implications for Therapy  1729 The Neonatal Intestine  1729 Development and Adaptation of Nutrient Digestion and Absorption  1729 Developmental Changes  1729 Triglyceride Digestion  1730 Carbohydrate Digestion and Absorption  1732 Protein Digestion and Absorption  1732 Effects of Bariatric Surgery on Normal Digestion and Absorption  1732

dynamics of several key carbohydrate, fat, peptide, amino acid, vitamin, and nutrient transporters are increasingly understood. In this chapter, integration of intestinal function with the dietary intake, digestion, and absorption of major nutrients (fat, carbohydrate, and protein) and essential micronutrients (vitamins and trace elements) is discussed. The evolving genetic and molecular basis of these functions also is evaluated.

DIGESTION AND ABSORPTION OF NUTRIENTS AN OVERVIEW OF THE DIGESTIVE PROCESS

The cerebral phase of digestion, whether triggered by the sight, smell, or thought of food, initiates the digestive process. Salivary and gastric secretory responses to this type of stimulus are mediated via the autonomic nervous system, and there is modest stimulation of pancreaticobiliary secretion via the vagus nerve.7 The further stimulus of nutrients in the mouth and upper gastrointestinal tract markedly potentiates secretion by humoral and local neural mechanisms (see Chapter 1).8 The rapidity with which food is normally chewed and swallowed affords little time for significant oral digestion of nutrients; however, good mastication and mixing with

1695

1696

Section X  Small and Large Intestine saliva initiates digestion of starch by salivary amylase. In infants, digestion of fat is begun in the stomach by gastric lipase. Gastric acid would soon switch off these enzymes were it not for the buffering capacity of food that allows some digestion to continue. The optimal pH for gastric lipases is 4.5 to 6.0, and it has been suggested that a considerable portion of dietary triglyceride may be digested by these lipases.9,10 Protein digestion begins in the stomach when gastric pepsinogens are converted to pepsins by gastric acid. Pepsins become increasingly active as intraluminal pH falls and therefore the digestive action of pepsins on proteins is restricted to the stomach. During ingestion of food, the stomach may become distended, but intragastric pressure rises little because of neurally mediated receptive relaxation. The mechanisms by which subjects perceive satiety and, therefore, cease eating are complex and explained only partly by the sensation of fullness. Although dozens of enzymes and hormones are secreted by the gastrointestinal tract in response to intraluminal food, only a few are able to influence food intake directly. Satiety signals are relayed to the hindbrain, either indirectly via nerves such as the vagus nerve or else directly via the blood. Most factors that influence how much food is eaten during individual meals act by changing the sensitivity to satiety signals.11 Cholecystokinin (CCK), gastrin-releasing peptide, and apolipoprotein A-IV (apo A-IV) have all been implicated as messengers that transmit the satiety signal to the central nervous system (CNS).12,13,14 They potentiate each other’s actions, and a combination of these agents may participate in the satiety signal. The administration of exogenous CCK or other satiety signals causes smaller meals to be consumed, whereas blocking the action of endogenous CCK and other satiety signals allows larger meals to be consumed.11,15,16 Additional peptides, known as the anorectic peptides, including peptide YY (PYY), pancreatic polypeptide (PP), glucagon-like peptide 1 (GLP-1), and oxyntomodulin also have been shown to decrease appetite and promote satiety in animal and human models.17 Apo A-IV is a glycoprotein synthesized by the enterocytes of human intestine and the hypothalamus, especially the arcuate nucleus. Intestinal apo A-IV synthesis is markedly stimulated by fat absorption and does not appear to be mediated by the uptake or re-esterification of fatty acids to form triglycerides. The local formation of chylomicrons acts as a signal for the induction of intestinal apo A-IV synthesis. Intestinal apo A-IV synthesis is also enhanced by a factor from the ileum (probably peptide tyrosine-tyrosine [PYY]), as well as neuropeptide Y (NPY) and pancreatic polypeptide (PP).18 Inhibition of food intake by apo A-IV is mediated centrally. The stimulation of intestinal synthesis and secretion of apo A-IV by lipid absorption are rapid; thus apo A-IV plays a role in the short-term regulation of food intake. There also is evidence suggesting that apo A-IV may be involved in the long-term regulation of food intake and body weight, because it is influenced by leptin and insulin. Chronic ingestion of a high-fat diet blunts the intestinal and the hypothalamic apo A-IV response to lipid feeding.19 Hypothalamic apo A-IV level is reduced by food deprivation and restored by lipid feeding.20,21 Leptin, a hormone released from fat cells, is an important peripheral signal from fat stores that modulates food intake by acting on receptors in the arcuate nucleus and hypothalamus.22 Leptin deficiency and leptin receptor defects produce massive obesity. Only one gastrointestinal signal, ghrelin, has been shown to increase appetite.11 The major digestive processes are initiated in the duodenum. Delivery of chyme from the stomach is delicately

Particle size

Acid Osmolarity Fatty acids

Calorie content Peptide YY

Nutrient

GLP-1 GLP-2

Ileal brake

Figure 100-1.  Some factors that delay gastric emptying. Receptors for acid, osmolarity, fatty acids, and other nutrients in the duodenum signal gastric delay via neurohumoral mechanisms. Food particles larger than 2 mm in diameter (large circles) are rejected by the antrum. Nutrients in the ileum and colon also influence gastric emptying by the ileal brake mechanism (see Chapter 48). GLP, glucagon-like peptide; peptide YY, peptide tyrosine-tyrosine.

adjusted so that it enters the duodenum at a controlled rate, thus allowing efficient mixing with pancreaticobiliary secretions. Control of gastric emptying is thus critical to ensuring optimal digestion. The characteristics of gastric contents that determine the rate at which the stomach empties include their consistency, pH, osmolality, and lipid and calorie content (Fig. 100-1).23 The pylorus is selective in that it allows rapid passage of liquids while retaining solid particles with diameters of 2 mm or larger.24 Thus, large particles are retained and progressively reduced in size by the gastric mill, a process referred to as trituration. Trituration ensures that particles will be small enough to allow them reasonably close apposition to digestive enzymes once the nutrient is allowed to enter the duodenum. Meals of high viscosity empty more slowly than do those of low viscosity. Duodenal mucosal receptors for pH and osmolality trigger a delay in gastric emptying when the gastric effluent is acidic or hyper- or hypotonic.25,26 When duodenal luminal contents are neutralized by pancreaticobiliary bicarbonate

Chapter 100  Digestion and Absorption of Nutrients and Vitamins and osmolality is adjusted by water fluxes, gastric emptying is encouraged once more. This careful titration in the duodenal lumen ensures that nutrients are presented optimally to the pancreatic enzymes, which function best at neutral pH. The total calorie content of meals also controls gastric emptying rates; on average, the human stomach delivers about 150 kcal/hr to the duodenum.27 An increase in the size or energy density of a meal leads to a corresponding increase in the rate of delivery. Receptors for fatty acids, amino acids, and carbohydrates in the duodenal mucosa are involved in this response, which probably is mediated by both neural and humoral feedback mechanisms.28 Gastric emptying additionally is controlled by a mechanism involving the ileum and colon. If much nutrient escapes digestion and absorption in the jejunum, its presence in the ileum and colon delays gastrointestinal transit, and this again provides more time for digestion and absorption.29,30 This ileal brake probably is mediated by a neurohumoral mechanism, for which various neuro­ transmitters and hormones have been implicated including peptide YY and the glucagon-like peptides-1 and -2 (GLP-1 and GLP-2).31,32 The GLPs are synthesized in and cosecreted from enteroendocrine cells in the small and large intestine in response to luminal carbohydrate and fat. GLP-1 promotes efficient nutrient assimilation by decreasing appetite, slowing gastric empting and enhancing glucose-induced insulin secretion. GLP-2 is cosecreted with GLP-1and also regulates energy absorption by its effects on nutrient intake, gastric acid secretion, gastric emptying, and nutrient absorption. Circulating levels of GLP-1 and GLP-2 are low in the fasted state and increase rapidly following ingestion of nutrients.33-37 The gallbladder is stimulated to contract and the pancreas to secrete simultaneously in response to the presence of nutrients in the duodenal lumen. A range of nutrient receptors stimulates the release of CCK and secretin from mucosal endocrine cells into the portal circulation, and these are largely responsible for this response. Exocrine pancreatic secretion is primarily controlled by cephalic mechanisms (the vagus nerve), gastric mechanisms (acid and pepsin secretion, and nutrients delivered into the duodenum by gastric emptying), and intestinal mechanisms (secretin and CCK) (see Chapter 56). CCK and other enterohormones stimulate the pancreas by excitation of sensory nerves and by triggering of long vagovagal or enteropancreatic reflexes. Numerous neurotransmitters, such as acetylcholine and nitric oxide, and certain neuropeptides, such as gastrin-releasing peptide (GRP), generated by neurons of the enteric nervous system, have been implicated in the regulation of the exocrine pancreas. CCK appears to act via vagal cholinergic pathways to mediate pancreatic enzyme secretion. Human pancreatic acini lack functional CCK-A receptors, explaining why a CCK infusion that produces plasma CCK levels similar to those seen postprandially stimulates pancreatic exocrine secretion by an atropine-sensitive pathway.38 Under physiologic conditions, cholinergic vagal afferent pathways rather than pancreatic acinar cells represent the primary targets on which CCK can act as a major mediator of postprandial pancreatic secretion.38 Serotonin (5-hydroxytryptamine, 5-HT) released from enterochromaffin cells in the intestinal mucosa and nerve terminals of the enteric nervous system and the intrapancreatic nerves may be involved in both stimulation and inhibitory mechanisms through its various receptor subtypes; 5-HT also mediates the actions of secretin and CCK.

A synergistic interaction between CCK and 5-HT at the level of the nodose ganglia might explain the robust postprandial pancreatic secretion despite a modest postprandial increase in plasma CCK. Peptides affecting appetite and originating from the intestine (e.g., leptin and ghrelin) or from the pancreas (e.g., PP and NPY) appear to modulate the exocrine pancreas via hypothalamic centers.39,40 Pancreatic juice provides both positive and negative feedback regulation of pancreatic secretion through mediation of both secretin- and CCK-releasing peptides. Three CCKreleasing peptides have been purified: monitor peptide from pancreatic juice, diazepam-binding inhibitor from porcine intestine, and luminal CCK-releasing factor from rat intestinal secretion. All have been shown to stimulate CCK release and pancreatic enzyme secretion. Pancreatic phospholipase A2 from pancreatic juice and intestinal secretions appears to function as a secretin-releasing peptide.41 The simultaneous release of bile salts, pancreatic enzymes, and bicarbonate provides optimal conditions for further nutrient digestion. The simultaneous release of entero­ peptidase (enterokinase) from duodenal mucosa is critically important to the activation of pancreatic proteolytic enzymes. Enterokinase releases trypsin from trypsinogen, thus encouraging proteolysis within the duodenal lumen rather than inside the pancreatic duct. These three factors— bile, pancreatic enzymes, and enteropeptidase—remain separate until they are mixed in the intestinal lumen, which ensures that they become operative only at the site of nutrient delivery. Adequate lipid digestion is critically dependent on the presence of bile salts and pancreatic lipase and colipase at nearly neutral pH,42 whereas digestion of carbohydrate and protein depends on the combined actions of intraluminal secreted enzymes and then enzymes sited on the brush border membrane and within the intestinal mucosa. The close physical relationship, at the brush border, between the sites for terminal digestion of protein and carbohydrate and the active absorption of digestive products provides a very efficient mechanism for dealing with these nutrients. Two other simultaneous phenomena encourage efficient digestion and absorption. Ingestion of a meal stimulates salt and water secretion by the jejunal mucosa, and this maintains luminal contents in a sufficiently fluid state for proper mixing and digestion (see Chapter 99).43 The other phenomenon is the motor response of the intestine. After feeding, the characteristic repetitive pattern of motility that occurs during fasting is disrupted. Instead, an apparently disordered pattern is seen, which, presumably, ensures that nutrients are well mixed and brought into close contact with intestinal mucosa (see Chapter 97). There is close integration of the neurohumoral control mechanisms involving the motor and secretory responses of the intestine.44 For rapidly absorbed molecules, intestinal blood flow may be the ratelimiting step.45 Efficient conservation and recycling mechanisms ensure that gastrointestinal secretions are not entirely lost. Gastric acid secretion is balanced to a large extent by pancreaticobiliary bicarbonate secretion, so that overall acid-base balance is not disturbed. Although intact digestive enzymes are reabsorbed only in trace amounts, the nitrogen they contain is reabsorbed after their digestion. Finally, efficient enterohepatic circulation recycles bile salts several times each day so that they may be used approximately twice for each meal.46 Although bile salts are passively reabsorbed throughout the small intestine, most reach the terminal ileum, where they are reabsorbed via specific active absorptive mechanisms. Thus, bile salts remain in the lumen where they are needed for lipid digestion, but they are

1697

1698

Section X  Small and Large Intestine largely reabsorbed at the last moment to avoid being lost by the colon (Fig. 100-2; see Chapter 64). Once intestinal chyme leaves the ileum and enters the colon, most nutrients have been digested and absorbed and colonic function largely serves to dehydrate luminal contents through absorption of salt and water and to store the

residuum. Dietary fiber may be digested by bacteria, with release of short-chain fatty acids, which are avidly absorbed; however, short-chain fatty acids do not usually have much nutritional significance, other than in areas of the world where the major source of energy is a high-fiber diet. Shortchain fatty acids, however, are the major source of nutrition for the colonocytes.

FAT

Total pool size 5g

DIETARY INTAKE

In the United States, fat intakes for all age groups and both sexes rose from approximately 34% of total energy consumption in the 1930s to 40% to 42% in the late 1950s to mid-1960s, and then fell steadily to approximately 36% (90 to 100 g/day) in 1984. Saturated and monounsaturated fatty acid intakes fell from 18% to 20% of total energy consumption in the early 1950s to 12% to 13% of energy in 1984; polyunsaturated fatty acid intakes rose from 2% to 4% of energy to 7.5%.47 The majority of fatty acids present in dietary triglyceride are oleate and palmitate (18 : 1 and 16 : 0, respectively).48 In animal triglyceride, most of the fatty acids are long-chained (i.e., longer than 14 carbon chains) and saturated (Fig. 100-3). Polyunsaturated fatty acids such as linoleic and linolenic acid are derived from phospholipids of vegetable origin and, because they cannot be synthesized de novo, they are considered essential fatty acids (Table 100-1). The average range of phospholipid ingestion is between 2 and 8 g/day. The most commonly ingested phospholipid is phosphatidyl choline (lecithin), and the predominant fatty acids in phospholipid are linoleate and arachidonate (see Fig. 100-3). More phospholipid is found in the duode-

Portal vein

Figure 100-2.  Enterohepatic circulation of bile salts. Active transport in the ileum retrieves most bile salts, and the small fraction lost into the colon and eliminated in the feces is compensated for by fresh hepatic synthesis (see Chapter 64).

Glycerol

Ester bonds

Fatty acids

O 1CH 2

C

O

(CH2)n

CH3

(CH2)n

CH3

(CH2)n

CH3

O 2CH

O

C O

3CH 2

A

O

C

Triglyceride

Glycerol

Hydrophilic portion Choline (CH3)3

N

O (CH2)2

O

P

O

Ester bonds O

Fatty acids

CH2

O

C O

(CH2)n

CH3

CH

O

C

(CH2)n

CH3

CH2 Hydrophobic portion

O

B

Phospholipid

Figure 100-3.  General molecular structure of triglyceride (A) and phospholipid (phosphatidyl choline or lecithin) (B).

Chapter 100  Digestion and Absorption of Nutrients and Vitamins nal lumen (10 to 22 g/day) than is ingested, most of which is derived from endogenous sources, particularly bile. Cholesterol intake varies widely but averages about 200 to 250 mg/day.48 Some people consume as much as 500 mg/day. Commercial hydrogenation of unsaturated bonds in the fatty acids of natural oils raises their melting points, thus allowing production of margarines and spreads of variable consistency. Hydrogenation, in addition to saturation, results in isomerization of cis to trans double bonds.49 Although many commercial products contain partially hydrogenated fats, the content of trans-fatty acids in some margarines exceeds 60%, thus raising concerns about their relationship to cancer induction.50

DIGESTION AND ABSORPTION

Most dietary lipid is absorbed by the upper two thirds of jejunum, although its rate and extent of absorption are influenced by the presence of other foods, particularly dietary fiber, which reduces the rate of absorption.51 The types of ingested fat also appear to influence the absorptive process, both by modifying the morphologic structure of the intestinal mucosa and by influencing its absorptive function for other nutrients such as carbohydrate.52 The insolubility of fat in water dominates the mechanisms that have evolved to digest and absorb lipid. Within the lumen, ingested fat has to be physically released and broken down into emulsion droplets. Following digestion, the products have to be transported across the bulk (lumen) water phase to the lipid epithelial cell membrane. Transfer across the lipid membrane is followed, within the epithelium, by reconstitution into larger lipid molecules, pre­ dominantly triglyceride, which then require specialized processing to permit export from the cell. Thus, to be used

Table 100-1  Common Dietary Fatty Acids FATTY ACID

CONFIGURATION*

Saturated Fatty Acids Butyric Caproic Lauric Myristic Palmitic Stearic Mono-unsaturated Fatty Acids Oleic Palmitoleic Polyunsaturated Fatty Acids Arachidonic Linoleic Linolenic

4 : 0 6 : 0 12 : 0 14 : 0 16 : 0 18 : 0 18 : 1 16 : 1 20 : 4 18 : 2 18 : 3

*By convention, the number of carbon atoms in the chain is given by the first figure and the number of double bonds in the chain is given by the second.

after ingestion, lipid must pass through three physical phases: water in the lumen, lipid in the epithelial membrane, and water in the lymphatics and bloodstream. Despite these potential barriers, more than 95% of ingested fat is absorbed by adults.

Triglyceride

Liberation of fatty acids from the glycerol backbone of triglycerides (lipolysis) is achieved by lipases acting at the surface of emulsified droplets (Table 100-2). This process occurs initially in the stomach, but most lipolysis is accomplished in the small intestine.53 Intragastric lipolysis might account for 20% to 30% of total intraluminal lipid digestion.54 Gastric lipase, which is of fundic origin, has been demonstrated in the gastric contents of premature neonates and in mucosal biopsy specimens from adults up to 80 years of age. Gastric lipase does not hydrolyze phospholipids or cholesterol esters. Human gastric lipase is a 379-amino acid protein that shares similar homology with rat gastric lipase but not human pancreatic lipase. For either gastric or small intestinal lipolysis to occur, two conditions are critical: First, a stable emulsion is required of fat droplets of such a size that they present a large surface area to the digestive enzyme; second, a mechanism is required for bringing enzyme and triglyceride into close apposition within the emulsion. Emulsification A number of factors assist in optimal production of an emulsion. Physical release of fat by mastication and the gastric milling of food produces a relatively unstable emulsion that is delivered into the duodenum. To permit its stabilization, the droplets in this emulsion have to be coated, and dietary phospholipid provides one such coat. The ratio of ingested phospholipid to triglyceride is about 1 : 30, and more phospholipid is added in the duodenum from bile.55 In breast milk, emulsion droplets are smaller and have proteins as well as phospholipid incorporated into their surface trilayer.56 Emulsification also is enhanced by the fatty acids liberated by intragastric lipolysis and, within the duodenum, by bile salts (Fig. 100-4). The final product in the duodenum is an emulsion consisting predominantly of triglyceride together with cholesterol esters and some diglyceride and coated by phospholipid, partially ionized fatty acids, monoglyceride, and bile salts. Lipase This stable emulsion is then presented to pancreatic triglyceride lipase. Unlike other soluble enzymes, which can act in a three-dimensional solution, lipase has to act at the two-dimensional surface of the emulsion droplet, and this requirement poses particular problems.57 Certain characteristics of the enzyme itself are important. Thus, the lipolytic zone of the molecule is hydrophobic and lies deep within it, shielded from the aqueous phase. It is revealed to the lipid only on close apposition to its surface. The presence of a coat on the lipid droplet thus poses a barrier to the

Table 100-2  Characteristics of Lipase Activity in Infancy ENZYME

OPTIMAL pH

site of lipase activity

OTHER

Milk-derived lipase Gastric lipase Pancreatic lipase

7.0 (inactivation by acid is reversible) 4.0-6.0 7.0

a-1, -2, and -3 ester bonds a-1 ester bond a-1 and -3 ester bonds

Stimulated by bile salts Inhibited by pancreatic proteolysis —

1699

1700

Section X  Small and Large Intestine Panel B

Product vesicles

Stomach

Emulsification Increased stability of lipid emulsion

TG PL

TG (Oil droplet)

Gastric lipolysis

Mixed micelles

TG PL DG Duodenal lipolysis

MG FA

Duodenum

Aqueous phase MG

Bile salts Lecithin

Lipase

PL

Colipase

BS Panel A

MG Lipid phase

TG

BS

DG Lipid phase

Aqueous phase Panel C FA 1 2 3

+ TG

MG

FA

–

Long section

Cross section

Lipase + HCO3

Figure 100-4.  Steps in lipolysis. The initial step in lipolysis is to increase the stability of the fatty emulsion. Gastric lipase acts on triglycerides to yield fatty acids and diglyceride (diglyceride enhances emulsification). This step is enhanced in the duodenum by bile salts and phospholipid (lecithin), which enable lipase, in the presence of colipase, to act at the surface of the emulsion droplet to bring it close to the triglyceride molecule, whereupon monoglyceride and fatty acids are released. Lipolysis in the duodenum yields fatty acids (from the α1 and α3 positions) and monoglyceride and occurs in a rapid and efficient manner at nearly neutral pH. In panel A are diagrammatic representations of bile salt molecules (top) oriented at an oil-water interface with its hydrophobic sterolic backbone in the oil phase and its hydrophylic hydroxyl and either taurine or glycine conjugates in the aqueous phase. At concentrations above critical micellar concentration, bile salts aggregate as simple micelles in water, with their hydrophylic groups facing into the water. In this diagram, three hydroxyl groups (cholate) are shown as open circles and an additional polar group represents either taurine or glycine. Panel B is a diagrammatic sketch of the dispersion of the products of lipolysis into lamellae at the surface of the oil phase, each about 4 to 5 nm thick, with water spacings up to 8 nm, and from there into vesicles of about 20 to 130 nm in diameter. In panel C, fatty acids and monoglyceride within the vesicles pass into mixed micelles. BS, bile salt; DG, diglyceride; FA, fatty acid; MG, monoglyceride; PL, phospholipid; TG, triglyceride.

action of lipase, and assistance is required to bring it into close contact with the triglyceride. The presence of colipase, cosecreted by the pancreas along with lipase in a molar ratio of 1 : 1, is critical in approximating lipase to triglyceride (see Fig. 100-4). Colipase attaches to the ester bond region of the triglyceride, lipase then binding strongly to colipase by electrostatic interactions.55 Phospholipase A2 digestion of the phospholipid on the surface of the lipid emulsion allows exposure of the triglyceride core to the colipase-lipase complex, further enhancing colipase-dependent anchoring of lipase to the lipid emulsion. Phospholipase A2 digestion requires

bile salts and Ca2+ for activation, which can further assist colipase-lipase–mediated triglyceride lipolysis by providing a mechanism for removing lipolytic products. In the absence of colipase, bile salts on the surface of the emulsion droplet inhibit lipase activity. The colipase gene is located on chromosome 6, and the amino acid sequence of the lipid-binding domain, the lipase-binding domain, and the activation peptide appear to be highly conserved.58 Colipase is secreted by the pancreas as pro-colipase,59 which is activated when trypsin cleaves a pentapeptide from its N-terminus after entering the small intestinal lumen. A valine residue at position 407 and a

Chapter 100  Digestion and Absorption of Nutrients and Vitamins leucine at position 412 are important for the interaction of lipase with colipase and the bile salt micelles.60 Interestingly, the pentapeptide cleaved from the pro-colipase by trypsin, called enterostatin, seems to be a specific satiety signal for the ingestion of fat.61 Because pancreatic lipase is most active at nearly neutral pH, secretion of bicarbonate by the pancreas and biliary tree is critically important and provides the necessary neutralization of gastric acid; however, luminal pH falls to about 6 in the jejunum, and here the fact that bile salts lower the pH optimal for lipase activity from 8 to 6 may be significant. In the presence of colipase and optimal pH, lipase activity releases fatty acids and monoglyceride extremely rapidly and efficiently (see Fig. 100-4). Pancreatic triglyceride lipase also binds strongly to the mucosal brush border membrane,62 where it may participate in lipolysis of cholesteryl esters or triglyceride, releasing fatty acids, monoglyceride, and free cholesterol in proximity to the brush border membrane, where they undergo rapid uptake. In addition to pancreatic triglyceride lipase and its protein cofactor, colipase, pancreatic acini also synthesize two pancreatic lipase-related proteins (PLRP-1 and PLRP-2), which have strong nucleotide and amino acid sequence homology to pancreatic triglyceride lipase. Although PLRP-1 has no known activity, PLRP-2 does have lipase activity and, like pancreatic triglyceride lipase, PLRP-2 cleaves triglycerides but with broader substrate specificity. PLRP-2 also hydrolyzes phospholipids and galactolipids, two fats that are not substrates for pancreatic triglyceride lipase. It is also different from pancreatic triglyceride lipase with respect to sensitivity to bile salts and in response to colipase. A further critical difference is that PLRP-2 mRNA appears before birth and persists into adulthood, whereas pancreatic triglyceride lipase mRNA first appears at the suckling-to-weaning transition. This suggests that PLRP-2 plays a critical role in the digestion of breast-milk fats.63,64 Micelles and Other Lipid-Containing Particles The products of lipolysis are distributed among the aqueous, oil, and intermediate phases in a number of forms prepared for transfer across the lumen to the mucosal brush border membrane. The shuttling of these products depends, in part, on the formation of micelles with bile salts. The concentration of bile salts secreted in bile is about 35 mmol and, in the duodenum, this is further decreased by dilution to 10 to 20 mmol; this concentration lies well above the critical concentration for micelle formation. Mixed micelle production depends on a number of other factors, including pH, presence or absence of lipids, and the types of bile salts that are secreted (see Chapter 64).42 Bile salts are capable of forming micelles because they have a particular three-dimensional structure and they are amphipathic; that is, their molecules have both watersoluble and lipid-soluble portions (see Fig. 100-4). They orient themselves at an oil-and-water interface and thus are ideal emulsifying agents. In addition, micelles are formed when bile salt levels are present above critical concentrations and thus are able to aggregate in disk-like particles with their hydrophobic sterolic backbones oriented toward each other and their hydrophilic polar groups facing outward into the aqueous phase. Bile salt micelles have the capacity to dissolve fatty acids, monoglycerides, and cholesterol, but not triglyceride.65 The mixed micelles thus formed are arranged so that the insoluble lipid is surrounded by bile salts that are oriented with their hydrophilic groups facing outward. Mixed micelles are about 50 to 80 nm in diameter and, unlike emulsion droplets, are too small to

scatter light; thus, micellar solutions are clear. The presence of phospholipid secreted in bile enlarges mixed micelles and makes them more efficient in the dissolution of fat. Other lipid-containing particles participate in the transfer of lipid to the mucosa. As the emulsion droplet shrinks during lipolysis, liquid crystalline structures are formed at its surface.66,67 These vesicular structures with multilamellar and unilamellar forms can be seen under the electron microscope, budding off the surface of emulsion droplets and occasionally close to the brush border membrane of the intestinal mucosa.68 This physical phase of lipid within the lumen might provide a significant mechanism for transfer of lipid to the mucosa, beyond that provided by bile salt micelles, and it could explain the observation that in the absence of bile salts, some 50% or more of dietary triglyceride may be absorbed. In the presence of adequate concentrations of bile salts, however, these vesicles undergo rapid spontaneous dissolution and release their lipid into micelles, which are likely to be the major route for lipid traffic (see Fig. 100-4); numerically, they are much more common than lipid vesicles. Importance of Intraluminal pH Lipid digestion and absorption are highly dependent on intraluminal pH at several steps in the process. Pancreatic lipase operates best in the presence of bile salts and at least pH 6. It therefore functions well at the pH of the luminal duodenum, where most lipid digestion occurs. Glycineconjugated bile salts precipitate below pH 5; fatty acids are in their protonated form below about pH 6 and have limited solubility in bile salt micelles. Thus, in conditions in which intraluminal pH becomes more acid, as for example in the Zollinger-Ellison syndrome, pancreatic lipase is inactive, bile acids precipitate out of solution, and fatty acid partitioning is reduced. It is not surprising, therefore, that steatorrhea (without any other nutrient or hematologic disturbances) is a feature of this syndrome. Biological characteristics of lipases, including effect of pH on activities, are detailed in Table 100-2. Unstirred Water Layer An unstirred water layer is present on the surface of the intestinal epithelium, which in humans is approximately 40  µm deep.69 This layer may be rate limiting for uptake of long-chain fatty acids but not for short- or mediumchain fatty acids, the limiting step for which occurs at the brush border membrane.55 The provision of a high concentration of fatty acid in the microenvironment adjacent to the epithelium depends on the diffusion of micelles into this region. The microclimate here is slightly acidic, owing to activity of a sodium-hydrogen (Na+/H+) exchanger at the brush border membrane, and at pH between 5 and 6, the solubility of fatty acids in micelles decreases, thus encouraging liberation of fatty acids close to the mucosa. The high concentration of fatty acids necessary for diffusion across the mucosal membrane is thus achieved; evidence for this model is increasingly persuasive.70 The low-microclimate pH also encourages the fatty acids to be presented in an undissociated, protonated form. Thus, the pH partition hypothesis predicts that fatty acids could diffuse passively into the cell as protonated species and, at the near-neutral intracellular pH, become trapped in the ionized form. A surfactant-like material has been discovered close to the brush border membrane, although its role in absorption, if any, is uncertain.71 It is secreted by enterocytes, contains phosphatidylcholine and alkaline phosphatase, and appears

1701

1702

Section X  Small and Large Intestine as flat lamellae or vesicles adjacent to the brush border membrane.

Other Lipids

Phosphatidylcholine, the major dietary phospholipid, is hydrolyzed by pancreatic phospholipase A2 (PLA2) to yield fatty acid from the 2-position and lysophosphatidylcholine. Pancreatic PLA2 is secreted as an anionic zymogen that is activated in the small intestine by tryptic cleavage of an N-terminal heptapeptide. It has a molecular weight of approximately 14 kd and requires calcium for activation and bile salts for its activity. It has multiple isoforms and apparently requires a 2 : 1 bile salt-to-phosphatidylcholine molar ratio for optimal activity. Although the bulk of intestinal PLA2 activity is derived from pancreatic juice, there is some contribution from the intestinal mucosa, where the enzyme is concentrated in the brush border.72 Cholesterol esters, in the presence of bile salts and calcium, are hydrolyzed by carboxyl ester lipase (CEL) (also known as pancreatic cholesterol esterase) to release the free sterol, in which form it is absorbed. Cell culture and other in vitro studies have thoroughly defined the potential functions of CEL in the digestion of cholesteryl ester, phospholipids, and triglycerides, but only its cholesteryl ester hydrolytic activity is unique to this enzyme in the digestive tract.73 CEL belongs to the α/β hydrolase family, is well conserved, and shares 78% homology in rats and humans.74 It is secreted primarily by the pancreatic acinar cells and lactating mammary glands. Using site-directed mutagenesis, the serine at position 194, the histidine at position 435, and the aspartic acid at position 320 have been shown to be important for CEL’s catalytic activity.73,75-77 The hydrolysis of water-insoluble substrates by CEL requires bile salt–containing 3α- or 7-α-hydroxy groups (e.g., cholate or chenodeoxycholate and their conjugates).78 The arginine-63 and arginine-423 sites play an important role in this bile salt–dependent process, but not in the bile salt– independent lysophospholipid hydrolytic activity of CEL.79 Both hydrolytic enzymes act on the emulsion phase at the surface of droplets, and the products of digestion are released into multilamellar and unilamellar vesicles and then to mixed micelles. Fatty acids and monoglycerides increase the solubility of cholesterol in micelles, thus encouraging its absorption. The products of phospholipid and cholesterol hydrolysis thus follow the same route to the brush border membrane as the fatty acids and monoglyceride, which originate from dietary triglyceride. Unabsorbed long-chain fatty acids that enter the colon are not absorbed by this organ, and they undergo a series of bacterial modifications, principally hydroxylation. In healthy persons, no undigested triglyceride is found in the stool, and the normal fecal fat estimate of approximately 7 g/day reflects the cumulative total excretion of saponification products (i.e., fatty acids) that arise principally from membrane phospholipid and bacteria.

Transfer across the Brush Border Membrane

Much of the current understanding of the micellar solubilization and uptake of dietary lipids comes from the work of Hofmann and Borgstrom, who described the uptake of lipid digestion products by enterocytes.80 Further work by Carey discovered the coexistence of unilamellar liposomes with bile salt–lipid mixed micelles in the small intestine.81 Although the uptake of lipid digestion products by enterocytes has been accepted as a passive process, recent work has raised the possibility that some lipids may be taken up by enterocytes via carrier-mediated processes that are energy dependent.82

Studies with brush border membrane vesicles suggest that linoleic acid uptake occurs by facilitated diffusion.83 Absorption of oleic and arachidonic acid also appears to occur by a saturable process, suggesting the possibility of active transport. Several membrane proteins that increase the uptake of long chain fatty acids when overexpressed in cultured mammalian cells have been identified, the most prominent and best characterized of which are FAT/CD36, long-chain fatty acyl-CoA synthetases (LACS), and fatty acid transport proteins (FATPs/solute carrier family 27).84-87 The FATPs are transmembrane proteins that have been shown to enhance the cellular uptake of long-chain and very-long-chain fatty acids. In humans, FATPs comprise a family of six highly homologous proteins, hsFATP 1-6, which are found in all tissues of the body that use fatty acids.84,88,89 Although hsFATP1 is the best characterized of the FATPs, hsFATP4 is the only FATP expressed in the small intestine; it is localized to the apical brush border of the epithelial cells, where it is responsible for absorption of dietary lipids. Studies with cell lines and isolated enterocytes that overexpress FATP4 demonstrated that FATP4 is both necessary and sufficient for efficient uptake of longchain and very-long-chain fatty acids.90 Detailed substrate studies based on 14C-labeled fatty acids have been presented for FATP1 and FATP4.87,90 Both studies showed that uptake of fatty acids shorter than 10 carbon atoms, such as butyric acid and octanoic acid, was unaffected by FATP expression, whereas uptake of common long-chain fatty acids, such as palmitate and oleate, was robustly enhanced.90 More recently, a wrinkle-free phenotype has been associated with the spontaneous autosomal recessive mutation of the gene for FATP4, resulting in a very tight and thickskinned phenotype.91 Nutrients, hormones, and cytokines have been reported to regulate FATP expression. Rats fed a high-fat diet showed increased FATP expression in the heart, but not the liver. Several reports have shown a positive regulation of mouse FATP by ligands that activate either PPAR-γ, PPAR-α, or PPAR-γ/RXR heterodimers in hepatoma cell lines, the liver, and the intestine. Further, a PPAR binding site was identified in the murine FATP1 promoter. TNF-α is a negative regulator of FATP expression and down-regulates FATP mRNA in liver and FATP1 and FATP4 proteins in adipocytes.88 The exact mechanism of FATP transport of long-chain fatty acids into the intestinal cell is unknown. It has been postulated that extracellular long-chain fatty acids might directly bind to FATP complexes and be transported into cells. Alternatively, long-chain fatty acids could bind first to CD36, which hands off the long-chain fatty acids to FATP dimers. Intracellular long-chain fatty acids are coupled to CoA by long-chain acyl-CoA synthetase (LACS), preventing their efflux, and fatty acid–binding proteins (FABPs) act as a cytoplasmic buffer for incorporated long-chain fatty acids (Fig. 100-5).88 Cholesterol, unlike β-sitosterol (plant sterol), is well absorbed by the proximal jejunum,92 although both are present in the human diet. The second-order kinetics of cholesterol absorption, its sterol specificity, and its inhibition by drugs such as ezetimibe all suggest that cholesterol absorption is mediated by specific transport proteins at the brush border membrane.93,94 Thurnhofer and colleagues first described the presence of a possible binding protein in the small intestinal brush border that facilitates the uptake of cholesterol by the small intestine95; however, this 14K protein was later identified as sterol carrier protein-2 (SCP2), which is an intracellular protein.96 The adenosine triphosphate (ATP)-binding cassette (ABC) A1 transporter and

Chapter 100  Digestion and Absorption of Nutrients and Vitamins LCFA

FATP CD36

Acyl CoA LACS

they are coregulated by the nuclear hormone receptor’s liver X receptor (LXR)105; however, these transporters provide the apparatus for efficient shunting of sterols away from the transfer pathway directing the production of cholesteryl esters by acyl-CoA cholesterol acyl transferase 2 (ACAT2) and do not appear to be involved in the initial uptake of cholesterol. Niemann-Pick C1-like 1 (NPC1L1) protein has been suggested as a key component of intestinal cholesterol transport. Niemann-Pick C1 (NPC1) is the defective gene in the cholesterol storage disease Niemann-Pick type C (NP-C); NPC1 protein is highly abundant in a variety of tissues, and functions in intracellular cholesterol trafficking.106,107 In contrast, NPC1L1 protein has 50% amino acid homology with NPC1 protein, has several predicted features of a plasma membrane–expressed transporter, and is expressed in high abundance in the small intestine, especially in the brush border membrane of enterocytes. NPC1L1 protein has been suggested as a target for the cholesterol absorption– reducing drug ezetimibe.108 NPC1L1-deficient mice show a substantial reduction in absorbed cholesterol that is unaffected by dietary supplementation of bile acids.109

Intracellular Processing and Absortion

ACBP

FABP Figure 100-5.  Mechanisms of long-chain fatty acid (LCFA) and verylong-chain fatty acid transport. Extracellular LCFA may directly bind to a fatty acid transport protein (FATP) dimer complex and be transported into the cell. Alternatively, LCFA may bind first to CD36, which transfers the LCFA to a FATP dimer. Intracellular LCFA is coupled to coenzyme A (CoA) by long-chain fatty acyl-CoA synthetase (LACS), preventing their efflux, and fatty acid binding protein (FABP) acts as a cytoplasmic buffer for LCFA. ACBP, acyl CoA binding protein. (Adapted, with permission, from Stahl A. A current review of fatty acid transport proteins (SLC27). Pflugers Arch 2004; 447:722-7.)

scavenger receptor type B1 (SR-B1) also were postulated to play a role in cholesterol absorption. Evidence supporting the role of scavenger receptors in cholesterol uptake includes decreased cholesterol uptake by use of scavenger receptor inhibitors such as ezetimibe,97,98 but targeted inactivation of these genes in mice had no effect on cholesterol uptake.99-101 CD36 also has been demonstrated to be important in promoting the uptake of cholesterol by the small intestine and to play an important role in packaging absorbed fatty acids as triglycerides.102 Evidence favoring a cholesterol membrane transporter also is seen in individuals with β-sitosterolemia, a condition in which the intestine fails to discriminate between cholesterol and β-sitosterol. Major findings in patients homozygous for sitosterolemia include xanthomatosis and accelerated, often fatal, premature atherosclerosis. The genetic defect of β-sitosterolemia is linked to chromosome 2p21.103 Seven different mutations in two adjacent genes have been described that are responsible for encoding new members of the ABC transporter family (ABCG5 and ABCG8) in patients with sitosterolemia. These defects cause an increased intestinal absorption and decreased biliary excretion of all sterols (plant sterols and cholesterol), leading to a 50- to 200-fold increase in plasma plant sterol concentrations. Feeding cholesterol to mice up-regulated these genes, suggesting that ABCG5 and ABCG8 work together to limit intestinal cholesterol absorption by cholesterol efflux from small intestinal epithelial cells.104 These two genes are expressed almost exclusively in the liver and intestine, and

Once within the cell, fatty acids bind to specific FABPs, which are found predominantly in the jejunum and more in villus cells than in crypt cells. The small intestine has three distinct proteins belonging to the intracellular lipidbinding protein family: the liver-type FABP (L-FABP), the intestinal FABP (I-FABP), and the ileal lipid-binding protein (ILBP)110,111; all have greater affinity for unsaturated fatty acids than for saturated ones and very little affinity, if any, for short-chain or medium-chain fatty acids.42 Based on nuclear magnetic resonance (NMR) binding studies, it has been suggested that the binding of I-FABP is involved in the intracellular transport of fatty acids, whereas the L-FABP is involved in the intracellular transport of monoglycerides and lysophosphatidylcholine.112 These binding proteins can assist transfer across the cytoplasm to the endoplasmic reticulum for triglyceride resynthesis as well as modulating intracellular lipid metabolism and regulating gene expression. In addition, two sterol carrier proteins, SCP-1 and SCP-2, have been isolated and characterized. SCP-1 is important in the microsomal conversion of squalene to lanosterol,113 whereas SCP-2 participates in the microsomal conversion of lanosterol to cholesterol as well as the intracellular transport of cholesterol from cytoplasmic lipid droplets to mitochondria.114 In the endoplasmic reticulum, during feeding, triglyceride is resynthesized by two processes (Fig. 100-6).115 In the first, monoglyceride is re-esterified with absorbed fatty acid after it has been activated to form acyl coenzyme A (CoA) (the monoglyceride pathway). Microsomal acyl CoA-ligase is necessary to synthesize acyl CoA from the fatty acid before esterification. Diglyceride and then triglyceride are formed sequentially in reactions that favor long-chain fatty acid absorption from the lumen. This route, involving monoglyceride esterification, accounts for the majority of the triglyceride synthesized during the absorptive phase, no more than 4% being formed by acylation of absorbed glycerol. It is thought that the synthesis of triglyceride from diglyceride is catalyzed by the enzyme acyl CoA:diglyceride acyl transferase.116 The gene for this enzyme has been isolated, although a knockout mouse model of this gene still synthesized triglyceride in the intestinal mucosa, suggesting that there may be yet another enzyme involved in forming triglyceride from diglyceride.117,118

1703

1704

Section X  Small and Large Intestine Glucose

Fasting

α-Glycerophosphate Phosphatidic acid

ATP Fatty acid

Fatty acid + CoA

Lipoproteins Triglyceride

Acyl CoA Diglyceride

Monoglyceride

Phospholipid

Triglyceride

Monoglyceride Feeding

Figure 100-6.  Metabolic fate of absorbed fatty acid and monoglyceride in enterocytes. During feeding, triglyceride is resynthesized largely from absorbed fatty acid and monoglyceride. During fasting, triglyceride and phospholipid are synthesized from α-glycerophosphate derived from glucose entering across the basolateral membrane of the enterocyte, and from fatty acids. Unsaturated fatty acids tend to form the phospholipid. ATP, adenosine triphosphate; CoA, coenzyme A.

During fasting, triglyceride (and phospholipid) is synthesized via the second route, which involves acylation of α-glycerophosphate with the formation of phosphatidic acid and, then, triglyceride or phospholipid (see Fig. 100-6). The α-glycerophosphate is synthesized largely in the cytoplasm, from glucose. The relative importance of the monoglyceride-pathway and the α-glycerophosphate pathway depends on the availability of 2-monoacylglycerol and fatty acid. During normal lipid absorption, when 2-monoacylglycerol is sufficiently present, the monoglyceride pathway facilitates the conversion of 2-monoacylglycerol and fatty acid to form triglyceride and aids in inhibiting the α-glycerophosphate pathway. Conversely, when the supply of 2-monoacylglycerol is lacking or insufficient, the α-glycerophosphate pathway becomes the major pathway for forming triglyceride. Some absorbed lysophosphatidylcholine is reacylated to form phosphatidylcholine. The remaining absorbed lysophosphatidylcholine is hydrolyzed to form glycero-3phosphorylcholine. The liberated fatty acids are used for triglyceride synthesis, whereas the glycero-3-phosphorylcholine is readily transported via the portal blood for use in the liver. Absorbed dietary cholesterol enters a free cholesterol pool within enterocytes that also contains cholesterol from endogenous sites (nondietary sources such as biliary cholesterol, cholesterol derived from plasma lipoproteins, and cholesterol synthesized de novo). Cholesterol is transported mainly as esterified cholesterol and almost exclusively by the lymphatic system. Cholesterol esterase and ACAT are thought to be predominantly responsible for cholesterol esterification. ACAT is stimulated by a high-cholesterol diet and appears to play a more important role in mucosal cholesterol esterification than cholesterol esterase does.119 Two ACAT proteins have been identified: ACAT-1 and ACAT-2.120,121 The role of ACAT-2 in intestinal cholesterol absorption is supported by resistance to diet-induced hypercholesterolemia due to defective cholesterol esterification and absorption by the small intestine in the ACAT-2 knockout mouse model.122 Once synthesized, triglyceride, cholesterol and its esters, and phospholipids are packaged for export in the form of chylomicrons and very-low-density lipoproteins (VLDLs). During fasting, VLDLs are the major triglyceride-rich lipoproteins that emerge from the epithelium; after feeding,

chylomicrons predominate. VLDL triglycerides have a different fatty acid composition from those in chylomicrons, different pathways being involved in their formation. Furthermore, the fatty acids derived from dietary tri­ glyceride go predominantly into forming chylomicrons, whereas those derived from phospholipid appear to be used in forming VLDL.55 The diameter of chylomicrons ranges between 750 and 6000 nm; their cores comprise triglycerides, and cholesterol ester and phospholipid form more than 80% of the surface coat. Forming a smaller portion of the surface of chylomicrons is an essential component, apolipoprotein. Apo A is an important apoprotein for all lipoproteins, including chylomicrons, VLDLs, and high-density lipoproteins (HDLs). It is synthesized in the small intestine and is found in bile.123 Apo B probably is synthesized in the Golgi cysterni and is found in the rough endoplasmic reticulum. After feeding, Apo B is found in association with the chylomicrons in the smooth endoplasmic reticulum. The absence of apo B prevents synthesis and secretion of chylomicrons; however, data suggest that the supply of apo B is not the rate-limiting step for chylomicron formation. For example, the apo B output in lymph does not change after intraduodenal infusion of lipid, even though lymphatic triglyceride output increases seven-fold to eight-fold.124,125 Apolipoproteinemia is a rare genetic disorder resulting in complete failure of the liver and intestine to make triglyceride-rich lipoproteins.126 Previously, it had been thought that abetalipoproteinemic patients have a problem synthesizing apo B. Actually, apo B synthesis is reduced, but not abolished, suggesting that failure of the intestine and liver to synthesize apo B might not be the reason abetalipoproteinemic patients do not produce chylomicrons and VLDL,127 a fact that has been confirmed by the finding that the abetalipoproteinemia results from mutations of the microsomal triglyceride transfer protein gene.125,128 This gene’s lipid transfer activity is primarily responsible for the lipidation of the primordial particle—the initial step in chylomicron formation whereby the addition of phospholipids to the apo B molecule is followed by the addition of small amounts of triglyceride. Anderson’s disease, also known as chylomicron retention disorder, is another disorder of formation or secretion of chylomicrons by the small intestine. There is no defect in genes that carry known apoproteins or microsomal triglyc-

Chapter 100  Digestion and Absorption of Nutrients and Vitamins eride transfer protein,129 suggesting that this disease is caused by an unknown factor central to secretion of chylomicrons. Once chylomicrons have formed in the smooth endoplasmic reticulum, they are transferred to the Golgi apparatus. Golgi-derived chylomicron vesicles are then incorporated into the basolateral membrane and secreted by exocytosis into the lymphatic circulation (Fig. 100-7). During absorption, lacteals distend and endothelial cells, which overlap each other in the fasting state, move apart and open gaps through which chylomicrons can readily pass.130 Medium-chain fatty acids are absorbed by way of the portal vein, but as the chain length of saturated fatty acids increases, they are increasingly absorbed via the lymphatics. Polyunsaturated fatty acids may pass directly across the basolateral membrane and into the portal circulation.

conversion of cholesterol into bile acids so as to promote the net excretion of cholesterol. LXR-α up-regulates the transcription of CYP7A1 by directly binding to an LXRE in the promoter of this gene.136,137 The liver-specific expression of CYP7A1 requires LRH-1 (liver receptor homolog-1, also

Intestinal lumen

Lysophospholipids Phospholipids Fatty acids Cholesterol

Brush border SER

Liver X Receptors and Lipid Homeostasis

The liver X receptors (LXR-α and LXR-β) are nuclear receptor transcription factors that are activated by certain derivatives of cholesterol.131 Hence, LXR activity may be up-regulated by cellular lipid load or dietary cholesterol intake. The identification of a large list of LXR target genes and their response to LXR activation (Table 100-3) indicate that the LXRs play an important role in the response to excess cholesterol and that their activation might protect against tissue cholesterol overload.132 LXR-α and LXR-β form obligate heterodimers with the retinoid X receptor (RXR) to result in transcription factors that can be activated by ligands (i.e., lipids) for either RXR or LXR. RXR-LXR heterodimers bind to a specific DNA sequence called the LXR response element (LXRE), which consists of two hexanucleotide sequences separated by four bases (Fig. 100-8).133 A specific group of LXR agonists has been identified: intermediates in cholesterol metabolic pathways. Furthermore, using mouse models lacking LXR-α, LXR-β, or both, the key role of LXRs has been shown in regulating the expression of genes involved in cholesterol catabolism, absorption, and transport, as well as fatty acid synthesis (see Table 100-3). Cholesterol 7α-hydroxylase (CYP7A1) is the rate-limiting enzyme of the classic pathway that converts cholesterol into bile acids.134,135 The soluble bile acids, primarily produced in the liver, promote the secretion of cholesterol into bile for excretion in feces. In some species (e.g., rodents), but not humans, expression of the CYP7A1 gene is induced in response to dietary cholesterol, thereby accelerating the

RER

Golgi

Nucleus

Basement membrane Lamina propria Lymphatic vessels Figure 100-7.  Pathway taken by lipids during passage across an enterocyte. Triglyceride and phospholipid are synthesized in the smooth endoplasmic reticulum (SER) and accumulate there as dense droplets (small blue dots). Apolipoproteins, synthesized in the rough endoplasmic reticulum (RER), assist in the formation of chylomicrons and very-low-density lipoproteins in the tubular endoplasmic reticulum and Golgi apparatus (yellow circles). Chylomicrons are finally released across the basolateral membrane by exocytosis to enter the lymphatics.

Table 100-3  Liver X Receptor Gene Targets and Functions TARGET GENE

TARGET TISSUES

FUNCTION

CYP7A1 ABCA1

Liver Macrophage Intestine Liver, intestine Macrophage Macrophage, adipocyte Liver Mouse macrophage — Liver, fat Intestine Macrophage Liver

Bile acid synthesis Cholesterol efflux

ABCG5/G8 ABCG1 APOE LPL CETP SREBP-1c LXR YZ-2

Sterol transport Cholesterol efflux Component of lipoprotein particles Triglyceride hydrolysis Cholesterol ester transfer Fatty acid synthesis Cholesterol sensor Pyrimidine metabolism(?)

1705

1706

Section X  Small and Large Intestine called CPF and FTF), a monomeric orphan nuclear receptor. The transcription of CYP7A1 also is regulated via feedback inhibition. Specifically, the bile acid receptor FXR (farnesoid X receptor) binds bile acids and induces the expression of SHP (small heterodimer partner), an orphan nuclear receptor that preferentially dimerizes with LRH-1 and represses a number of enterohepatic genes involved in synthesis and transport of bile acids, including CYP7A1.138 In this way, LXR and FXR act together to tightly regulate bile acid homeostasis, respectively functioning, with a cholesterol precursor and cholesterol metabolite, to up-regulate and down-regulate CYP7A1. Dietary and secreted biliary cholesterol enter the intestinal lumen and are absorbed by proximal enterocytes. In the mouse, LXR and RXR agonists, which are metabolic indicators of active cholesterol turnover, decrease cholesterol absorption.139 This net reduction in cholesterol uptake is associated with an RXR- or LXR-mediated up-regulation of LXRE-containing genes that encode ABC transporters.140 In the small intestine, at least three ABC transporters are transcriptionally regulated by LXRs and putatively limit cholesterol absorption by pumping cholesterol back into the lumen of the gut.139 ABCA1 is a full transporter protein containing two symmetric halves, each composed of a six-transmembrane domain and an ABC. Mutations in the ABCA1 gene result in Tangier disease, which is characterized by a low concentration of plasma HDL and the accumulation of cholesterol esters in tonsils, liver, spleen, intestinal mucosa, and macrophage foam cells.141-143 ABCA1 is up-regulated by LXRs in both intestine and macrophages.144 ABCG5 and ABCG8 are half-transporters, each composed of a single transmembrane domain and ATP-binding cassette, and are coexpressed exclusively in the liver and intestine.104 Mutations in either gene cause a rare autosomal recessive disorder called sitosterolemia (see earlier), which

LXRE

LXRE

(A/G)G(G/T)T(C/T)Annnn(A/G)G(G/T)T(C/T)A (n = any nucleotide) Figure 100-8.  The consensus DNA sequence that defines the liver X receptor (LXR) response element (LXRE) to which functional LXRs bind to promote transcription. The LXRE consists of a direct repeat of the indicated hexanucleotide sequence separated by four bases. A/G and G/T signify that either base (A or G and G or T) may be used in the hexanucleotide sequence of the LXRE.

H 4

O

is characterized by increased absorption of cholesterol and toxic plant sterols and by decreased biliary sterol secretion. In vivo and in vitro experiments indicate that ABCG5 and ABCG8 are direct targets of LXRs. These findings strongly support the hypothesis that LXRs promote cholesterol loss by increasing biliary cholesterol secretion and limiting cholesterol absorption.104

CARBOHYDRATE DIETARY INTAKE

In Western societies, about 45% of total energy requirement is provided by carbohydrate, making it the major source of calories at all stages of life.145 The volume of carbohydrate ingestion appears to be declining, however, owing, in part, to a reduction in the intake of purified sugar.146,130 Overall, total calorie intake also is on the decline because of reductions in dietary fat and carbohydrate by affluent and dietconscious Western societies. The proportion of carbohydrate ingested as fruit and vegetables is rising as the intake of raw fiber increases. In adults who consume a Western diet, the amount of glucose produced by digestion is about 180 g/day (~1 mol). A growing amount of fructose had been increasingly added to our diets (often in excess of 50 g/day) through the widespread use of corn syrup as a sweetener, although recently there has been a move to limit this use and return to sugar as the preferred sweetening agent. All ingested glucose and galactose is absorbed normally, but the capacity to absorb fructose is limited in both young children and adults. This was evident in healthy, young adults (medical students in the United States and the United Kingdom) in whom the ingestion of 50 g of fructose produced abdominal pain, bloating, borborygmi, flatus, and a positive hydrogen breath test in 70% of subjects.147 It has been noted that two 12-ounce cans of some popular soft drinks contain about 50 grams of fructose in the form of corn syrup. About half of the digestible carbohydrate in an average Western diet is starch that is derived from cereals and plants and of which it is the major storage form of carbohydrate. Starch (as either amylose or amylopectin) is made up of long chains of glucose molecules. Amylose, a linear polymer in which each glucose molecule is coupled to its neighbor by α-1,4 linkage, has a molecular weight 106. Amylopectin, by contrast, is a branched-chain polymer in which α-1,6 links provide the angulations between adjacent chains of α-1,4 linked glucose molecules (Fig. 100-9); it has a molecular weight greater than 109. The amylose-to-amylopectin ratio

CH2OH O H OH H H

1

O H OH H

4

O

H

OH H 4

Figure 100-9.  Part of an amylopectin molecule indi­ cating the disposition of α−1,4 and α−1,6 linkages between glucose molecules.

CH2OH

H

H

O

CH2OH O

α–1,6 link 1

O

OH

H

H OH H

H

H

1

4

O

6 CH2 O

H

H OH H

H 1

4

O

6 CH2OH 5

OH

H

OH

H

H 1 OH H 3

H

O

α–1,4 link H

2

OH

α O

Chapter 100  Digestion and Absorption of Nutrients and Vitamins + Amylose

Maltotriose

Maltose

α amylase

Amylopectin

α-limit dextrins

Figure 100-10.  Action of pancreatic α amylase on amylose and amylopectin molecules. Because the α−1,6 link in the latter is resistant to amylase, the products include α-limit dextrins. Brown circles represent glucose units, and blue circles represent reducing glucose units. Sugars are classified as reducing or non-reducing based on their reactivity with Tollens’, Benedict’s, or Fehling’s reagents; if a sugar is oxidized by these reagents, it is called reducing since the oxidant (Ag+ or Cu2+) is reduced in the reaction. (From Gray GM. Carbohydrate absorption and malabsorption. In: Johnson LR, editor. Physiology of the Gastrointestinal Tract. 1st ed. New York: Raven Press; 1981. p 1064.)

varies widely, but most starches usually contain more amylopectin than amylose. Although starches are relatively easily digested, food preparation can influence their biologic availability. Use also may be determined by proteins associated with the starch, particularly gluten.1 Other major sources of dietary carbohydrate include sugars derived from milk (lactose), contained within the cells of fruits and vegetables (fructose, glucose, sucrose), or purified from cane or beet sources (sucrose). Processed foods form a major source of dietary sugars, particularly fructose and corn syrup; the latter contains not only fructose but also oligosaccharides and polysaccharides. The sugar alcohol sorbitol is used widely in the manufacture of diabetic sweets and preserves. Sorbitol is formed when the aldehyde group of glucose is hydrogenated to an alcohol group during manufacture, which slows its rate of absorption and thus diminishes its effect on blood sugar concentrations.48 Glycogen is the major storage form of polysaccharide in animals, but the amounts ingested in a normal diet are small. The structure of glycogen is similar to that of amylose and composed of straight chains of α-1,4-linked glucose monomers. Nonstarch polysaccharides form the majority of unavailable carbohydrates. The dietary fiber component of unavailable carbohydrate is found most abundantly in cereals, peas, beans, carrots, and peanuts. In the United Kingdom, 10 to 15 grams of dietary fiber, consisting predominantly of celluloses and hemicelluloses, is consumed by each person every day.146 Cellulose is made up of β-1,4-linked glucose molecules in straight chains, and hemicelluloses are pentose and hexose polymers with both straight and branched chains. Both forms are resistant to digestion in the small intestine because the β-1,4 bond, unlike the α bond in starch, is resistant to amylases. They are, however, broken down to some extent by colonic bacteria to yield short-chain fatty acids, which are avidly absorbed by colonic mucosa.148 The quantity of cellulose and hemicelluloses in vegetables and fruit varies markedly and depends on their age and ripeness. Other unavailable carbohydrates include pectins, gums, and alginates, which are only partially metabolized in the colon. Lignins, elaborated by plants in the process of becoming woody, are completely indigestible.48 It is well recognized that an increased intake of dietary fiber can ease constipation by increasing fecal bulk, mainly as a result of the increase in the mass of fecal flora. Dietary fiber has other roles, however, and also has effects on the absorption of other nutrients. Thus, for example, fiber

delays absorption of sugars and fats and curtails the insulin response to a carbohydrate meal. Some fiber, such as lignins, can lower serum cholesterol by binding bile salts. It may be these effects that have led to the widespread recommendation of a high-fiber diet for management or prevention of such diseases as diabetes mellitus and atherosclerosis. Finally, satiety is achieved more rapidly from a diet rich in fiber than from a low-fiber diet; it also takes longer to ingest a high-fiber meal. Advantage of this is taken in the management of obesity (see Chapter 6).

DIGESTION AND ABSORPTION Salivary and Pancreatic Amylase

Salivary and pancreatic amylases are endoenzymes; that is, they cleave the α-1,4 links internal to, or at the second or third bond from, the end of the polysaccharide chain. The products of amylase digestion therefore are short, linear oligosaccharides of maltotriose and maltose (Fig. 100-10). Because α-1,6 links, and the adjacent α-1,4 bonds, in the branched chains of amylopectin are not hydrolyzed by amylase, the products of amylopectin digestion include short, branched oligosaccharides, termed α-limit dextrins. Amylase proteins are encoded by a clustered gene family located on chromosome 1 of the human genome.149 In humans, the AMY1 gene is expressed in the parotid gland, and the AMY2 gene is expressed in the pancreas.150 The sequences of the pancreatic and salivary complementary DNAs are 94% similar, encoding for polypeptides with the same number of amino acids.151 Salivary amylase depends for its effect on its proximity to the ingested starches and the time they spend within the mouth. Thus, careful, slow chewing affords a good start to digestion, whereas rapid swallowing of poorly chewed foods—often a problem for edentulous persons—can cause suboptimal salivary amylase action. Salivary amylase is rapidly inactivated by gastric acid, but some activity may persist within the food bolus; shortchain oligosaccharides offer further protection for the enzyme against inactivation at acid pH. It is uncertain what fraction of dietary starch is digested before it reaches the duodenum. Pancreatic amylase is the major enzyme of starch digestion and, as with salivary amylase, produces short oligosaccharides, maltotriose, maltose, and α-limit dextrins; glucose monomer is not produced. Most of this hydrolysis occurs within the intestinal lumen, but because amylase also attaches itself to the brush border membrane of enterocytes, some digestion can occur at this site as well. Amylase

1707

1708

Section X  Small and Large Intestine Table 100-4  Characteristics of Brush Border Membrane Carbohydrases ENZYME

SUBSTRATE

PRODUCTS

Lactase

Lactose

Glucose Galactose Glucose

Maltase α–1,4 Linked oligosaccharides (glucoamylase) up to 9 residues Trehalase Trehalose Sucrase-isomaltase (Sucrose-α-dextrinase)   Sucrase Sucrose   Isomaltase   Both enzymes

a-limit dextrin, α–1,6 link α-limit dextrin, α–1,4 link at nonreducing end

Maltase or Isomaltase

Glucose Glucose Fructose Glucose Glucose

Maltase or Isomaltase

Isomaltase

concentration becomes limiting for starch hydrolysis only in severe cases of pancreatic insufficiency, in which luminal amylase activity levels are reduced to less than 10% of normal.152 Human milk contains amylase activity, which may be important for carbohydrate digestion in infants.153

Brush Border Membrane Hydrolases

The terminal products of luminal starch digestion, together with the major disaccharides in the diet (sucrose and lactose), cannot be absorbed intact and are hydrolyzed by specific brush border membrane hydrolases that are maximally expressed in the villi of the duodenum and jejunum. Several types have been identified (Table 100-4).154 Lactase hydrolyzes lactose to produce one molecule of glucose and one of galactose. Sucrase-isomaltase (SI, sucrase-α-dextrinase) possesses two subunits of the same molecule, each with distinct enzyme activity. Sucrase hydrolyzes sucrose to yield one molecule of glucose and one of fructose. Both sucrase and isomaltase remove glucose molecules from the nonreducing end of α-limit dextrins. Critically important is the ability of isomaltase (debrancher enzyme) in hydrolyzing the 1-6 glycosidic linkage in α-limit dextrins. The concerted action of sucrase and isomaltase thus yields monomeric glucose molecules from sucrose and α-limit dextrins (Fig. 100-11). In addition, two other carbohydrases participate in terminal hydrolysis of starch products: maltase-glucoamylase and α-limit dextrins. Maltase-glucoamylase acts on 1-4linked oligosaccharides containing as many as nine glucose residues, liberating glucose monomers. The human maltaseglucoamylase gene (MGAM) is located on chromosome 7 and has a structural homology similar to that of the SI gene.155 The maltase-glucoamylase enzyme does not undergo intracellular or extracellular proteolytic cleavage and is expressed in the brush border membrane as a monomeric protein. Maltase-glucoamylase is expressed prenatally with similar levels after birth and into adulthood.156 It has been suggested that isomaltase hydrolyzes the smallest α-limit dextrin, and another enzyme, α-limit dextrinase, is responsible for rapid hydrolysis of penta- and hexa-α-limit dextrins.157 The combination of SI, maltase and α-limit dextrinase serves to liberate glucose monomers very rapidly and close to hexose carriers, thus encouraging efficient absorption. Because free hexoses are found in the intestinal lumen, it is likely that the transport process is the rate-limiting step for uptake of monomers into the epithelium rather than the actions of the carbohydrases.

Sucrase or Maltase Sucrase or Maltase

+ Figure 100-11.  Actions of brush border membrane hydrolases. The combined actions of maltase, isomaltase, and sucrase yield glucose molecules from α-limit dextrins. Isomaltase is necessary to split the α−1,6 link. Brown circles represent glucose units, and blue circles represent reducing glucose units. Sugars are classified as reducing or non-reducing based on their reactivity with Tollens’, Benedict’s, or Fehling’s reagents; if a sugar is oxidized by these reagents, it is called reducing since the oxidant (Ag+ or Cu2+) is reduced in the reaction.

Trehalose is a disaccharide found predominantly in mushrooms, and so it is an insignificant element of the normal diet; nevertheless, there is a specific brush border enzyme, trehalase, for its hydrolysis to its two glucose molecules. Isolated trehalase deficiency has been reported in Greenland and can result in severe diarrhea after ingestion of mushrooms.156

Disaccharidase Biosynthesis and Regulation

Much has been learned about the gene regulation, biosynthesis, and processing of the disaccharidases.158,159,160,161 The human trehalase gene (TREH) is located on chromosome 11 and encodes a 583-amino acid protein with a molecular mass of about 75 kDa. SI is encoded by a single gene in the human,162 located on human chromosome 3 at locus 3q2526.163 The 5′-flanking region of the SI gene has a number of DNA regulatory regions that control initiation of gene transcription.164,165 Using mouse genetics, all four epithelial cell types in the small intestinal mucosa have the transcriptional machinery to express the SI gene.166 The elements necessary to direct intestinal epithelial cell–specific expression are embodied in a 201-nucleotide, evolutionarily conserved, 5′-flanking regions of the gene.167 At least three types of transcriptional proteins are involved in SI promoter trans­ cription, including hepatocyte nuclear factor 1 (HNF1),168,169 GATA-type zinc-finger transcription factor family members (GATA 4 and 5),170 and caudal-related homeodomain proteins (Cdx).171 The interaction of tissue-specific and tissue-

Chapter 100  Digestion and Absorption of Nutrients and Vitamins N

Transfer vesicle

N

Golgi apparatus N

Rough endoplasmic reticulum

Isomaltase site Sucrase site P Pancreatic protease High mannose oligosaccharides Complex type oligosaccharides

N

P

Plasma membrane

restricted transcription factors facilitates the transcription of genes in a single cell type. It has been suggested that the ratio of HNF-1α to HNF-1β might determine the degree of transcription induced by HNF-1.168,169 Congenital SI deficiency (CSID) is an autosomal recessive intestinal disease that is characterized by the absence of the sucrase and most of the maltase digestive activity within the SI enzyme complex; the isomaltase activity varies from absent to normal. Clinically, the disease is manifested as an osmotic-fermentative diarrhea upon ingestion of disaccharides and oligosaccharides. Analysis of this disorder at the molecular and subcellular levels has unraveled a number of phenotypes of CSID, which are characterized by perturbations in the intracellular transport, polarized sorting, aberrant processing, and defective function of SI.172,173 Changes in diet have a marked effect on the expression of SI. Starvation leads to a decline in brush border proteins and SI activity; this decline in sucrase-isomaltase activity is restored rapidly after refeeding. The type of carbohydrate ingested is important for regulation of SI expression. Starch and sucrose both induce SI activity, although sucrose is a more potent inducer.174 Study of the intestinal cell line Caco-2 has shown that a promoter region of the human sucrase gene (nucleotides −370 to +30) can down-regulate SI transcription in the presence of glucose.175 The human lactase gene is approximately 55 kb long; it has 17 exons and is located on the long arm of chromosome 2.176,177 Studies in intestinal cell lines have identified functional DNA elements in the lactase gene promoter that interact with nuclear transcription factors.178 Cdx proteins, GATA 5, and HNF-1α all have been shown to interact with the human lactase gene promoter and to activate transcription.179 Lactose intolerance is the most common manifestation of disaccharidase deficiency and results from an absence or drastically reduced level of lactase. In humans, lactase is expressed in fetal small intestine at a time in gestation just after the onset of expression of SI. Lactase expression is

Figure 100-12.  Biosynthesis of sucrase isomaltase. The nascent polypeptide (N) is translocated across the rough endoplasmic reticulum membrane after ribosomal mRNA translation. Oligosaccharide side chains join the polypeptide to be transferred to the Golgi apparatus for further processing. After incorporation in the plasma membrane, luminal proteases cleave the molecule into its active subunits. (From Lloyd ML, Olsen WA. Intestinal carbohydrases. Viewpoints Dig Dis 1991; 3:13-8.)

maintained throughout development and during childhood, although sometime during childhood, lactase activity declines to 5% to 10% of early childhood levels in most of the world’s populations. This decline occurs at the same time that intestinal SI activity is increasing. Ingestion of milk or milk products by persons with diminished lactase activity leads to flatulence, abdominal cramping, and diarrhea. This pattern of reduction of lactase activity has been termed late-onset lactase deficiency or adult-type hypolactasia. It initially was thought that the regulation of lactasephlorizin hydrolase (LPH) was post-translational and associated with altered structural features of the enzyme; it is now believed that the major mechanism of regulation of LPH is transcriptional (see later). Other forms of lactose intolerance include the rare congenital lactase deficiency and secondary forms, such as those caused by mucosal injury resulting from infectious gastroenteritis, parasitic infection, celiac disease, drug-induced enteritis, and Crohn’s disease. Differential activation of both the lactase and the SI promoter is effected by multiple similar transcription factors including GATA factors, HNF-1α, and Cdx-2, alone and in combination. This synergistic activation may be a method whereby higher levels of tissue-specific expression might be possible.180 Disaccharidase synthesis occurs within the endoplasmic reticulum, and the proenzymes then follow the path for secretory proteins through the Golgi complex before being inserted into the brush border membrane. All are glycoproteins and all undergo extensive intracellular processing with removal of redundant segments of the molecule. In the case of SI, final processing occurs on insertion into the brush border membrane after exposure to luminal pancreatic proteases (Fig. 100-12), at which point it is cleaved into its two active subunits; in contrast, lactase is already completely processed before its insertion. In their final active form, the carbohydrases project into the intestinal lumen, forming part of the glycocalyx, and they are attached to the membrane by a hydrophobic anchor

1709

1710

Section X  Small and Large Intestine that represents about 10% of the total mass of the molecule. Evidence suggests that MYO1A (brush border myosin I), a group of monomeric actin-based motors that are known to associate with membranes in intestinal villi, are involved in retaining SI within the brush border.181 Disaccharidases are synthesized by both crypt and villus cells but are expressed only on the latter. The expression of these genes in the intestine exhibits a complex spatial pattern along the vertical (crypt-to-villus) and horizontal (proximal-to-distal) axes.182 There is little SI activity in the crypts and villus tip cells, with maximal activity in lowerand mid-portions of the villus.183 The major mechanism for regulating the expression of the SI protein along the cryptvillus axis is the steady-state level of SI mRNA; however, post-transcriptional and post-translational regulation also likely play a role in the expression of the functional SI protein along the intestinal crypt-villus axis.184 A functional difference also exists between the jejunum and the distal ileum that reflects differences in the expression of different genes, or gradients of gene expression, along the proximal-distal axis of the intestine. For example, SI activity is four- to five-fold greater in the jejunum than in the ileum,185 although SI mRNA appears to be similar in the two areas. Although there are minor differences in the pattern of glycosylation in the Golgi apparatus, the major difference in regulation between the jejunum and ileum appears to be at the level of mRNA translation.186 Pancreatic proteolytic enzymes shorten the half-life of the carbohydrases.187 SI half-life can drop to as low as 4.5 hours after meals, compared with more than 20 hours during fasting. Presumably, proteolysis, as largely determined by meals, is responsible for the diurnal variation in carbohydrase activity.48 The levels of SI and other saccharidases also can decrease with infection and inflammation. In some cases, a decline in enzyme activity leads to malabsorption of carbohydrates with resultant diarrhea, flatulence, and weight loss. In most disease processes, however, the diminished levels of SI are associated with global dysfunction of the small intestinal mucosa.

Transport across the Mucosa

The three major diet-derived monosaccharides, glucose, galactose, and fructose, are absorbed by saturable carriermediated transport systems located in the brush border membrane of enterocytes in the proximal and mid small intestine.188 The active transport of glucose and galactose is achieved by the same transport protein that acts as a sodium cotransporter (primarily SGLT1)189; active glucose transport is driven by the sodium gradient across the apical cell membrane (Fig. 100-13). First, a low intracellular sodium concentration is generated by the Na+,K+-ATPase (adenosine triphosphatase) pump located in the basolateral membrane of the enterocyte, which transports 3 Na+ out of the cell and 2 K+ into the cell, resulting in a low intracellular Na+ concentration. Then two Na+ ions bind to the outer face of the transporter, producing a conformational change that permits subsequent sugar binding. The two Na+ ions and the glucose molecule then are transferred to the cytoplasmic face of the membrane through another conformational change involving a coordinated rotation or tilt (or both) of transmembrane helices.190 At the cytoplasmic surface, glucose dissociates first, and then the two Na+ ions dissociate into the cytosol to produce a ligand-free transporter. The low affinity of the cytosolic sites for glucose and Na+, and the low intracellular Na+ concentration relative to the extracellular concentration (10 vs. 140 mEq/L), promote these dissociations. The ligand-

Glucose Galactose

Na+

SGLT1

Na+

~

Na+

Fructose

GLUT5

K+

GLUT2

GLUT5

Glucose Fructose Galactose Figure 100-13.  Sodium gradient hypothesis of glucose absorption. The sodium pump (Na+,K+-ATPase) at the basolateral membrane generates a low intracellular sodium concentration. Sodium passes down the concentration gradient so created across the apical membrane coupled to glucose on a common carrier. The sodium pump thus generates the energy for this system. Glucose leaves the cell via facilitated diffusion across the basolateral membrane. GLUT2, glucose transporter; GLUT5, fructose transporter; SGLT, sodium-glucose cotransporter.

free transporter then relaxes to the outward-facing conformation to complete the cycle. The complete enzymatic turnover of the transporter occurs about 1000 times a second at 37°C. Although some of this glucose fuels cellular metabolism, a sizable fraction passes out of the cell across the basolateral membrane by facilitated diffusion (uniport). The net result is that for every glucose molecule that is transported across the brush border, Na+ ions (and two accompanying anions) also are transported across the epithelium. This, in turn, draws about 1100 water molecules across the epithelium to maintain iso-omolarity of the absorbate. Ion and nutrient absorption across the intestine do not increase the osmolarity of the fluid remaining in the intestinal lumen. The coupling among glucose, salt, and water absorption provides the explanation for the finding that water absorption across the upper and mid intestine is glucose dependent, and it is the rationale for the oral rehydration therapy (ORT) used so effectively to treat patients with secretory diarrhea (see Chapter 107).191 The prevailing opinion is that two types of glucose transporters are found across brush borders: One is a high-affinity Na+-dependent, phlorizin-sensitive transporter (SGLT1), and the other is a low-affinity transporter that might or might not be Na+-dependent and phlorizin-sensitive. Candidates for the latter role in humans include GLUT2, SGLT4, and SGLT6.192 The sodium-glucose cotransporter (SGLT1) has been characterized extensively.193,194,195 Activity of this 73-kDa

Chapter 100  Digestion and Absorption of Nutrients and Vitamins cotransporter in the intestinal brush border membrane rests with the presence of four independent, identical subunits arranged in a homotetramer. SGLT1 resides on chromosome 22 and has been cloned and sequenced. The cloned cDNA encodes for transport activity with the same relative specificity as the previously characterized native transport system: d-glucose > α-methyl-d-glucose > d-galactose > 3-O-methyl-d-glucopyranose >>> l-glucose.196 The cDNA encodes a 662-amino acid protein with a predicted molecular weight that correlates well with the biochemically defined size. SGLT1 is predicted to have 14 membranespanning domains, with one asparagine-linked carbohydrate group on the third extracytoplasmic loop.197 The expression and activity of glucose transport in the intestinal brush border are regulated by short-term and longer-term processes. In the short term, activity of glucose transport is increased by both protein kinase A- and C-dependent processes.198 The mechanism of this enhanced activity is an increase in the number of membrane transporters, mediated by changes in exocytosis and endocytosis of membrane vesicles that contain the transport protein. Longer-term regulation of glucose transport is mediated by changes in the expression of SGLT, which is controlled by changes in the nutrient environment.199 Glucose-galactose malabsorption is characterized by the neonatal onset of severe diarrhea upon the newborn’s ingestion of breast milk or regular infant formula. Several distinct mutations in the SGLT1 gene have been identified (e.g., a missense mutation resulting in a change of amino-acid residue 28 from an aspartate to an asparaginase),9 and most of these mutations are responsible for defective passage of the SGLT1 through the biosynthetic machinery from the endoplasmic reticulum or for poor trafficking from the Golgi apparatus to the brush border membrane. Rarely do mutant SGLT1 proteins reach the brush border at a normal rate, in which case, the glucose transport is defective.192 Fructose absorption occurs by facilitated diffusion: Transport occurs not against a concentration gradient but with a carrier protein to achieve transport rates greater than one would expect from simple diffusion. This process is completely independent of glucose absorption. Studies in humans have shown that there is a saturable, facilitative transport system for fructose in the intestinal epithelium that has a lower activity than that for transport of glucose and galactose. The protein responsible for most apical membrane fructose transport is a member of the facilitative monosaccharide transporter family called GLUT5, encoded by the gene SLC2A5. This 501-amino acid protein in humans has 12 membrane-spanning domains, as do other GLUT molecules, and transports fructose exclusively.200 GLUT2, however, might assist in absorption of excess luminal fructose. Little fructose is metabolized in the enterocytes, and fructose is transported across the basolateral membrane (by GLUT2 and GLUT5) and is taken up and metabolized rapidly by the liver, resulting in low postabsorptive blood levels of fructose. There may be more than one type of fructose transport system. Malabsorption of fructose in humans can be prevented by the simultaneous administration of glucose, suggesting that another glucose-responsive system may be present in the enterocytes. No inherited disorders of fructose transport (GLUT5) have been reported.201 Overall, fructose is not as well absorbed as glucose. High levels of dietary fructose can lead to dietary intolerance. Ingestion of large amounts of fructose can cause diarrhea, excessive intestinal gas, and recurrent abdominal pain. Fructose malabsorption has been associated with similar symptoms.202

Debate has developed over the mechanism of the passive or diffusive component of intestinal glucose absorption and, indeed, whether it exists.203 Pappenheimer and colleagues proposed that paracellular solvent drag contributes a passive component, which, at high concentrations of sugars similar to those in the jejunal lumen immediately after a meal, is several-fold greater than the active component mediated by the Na+-glucose cotransporter SGLT1.204 Other investigators have argued that the kinetics of glucose absorption can be explained solely in terms of SGLT1 and that a passive or paracellular component plays little, if any, part.205 Morerecent data suggest that the passive component of glucose absorption exists but that it is facilitated because it is mediated by the rapid glucose-dependent activation and recruitment of the facilitative glucose transporter GLUT2 to the brush-border membrane. This is regulated through a protein kinase C–dependent pathway activated by glucose transport through SGLT1 and also involves mitogen-activated protein kinase (MAP kinase) signaling pathways.206

Exit from the Epithelium

Most hexoses are exported from the epithelial cell by way of the basolateral membrane, although small amounts are used for intracellular metabolism. Exit across the basolateral membrane depends on facilitated diffusion (not requiring energy) via a specific carrier. Two genes that are expressed in the small intestine—GLUT2, the basolateralmembrane–associated glucose transporter, and GLUT5, an apical membrane fructose transporter—encode these facilitative sugar transport proteins.207 GLUT2 has molecular structural characteristics similar to those of the other members of this family of genes. The protein has 500 amino acids with many hydrophobic residues that predict a total of 12 membrane-spanning domains. There is one long extracellular loop between membrane-spanning domains 1 and 2 that contains an asparagine that is N-glycosylated and one long cytoplasmic loop between membrane-spanning domains 6 and 7. Once the hexoses have entered the interstitial space, they pass onward by diffusion into the portal circulation. A congenital defect in glucose transport by GLUT2 has been identified (Fanconi-Bickel syndrome). Because GLUT2 is normally expressed in the liver, pancreas, and kidney as well as in the intestine, defects in this transporter are expected to have a widespread effect on glucose homeostasis. Indeed, patients with the Fanconi-Bickel syndrome exhibit tubular nephropathy, fasting hypoglycemia, rickets, stunted growth, and hepatomegaly secondary to glycogen accumulation.208,209 The accepted dogma of intestinal glucose absorption at the basolateral membrane by glucose transporters has been challenged by studies of intestinal glucose absorption in GLUT2 null mice and in patients with GLUT2 deficiency; in both cases, glucose absorption was not impaired. Additional work has suggested that there are two separate pathways for the exit of sugar from enterocytes: one that involves GLUT2 and another that requires glucose phosphorylation, the transfer of glucose-6-phosphate into the endoplasmic reticulum and the release of free glucose into the blood. The release mechanism is unclear, but it has been proposed to involve vesicle trafficking. This postulate is supported by oral tolerance tests on a patient with congenital deficiency in glucose-6-phosphate translocase1, in whom glucose absorption was impaired but not eliminated.209 Not all potentially digestible carbohydrate is absorbed in the small intestine. As much as 20% of dietary starch escapes into the colon, particularly that derived from cereals

1711

1712

Section X  Small and Large Intestine and potatoes.2 Most of this unclaimed carbohydrate, however, is metabolized by colonic bacteria, and the shortchain fatty acids thus derived are readily absorbed; hydrogen and methane also are generated and contribute to production of flatus.

PROTEIN DIETARY INTAKE

Dietary proteins are the major source of amino acids, and in the average Western diet they provide about 10% to 15% of energy intake. Affluent populations ingest more protein than they need to maintain their normal balance. An average adult in a Western country consumes 70 to 100 g of protein per day, whereas the poor in Asia and Africa consume 50 g or less per day.48 Recommended dietary requirements vary from 0.75 to 1 g/kg of body weight per day, but deficiency states are rare even with intakes of 0.5 g/kg per day or less. In the United Kingdom, protein intake has remained fairly steady since the mid-1970s, but with the marked decline in fat and carbohydrate ingestion, the ratio of protein to nonprotein energy intake has risen.48 Little harm appears to occur in the unusual subgroups of society who consume very large amounts of protein, although renal function can be impaired by this dietary habit. The Masai tribes of Africa and the Gaucho of South America, who consume 250 to 300 g (largely of animal origin) per day, suffer no obvious untoward effects from this consumption.48 The variety of types of animal and plant proteins is enormous. Generally, plant proteins are less digestible than those derived from animals, but some fibrous animal proteins, such as keratin and collagen, also are relatively indigestible. High-proline proteins such as the glutenins are less thoroughly digested than are others. The quality of proteins depends largely on their amino acid composition; proteins rich in essential amino acids are regarded as being of high quality. Proteins from animal sources have a high content of essential amino acids, unlike proteins from certain specific plant sources, which are said to be incomplete because they lack or contain only certain essential amino acids. Such deficiencies in essential amino acids typically are overcome in a mixed diet; however, the relative contribution of animal and plant protein varies according to geographic region. In developed countries, such as in North America and Europe, animal protein contributes about 70% of the total protein compared with developing nations in the Middle East and Africa where the animal protein contribution can be as low as 20%. Food processing, by heat for example, can cause interand intramolecular bonding in the proteins to produce polymeric forms that are relatively resistant to hydrolysis.1 Other constituents of the diet also can interfere with protein digestion; for example, starch and reducing sugars have the potential to impair digestion.48 Despite these interferences, digestion and absorption of proteins are remarkably complete, and only about 3% to 5% of ingested nitrogen is lost in the stool, probably because of the resistance of some peptide bonds to hydrolysis.1 A few selected proteins are resistant to proteolysis in the small intestine, including secretory immunoglobulin (Ig) A and intrinsic factor. Among the 20 common amino acids that form animal and plant proteins, 8 cannot be synthesized by animals: leucine, isoleucine, lysine, methionine, phenylalanine, threonine, tryptophan, and valine. These eight essential amino acids have to be ingested, usually in plant-derived foods. Histidine also is required for growth in infants.

PROTEINS FROM ENDOGENOUS SOURCES

Almost half of all protein that enters the intestine is derived from endogenous sources. Of this, about 20 to 30 g/day are derived from secretions of salivary, gastric, biliary, pancreatic, and mucosal origin. Another 30 g/day of protein are provided by epithelial cells desquamated from the villus tips, and 2 g of plasma proteins are delivered into the intestinal lumen each day.

DIGESTION AND ABSORPTION Pepsins

Digestion of proteins begins in the stomach with the action of pepsins. Pepsins are released from their precursor pepsinogens with the loss of a small basic peptide by auto­ activation in an acid pH. At neutral or alkaline pH, the pepsinogen amino-terminal region is folded in such a way as to mask a catalytic site. In the acidic environment of the stomach, the catalytic site is uncovered and then proceeds to remove the amino-terminal region, which consists of 40 amino acids, thereby generating the active form of the molecule, pepsin. Pepsinogen release from chief cells is stimulated by gastrin, histamine, and cholinergic stimulation and closely mirrors acid secretion.210 The pepsins are a family of endoproteases that hydrolyze internal peptide bonds in proteins. They act preferentially on peptide bonds formed by the aromatic amino acids phenylalanine and tyrosine and by the branched-chain amino acid leucine. There are two immunologically distinct groups of pepsins (groups 1 and 2), although eight fractions are identified electrophoretically. Both of the immunologically separated species are secreted by chief cells, but group 2 isoforms also are present in the mucus cells in the oxyntic and pyloric areas of the stomach and in Brunner’s glands of the duodenum. Their substrate specificities vary little, but their pH optima differ slightly (between 1.8 and 3.5); all are irreversibly inactivated in alkali. Pepsins remain active at the acid pH of gastric contents to produce a mixture of peptides with a small portion of amino acids. Pepsin activity is therefore confined to the stomach. The completeness of gastric proteolysis depends, in part, on the rate of gastric emptying, the pH of intragastric contents, and the types of protein ingested. Moreover, the products of protein digestion by pepsins in the stomach can further influence acid and pepsinogen secretion as well as gastric emptying. Subjects who are achlorhydric or who have lost control of gastric emptying as a result of pyloroplasty or partial gastrectomy do not appear to have a problem with assimilation of protein, suggesting that gastric proteolysis is not an essential component of digestion.

Pancreatic Proteases

In contradistinction to amylase and lipase, which are secreted in their active forms, each of the pancreatic pro­ teases is secreted as a proenzyme and therefore must be activated within the intestinal lumen. Enterokinase (enteropeptidase) plays a key role in proteolysis. It is liberated from its superficial position in the brush border membrane by the action of bile acids,211 its action being to convert trypsinogen to trypsin by removing its hexapeptide NH2 terminus. Trypsin in turn activates the other proteases and continues to split more trypsin from trypsinogen (Fig. 100-14). The proteases are classified as endo- and exopeptidases, according to the sites of the peptide bonds against which they are most active. Endopeptidases include trypsin, chymotrypsin, and elastase, and exopeptidases include carboxypeptidase A and B (Table 100-5).

Chapter 100  Digestion and Absorption of Nutrients and Vitamins Chymotrypsinogen

Chymotrypsin

Proelastase Trypsinogen

Elastase

Trypsin Procarboxypeptidase A

Carboxypeptidase A

Enterokinase Procarboxypeptidase B

Carboxypeptidase B

3000

ACTION

PRODUCTS

Trypsin

Endopeptidase; cleaves internal bonds at lysine or arginine residues; cleaves other pancreatic proenzymes Endopeptidase; cleaves bonds at aromatic or neutral amino acid residues Endopeptidase; cleaves bonds at aliphatic amino acid residues Exopeptidase; cleaves aromatic amino acids from carboxy terminal end of protein and peptides Exopeptidase; cleaves arginine or lysine from carboxy terminal end of proteins and peptides

Oligopeptides

Chymotrypsin

Elastase Carboxypeptidase A

Carboxypeptidase B

Oligopeptides

Oligopeptides Aromatic amino acids and peptides

Absorption rate (µmol/min/30 cm jejunum)

Table 100-5  Pancreatic Proteolytic Enzymes, Their Sites of Action, and Their Products ENZYME

Figure 100-14.  Activation of pancreatic proteolytic enzymes. Enterokinase (enteropep­ tidase) plays a critical role in activating trypsinogen to form trypsin. Trypsin in turn activates not only more trypsinogen but also the other proteolytic enzyme precursors. Products of enzymatic actions are shown in italics.

2400

Triglycine

1800

Diglycine 1200

600 Free glycine

Arginine, lysine, and peptides

0 0

100

200

300

Amount of glycine in perfusate (mmol)

Trypsin, chymotrypsin, and elastase have specificity for peptide bonds adjacent to certain specific amino acids. They split peptide bonds within the protein molecule, whereas exopeptidases remove a single amino acid from the carboxyl terminal end of the peptide. Trypsin produces short-chain oligopeptides that are further hydrolyzed by the exopeptidases, carboxypeptidase A acting on aromatic and aliphatic carboxyl terminals, and carboxypeptidase B acting on peptides containing basic carboxyl terminals. The final products of intraluminal digestion thus are produced by cooperative activity of endo- and exopeptidases and consist of a number of neutral and basic amino acids together with peptides of two to six amino acids in length. About 30% of luminal amino nitrogen is found in amino acids and about 70% is found in oligopeptides.212 In addition to nutrient protein hydrolysis, pancreatic proteases have other functions: They split vitamin B12 from the R protein to which it is linked so that it then can bind intrinsic factor; they increase the turnover of brush border membrane hydrolytic enzymes, and, as discussed earlier, they initiate the final steps in the processing of the SI complex; finally, they may have a role in the inactivation of some organisms.1

Figure 100-15.  Rates of glycine absorption (mean ± standard error of mean) from perfusion solutions containing equivalent amounts of glycine in free or peptide (diglycine, triglycine) form. Results are from studies in the jejunums of four normal humans. (From Adibi SA, Morse EL, Masilamani SS, Amin PM. Evidence for two different modes of tripeptide disappearance in human intestine. Uptake by peptide carrier systems and hydrolysis by peptide hydrolases. J Clin Invest 1975; 56:1355-63.)

Digestion at the Brush Border Membrane and in the Cytoplasm In contrast to the absorption of carbohydrate, which is largely restricted to uptake of hexose monomers across the brush border membrane, amino acids can be absorbed either as monomers or as di- or tripeptides. Indeed, amino acid absorption is achieved more efficiently in the form of peptides than as single amino acids (Fig. 100-15).213 The fact that the vast majority of the end-products of protein digestion that reach the portal circulation are amino acids, however, speaks strongly in favor of the presence of peptidases in the epithelium. Patients with cystinuria and Hartnup’s disease, who have specific defects in the absorption of basic and neutral amino acids, respectively, do not develop protein deficiency states

1713

1714

Section X  Small and Large Intestine Table 100-6  Peptidases Found on the Brush Border Membrane and in the Cytoplasm of Villus Epithelial Cells PEPTIDASE Brush Border Membrane Peptidases Amino-oligopeptidases (at least two types) Aminopeptidase A Dipeptidase I Dipeptidase III Dipeptidyl aminopeptidase IV Carboxypeptidase P Gamma-glutamyl transpeptidase Folate conjugase Cytoplasmic Peptidases Dipeptidases (several types) Aminotripeptidase Proline dipeptidase

ACTION

PRODUCTS

Cleave amino acids from carboxy terminus of 3-8 amino acid peptides Cleaves dipeptides with acidic amino acids at amino terminus Cleaves dipeptides containing methionine Cleaves glycine-containing dipeptides Cleaves proline-containing peptides with free α-amino groups Cleaves proline-containing peptides with free carboxy terminus Cleaves γ-glutamyl bonds and transfers glutamine to amino acid or peptide acceptors Cleaves pteroyl polyglutamates

Amino acids and dipeptides

Gamma glutamyl amino acid or peptide Monoglutamate

Cleave most dipeptides Cleaves tripeptides Cleaves proline-containing dipeptides

Amino acids Amino acids Proline and amino acids

Table 100-7  Distribution of Peptidase Activity SUBSTRATE Dipeptides Tripeptides Tetrapeptides Higher peptides

BRUSH BORDER MEMBRANE (%) 5-10 10-60 90 98

CYTOPLASM (%) 80-95 30-60 1-10 0

because the absorption of peptides in these patients is normal.214 The discovery that di- and tripeptides are actively transported by the brush border membrane of enterocytes has been valuable in explaining this observation, and it emphasizes the need for critical evaluation of the supposed nutritional advantage provided by elemental diets that consist only of free amino acids. A range of peptidases is present in the brush border membrane and in the cytoplasm of villus epithelial cells for the hydrolysis of oligopeptides up to approximately eight amino acid residues in length (Table 100-6).215-217 The peptidases on the brush border membrane differ in several important respects from those within the cytoplasm (Table 100-7). About 90% of the dipeptidases are found in the cytoplasm and only about 10% in the brush border, whereas the distribution of hydrolases for tetrapeptides is the reverse of this. Peptidases for pentapeptides and larger molecules are confined almost entirely to the brush border membrane. Cytoplasmic enzymes are much more heat labile than those in the brush border, and there are differences in the electrophoretic mobility patterns for the two sets of enzymes.75 Most oligopeptidases appear to be aminopeptidases; that is, they act by removing residues from the amino terminus of the peptide. The chain length of the peptides is an important factor that determines not only whether the site at which hydrolysis occurs is at the brush border or within the cell but also its rate. Thus, rates of brush border membrane hydrolysis for tripeptides are most rapid and for dipeptides is least rapid, whereas tetra- and pentapeptide hydrolysis rates occupy an intermediate position.213

Amino acids Amino acids Amino acids Peptides and amino acids Peptides and amino acids

Distinct from the amino oligopeptidases are at least three other peptidases. Aminopeptidase A has specificity for peptides with acidic amino acids at their amino termini. Aminopeptidases 1 and 3 (distinguished on electrophoretic mobility) have specificities for different substrates with different amino acid peptide bonds.1 Proline-containing oligopeptides are not readily hydrolyzed by most proteases, although many proteins— including collagen, gliadin, and casein—are rich in proline. Two proline-specific carboxypeptidases, however, have been demonstrated in the brush border membrane. They have slightly different substrate specificities218 and together with a cytoplasmic proline dipeptidase, they are likely to be responsible for hydrolysis of proline-rich peptides. A number of other brush border membrane peptidases need to be mentioned. Gamma glutamyl transpeptidase hydrolyzes gamma glutamyl peptide bonds, with the transfer of the gamma glutamyl group to another amino acid to form a gamma glutamyl amino acid or peptide derivative.1 The role of this brush border membrane in the intestine is not yet clear. Folate conjugase, an enzyme concerned with hydrolysis of dietary folate, will be considered later. The recent demonstration of angiotensin I-converting enzyme (ACE) in intestinal mucosa suggests that it, too, might hydrolyze dietary peptides.219 Indirect evidence suggests that endopeptidases also may be present on the brush border membrane, because protein digestion occurs, even in the complete absence of pancreatic function; these enzymes have yet to be isolated. As with other proteins, synthesis of each specific peptidase occurs in the rough endosplasmic reticulum, and following transfer to the Golgi apparatus, the proteins are transported to the brush border membrane, where they are inserted by exocytic fusion.220,221 They are attached to the basement membrane by short anchoring pieces in a manner analogous to the attachment of disaccharidases222; however, unlike the latter enzymes, there is little post-translational processing, either within the cytoplasm or by pancreatic enzymes on the brush border. Of the cytoplasmic dipeptidases, the most abundant appears to be one with broad specificity for neutral amino acid-containing dipeptides. The tripeptidase isolated has broad specificity for amino-terminal residues and high spec-

Chapter 100  Digestion and Absorption of Nutrients and Vitamins ificity toward tripeptides containing proline as the aminoterminal residue, which distinguishes it from the brush border membrane amino oligopeptidase. Other characteristics of the tripeptide that are required for rapid hydrolysis include a free α amino group, an α carboxyl group, and an l-configuration for the two amino acid residues.223

Absorption of Peptides

Substrate inhibition studies indicate that tri- and dipeptides inhibit uptake of either peptide from the lumen, but neither is affected by single amino acids. Such evidence suggests that small peptides use a separate transporter system from those used by single amino acids. By contrast, tetrapeptide absorption is inhibited by single amino acids but not by di- and tripeptides, suggesting that tetrapeptides are split before absorption. The advantage of dipeptide absorption over single amino acid absorption has been largely demonstrated experimentally with single peptides containing a single amino acid, usually glycine.213 Several studies, however, have demonstrated the kinetic advantage of peptides over amino acids, even in complex mixtures of partial digests of proteins.224,225 Absorption was greater from tryptic hydrolysates of proteins than from a mixture of amino acids. Furthermore, the wide variation in rates of absorption seen with different individual amino acids was reduced when they were presented as a tryptic hydrolysate. A number of other factors influence digestion and absorption. The presence of amino acids in the lumen inhibits peptide hydrolysis (product inhibition), whereas luminal glucose and luminal acidification each inhibit amino acid and peptide absorption.213 There is good evidence to suggest that di- and tripeptides are taken up by a single type of transporter with some stereospecificity because the length of the amino acid side chains on the di- or tripeptides is important; the longer the side chain the more preferred the substrate for the absorption site (Table 100-8).226 The l-isomers of the amino acids in dipeptides are much preferred to the d-isomers, whereas the presence of acidic and basic amino acid residues in dipeptides reduces affinity for the transport system, compared with neutral amino acid residues. Affinity is also greater for dipeptides than for tripeptides, at least in the example of peptides that contain glycine. The transporter for peptides is not dependent on sodium, but cotransport with protons may occur instead.227 The peptide transporter for human small intestine has been cloned228,229 and is a member of a superfamily of H+coupled peptide transporters. The human protein consists of 708 amino acids, with a predicted core molecular size of 79 kDa that contains 12 transmembrane domains. The gene is located on chromosome 13. In humans, it is expressed in the small intestine (duodenum, jejunum, and ileum), but not in the esophagus, stomach, or colon. In the small intestine it is expressed only on absorptive epithelium. It recog-

Table 100-8  Relative Specificities of Intestinal Peptide Transporters DIPEPTIDES

TRIPEPTIDES

L-form

d-form

of amino acids Neutral amino acids Long side chains

Acidic or basic amino acids Short side chains

nizes a variety of neutral, anionic, and cationic dipeptides as substrates,230,231 which explains the broad substrate specificity of the intestinal peptide transport system. The most interesting feature of this transport process is that it uses a transmembrane electrochemical H+-gradient rather than a transmembrane electrochemical Na+-gradient as its driving force.232 There is an acid pH microclimate on the luminal surface of the intestinal brush border membrane that creates a H+-gradient across the brush border membrane in vivo. This acid pH microclimate is generated and maintained by the combined action of the Na+-H+ exchanger in the brush border membrane and Na+,K+-adenosine triphosphatase (ATPase) in the basolateral membrane of the enterocyte. The mechanism of the transport process is a simultaneous translocation of H+ and peptide substrate involving a single H+ binding site on the protein (Fig. 100-16).233,234 A multitude of processes are involved in the absorption of peptides. The well-established processes include a Na+H+ exchanger located in the brush-border membrane that maintains an intracellular alkaline pH, a Na+,K+-ATPase located in the basolateral membrane that maintains an inside negative membrane potential, and several cytoplasmic peptidases that prevent intracellular accumulation of absorbed peptides. These enzymes convert most of the

Lumen Dipeptides, tripeptides

H+

H+

Na+-H+ exchanger

Brush-border membrane

~

Pept-1

Peptides

H+

Na+

Peptidase

Na+ Amino acids

Peptides

~ Basolateral membrane Amino acids

ATPase Peptides

K+

Blood Figure 100-16.  Peptide transport across the intestinal epithelium. This transport process uses a transmembrane H+-gradient rather than a transmembrane electrochemical Na+-gradient as the driving force. The acid pH microclimate on the luminal surface of the intestinal brush border membrane is generated and maintained by the combined action of the Na+-H+ exchanger in the brush border membrane and Na+,K+-ATPase (adenosine triphosphatase) in the basolateral membrane of the enterocyte. The mechanism of the transport process is a simultaneous translocation of H+ and peptide substrate involving a single H+ binding site on the protein, Pept-1.

1715

1716

Section X  Small and Large Intestine absorbed oligopeptides to amino acids that are either used by the absorbing cells or are released into the portal circulation via the amino acid transporters located on the basolateral membrane of these cells. The oligopeptides that escape hydrolysis by the cytoplasmic peptidases are transported across the basolateral membrane into the portal circulation by a peptide transporter that appears to be different from the Pept-1 transporter. Oligopeptide transport could be regulated by alteration in activity or abundance of Pept-1, Na+H+ exchanger, Na+,K+-ATPase, cytoplasmic peptidases, and basolateral oligopeptide transporter.235,236 Studies of individual substrates and hormones in cell culture have shown that the membrane population of Pept-1 is increased by dipeptides, certain amino acids, insulin, and leptin and decreased by epidermal growth factor (EGF) and triiodothyronine. In the case of dipeptides, EGF, and thyroid hormone, there are parallel changes in the gene expression brought about by alteration of transcription or stability of Pept-1 mRNA. In contrast, treatment with insulin and leptin does not induce any alteration in the Pept-1 gene expression, and the mechanism of increased protein expression appears to be increased trafficking from a preformed cytoplasmic pool to the apical membrane.235,236

Transport of Amino Acids

Whereas there appears to be only one type of dipeptide transporter in the brush border membrane for the 400 possible dipeptides, there is a multiplicity of transport mechanisms for the 20 amino acids. In adults, these are situated on villus enterocytes and involve carrier-mediated active transport or facilitated diffusion processes, which typically depends on the Na+ gradient as the driving force; a small portion may be absorbed by simple diffusion, independent of any ion gradient. There has been some difficulty in defining the number and types of transporters because of their overlapping specificities; several amino acids use a number of different transport systems (Table 100-9). On the basis of kinetic studies, at least four active processes have been identified for transport of neutral amino acids across the

Table 100-9  Major Amino Acid Transport Systems Detected in Intestinal Epithelial Cells TRANSPORT SYSTEM Brush Border Membrane Neutral Amino Acids SLC6A19 SLC36A1 SLC6A20 SLC6A14 SLC1A5 SLC7A9/SLC3A1 Basic Amino Acids Acidic Amino Acids SLC1A1 (X-GA−) Basolateral Membrane L A SLC1A5 (ASC) N

apical cell membrane. Each is electrogenic and sodium dependent. One has broad specificity for a number of neutral amino acids (NBB system); a second provides another route for phenylalanine and methionine (PHE system); a third provides a mechanism for imino acid absorption (IMINO system); and the fourth transports beta amino acids. Separate sodium-dependent, active transport processes for basic and acidic amino acids also have been demonstrated, and some evidence suggests that facilitated diffusion of these types of amino acids also occurs, although this is likely to be a minor pathway. Genomic advances have allowed most mammalian amino acid transport functions to be attributed to specific gene products: At least 52 amino acid transporter-related gene products are grouped within 12 solute carrier families, with their own new nomenclature.237 The classic Na+-dependent imino acid transporter has been identified as the human PAT1 (human proton-coupled amino acid transporter 1) or solute carrier SLC36A1. This high-capacity imino acid carrier has been localized to the small intestinal luminal membrane and transports imino and amino acids (glycine, proline, alanine, taurine).238,239 Human PAT1 mediates 1 : 1 symport of protons and small neutral amino acids. The acid microclimate of the brush border membrane drives transport of the amino acids into the cytosol. Transport activity is independent of Na+ and Cl− (Fig. 100-17). In addition, the IMINO system is a Na+-dependent transporter with specificity toward the imino acids proline and hydroxyproline. The protein responsible for this transport activity is SIT1(Na+coupled imino acid transporter 1).240,241 System B0,+, also present on the brush border membrane, mediates the Na+ and Cl− coupled electrogenic transport of neutral as well as cationic amino acids across the brush border membrane. The gene encoding the protein respon-

Apical

Basolateral

Gly, Ala, Pro PAT-1 H+

H+

SUBSTRATES

Neutral amino acids Imino acids; proline, hydroxyproline Imino acids Neutral and cationic amino acids Alanine, serine, cysteine, glycine, asparagine Neutral amino acids, cationic amino acids, cysteine Lysine, cysteine, basic amino acids Glutamate, aspartate Broad selectivity Broad selectivity Neutral amino acids, alanine, serine, cysteine Glutamine, histidine, asparagine

H+

H+ Na+-H+ exchanger

Na+

Enterocyte Figure 100-17.  Intestinal amino acid transport. The human proteincoupled amino acid transporter (PAT-1) is involved in the absorption of small amino acids across the apical membrane. The acid microenvironment generated by a Na+-H+ exchanger provides the electrochemical proton gradient that drives amino acids to the cytosol. Ala, alanine; Gly, glycine; Pro, proline. (Boll M, Daniel H, Gasnier B. The SLC36 family: Proton-coupled transporters for the absorption of selected amino acids from extracellular and intracellular proteolysis. Pflugers Arch 2004; 447[5]:776-9).

Chapter 100  Digestion and Absorption of Nutrients and Vitamins sible for this activity, ATB0,+ is located on human chromosome X. A separate Na+-dependent transport system, X−AG, (SLC1A1) is specific for anionic amino acid, aspartate, and glutamate. The glutamate transporter expressed in the intestinal brush border membrane is known as the EAAT3. This transporter is defective in the inherited amino acid transport defect known as dicarboxylic aciduria. Hartnup’s disease is a disorder of renal and gastrointestinal neutral amino acid transport that is inherited as an autosomal recessive trait. The gene causing Hartnup’s disease has been localized to chromosome 5p (it had pre­viously been localized to chromosome 19), and a new gene, SLC6A19, a sodium-dependent and chlorideindependent neutral amino acid transporter, has been suggested as the defective gene by two separate groups.242,243 This transporter has been shown to be expressed in the intestine and has properties of system B0. System B0 refers to a broad range of amino acids with neutral (0) charge. SLC1A5 is the proposed ASC carrier for the neutral amino acids alanine, serine, and cysteine.244 Whereas the SLC1A1 carrier cotransports 3 Na+ and 1 H+ with countertransport of 1 K+, the SLC1A5 transporter mediates Na+-dependent transport. The B0,+ is a Na+-independent transport system that recognizes neutral and cationic amino acids in addition to the disulfide amino acid cystine. It is a heterodimer, consisting of a light and heavy chain, with the genes responsible for each chain being found on chromosomes 2 and 19, respectively. This transporter is defective in cystinuria. Several hormones have been shown to alter the amino acid and peptide transport process in the intestine. Somatostatin and vasoactive intestinal polypeptide decrease these transport processes, whereas EGF, neurotensin, cholecystokinin, and secretin enhance them. Human Pept1 appears to be inhibited by protein kinase C245 and cyclic adenosine monophosphate (cAMP).246 The expression of the intestinal peptide transporter is also modulated by dietary protein content.247 Even though the peptide transporter is expressed along the entire small intestine, diet-induced changes in the expression of the transporter are specific to certain regions. A high-protein diet increases the steady state levels of the transporter-specific messenger RNA in the middle and distal regions of the small intestine. The expression of the brush border peptidases dipeptidylcarboxypeptidase and dipeptidylaminopeptidase IV, which release dipeptides from oligopeptides, also are enhanced by a highprotein diet.

Exit from the Epithelium

Exit through the basolateral membrane operates via a number of different mechanisms that involve active transport and diffusion of both facilitated types.248 Active, sodium-dependent processes exist at this membrane for the uptake of neutral amino acids, which presumably supply nutrients for crypt cells and for villus enterocytes during fasting when a luminal source is unavailable (see Table 100-9). Villus enterocytes normally receive the amino acids necessary for production of their own protein from luminal nutrients; crypt cells obtain their supply from the portal circulation. Of all the amino acids, glutamine appears to be a unique and major source of energy for enterocytes; ammonia is an important metabolic byproduct of this process. Active uptake of glutamine at the basolateral membrane, as well as via apical membrane processes, is therefore of particular importance. It has been estimated that approximately 10% of amino acids are used in the production of enterocyte protein. Some

of these proteins are secreted across the basolateral membrane specifically by villus enterocytes, including apo A-I and apo A-IV, secretion of which increases many-fold after a fatty meal.48 The intestinal basolateral membrane possesses a set of amino acid transport systems that differ from those in the brush border membrane. The amino acid transport systems in the basolateral membrane function to export amino acids from the enterocytes into the portal circulation during feeding. They also participate in the import of amino acids from the portal circulation into the enterocyte for cellular metabolism when amino acids are not available from the intestinal lumen, such as between meals. The intestinal basolateral membrane also possesses a peptide transporter system that is probably identical to that in the brush border membrane and that facilitates the exit of hydrolysisresistant small peptides from the enterocyte into the portal circulation. Several well documented amino acid transport systems have been described in the basolateral membrane. System y+L is the amino acid exchanger that permits Na+independent efflux of cationic amino acids from intestinal cells into the blood coupled to the Na+-dependent influx of neutral amino acids from the blood into intestinal cells. System A is a Na+-coupled transport system for neutral amino acids, including glutamine, that plays a role in the entry of amino acids from the blood into intestinal cells for cellular metabolism. This Na+-coupled neutral amino acid transporter (SNAT) consists of three subtypes, SNAT1, 2, and 4; SNAT2 is expressed in the small intestine.249 Very small amounts of dipeptides have been detected in the portal circulation after a meal, but the great majority of absorbed products of protein digestion that reach the circulation are in the form of single amino acids. A somewhat surprising finding is that digestion of protein continues into the ileum, with approximately 40% of ingested protein undergoing transport in this segment of small intestine.250

VITAMINS WATER-SOLUBLE VITAMINS

Although it had been thought that the absorption of watersoluble vitamins depended simply on passive diffusion across the intestinal mucosa, there has been increasing recognition of the importance of specific carrier-mediated processes for this process (Table 100-10). Several of these vitamins are present in the diet as conjugates or coenzymes that require hydrolysis before or during their absorption.

Ascorbic Acid (Vitamin C)

Although most species synthesize all their vitamin C requirements, primates, guinea pigs, and some birds have lost this capacity and thus depend on diet for their needs. Vitamin C is found in a wide range of foods, but the most abundant sources are fresh fruits and fruit juices. Black currants are particularly rich (200 mg/100 g) in vitamin C; apples and pears less so (5 mg/100 g). Of animal sources, raw liver contains about 20 mg/100 g, and milk contains about 2 mg/100 g; fresh meat contains only traces of vitamin C. The ingestion of as little ascorbic acid as 10 mg/day prevents scurvy, and the recommended daily intake is 40 mg.145 Absorption of ascorbic acid decreases with increased intake and varies from 16% at high (more than 10 g) to 98% at low (less than 20 mg) intakes. From what is known, 80% to 90% of dietary ascorbic acid is absorbed.

1717

1718

Section X  Small and Large Intestine Table 100-10  Water-Soluble Vitamins VITAMIN

Recommended dietary allowances*

Ascorbic acid Folic acid

75-90 mg/day 400 mg/day

Cobalamin (B12)

2.4 µg/day

Thiamine Riboflavin Pantothenic acid Biotin

1.1-1.2 mg/day 1.1-1.3 mg/day 5 mg/day 30 µg/day

Pyridoxine Niacin

1.3-1.7 mg/day 14-16 mg/day

TRANSPORT MECHANISM Active; Na+-dependent process at brush border membrane Hydrolysis of dietary polyglutamates by folate conjugase at brush border membrane; Na+-dependent active transport or facilitated diffusion of monoglutamate at brush border membrane Intrinsic factor binding; uptake of intrinsic factor–B12 complex at ileal brush border membrane by way of specific receptor Na+-dependent active transport Absorption includes hydrolytic and phosphorylation steps Absorption includes hydrolytic and phosphorylation steps Saturable facilitated diffusion (at low concentrations); nonsaturable linear diffusion (at high concentrations) Simple diffusion ?

*Daily Reference Intakes (DRI) were established by the Institute of Medicine between 1997-2001. They are quantitative estimates of nutrient intakes to be used for planning and assessing diets for healthy people. The DRIs include both recommended intakes and tolerable upper intake levels. The RDAs (Recommended Dietary Allowances) are a component of the DRIs and are defined as the daily intake of a nutrient considered sufficient to meet the requirements of 97% to 98% of adults.

Cooking destroys some of the ascorbic acid contained in food, but such destruction can be minimized by shortening cooking times and not keeping foods hot for prolonged periods before they are eaten. Prolonged storage of foods also depletes vitamin C content. With loss of the capacity for hepatic synthesis of vitamin C, a specific absorptive mechanism has developed in humans (and guinea pigs). Transport across the apical membrane of small intestinal enterocytes occurs by an active Na+-dependent process.251 The active absorption mechanism becomes saturated when the mucosal concentration of vitamin C is greater than 6 mmol/L; this might account for the decreased fraction of dietary vitamin C absorbed with increasing intake. Electrically neutral, uphill transport of vitamin C probably occurs in the form of sodium ascorbate, via a concentrative, carrier-mediated, Na+-dependent mechanism. Two distinct isoforms of these Na+-dependent vitamin C transporters have been described in humans: hSVCT1 and hSVCT2.252,253 The intestinal ascorbic acid uptake process is regulated by extracellular substrate levels and by an intracellular protein kinase C–mediated pathway.254 Flavanoids in fruit and vegetables can decrease vitamin C absorption by inhibiting the SVCT1 transporter.255 A variable proportion of luminal vitamin C is present in the oxidized form as dehydroascorbic acid, which also is actively absorbed. Within the intestinal cell, dehydroascorbate is rapidly reduced back to ascorbic acid by the enzyme dehydroascorbate reductase, which requires reduced glutathione. It is through this mechanism that the intracellular level of dehydroascorbate is believed to be maintained at low, nontoxic levels. Ascorbic acid then circulates unbound in plasma and is transported against a gradient into target tissues by SVCT2.

Folic Acid

Folic (pteroyl monoglutamic) acid consists of the complex pterin molecule conjugated to para-aminobenzoic acid and glutamic acid. Although considerable amounts of dietary folate are in the form of polyglutamates with at least six glutamic acid residues, much is present as formyl- and methylhydrofolate. The folates are distributed widely in the diet, particularly rich sources being spinach (200 mg/100 g), liver (140 mg/100 g), and peanuts and beans (100 mg/100 g). Meat, chicken, potatoes, and fruit (except orange juice) are poor sources (less than 15 mg/100 g). Prolonged cooking of

food destroys its folate. Recommended dietary intakes are on the order of 200 mg/day in adults and 400 mg/day during pregnancy.145 Unlike vitamin B12, there is no vast storehouse of folate. The adult body contains only 2 to 3 mg of folate; therefore, in cases of poor intake (or malabsorption of folate) for just a short period, folate depletion occurs relatively rapidly. Absorption of dietary polyglutamates depends on hydrolysis to monoglutamate at the brush border membrane followed by transport into the cytoplasm.256,257 The apical membrane hydrolase (conjugase) in human intestine is expressed predominantly in the proximal jejunum and is a folylpoly-γ-glutamate carboxypeptidase (GCP2, or folate hydrolase) that cleaves off a single glutamic acid residue at a time. This brush border form of folate hydrolase has been cloned and has been shown to be up-regulated in dietary folate deficiency.258,259 GCP2 is a 120-kDa membrane protein that contains a single 5′ membrane-spanning domain. It is an exopeptidase with an optimal pH of 6.5 that releases intermediate products of folylpolyglutamate ending in the folylmonoglutamate derivative.258 Also recognized is a cytoplasmic folate hydrolase, an endopeptidase prominent in several species and present in humans; its role in humans is uncertain. Uptake is achieved by the reduced folate carrier, a specific concentrative, carrier-mediated, Na+-dependent, pH-sensitive process that is active at acid pH.260 It is inhibited by diphenylhydantoin and sulfasalazine, which also reduces hydrolysis. Prolonged exposure to ethanol inhibits hydrolysis (but not uptake), and this may be relevant to the folate deficiency sometimes found in alcoholics. As folate is transported through the enterocyte, pteroylglutamate is reduced and methylated, and 5-methyltetrahydrofolate (5-MTHF) is transported across the intestinal basolateral membrane by the reduced-folate carrier to the portal vein and the liver. The intestine also is exposed to a second source of folate: the folate that is synthesized by the normal microflora of the large intestine. Significant amounts of this folate source have been shown to exist in the absorbable monoglutamate form. The colon is capable of absorbing some of this folate.261

Cobalamin (Vitamin B12)

Cobalamin exists largely as hydroxycobalamin, methylcobalamin, and adenosylcobalamin, and these are found almost entirely in animal sources: Liver, kidney, beef, fish,

Chapter 100  Digestion and Absorption of Nutrients and Vitamins Salivary R Food-bound protein vitamin B12 Gastric parietal cell HCI

Stomach

IF Pancreatic zymogen cell

Pancreas

Trypsin

Liver Ileal enterocyte

Terminal ileum

Portal venous system Transcobalamin II Figure 100-18.  Steps in the chain leading to the binding of vitamin B12 to intrinsic factor (IF). Food-bound B12 is released by pepsin working at an acid pH and is picked up preferentially by salivary R protein in the stomach. Proteolysis of R protein by duodenal trypsin releases B12 for binding to IF. The subsequent binding and uptake of the IF-B12 complex occurs via a specific receptor-mediated process on the brush border membrane of ileal enterocytes. Vitamin B12 is released at an intracellular site, transported across the basolateral membrane, and there taken up by transcobalamin II for transport into the portal circulation.

eggs, and milk provide most of the cobalamin in a normal diet.48 Vegetables are almost entirely lacking in vitamin B12, and therefore the strict vegan’s dietary intake of cobalamins may be inadequate. About 10 to 20 µg is ingested per day in an average diet, and of this, about 1 to 2 µg/day is required to provide for normal needs.262 Because cobalamin absorption declines with age, persons older than 50 years of age are advised to take supplemental vitamin B12. Three types of binding proteins are concerned with the absorption of cobalamin: one in saliva, one in gastric juice, and one in the circulation.262 The vitamin is released by mastication and by gastric acid from the various dietary proteins with which it is associated (Fig. 100-18), after which the first specific binding protein secreted into saliva and also by parietal cells, the R protein (haptocorrin), takes up the free cobalamin and binds it with strong affinity. At intragastric pH values below 3, intrinsic factor (IF) has much weaker affinity for the vitamin than does R protein.263 It is only in the duodenum, where the R protein is hydrolyzed by pancreatic enzymes, that IF can bind the cobalamin that has been released.264 In humans, IF is secreted by parietal cells; in rats and mice it is secreted by chief cells. IF release is stimulated in response to the same agonists that stimulate acid secretion: histamine, gastrin, and cholinergic agonists. Unlike R proteins, which can bind a wide variety of cobalamin analogs, IF is much more selective and specific for cobalamin. It has been suggested that the nonspecificity of binding to the R protein that exists in plasma might offer an advantage in

binding potentially harmful compounds.265 IF has a very strong affinity for cobalamin and binds it tightly by enclosing the vitamin in its cup-like interior. This complex resists pancreatic proteolysis by undergoing molecular conformational changes and glycosylation, passes down the intestine to the terminal ileum, and there binds to specific receptors on ileal enterocytes, called cubulin-amnionless (AML). Distribution of ileal cubulin-AML receptors is patchy,266 and estimates suggest that there are about 300 to 400 receptors per enterocyte, or one per microvillus, located deep between microvilli.267 The number of receptors available determines how much vitamin can be absorbed; absorption doubles during pregnancy by a doubling of the number of available receptors.268 After binding to the receptor, the IF-cobalamincubulin-AML complex probably enters the cell intact by translocation (see Fig. 100-18). B12 accumulates in the mitochondria, and the complex is split at some point within the enterocyte. Free cobalamin leaves the base of the cell, where it is immediately bound to an ileal pool of trans­ cobalamin II, a 43-kDa protein that is synthesized in the enterocyte, and which transports it into the portal circulation. The transcobalamin II–cobalamin complex is essential for the transport of cobalamin into all cells in the body, where, after uptake by endocytosis, the cobalamin is enzymatically released. It is clear that this complicated series of events can be interrupted at a number of different points in the pathway. Lack of pancreatic proteolysis leads to a defect in the release of the vitamin from the R protein for subse-

1719

1720

Section X  Small and Large Intestine quent IF binding, lack of IF fails to provide the complex necessary for binding and absorption at the ileal mucosa, loss of ileal receptors prevents absorption, and defects within ileal enterocytes could prevent release of the vitamin into the circulation.

Other Water-Soluble Vitamins48,269

Thiamine Thiamine (vitamin B1) is distributed widely, but the only important dietary sources are seeds of plants. Germs of cereals, nuts, peas, beans, and lentils are major sources, and green vegetables and fruit are relatively poor ones. White flour or bread and purified rice have virtually no thiamine. Thiamine is water soluble and is readily lost in water used for cooking. Thiamine occurs almost entirely in phosphorylated form in animal foods, but it exists in its free form in foods of plant origin and enriched cereal. Dephosphorylation results from the action of phosphatases, present in the intestinal lumen, before it is absorbed. The rate of thiamine absorption approaches a limit as the dosage increases. This absorption is greatest in the jejunum and ileum and can be active or passive, depending upon concentration of the vitamin. With low intraluminal concentrations, thiamine is absorbed by a Na+-independent active process. Two human thiamine transporters, SLC19A2 and SLC19A3, have been cloned.270-273 At high concentrations (more than 8 mg in a single dose), thiamine is absorbed by means of passive diffusion.274 After absorption, thiamine is phosphorylated in the enterocyte and then transported out of the cell, possibly by the basal membrane Na+,K+ATPase. This exit step is inhibited by ethanol, which, in addition to poor intake, contributes to thiamine deficiency in alcoholics. Niacin Niacin (nicotinic acid) and nicotinamide are distributed widely in foods, but their availability in foods varies. About half of the North American and European dietary intake of niacin is supplied in meat and fish. The niacin content of legumes, however, is largely bound and unavailable, although it can be released by treatment with alkali; unfortunately, the food preparation methods widely employed in Asia and Africa, such as for maize, do not render the niacin available. Nicotinic acid can be synthesized in humans from tryptophan, 60 mg of tryptophan being required for the synthesis of 1 mg of niacin. The concept of the nicotinic acid equivalent has thus arisen, and foods lacking in niacin nonetheless can remain valuable in preventing pellagra because of high tryptophan content; such is the case with milk and eggs. Both nicotinic acid and nicotinamide are absorbed rapidly from the stomach and the small intestine. At low concentrations, absorption occurs through Na+-dependent facilitated diffusion, but at higher concentrations, passive diffusion predominates. Niacin occurs mostly in its coenzyme forms, which are hydrolyzed in the intestinal lumen by pyrophosphatase to nicotinamide; nicotinamide seems to be absorbed as such, without further hydrolysis in the gastrointestinal tract. Riboflavin Riboflavin (vitamin B2) is linked with phosphoric acid in most animal and plant tissues to form flavin mononucleotide and with adenosine monophosphate (AMP) to form flavinadenine dinucleotide. The richest dietary sources are liver, eggs, milk, green vegetables, and beer. Riboflavin also is synthesized by colonic bacteria, but its availability from the colon is uncertain. Cooking does not destroy much ribo-

flavin, but exposure to sunlight might. Riboflavin is presented to the mucosa in the form of coenzymes, so it is necessary for these to be hydrolyzed by intestinal phosphatases at the brush border membrane before active transport occurs into the cell. Studies suggest that riboflavin uptake occurs mainly in the proximal portion of the small intestine and involves a specialized Na+-independent carrier-­mediated system. This system is regulated by extracellular substrate levels and by specific intracellular protein kinase–mediated pathways.275,276 Once within the cell, rephosphorylation occurs. Biotin Biotin is so widely available that spontaneous deficiency states in humans have been described only rarely. Many yeasts and bacteria contain biotin and can provide a sufficient supply in normal foods. Liver, legumes, nuts, and vegetables are reasonable sources. Little is known about the mechanism of absorption of biotin in humans. Animal studies have indicated that most absorption of the free vitamin occurs in the jejunum and upper ileum by two mechanisms: a saturable facilitated diffusion process that operates at low concentrations and a nonsaturable linear diffusion process operating at higher luminal concentrations of the vitamin. Further studies have shown that the intestinal biotin transport system is also used by two unrelated micronu­ trients: the vitamin pantothenic acid and the metabolically important substrate lipoate. For these reasons, the biotin transport system now is referred to as the Na+-dependent multivitamin transporter (SMVT).277 It is unclear how cells regulate transport of the individual vitamin via this common transport system and how the substrate level of the individual substrate affects SMVT function in these cells. Pantothenic Acid Pantothenic acid usually is found as its calcium salt and is derived largely from animal tissues, especially liver, kidney, egg yolk, wheat germ, and peanuts. It is almost completely lacking in many processed foods but is not lost in normal cooking. The intestine also is exposed to bacterial sources of pantothenic acid. In the diet, pantothenic acid exists mainly in the form of coenzyme A, which is hydrolyzed to free pantothenic acid in the intestinal lumen before absorption. Hydrolysis is followed by transport of free pantothenic acid into the absorptive cells via the SMVT (see earlier).277 Pyridoxine Pyridoxine (vitamin B6) occurs in the diet in one of three forms: pyridoxamine phosphate, pyridoxal phosphate, and pyridoxine phosphate. Its presence is widespread in plant and animal tissues; cereals, peanuts, bananas, and liver are good sources. The phosphorylated form must be dephosphorylated before absorption can occur, and absorption is achieved by means of a membrane-bound alkaline phosphatase found in the intestinal brush border. All three forms of pyridoxine are freely absorbed by passive diffusion in the jejunum and ileum.

FAT-SOLUBLE VITAMINS

Vitamins A, D, E, and K are structurally different from one another, but all can be classified as polar, nonswelling, insoluble lipids. Although their chemical structures are known, the retention of a letter to signify their individuality is useful because each consists of a number of closely related compounds with similar properties (Table 100-11).48

Chapter 100  Digestion and Absorption of Nutrients and Vitamins β carotene

Vitamin A

Retinol R1 OH

CH3

R2

CH3

CH3

O R3

CH3 α-tocopherol

CH3 Vitamin E

CH3 7 dehydrocholesterol

O CH3

OH Ultraviolet light

CH2

CH

O

C

CH2

n

H

n = 4 to 13

CH3

Vitamin K2 (Menaquinones) O CH3 CH2

CH3

O

CH

C

CH2

CH2

CH2

CH2

CH2

CH2

3

H

CH2

Vitamin K1 (Phytomenadione)

OH

Vitamin K

Vitamin D

Figure 100-19.  Structural formulae of the fat-soluble vitamins A, D, E, and K. Vitamin D is shown as cholecalciferol (D3).

Table 100-11  Absorptive Mechanisms for Fat-Soluble Vitamins VITAMIN

Recommended dietary allowances*

MECHANISM OF ABSORPTION

A (retinol) D3 (cholecalciferol) E (α-tocopherol) K (phytomenadione [K1]284 and menaquinones [K2])

700-900 µg/day 10-15 µg/day 15 mg/day 90-120 µg/kg/day

Passive diffusion Passive diffusion Passive diffusion K1: carrier-mediated uptake K2: passive diffusion

*Daily Reference Intakes (DRI) were established by the Institute of Medicine between 1997-2001. They are quantitative estimates of nutrient intakes to be used for planning and assessing diets for healthy people. The DRIs include both recommended intakes and tolerable upper intake levels. The RDAs (Recommended Dietary Allowances) are a component of the DRIs and are defined as the daily intake of a nutrient considered sufficient to meet the requirements of 97% to 98% of adults.

Vitamin A

Vitamin A (retinol) is found in the diet in milk and milk products, egg yolk, and fish oils. Carotenoids are defined by their chemical structure: The hydrocarbon carotenoids are known as carotenes, and the oxygenated derivatives of these hydrocarbons are referred to as xanthophylls.

Beta carotene, the principal carotenoid in carrots, consists of two conjoined molecules of retinol and is a precursor of the active vitamin (Fig. 100-19). There are many other carotenoids in the diet, but these contain only one retinol molecule. Carotenoids are found predominantly in green vegetables and carrots, and in the United States and Europe

1721

1722

Section X  Small and Large Intestine these sources account for about half the dietary intake. Retinol and the carotenes are stable in normal cooking. Lycopene is the carotenoid that gives tomatoes and other red fruits and vegetables their color. Although it is similar in structure to carotene, lycopene lacks a β-ionone ring structure and thus cannot form vitamin A or retinal. Its biologic effects in humans are attributed to mechanisms other than those for vitamin A, Retinol and carotene are absorbed in the small intestine, carotene less readily than retinol.278,279 Dietary retinal (vitamin A aldehyde) esters first are hydrolyzed to retinol in the intestinal lumen before they are absorbed into the intestinal mucosa. Carotenes, by contrast, are converted into retinol, primarily in the enterocytes. Before entering the mucosal layer of the intestine, however, carotenes first are solubilized into micellar solution along with other fatsoluble compounds. Hence, bile salts and dietary fats are needed for absorption of carotenes in the upper region of the small intestine. The biosynthetic process for retinol from carotene in the enterocytes involves two soluble mainly cytosolic enzymes. The first enzyme is responsible for the oxidative split of the beta carotene molecule to produce the cleavage product of retinal, which is then reduced to retinol by a second enzyme that depends on reduced nicotinamide adenonine dinucleotide (NADH). Transport across the apical membrane appears to occur by passive diffusion, but facilitated diffusion cannot be excluded. Free retinol (from retinyl ester and carotenes) in the mucosal cells is re-esterified mainly with palmitic acid before incorporation into chylomicrons, which is how vitamin A mostly leaves the mucosa.

Vitamin D

Vitamin D comprises a group of sterols that have antirachitic properties, but the only two nutritionally important members are vitamins D2 (ergocalciferol) and D3 (cholecalciferol). Both are produced by ultraviolet irradiation of their precursor sterols, ergosterol and 7-dehydrocholesterol, respectively (see Fig. 100-19). Ergosterol, found in fungi and yeasts, is an unusual constituent of the normal diet, whereas vitamin D3 is the major dietary form. Vitamin D3 is found in a restricted range of foods, predominantly the oils of fatty fish, which themselves ingest it in plankton found near the surface of the sea.48 Human breast milk contains sufficient vitamin D to prevent rickets, but cow’s milk is a poor source of this vitamin. Most of a person’s requirement for vitamin D, however, is supplied by endogenous synthesis in the skin during exposure to sunlight, and dietary intake becomes critical only when such exposure is inadequate (see Table 100-11). As with vitamin A, vitamin D absorption occurs by simple passive diffusion in the small intestine.280 Bile salts are unnecessary, but luminal pH influences absorption. Absorption is reduced at neutral pH and increased in an acidic milieu.281 Most absorbed vitamin D passes into the lymphatics unchanged in chylomicrons. Both vitamins D2 and D3 are converted in the liver to 25-hydroxycholecal­ciferol, which is then further converted to 1-α,25-dihydroxycholecalciferol (1,25(OH)2D3) by renal 1-hydroxylase. 1,25(OH)2D3 functions to further promote intestinal calcium absorption.

Vitamin E

Vitamin E is still seeking a role in humans. It comprises a group of eight or so tocopherols, the most potent of which in animals is α-tocopherol (see Fig. 100-19).106 It is distrib-

uted widely in the diet: vegetable oils, cereals, eggs, and fruit are good sources. Margarines are particularly rich in vitamin E, and breast milk contains much more than cow’s milk. Although a variety of diseases can result from deficiency of vitamin E in a number of animal species, it has proved difficult to ascribe a human disease to vitamin E deficiency. Vitamin E is absorbed passively across the intestinal mucosa.282 The ester form, in which many vitamin preparations are presented, is hydrolyzed by pancreatic and/or duodenal esterases before absorption, but the ester can be absorbed intact.283 After incorporation into micelles, vitamin E is transported into enterocytes and incorporated into chylomicrons for transfer into lymphatics.

Vitamin K

Vitamin K is found in two forms: K1, derived largely from plants, is phytomenadione; K2 comprises a group of bacteria-produced compounds, the multiprenyl menaquinones. K1, the major dietary form, is found in green vegetables, but beef liver is another good source. K2 is produced by colonic bacteria, and although some K2 may be absorbed from the colon, this alone is an inadequate source if K1 absorption is impaired. Absorption of K1 from the small intestine is dependent on luminal bile salts, and uptake is achieved by a carrier-mediated process,284 whereas K2 absorption is entirely passive.285

MINERALS AND TRACE ELEMENTS Various divalent ions are essential nutrients; some are absorbed in milligram amounts and are major constituents of the body; others are necessary only in trace amounts. Iron, calcium, magnesium, phosphorus, and sulfur are in the former category, and specialized absorptive mechanisms are concerned with their assimilation.

CALCIUM

Milk and other dairy products are the most valuable sources of calcium, accounting for up to 75% of dietary calcium intake; cereals, legumes, and other vegetables contribute lesser amounts. Phytic acid or oxalate in vegetables binds strongly to calcium, thus reducing its availability. Dietary fiber also binds calcium and can interfere with its absorption; by contrast, dietary lactose enhances its absorption.286 Fractional, or true, absorption is only about 20% to 30% of total dietary calcium, the remainder being excreted in stool. Absorption of calcium across the intestinal mucosa is achieved by two parallel processes: an active, transcellular transport process, which dominates with lower levels of calcium intake, and a passive, paracellular diffusive process, which becomes more important at higher levels of calcium intake.287-289 Under normal dietary conditions, the duodenum is the major site for active calcium transport, whereas passive, paracellular transfer occurs throughout the small intestine. Despite this localization of the active transport site, quantitatively more calcium may be absorbed in the jejunum and ileum than in the duodenum because of the relative amounts of time luminal contents spend in these regions of intestine. The human jejunum absorbs calcium faster than does the ileum, and absorption rates in both regions are increased by treatment with vitamin D.290 The paracellular route, via the tight junctions, may be capable of modifying calcium transport because passive transport increases in response to treatment with vitamin

Chapter 100  Digestion and Absorption of Nutrients and Vitamins 1,25-(OH)2 vitamin D

Ca2+ Protein transcription

ADP ATPase ATP

Ca2+ Ca2+

Ca2+- Calbindin

Figure 100-20.  Mechanisms of calcium transport across the intestinal epithelium. A paracellular route allows bidirectional flux. Transport into the epithelial cell occurs via specific channels down an electrochemical gradient. A critical step is the binding to calbindin, which then presents calcium for export via a calcium-dependent adenosine triphosphatase (ATPase) on the basolateral membrane. Each of these processes appears to be influenced by 1,25-(OH)2 vitamin D, although its maximal effect is on synthesis of fresh calbindin. ADP, adenosine diphosphate; ATP, adenosine triphosphate.

D.291 Furthermore, there is evidence suggesting that tight junctional permeability increases during sugar transport, and this might provide another mechanism for control of paracellular transport.292 The transcellular route involves transport across the apical membrane, transfer across the cytoplasm, and exit across the basolateral membrane (Fig. 100-20). Entry pro­ bably occurs via specific non–voltage-gated calcium channels in the apical membrane and down the prevailing electrochemical gradient. Within the cytoplasm, binding to a calcium-binding protein, calbindin D9K (or calbindin 3), is a key step.293 Maximal transport rates correlate closely with calbindin concentrations. This protein, present in concentrations of 0.1 to 0.2 mmol, must rapidly take up the calcium entering the cell because intracellular free calcium concentrations are carefully maintained at very low values (about 10−7 M). Transient rises in intracellular calcium act as key second messenger signals for secretory responses in enterocytes. Absorbed calcium thus is pre­ sumably segregated from the calcium concerned with cell signaling, and calbindin D9K plays a vital role here by bringing calcium to the transporter at the basolateral membrane.294 Another calcium-binding protein, calbindin D28K (CALB1) is induced by vitamin D and binds four calcium ions compared with calbindin D9K, which binds two calcium ions. An active mechanism is then necessary to drive calcium uphill against the electrochemical gradient, for which a calcium-dependent ATPase is responsible.293 Calcium arrives at the basolateral pole bound to a site at the cytoplasmic aspect of the calcium-dependent ATPase that spans the basolateral membrane. There follows a phosphorylation-induced change in the conformation of the calcium-dependent ATPase, and the calcium ion is extruded through the channel formed by the enzyme transmembrane elements.295 The rate-limiting step in the absorption process of calcium is the intracellular calbindin concentration, which is regulated by a metabolite of vitamin D, 1,25-dihydroxyvitamin

D (1,25-[OH]2D), produced in the kidneys from 25-hydroxyvitamin D (25-[OH]D); the latter is converted from absorbed vitamin D by the liver.288 The vitamin also has a modest effect on the calcium-entry step and enhances activity of the basolateral calcium ATPase. Up-regulation of the calbindin gene in response to vitamin D occurs largely in villus cells.296 Some evidence supports colonic absorption of calcium, which also can be enhanced in response to vitamin D.297 Although normally the colon can account for only up to 7% of total calcium absorption, it becomes an important route for calcium absorption in patients with short bowel syndrome.298 Active duodenal calcium absorption is increased in calcium deficiency states and reduced in calcium repletion states. Increased production of the active 1,25-(OH)2 vitamin D metabolite in response to a small drop in plasma calcium concentration is responsible for increasing calcium absorption, and this change occurs within a day of changing from a high-calcium to a low-calcium diet.299 This same mechanism is likely to be the cause of the enhanced calcium absorption seen during late pregnancy and lactation. At birth, the active vitamin D–dependent absorptive mechanisms already are present in the human duodenum. Ingestion of large amounts of calcium, together with lactose, in breast milk ensures adequate intake at this critical stage of life. Calcium absorption declines with age, but this might result in part from a lack of vitamin D, or decreased responsiveness of the intestine to vitamin D.286

MAGNESIUM

An average diet provides about 300 to 500 mg of magnesium per day from a wide range of vegetables. Magnesium absorption has been less thoroughly investigated than that of calcium, but it seems likely that the mechanisms involved are different. In contrast to calcium absorption, magnesium absorption in the basal state is greater in the human ileum than in the duodenum or jejunum.290 Jejunal absorption of magnesium is increased by vitamin D, whereas ileal absorption is not. Ileal transport involves both a paracellular, diffusive pathway and a transcellular, carrier-mediated, saturable process.300 There is some competition from calcium for the diffusive pathway but not for the saturable, presumably carrier-mediated process.300 Quantitatively, magnesium fluxes across ileal mucosa are several-fold greater than those for calcium, but the overall efficiency of magnesium absorption after normal dietary intake ranges from 21% to 27%.

IRON

Meat-eating, affluent societies ingest about 20 to 30 mg of iron per day, largely as myoglobin or hemoglobin. Vegetarian societies in poor countries ingest much less than this in wheat and vegetables, and iron in these foods is less readily available for absorption. A careful balance of absorption and loss is maintained in normal adults: each is about 1 mg/day. Developing children and adolescents need to absorb about 0.5 mg/day more, to build up total body iron to adult values. Iron is present in breast milk in the form of lactoferrin, for which a specific brush border membrane receptor has been demonstrated301,302 that facilitates iron absorption in neonates. During reproductive life, normal women need to compensate for menstrual losses, which are on the order of 5 to 50 mg per month and approximately 500 mg for each pregnancy. Because dietary intake often markedly exceeds the body’s need for iron, it is necessary to absorb only a small portion

1723

1724

Section X  Small and Large Intestine of that ingested. Overall, there is a positive and linear relationship between the amount ingested and that absorbed, but the fraction absorbed decreases as more is taken in.303 Total body iron content is regulated by controlling the level of iron absorbed from the diet. Under normal circumstances, only about 10% (1 to 2 mg/day) of dietary iron is absorbed. Most absorption occurs in the proximal small intestine, and the ferrous (Fe2+) form is absorbed better than the ferric (Fe3+) form. The latter is insoluble at pH values greater than 3, and gastric acid and some sugars and amino acids render it more available for absorption. The presence of some anions, such as oxalate, phosphate, and phytate, precipitate iron out of solution and reduce its absorption. The presence of bile enhances absorption, but the mechanism of this enhancement is unclear. Dietary iron is predominantly found in the ferric form, but Fe3+ is highly insoluble under physiologic conditions. Therefore, during uptake, Fe3+ is converted to the Fe2+ form at the apical cellular membrane before it attaches to an acceptor protein in the membrane. The ability of intestinal mucosa to reduce Fe3+ to Fe2+ has been documented,304 and a ferrireductase activity has been characterized for the intestinal Caco-2 and HuTu-80 cell lines.304,305 A functional role for Fe3+ reduction in iron transport across the brush border is implicated by the fact that inhibition of ferrireductase activity reduces Caco-2 cell apical iron uptake. Increased ferrireductase activity correlates with enhanced iron uptake induced by iron deficiency and hypoxia.305,306 Iron uptake into the body occurs at two interfaces of the intestinal epithelium: the apical and basolateral plasma membranes.307 The apical plasma membrane of the differentiated enterocyte is specialized for transport of heme and ferrous iron into the cell. Three major pathways of iron transport across the apical membrane have been proposed. The best-characterized pathway is via the divalent metal transporter 1 (DMT1, also known as Nramp2 or DCT1),308 which is located in the brush border membrane, primarily in the villus tip cells. There are two splice variants of DMT1 that yield two messenger RNAs (mRNAs): one containing an iron-responsive element (IRE) termed DMT1 (IRE) mRNA, and another without an IRE designated DMT1 (non-IRE) mRNA. DMT1 is a proton symporter that transports ferrous iron and other divalent metals from the intestinal lumen into the enterocyte (Fig. 100-21). It is up-

Fe 2+

Fe 2 + DMT1 Fe 2 +

Fe-BP F e 2+

Fe 3 + Heme

Fe 2+ + Transferrin

Heme-Oxygenase Heme

Figure 100-21.  Mechanisms of iron transport in the intestine. A small amount of inorganic iron may pass through the paracellular route. Inorganic iron (Fe3+) is converted into its ferrous form (Fe2+) at the brush border membrane before transport into the cell. Heme iron is transported into the cell by a separate mechanism. Within the cell, one or more ironbinding proteins take up iron and transfer it to the basolateral membrane for delivery across the membrane and subsequent binding to transferrin. DMT1, divalent metal transporter 1; Fe3+, ferric; Fe2+, ferrous; Fe-BP, ironbinding protein.

regulated during iron deficiency and down-regulated in cases of iron excess. In order of substrate preference, DMT1 can mediate import of Fe2+, Zn2+, Mn2+, Co2+, Cd2+, Cu2+, Ni2+, and Pb2+. The idea that the transporter responsible for dietary iron absorption recognizes other divalent cations agrees well with observations that Zn2+, Mn2+, Cd2+ and Cu2+ all can inhibit this process. DMT1 mRNA is found in many different tissues, but the protein and its mRNA are most abundant in the proximal duodenum, with decreasing absorption along the distal axis, consistent with a function in intestinal iron absorption.309 Iron depletion results in increased DMT1 mRNA levels in the intestine, which suggests that IREs in its 3′ untranslated region bind and stabilize the DMT1 mRNA.309 Although the major route for dietary iron absorption likely is mediated by DMT1, this transporter is found only in the apical surface of enterocytes. Thus, other factors must be involved in the transfer across the intestinal epithelium. DMT1 may be involved in the pathogenesis of hereditary hemochromatosis. HFE is the gene responsible for hereditary hemochromatosis. HFE protein is found in the crypt cells of the duodenum associated with β2-microgobulin and transferrin receptor. It is hypothesized that HFE protein facilitates transferrin receptor–dependent iron uptake into crypt cells and that mutant HFE protein might lose this ability, leading to a relative iron deficiency in duodenal crypt cells. In turn, this might lead to an increase in the expression of DMT1, resulting in increased iron absorption in hereditary hemochromatosis. Up-regulation of DMT1 expression has been confirmed in the HFE-knockout mouse and in humans with hereditary hemochromatosis (see Chapter 74).310 Iron also can be absorbed in the form of heme iron (in hemoglobin and myoglobin), which is readily transported across the brush border of the enterocytes as an intact heme moiety. It is the presence of globin, which increases the absorption of iron in this form. Heme-carrier protein 1 (HCP1) has been isolated from the mouse duodenum. This large hydrophobic transporter is present in the apical membrane during iron deficiency and in the cytoplasm during iron overload, making it a putative candidate for heme-iron transport.311 Once within the cell, heme is broken down by heme oxygenase, and the iron is released into the nonheme pool for incorporation into intracellular ferritin and export out of the cell.312,313 Another iron absorptive pathway has been proposed involving intestinal mucins, a 56-kDa protein designated mobilferrin, an integrin, and a ferric reductase. This pathway is regulated and depends on metabolic energy, and it appears to be encouraged by nonessential fatty acids.305,314,315 A small proportion of the iron crossing the mucosa uses a paracellular route by simple diffusion (see Fig. 100-21). Once within the enterocyte, the expression of the iron-storage protein ferritin is regulated by the intracellular concentration of iron. Ferritin synthesis increases when iron is present in excess and decreases when iron level is low. Iron transport across the basolateral plasma membrane of villus enterocytes involves at least two proteins: a ferroxidase called hephestin, which is associated with the basolateral membrane, and a basolateral iron transporter termed ferroportin 1 (FPN1), iron-regulated protein 1, or metaltransport protein 1. These proteins may work in close conjunction with each other, with ferroportin 1 transporting ferrous iron out of the cell and hephestin oxidizing the ferrous iron to ferric iron, which permits the avid incorporation into circulating apotransferrin (apoTF).307 The transferrin receptor, in combination with the hemochromatosis

Chapter 100  Digestion and Absorption of Nutrients and Vitamins protein (HFE), allows the binding of apoTF-bound iron and its reuptake back into intestinal cells.316 FPN1 is a membrane-bound protein containing iron-responsive elements that is up-regulated during iron deficiency and down-regulated during iron excess. Hepcidin, which is produced in the liver, might influence enterocyte iron transport by binding to FPN1, resulting in the internali­ zation and degradation of FPN1, and so reducing iron absorption.317 Intestinal iron absorption is regulated in at least three ways. The longstanding concept of mucosal block is based on the observation that after a large oral iron dose, enterocytes do not absorb additional iron for several days. A second regulatory mechanism of iron absorption is termed the stores regulator. It acts on a pathway that facilitates a slow accumulation of nonheme dietary iron. The functioning of the stores regulator is of great physiologic importance, because it prevents iron overload after ensuring iron needs are met. The exact molecular mechanism of the stores regulator has not been established, but it has been proposed to involve soluble factors such as transferrin-bound iron, serum ferritin, serum transferrin, or hepcidin. The erythropoietic regulator is a third regulatory mechanism that adjusts intestinal iron absorption in response to the demands of erythropoiesis, independent of body iron stores. This regulator must signal directly between the hematopoietic bone marrow and the duodenum. Although the erythropoietic regulator has been proposed to be a soluble component of the plasma, it is distinct from the stores regulator. This is evidenced by the rate of iron uptake in anemic persons that is much greater (20 to 40 mg/day) than could be produced by the stores regulator alone. The stores and erythropoietic regulators are circulating factors that maintain iron homeostasis of the entire organism. Within individual cells, the iron regulatory proteins (IRPs) IRP-1 and IRP-2 act to control iron availability by translational control of the synthesis of proteins such as transferrin and ferritin. IRPs are cytoplasmic RNA-binding proteins that function on mRNAs that contain IREs. Functional IREs are present in the 3′ untranslated region of mRNAs for transferring in one of the two isoforms of DMT1 (DMT1 IRE) and in the 5′-untranslated region of mRNAs for ferritin, ferroportin 1, mitochondrial aconitase, and the erythroid-specific form of δ-aminolevulinic acid synthase. IRPs functionally connect intracellular iron availability with cellular iron utilization; IRP function also can be altered by inflammation and oxidative stress. Basolateral iron uptake from the plasma by cryptal enterocytes plays an important role in sensing body iron stores. Indeed, there is considerable evidence that the iron concentration within the cryptal enterocyte is an important determinant of iron absorption. The mechanism by which the intracellular iron concentration can respond to body iron needs is poorly understood. It is clear, however, that cells in the crypts of Lieberkühn always express transferrin, and the endocytic mechanism imparts information about body iron storage based on plasma transferrin saturation. It is also well recognized that acute changes in body iron status, whether overload or deficiency, are not reflected by changes in iron absorption for a period of two to three days. This lag response time probably correlates with the migration time for proliferating cells in the crypts to differentiate and migrate into functional, mature enterocytes of villi. Thus, the luminal epithelial cells may be preprogrammed in the crypts based on body iron needs. This preprogramming would, in turn, initiate synthesis of iron transport proteins that are required for dietary iron uptake across the membranes of the villus enterocyte.

TRACE ELEMENTS

The importance of zinc, copper, and iodine in human nutrition has long been recognized, and they have received increasing attention in recent years as their roles in defined enteral and parenteral forms of nutrition have been demonstrated. The value of selenium also has been emphasized, and the need for manganese and chromium is receiving attention. Despite this interest, surprisingly few systematic studies of their absorption have been undertaken.

Zinc

Zinc is present in the body in about half the amount of iron (about 2 g), and largely in a wide variety of enzymes. It also plays important roles in maintaining configuration of gene transcription proteins and the integrity of membranes. It is found particularly in meat, shellfish, cereals, and legumes. Daily requirements are approximately 12 to 15 mg/day in adults. Persons who consume a low-energy diet might take in marginal amounts of zinc, and requirements are increased during pregnancy and lactation. Absorption is impaired by phytates and oxalates in the diet through their chelating properties, and food processing can render zinc less available for absorption.48,318 The protein content of a diet is positively correlated with zinc absorption, likely because of amino acids or small peptides that facilitate enterocyte uptake of zinc. Overall, the efficiency of zinc absorption from regular diet is 15% to 35%. There is enterohepatic circulation of zinc, and reabsorption appears to be maximal in the distal small intestine.319 Studies with vesicles of porcine jejunal brush border membranes have identified two uptake processes: an active saturable, carrier-mediated process (which dominates at low or normal intake) and a nonsaturable, diffusive process (which contributes more to absorption at higher intake).320 The relative importance of each is not known. There are nine members of the ZNT family of zinc transporters, with ZNT1, 2, and 4 occurring mainly in the villus.321,322 ZNT1 is a ubiquitously expressed protein that is present in the villi of the proximal small intestine. In response to manipulation of dietary zinc, however, expression in rats was increased in response to zinc supplemen­ tation but not to zinc restriction.323 These and other observations have led to the consensus that ZNT1 functions mainly as a zinc exporter and might play a role in zinc homeostasis for zinc acquisition and elimination under conditions of excess zinc.321 ZNT2 and ZNT4 are involved in intracellular transport of zinc by the enterocyte. The role of metallothionein (MT), an intracellular metalbinding protein, in the regulation of zinc absorption, particularly in conjunction with the zinc transporters, also remains unclear. This binding protein may be concerned with zinc absorption because changing dietary loads of zinc rapidly affect protein synthesis and alter binding capacity.324 Subjects on a low-zinc diet respond by decreasing their urinary excretion rate of zinc and by increasing its absorption rates.319,325 Absorption increases in pregnancy and during lactation.326 In experiments with knockout and transgenic mice, the rise in serum zinc after a single dose of zinc was much greater than in the control animals. By contrast, the serum zinc response of the MT transgenic animals was blunted compared with that of the control animals. The expression of ZNT1 also was measured and found to be directly related to serum zinc levels but unaffected by MT levels.327 Thus, MT might function in cellular responses to limit free zinc concentrations within narrow ranges and function as a zinc pool.321,327 Another transporter potentially involved in zinc and other metal uptake is DCT1, a transmembrane polypeptide

1725

1726

Section X  Small and Large Intestine that is found in the duodenum in the crypts and lower villi and may be available for the uptake of several metal ions.322 The ZIP (Zrt-, Irt-like protein) family of proteins also are believed to be involved in zinc transport.328 ZIP4 and ZIP5 likely exist on the apical side and basolateral membrane of the enterocyte, respectively, and may be responsible for zinc transport into the circulation.329

Copper

Copper is found in green vegetables and fish, and the average Western diet provides 1 to 3 mg/day, which is adequate for a daily need of about 1 mg/day. Dietary copper is absorbed very efficiently from the stomach and small intestine, especially the duodenum. Although the precise mechanisms involved in copper absorption remain incompletely known, within physiologic ranges of intake, absorption is probably by active transport. Competition for absorption between copper and zinc or iron may be demonstrable with large doses of these elements but not with normal dietary intakes.330 The uptake of copper might increase in pregnancy.331 Active transport and passive diffusion are both responsible for copper absorption in humans. A putative high affinity protein copper transport protein, denoted hCtr1, has been identified by functional complementation of the respiratory defect in yeast cells defective in copper transport because of inactivation of both the CTR1 and CTR3 genes.332 Human Ctr1 is a 190–amino acid protein with three transmembrane domains and significant homology to yeast Ctr1 and Ctr3, suggesting that mammalian high-affinity copper transporters may have evolved from Ctr1 and Ctr3. RNA blotting analysis has demonstrated that that hCtr1 is expressed in all organs and tissues examined: Liver, heart, and pancreas exhibited the highest levels of expression; intestine had intermediate levels of expression; and expression in brain and muscle was low. Whether hCtrl1 plays an important role in copper uptake into intestinal mucosal cells has yet to be firmly established.333,334 Two putative low-affinity mammalian copper transporters, hCtrl2 and Nramp2, have also been identified. It is unclear what role hCtr2 plays in copper homeostasis, because its mRNA levels are highest in the placenta and very low in liver, intestine, and colon.332,334 The Nramp2 protein also has been identified as a proton-coupled metal ion transporter that transports a broad range of metal ions.309 Acting as a permease or by endocytosis, Ctr1 delivers Cu2+ within cells.335 The mechanism for reduction of copper ion before uptake remains unknown. If uncontrolled, this pool of cuprous ions could lead to generation of reactive oxygen species; however, very few, if any, free copper ions exist in the cytoplasm. The delivery of copper to target cuproenzymes depends on an elegant metallochaperone system. Several cytoplasmic chaperones have been described (Atox1, CCS [copper chaperone for Cu, Zn superoxide dismutase], and Cox17) as well as membrane-associated copper-transporting ATPases (ATP7A and ATP7B).336 The copper-transporting Menkes ATPase ATP7A (MNK) is responsible for copper export from the enterocyte, and may be defective in patients with Menkes disease in whom copper accumulates in intestinal cells.337,338 Once entering the plasma, copper is bound with albumin and histidine in the portal blood and rapidly deposited in the liver, where hCtr1 may play a role in this process. Ceruloplasmin, a major copper-containing protein in plasma, is synthesized in the liver with incorporation of

copper by the Wilson disease protein, which has a high homology with MNK and is defective in Wilson disease patients who suffer from copper accumulation in the liver (see Chapter 75).

Iodine

Iodine is present in varying amounts in a wide range of foods, depending on the soil content in the region where animals were reared and vegetation was grown. Seafood is particularly rich in iodine. Iodine is absorbed largely as inorganic iodide, but some iodine also is transported as amino acid complexes.339

Selenium

Selenium is found predominantly in association with amino acids, and about 60% of dietary selenium is absorbed. Selenium deficiency states have been reported from China (Keshan disease), where there is very little selenium in soil and water,340 but not in New Zealand, where intake is equally sparse.341 Absorption of selenium occurs rapidly when it is associated with amino acids, as in selenomethionine, probably by active transport mechanisms operative for the amino acid.342 Inorganic selenium is absorbed more slowly, possibly by simple diffusion.

Others

The mechanisms underlying the absorption of other trace elements, including manganese and chromium, are largely unknown.339 Deficiencies of trace elements are rare in normal persons, even in those with poor protein and calorie intake. Exceptions occur when local geographic availability is suboptimal, as can occur with iodine and possibly with selenium.

ADAPTATION TO CHANGES IN NEED OR LOAD One of the most fascinating aspects of intestinal function is the phenomenon of adaptation. Two specific forms of intestinal adaptation have been identified in the intestine: mucosal hypertrophy6 (leading to a global increase in absorption of all nutrients) and an increase in specific transport mechanisms induced in response to specific dietary needs or availability.343

MUCOSAL HYPERTROPHY

Resection of more than 50% of the human intestine results in increased fecal nitrogen losses, which subsequently slowly return toward normal, thus implying that mucosal adaptation has occurred. Normalization is explained largely by hypertrophy of intestinal mucosa, which manifests with increases in the number of villus enterocytes and in villus height without obvious increase in the absorption rate per individual cell.153 Absorption increases for all nutrients, and absorptive capacity may be enhanced up to five-fold in response to intestinal resection. Jejunal adaptation following ileal resection appears to be less efficient than ileal adaptation in response to jejunal resection. While hypertrophy in response to resection is the bestcharacterized example of adaptation, other causes also have been discerned, at least in experimental animals. Thus, during lactation and pregnancy,344 in diabetes,345 and in the physiologic response to extreme cold,346 hypertrophy is evident, but this may be a result, at least in part, to the hyperphagia that accompanies these conditions.

Chapter 100  Digestion and Absorption of Nutrients and Vitamins High-CHO diet

Amylase output

The mechanisms by which hypertrophy occurs have been the subject of much study. Signals of adaptation might relate to various hormone levels, transcription factors, ATP levels, or changes in the concentration of luminal solutes.347 The signals and mechanisms of the adaptive process may be different for the jejunum and ileum, as well as in the crypt and villus tip, explaining the sitespecific alterations and differences between crypt and villus enterocytes.347,348 The presence of luminal nutrition is a major stimulus to growth,6 in addition to intraluminal pancreaticobiliary secretions349; certain peptide hormones also have been implicated, particularly enteroglucagon and glucagon-like peptides.350 Gastrin, CCK, and other trophic factors (e.g., epidermal growth factor and insulin-like growth factors) display trophic effects on the gastrointe­ stinal tract,351-353 but it is uncertain whether these factors act as local paracrine mediators or as circulating hormones. Polyamines are other important local mediators of mucosal hypertrophy,354 because epithelial production of polyamines follows intestinal resection, and inhibiting their synthesis prevents the hypertrophy usually associated with resection. Polyamines also might play a role in maintaining normal mucosal structure, because their mucosal level in the intestines of experimental animals decreases rapidly in response to a 24-hour fast and increases within a few hours of refeeding.355 Although certain prostaglandins have been shown to enhance cell proliferation in the stomach and intestine, their role in adaptation is uncertain.356

Low-CHO diet

–3

–2

–1

Meal

1

2

3

Hours Figure 100-22.  Pancreatic enzyme adaptation to dietary manipulation. Diagram of typical pancreatic amylase outputs in normal subjects fed either a high- or a low-carbohydrate (CHO) diet for two weeks. Greater amylase secretion rates occur, both during the interdigestive phasic periods (↓) between meals and in response to a standard meal, in subjects given a high-CHO diet.

8

It has long been known that the digestive capacity of pancreatic juice can be altered by changes in nutritional intake, but it is now clear that specific responses occur after different types of manipulation of the diet.357 A high-protein diet enhances proteolytic enzyme production; a highcarbohydrate diet enhances amylase secretion; and a highfat diet stimulates lipase secretion (Fig. 100-22). In part, these changes appear to depend on specific polypeptide hormone release. Prolonged administration of cerulein (an analog of CCK) stimulates trypsinogen and inhibits amylase secretion,357 whereas secretin stimulates lipase secretion (Fig. 100-23).358 Insulin released from pancreatic beta cells, in response to carbohydrate ingestion, appears to be involved indirectly in enhancing amylase secretion.357 A high-fat diet also induces increased capacity to secrete gastric lipase, but the mechanisms involved are not known.359 The underlying molecular biologic events that lead to pancreatic adaptation have been studied, and, as might be expected, responses depend on the period over which a dietary stimulus is applied. Responses to short-term stimulation, as after a single meal, appear to depend on enhanced translation of mRNAs for enzymes.357 Protein synthesis increases within the first two hours of stimulation of hormones and appears to be a result of translational events; however, prolonged stimulation—over several days—increases mRNA production by increased transcription, leading to enhanced biochemical commitment to enzyme secretion.360 A single stimulus after a prolonged period of high intake of protein, therefore, results in much greater proteolytic enzyme output than in persons whose protein intake is low. The polypeptide hormones secretin, CCK, and possibly insulin, liberated in response to a meal, not only cause immediate release of pancreatic enzymes but also stimulate

Absolute rates of lipase synthesis (cpm x 10–2/μg DNA)

SPECIFIC REVERSIBLE ADAPTATION Secretion of Pancreatic Enzymes

6 Secretin

4

Cerulein

2

Saline

3

6

12

24

Infusion time (hr) Figure 100-23.  Pancreatic enzyme responses to prolonged hormone stimulation in rats. The lipase synthesis rate rose in response to secretin (16 units/kg/hr, orange line), but not to the cholecystokinin (CCK) analogue, cerulein (0.25 µg/kg/hr, blue line) or saline (purple line), infused for up to 24 hours. (From Rausch U. Lipase synthesis in the rat pancreas is regulated by secretin. Pancreas 1986; 1:522-8.) cpm, counts per minute.

gene expression over the longer term and thus increase secretory capacity.

Mucosal Responses

Adaptive responses to changes in dietary intake influence mucosal digestive and absorptive processes. Activity of the

1727

Section X  Small and Large Intestine disaccharidase enzymes sucrase and maltase increases in response to high carbohydrate intake over several days but not to manipulation of protein intake.361 Sucrase levels increase first in crypt cells about 24 hours after refeeding sucrose following a period of starvation. Thus, synthesis of the disaccharidases is stimulated, but their breakdown is also diminished. Conversely, lactase is an enzyme that appears not to respond to manipulation of dietary intake of lactose.362 Absorptive function also adapts to dietary manipulation.205 It has been widely assumed that there is a considerable reserve of absorptive function under normal circumstances, but Diamond and colleagues argued eloquently that it would be inefficient and costly in biosynthetic energy for the intestine to have a large spare capacity.205,363 Furthermore, a fairly close match has been demonstrated between absorptive capacity for many nutrients and dietary load. There is a clear need, therefore, for adaptation to occur in response to changes in load, and there is good evidence to suggest that most nutrients regulate their specific mucosal transporter. Three types of adaptive responses are discernible in the mucosa (Fig. 100-24).205 In the first, as exemplified by sugars, peptides, and nonessential amino acids, transport activity rises in response to increased dietary loads. Experimental animals fed diets high and low in glucose increase or decrease their maximum capacity for glucose transport, respectively, over a two-fold range, probably by changing the number of transporters. An analogous response to increased dietary load is seen with protein ingestion, in which peptide transporters and some amino acid transporters are increased. In the second type of mucosal response, as exemplified by a number of vitamins and trace elements, absorptive mechanisms are switched on by low dietary loads and switched off by a large load. Here absorption is enhanced

A

Peak

Relative transporter activity

1728

C

Basal

B

Deficient

Average

Excessive

Nutritional intake Figure 100-24.  Diagram of three types of adaptive responses of intestinal transporters to variation in nutritional intake. The first type of response (A) characterizes hexose and nonessential amino acid transport; the second type of respone (B) typifies elements such as iron and calcium and some vitamins; the third type of respone (C) is a mixed pattern seen with some essential amino acids. (Adapted from Ferraris RP, Diamond JM. Specific regulation of intestinal nutrient transporters by their dietary substrates. Annu Rev Physiol 1989; 51:125-41, with permission.)

in nutrient deficiency but inhibited with nutrient excess, when potentially toxic effects may result. A third and mixed pattern is seen with other types of nutrients, as with essential, predominantly basic, and neutral amino acids, when absorption, presumably in an effort to ensure adequate intake, is enhanced at very low dietary levels. Absorptive mechanisms are at their minimum with average dietary intake, but they rise as dietary ingestion increases above that range.205 The signal for up-regulation of brush border membrane glucose absorption is glucose itself, although an actively transported but unmetabolized sugar also stimulates glucose uptake.363 Fructose stimulates its own absorption by a mechanism separate from the one that stimulates glucose transport.364 Regulation of mucosal transport of acidic amino acids and imino acids follows the same pattern as that for sugars; that is, an increase in dietary load up-regulates transport. There is an interesting cross-induction of transport mechanisms by one type of amino acid with another. Thus, the basic amino acid arginine up-regulates acidic as well as basic amino acid transport, and the acidic amino acid aspartate induces maximal transport of basic amino acids as well as acidic ones.365 This cross-stimulation occurs between amino acids and peptides, each of which stimulates the absorption of the other. Enhanced absorption could result from increased numbers of transporters or from increased activity of each transporter.205 There is evidence in favor of both mechanisms, although the former probably predominates. It is uncertain whether an increase in transporter number is caused by increased synthesis, decreased degradation, or an increase in the insertion of preformed transporters into the brush border membrane. Because dietary regulation of glucose and amino acid transporters takes two to three days, it is likely that regulation occurs at the level of crypt cells. In diabetes mellitus, the persistent hyperglycemia stimulates both basolateral and apical membrane glucose transport, which can be inhibited by protein synthesis inhibitors such as cyclohexamide, suggesting a role for increased synthesis of new glucose transporters.366

VITAMINS AND TRACE ELEMENTS

Deficiencies of vitamins and trace elements are associated with up-regulation of their absorptive mechanisms. This is seen, for example, with biotin, thiamine, and ascorbic acid and with iron, calcium, zinc, and phosphate (Fig. 100-25).205 Because some of these moieties are potentially toxic, most down-regulate their transport mechanisms when present in higher concentrations. Low body stores of iron, zinc, calcium, and phosphate signal enhanced absorptive mechanisms. Zinc deficiency enhances zinc absorption approximately five-fold by increasing transport capacity. Dietary calcium deficiency stimulates calcium uptake in the proximal intestine by a vitamin D–dependent mechanism involving increases in a cytosolic calcium-binding protein and by stimulating transport across the brush border membrane and at the basolateral membrane. These changes occur within a few hours, suggesting that mature enterocytes on the villi are capable of being regulated. By contrast, the effects of dietary regulation of glucose and amino acids take two to three days. The difference in time scales over which gastrointestinal responses to stimulation by various nutrients occur provides fascinating insights into the molecular and biologic events underlying these phenomena. Immediate responses seen within seconds to a few minutes after exposure are likely to involve release, or activation, of preformed proteins; adaptive responses found within two to three hours

Chapter 100  Digestion and Absorption of Nutrients and Vitamins 5 Iron, zinc

Relative absorptive capacity

4

Calcium

3

2 Phosphate 1

0 Deficient

Average

Excessive

Nutritional intake Figure 100-25.  Changes in intestinal absorptive capacity for some nutrients in response to nutritional intake, ranging from deficient to excessive. (Adapted from Ferraris RP, Diamond JM. Specific regulation of intestinal nutrient transporters by their dietary substrates. Annu Rev Physiol 1989; 51:125-41, with permission.)

increase in small-bowel length and function per unit length.372 In hypophysectomized rats, mucosal hypoplasia of the small intestine and a reduced adaptive response develops after resection, and they are restored by GH. Transgenic mice expressing elevated levels of GH experience hypertrophy of the small intestine.372 IGF-1 expression in the small intestine is regulated by GH and is believed to induce enterotrophic effects following resection. Human and rabbit studies have indicated that increased nutrient transport activity devoid of morphologic changes may be the method of GH-induced intestinal adaptation.373 In patients who have short bowel syndrome and are dependent on home parenteral nutrition (HPN), high-dose recombinant human GH (0.4 mg/kg/day) in controlled and uncontrolled studies has led to variable results.374-376 GLP-2 increases the adaptive response to massive intestinal resection in rats.377 The non–placebo-controlled study by Jeppesen and colleagues in eight patients with short bowel syndrome and an end-enterostomy type of anastomosis (six had Crohn’s disease and four were not receiving hyperalimentation) showed a significant increase in the absolute and percentage intestinal absorption of net weight and a significant increase in the percentage but not the absolute absorption of energy and nitrogen in response to the use of GLP-2.378

THE NEONATAL INTESTINE of stimulation are probably due to increased translation; responses that take several hours or days are likely to be due to increased transcription and production of more mRNA.367

SIGNALING FOR INTESTINAL ADAPTATION AND IMPLICATIONS FOR THERAPY

A variety of signals have been described as possibly playing a role in the process of intestinal adaptation. These include growth hormone (GH), epidermal growth factor (EGF) and EGF receptor, prostanoids, uncoupling proteins, peroxisome proliferation-activated receptor-α (PPAR-α), insulinlike growth factor (IGF-1) receptors and IGF-binding proteins, transforming growth factor-α, SPARC (secreted protein, acidic and rich in cystine), Bcl-2, endothelin-1, erythropoietin, keratinocyte growth factor, the GATA family of zinc-finger transcription factors, hepatocyte growth factor, the early-response genes (ERGs), ornithine decarboxylase (ODC), PC4/TIS7, epimorphin, and AP-1, a transcription factor composed of Fos and JUN family proteins.347 Several of these signals might lend themselves to be modified in a clinical setting to enhance the intestinal adaptive response. For example, evidence from animal intestinal resection models suggest that the enhancing influence of glucocorticoids on sugar uptake may be achieved by post-translational processes involving signaling with c-jun, ODC and proglucagon, or other as yet unknown signals.368,369 In an animal model of extensive intestinal resection (50% enterectomy), prednisone had no effect on the uptake of glucose or fructose. By contrast, the locally acting steroid budesonide increased the value of the jejunal maximal transport rate for the uptake of glucose by more than 120% and increased the ileal uptake of fructose by more than 150%.368,370 GH might possess such proadaptive properties.371 In animal models, the administration of GH results in an

DEVELOPMENT AND ADAPTATION OF NUTRIENT DIGESTION AND ABSORPTION

Nutrient requirements vary markedly during early postnatal development, and this is mirrored by alterations in digestive and absorptive capacities. Some of these changes are genetically determined and programmed and do not appear to be greatly influenced by changes in dietary load.5 Thus, for most of the world’s population, excluding whites, the decline in activity of the disaccharidase lactase, which occurs after infancy, cannot be prevented by maintaining a high milk intake.205 Some early postnatal responses and most responses in adult life, however, appear to be purposive and reversible, paralleling changes in dietary intake, for digestive enzyme production and absorptive capacity. Such adaptations may occur in response to changes in dietary load or altered body needs.205,357

DEVELOPMENTAL CHANGES

Approximately 50% of the total calorie requirement of infants is provided by the fat in milk. Breast milk contains 3.5% to 4% lipid, of which 95% is in the form of triglyceride.176 The fatty acid composition is a mixture of mediumand long-chain fatty acids. In neonates, pancreatic lipase secretion is low and the digestion of triglyceride in milk relies on the other lipases present in milk or secreted by the tongue or gastric mucosa. Pancreatic lipase secretion rises after weaning because milk-derived lipase is no longer available (Fig. 100-26).367 Protein digestion is incomplete in infancy, and many proteins, such as human milk protein (whey), partially escape digestion.379 This relative immaturity also might have advantages for the infant because some biologically important peptides and immunoglobulins remain intact. Proteolytic enzymes are derived from a variety of sources, which also are changing during early neonatal life. Thus, there are several specific proteases, including trypsin and elastase, in breast milk.379 Gastric acid and pepsin are

1729

Section X  Small and Large Intestine Ileal bile salt absorption

100

Pancreatic enzymes Milk-derived lipase

Relative activity (%)

1730

Lactase 0 Birth Weaning 1 year

Adult life

Figure 100-26.  Diagram of the relative activities of various digestive enzymes and processes from birth to adulthood.

secreted at birth and increase toward adult values over the following three to four months. Nonetheless, little protein digestion appears to occur in the stomach during the first few weeks of life, possibly owing to the presence of protease inhibitors in milk. Likewise, luminal proteolytic machinery in the small intestine is not fully developed at birth, although enterokinase and pancreatic proteolytic enzymes are detectable. Rates of chymotrypsin and trypsin secretion are slower in infancy than in adult life, and responses to stimulation with CCK are depressed.380 Low rates of pancreatic enzyme secretion at birth may be attributable in part to the retarded display of polypeptide hormone receptors on the basolateral membrane of acinar cells. Digestive enzymes appear at different times after birth, suggesting that the genes that code for these enzymes may be activated at different times during development.380 One of the most characteristic changes in the postnatal period is the decline in lactase activity seen in most of the world’s populations, apart from white persons.381 The human population can be divided into two groups with respect to the ability to digest lactose: lactase-persistent persons, who are able to digest lactose throughout their lifespan, and lactase-nonpersistent (lactose intolerant, adult-type hypolactasia) persons. Lactase nonpersistence is caused by a decrease in the expression of the small intestinal–specific enzyme lactase–phlorizin hydrolase (LPH) at approximately age 5 to 10 years. Persons with lactase nonpersistence have a limited ability to digest lactose as adults; however, both lactase-nonpersistent and lactose-persistent persons have a high LPH expression after birth. The lactasepersistent phenotype is found most often among Northern Europeans, Indians from Punjab, Bedouins, and some nomadic tribes in Africa.381,382 The mechanism behind the developmental downregulation of LPH expression has been shown to be mainly transcriptional in humans, with several transcription factors (Cdx-2, GATA factors, and hepatocyte nuclear factor 1) mutually interacting and activating LPH expression.383,384 The binding sites for these factors are clustered within 100 bp upstream of a TATA box. No differences have been described in the DNA sequence in the LPH gene that are correlated to the lactase phenotype; however, a T/C poly-

morphism at position −13910 and an A/G polymorphism at position −22018 from the start codon of the LPH gene have been identified. Although these nucleotide variants are located in introns 8 and 13 of the neighboring MCM6 gene, the −13910C polymorphism associates 100% and the −22018G polymorphism associates approximately 97% with the lactase-nonpersistent phenotype in the Finnish population.385 It has been shown that the two single nucleotide polymorphisms are associated with the transcriptional regulation of the LPH gene.386 The −13910 T/C polymorphism is located in a transcriptional enhancer sequence, which strongly activates the LPH promoter activity. Furthermore, the −13910T variant isolated from a lactase-persistent person has a significantly higher enhancer activity than the corresponding −13910C variant isolated from a lactase-nonpersistent person. Analyses of the binding of nuclear factors to the −13910 polymorphic sequences show that the −13910T sequence binds nuclear factors with higher affinity than the −13910C sequence. The −22018 region does not possess enhancer activity and, in fact, results in a very small but significant reduction in reporter gene activity. The reduction of the reporter gene level by the −22018 region is greater in the presence of the enhancer of the −13910 region. It has been suggested that the lactase-persistent phenotype is caused by a mutation in the −13910 position, creating a strong enhancer that is able to keep the LPH gene active during adulthood (Fig. 100-27).382 Based on these results, a model explaining the mechanisms behind the postweaning down-regulation of LPH expression and adult-type hypolactasia has been proposed.387,388 Transcription factors necessary for LPH expression are present in excess during childhood and before weaning in mammals. After the weaning period, the expression of some intestinal transcription factors is changed (e.g., HNF1α). Also, the availability of some of these factors may be decreased because genes necessary for digestion of a starch-rich diet (e.g., SI), are up-regulated after weaning. Many of these intestine-specific genes depend on the same transcription factors (Cdx-2, HNF1, and GATA factors) as the LPH promoter. The competition for these transcriptional activators is higher after weaning. These changes result in a lower LPH expression owing to the weak nature of the LPH promoter; however, the strong enhancer effect of the −13910T variant compensates for these changes and can keep the LPH gene active throughout adulthood, giving the phenotype of lactase persistence (see Fig. 100-27). Changes in epithelial membrane transport of nutrients take place when the intestine is suddenly and rapidly expected to assume the role of the placenta in providing nutrients at the moment of birth and immediately there­ after. Brush border membrane glucose and most amino acid transporters are present in the human fetal intestine well before birth,5 when, in contrast to adult intestine, they are found throughout the crypt-villus axis. Fructose absorptive capacity rises rapidly after weaning when this sugar is presented in the diet. Transporters for bile salts are not programmed to appear on ileal enterocytes until weaning.343

TRIGLYCERIDE DIGESTION

Some differences have been recognized between the way infants digest triglyceride, which they receive in milk, and how adults digest this same fatty substrate. In contradistinction to other dietary sources, the triglyceride in milk is

Chapter 100  Digestion and Absorption of Nutrients and Vitamins HNF1/GATA/Cdx

Child

High level of transcription MCM6 gene A

LPH gene T

G

C

–22018

–13910

Lactase-persistent adult –

+ High level of transcription

A –22018

T –13910

Lactase-nonpersistent adult –

+ Low level of transcription

G

C

–22018 –13910 Figure 100-27.  Schematic model showing the interaction between the promoter of the lactase phlorizin hydrolase (LPH) gene and polymorphic −13910 and −22018 regions in the MCM6 gene in children and lactase-persistent and -nonpersistent adults. During childhood and before weaning in other mammals, the level of LPH expression is high because the transcription factors (HNF1α, GATA factors, Cdx-2) known to regulate LPH expression are available in excess. The expression of LPH is therefore not dependent on the −13910 enhancer activity. In adulthood, accessibility to the transcription factors is reduced. The enhancer activity of the −13910T variant ensures an active LPH gene throughout life (lactase persistence). The lower activity of the −13910C variant fails to activate or recruit the transcription factors, which results in low LPH gene activity in lactase-nonpersistent adults. Although the −22018 region represses LPH transcription, the role of the −22018 region is unclear; the repression does not seem to be related to the A/G polymorphism. (From Troelsen JT. An upstream polymorphism associated with lactase persistence has increased enhancer activity. Gastroenterology. 2003; 125:1686-94.)

packaged in smaller emulsion droplets, each surrounded by a trilaminar membrane that includes both phospholipid and proteins (albumin and β-lactoglobulin).367 At a time in its life when the newborn infant relies for more than half of its energy requirements on milk-derived triglyceride, pancreatic lipase secretion rates are only about half those of adults when expressed in terms of body surface area. Only at weaning does pancreatic lipase secretion begin to rise to adult levels.389 Two other lipases are important at this stage in life (see Table 100-2). The first is secreted by the mammary gland but is inactive in milk and requires the presence of bile salts to activate it. It begins to function, therefore, upon entering the duodenum.367 The second lipase is secreted either from serous glands at the base of the tongue or from gastric mucosa (or both), depending on the species. In humans, this preduodenal lipase is largely, if not entirely, derived from chief cells in the gastric mucosa390; in the rat it is derived from the tongue. This lipase has optimal pH of 4 to 6, which is lower than that of pancreatic lipase. It preferentially acts at the α1 position on triglyceride and releases fatty acid and diacylglycerol, as opposed to the monoglyceride and two fatty acids produced by pancreatic lipase. It is not stimu-

lated by bile salts and is released during feeding by autonomic nervous stimulation. Gastric lipase appears to be particularly active at the surface membrane of droplets derived from milk; pancreatic lipase is less so. Although the amount of lipolysis that occurs within the stomach is relatively small, the release of even modest amounts of fatty acids, particularly from the shorter- and medium-chain triglycerides, may be important in the emulsification of fat in the duodenum and in enhancing pancreatic lipase activity. Gastric lipase is rapidly inactivated in the duodenum by proteolytic enzymes. There is, however, conflicting evidence about the importance of intragastric lipolysis, and although some have suggested that as much as 30% of fat may be digested here,56 the rapid product inhibition of gastric lipase activity by released fatty acids makes this unlikely. Each of the three lipases has a different specificity for the ester bonds in triglyceride. Gastric lipase preferentially cleaves the bond in the 1α position, and pancreatic lipase the bonds in the 1α and 3α positions; milk-derived, bile salt–stimulated lipase is nonselective and splits the bonds at the 1-, 2-, and 3α positions. Thus, newborns’ luminal contents contain more fatty acids and less monoglyceride

1731

1732

Section X  Small and Large Intestine and diglyceride than those of adults, and this is probably advantageous for absorption. Smaller amounts of bile salts are available in newborns than in adults, at least in part because active ileal reabsorption is immature.367 Under these conditions, fatty acids are likely to be absorbed more readily than monoglyceride. In addition, the low availability of bile salts makes it likely that transfer of fat to the brush border membrane depends more on unilamellar liquid crystalline vesicles. Gastric lipase persists into adult life, and the amounts found in biopsies of adult gastric mucosa are similar to those found in infant gastric mucosa. Studies suggest it might hydrolyze as many as one in four triglyceride acyl chains during digestion of a meal.10

CARBOHYDRATE DIGESTION AND ABSORPTION

Lactose is the major carbohydrate in breast milk, and the need for an amylase before weaning is therefore minimal. An α amylase is present in milk, however, and an amylase also is secreted in saliva at birth.381 Both of these amylases are inactivated by acid in the stomach, but they may resume their activity at nearly neutral pH on reaching the duodenum. It has been estimated that 15% to 40% of amylase activity in the duodenum of infants is of salivary origin.389 Pancreatic amylase secretion is low, and stimulation with exogenous agonists produces little response, indicating the prematurity of the pancreas at this stage. In any event, the need for any amylase in neonates is minor unless starch is introduced early. Most infant formulas do not contain starches, but some contain glucose polymers. Digestion of any starch ingested during the first two or three months of life relies on salivary amylase and mucosal α glucosidases as well as colonic salvage by fermentation of undigested carbohydrate by bacteria. Mucosal lactase is present at birth in high concentration, as are the other glucosidases.391 Nevertheless, lactose absorption may be incomplete in neonates, particularly in premature neonates. Estimates of the amount of lactose that reaches the colon vary and have been based on indirect measurements of breath hydrogen concentrations. Probably less than 20% of ingested lactose reaches the colon, but a much smaller proportion is lost in the stool because of bacterial hydrolysis and absorption of the products. The glucose/galactose transporter in the apical membrane of villus enterocytes is well developed in full-term infants, but it may be suboptimal in premature ones. This is unlikely, however, to pose a significant barrier to nutrition.

PROTEIN DIGESTION AND ABSORPTION

Although acid and pepsinogens are secreted in neonates, little intragastric proteolysis occurs during the first two to six weeks of life. A renin-like protease is secreted during the first 10 days of life, which causes protein precipitation.379 Pancreatic proteolytic enzymes are secreted at birth, although at slower rates than in adults. Trypsinogen secretion is low and, particularly in preterm infants, does not respond to feeding. Stimulation with pancreozymin has little effect on pancreatic enzyme secretion for the first one to two months of life.177 Enterokinase (enteropeptidase) is present at birth and is capable of activating trypsinogen. Despite the apparent immaturity of the proteolytic machinery, it has been estimated that duodenal proteolysis can cope with as much as 3 or 4 grams of protein per kilogram of body weight of casein, and infants seem not to be prone to defective nitrogen nutrition. A number of proteases have been found in breast milk, including plasmin, which is most active against casein.392

The overall nutritional importance of these milk-derived proteases or the protease inhibitors also found in milk393 is not known. Transport systems for amino acids and small peptides appear to be well developed in newborns. The infant intestine has greater capacity than the adult intestine to absorb intact macromolecules, including proteins. Transport by pinocytosis or receptor-mediated endocytosis probably accounts for the ability of infants to take in biologically important whole proteins, such as the immunoglobulins, during this phase of life. This mechanism disappears after the first three months of life, when closure is said to have occurred394; however, uptake of intact proteins continues throughout life, albeit in trace amounts, and the role of M cells on Peyer’s patches is of major importance in this process. It is likely to be an important mechanism by which dietary antigens are presented for immune surveillance later in life, but it is of little nutritional significance.

EFFECTS OF BARIATRIC SURGERY ON NORMAL DIGESTION AND ABSORPTION The total volume of bariatric surgical procedures in the United States continues to grow rapidly (see Chapter 7). The American Society for Bariatric Surgery estimates that such operations have increased 1431% since 2000 to more than 250,000 annually. At present, the Roux-en-Y gastric bypass (RYGB) is the most widely performed weight loss procedure in the United States, followed by vertical banded gastroplasty or laparoscopic gastric banding and biliopancreatic diversion with or without duodenal switch.395,396 The exact mechanisms of weight loss associated with each treatment modality, however, and how they affect intestinal digestion and absorption is not clearly understood. Operations for weight loss have generally been classified as malabsorptive, restrictive, or both, based on the proposed mechanism for the induction of weight loss. The jejunoileal bypass (JIB) and its more recent modifications, the bilio­ pancreatic diversion (BPD) and the distal gastric bypass with duodenal switch (DGB-DS), are classified as primarily malabsorptive procedures. The gastroplasties (horizontal, vertical, and vertical banded gastroplasty [VBG]) are restrictive. RYGB represents a combination of these two mechanisms. The results of surgical treatment of severe obesity differ considerably based on the mechanisms of weight loss. Malabsorptive procedures, especially the JIB, produce the greatest degree of weight loss but can be associated with serious and potentially life-threatening metabolic and nutritional complications.397 BPD and DGB-DS produce significant weight loss that persists and causes fewer metabolic complications than JIB.396,397 Restrictive procedures produce moderate degrees of weight loss and are associated with the lowest incidence of metabolic and nutritional complications.396 RYGB has consistently been shown to produce greater and more sustained weight loss than VBG and avoids the severe metabolic and nutritional consequences of intestinal bypass. The mechanism of weight loss following VBG appears to be similar to that of other forms of caloric restriction, although the weight loss tends to be more pronounced and to persist for a longer time. The role of the regulation of central satiety mechanisms has been studied for restrictive and malabsorptive types of surgery. Ghrelin is a gastric peptide with potent orexigenic effects. Circulating ghrelin concentrations are high in obese subjects, and they increase

Chapter 100  Digestion and Absorption of Nutrients and Vitamins after weight loss. In patients undergoing RYGB (compared with patients undergoing VBG or BPD), however, a decrease in ghrelin levels has been reported and appears to depend on the surgically induced bypass of the ghrelin-producing cell populations of the gastric fundus.398 Several mechanisms have been suggested to account for the more substantial weight loss after JIB, RYGB and BPD. These can be divided into three groups: caloric restriction, changes in energy metabolism, and alterations in gastrointestinal hormones and nutrient absorption. Caloric restriction seems to play a prominent role in JIB and RYGB, and numerous studies document significantly decreased caloric intake after each type of surgery. Animal studies have shown that rats subjected to intestinal bypass maintained lower weights despite caloric intakes comparable with those of sham-operated controls, suggesting an increase in energy expenditure. Human data suggest a relative increase in the mean resting energy expenditure in patients undergoing RYGB.399,400 Malabsorption leading to a loss of lean body mass, diarrhea, and vitamin, mineral, and electrolyte deficiencies prominently contributes to weight loss after JIB, but less so after BPD and DGB-DS. After JIB, malabsorption results from the drastically reduced intestinal surface in contact with nutrients (common channel) and rapid transit time. It has been suggested that the enhanced weight loss after BPD is related to the combination of a short common channel (approximately 50 cm), which limits fat absorption, and a long afferent limb, which is not in direct contact with food. Therefore, the effect of the alimentary (Roux) limb, where protein and carbohydrates are absorbed, is limited. The documented changes in absorption might result from the alterations in the hormonal milieu of the gastrointestinal tract. Dumping syndrome, evidenced by symptoms of lightheadedness, sweating, palpitations, and diarrhea following a test meal containing a high-carbohydrate load, results from operations that bypass the normal pyloric functions and has been documented in patient with RYGB.401,402 These symptoms are temporally associated with changes in intestinal peptides, especially an increase in enteroglucagon, which is secreted by the ileum. Increase in enteroglucagon and neurotensin following ileal transposition in rats is associated with weight loss equivalent to that achieved with caloric restriction. These findings suggest that rapid transport of undigested and unabsorbed nutrients into the ileum after these procedures stimulate the secretion of intestinal peptides, (e.g., enteroglucagon), which then produce or

serve as a marker for full-blown or subclinical dumping, which in turn contributes to the observed weight loss.401,403 Peptide YY (PYY) is a hormone secreted by the L cells lining the terminal small intestine and colon in response to intestinal nutrients, especially long-chain fatty acids.21,404 PYY decreases gastric and intestinal emptying and improves intestinal absorption. This mechanism is the so-called ileal brake.32 PYY is also a potent anorexigenic hormone that acts by modulating appetite circuits in the hypothalamus, as well as stimulating apo A-IV, which plays an important role in upper gastrointestinal function and satiety.21,405 It has been shown that PYY serum levels are higher in patients who have undergone JIB.406

KEY REFERENCES

Altmann SW, Davis HR, Jr., Zhu LJ, et al. Niemann-Pick C1 Like 1 protein is critical for intestinal cholesterol absorption. Science 2004; 303:1201-4. (Ref 109.) Drobnik W, Lindenthal B, Lieser B, et al. ATP-binding cassette transporter A1 (ABCA1) affects total body sterol metabolism. Gastro­ enterology 2001; 120:1203-11. (Ref 100.) Kuokkanen M, Enattah NS, Oksanen A, et al. Transcriptional regulation of the lactase-phlorizin hydrolase gene by polymorphisms associated with adult-type hypolactasia. Gut 2003; 52:647-52. (Ref 386.) Nemeth E, Tuttle MS, Powelson J, et al. Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science 2004; 306:2090-3. (Ref 317.) Owyang C, Logsdon CD. New insights into neurohormonal regulation of pancreatic secretion. Gastroenterology 2004; 127:957-69. (Ref 38.) Parkkila S, Niemela O, Britton RS, et al. Molecular aspects of iron absorption and HFE expression. Gastroenterology 2001; 121:1489-96. (Ref 307.) Peet DJ, Turley SD, Ma W, et al. Cholesterol and bile acid metabolism are impaired in mice lacking the nuclear oxysterol receptor LXR alpha. Cell 1998; 93:693-704. (Ref 137.) Ritz V, Alfalah M, Zimmer KP, et al. Congenital sucrase–isomaltase deficiency because of an accumulation of the mutant enzyme in the endoplasmic reticulum. Gastroenterology 2003; 125:1678-85. (Ref 173.) Santer R, Hillebrand G, Steinmann B, Schaub J. Intestinal glucose transport: Evidence for a membrane traffic-based pathway in humans. Gastroenterology 2003; 124:34-9. (Ref 209.) Stahl A. A current review of fatty acid transport proteins (SLC27). Pflugers Arch 2004; 447:722-7. (Ref 88.) Wang DQ. Regulation of intestinal cholesterol absorption. Annu Rev Physiol 2007; 69:221-48. (Ref 144.) Woods SC. Gastrointestinal satiety signals I. An overview of gastrointestinal signals that influence food intake. Am J Physiol Gastrointest Liver Physiol 2004; 286:G7-13. (Ref 11.) Full references for this chapter can be found on www.expertconsult.com.

1733

CHAPTE R

101 Maldigestion and Malabsorption Christoph Högenauer and Heinz F. Hammer

CHAPTER OUTLINE Etiology and Pathophysiology  1736 Fats  1737 Defective Mixing  1737 Reduced Solubilization of Fat  1738 Decreased Lipolysis  1738 Decreased Mucosal Absorption and Chylomicron Formation  1738 Defective Lymphatic Transport of Chylomicrons  1738 Proteins and Amino Acids  1738 Defective Intraluminal Proteolysis  1738 Defective Mucosal Hydrolysis of Peptides and Decreased Absorption of Oligopeptides and Amino Acids  1738 Carbohydrates  1739 Defective Intraluminal Hydrolysis of Carbohydrates  1739 Mucosal Defects of Carbohydrate Digestion and Absorption  1739 Vitamins  1739 Fat-Soluble Vitamins  1739 Water-Soluble Vitamins  1739 Minerals  1739 Calcium  1739 Magnesium  1740 Iron  1740 Zinc  1740 Others  1740 Mechanisms That Compensate for Malabsorption  1740 Role of the Colon  1740 Role of Intestinal Transit in the Salvage of Malabsorbed Nutrients  1741

In the past, it was believed that most malabsorptive diseases manifested clinically with diarrhea and steatorrhea. It is now recognized, however, that many malabsorptive dis­ orders, such as celiac disease, might have subtle clinical presentations, such as bloating or changes in bowel habits, or mainly extraintestinal manifestations, such as anemia, bone loss, or menstrual disturbance, that lead to erroneous diagnoses. Awareness also is increasing that subtle malab­ sorption of single nutrients, such as calcium or vitamin B12, can, if unrecognized, lead to complications that may be difficult to reverse or even irreversible. Therefore, the clini­ cal challenge today is to recognize and treat malabsorption despite its subtle manifestations. Classically, maldigestion is defined as defective hydro­ lysis of nutrients, and malabsorption is defined as defective mucosal absorption. Although this distinction may be

Clinical Features and Evaluation  1741 Suspecting and Confirming the Presence of Malabsorption  1742 Diagnostic Approach  1742 Anatomic Investigations  1746 Endoscopy, Biopsy, and Duodenal Aspiration  1746 Abdominal Imaging  1749 Noninvasive Evaluation of Gastrointestinal Digestive and Absorptive Function  1749 Malabsorption in Specific Situations and Disease States  1754 Lactose Malabsorption and Intolerance  1754 Incomplete Absorption and Intolerance of Fructose  1755 Ileal Bile Acid Malabsorption  1755 Amyloidosis  1756 Malabsorption Caused by Drugs and Food Supplements  1756 Malabsorption after Gastric Resection or Bariatric Surgery  1756 Malabsorption in the Elderly  1758 Connective Tissue Diseases  1758 Congenital Defects That Cause Malabsorption  1758 Primary Immunodeficiency Diseases  1763 Neurofibromatosis Type 1 (von Recklinghausen’s Disease)  1764 Nongranulomatous Chronic Idiopathic Enterocolitis and Autoimmune Enteropathy  1764 Endocrine and Metabolic Disorders  1765 General Approach to Management  1767

useful on pathophysiologic grounds, the clinical presenta­ tion and complications of maldigestion and malabsorption are similar. Moreover, physiologic processes other than digestion and absorption, such as solubilization, intestinal motility, or hormone secretion, contribute to the normal absorption of nutrients, vitamins, and minerals. Therefore, the classic definitions of maldigestion and malabsorption do not cover the pathophysiologic spectrum of the malab­ sorption syndrome. In this chapter, the terms digestion and absorption, or maldigestion and malabsorption, are used separately only in the discussion of pathophysiology. When the distinction between these terms is not of clinical relevance, only the terms absorption and malabsorption are used. Malabsorption can be caused by many diseases of the small intestine and also by diseases of the pancreas, liver,

1735

1736

Section X  Small and Large Intestine Table 101-1  Diseases That Cause Nutrient Malabsorption Gastric Diseases Atrophic gastritis Autoimmune gastritis (pernicious anemia) Gastric resection or bypass surgery Pancreatic Diseases Congenital pancreatic enzyme deficiencies Colipase deficiency Lipase deficiency Trypsinogen deficiency Pancreatic insufficiency Chronic pancreatitis Cystic fibrosis Johanson-Blizzard syndrome Pearson’s marrow-pancreas syndrome Shwachman’s syndrome Pancreatic tumors Liver Diseases Inborn errors of bile acid biosynthesis and transport Cirrhosis and other liver diseases Portal hypertension Obstructive Biliary Diseases Biliary tumors Primary and secondary sclerosing cholangitis Intestinal Diseases Amyloidosis Autoimmune enteropathy Celiac disease Collagenous sprue Congenital intestinal defects (see Table 101-14) Crohn’s disease Enteroendocrine cell deficiency Autoimmune polyglandular syndrome type 1 Enteric anendocrinosis Enterokinase deficiency Eosinophilic gastroenteritis Fistulas Food allergy Graft-versus-host disease Hypolactasia Ileal bile acid malabsorption Intestinal infections AIDS (HIV infection): Cryptosporidiosis, Mycobacterium avium complex infection, viral infections Giardiasis

Helminthic infections Tuberculosis Whipple’s disease Immunoproliferative small intestinal disease Intestinal ischemia Intestinal lymphoma Intestinal resections or bypass Mastocytosis Nongranulomatous chronic idiopathic enterocolitis Postinfection malabsorption Primary immunodeficiency diseases Radiation enteritis Refractory sprue Sarcoidosis Small intestinal bacterial overgrowth Tropical sprue Lymphatic Diseases Primary intestinal lymphangiectasia Secondary intestinal lymphangiectasia Lymphoma Solid tumors Thoracic duct trauma, damage, or obstruction Neuroendocrine Tumors Carcinoid syndrome Glucagonoma Somatostatinoma Zollinger-Ellison syndrome Cardiac and Vascular Diseases Congestive heart failure Constrictive pericarditis Endocrine Causes Addison’s disease Diabetes mellitus Hyperthyroidism Systemic Diseases Cronkhite-Canada syndrome Mixed connective tissue disease Neurofibromatosis type 1 Protein-calorie malnutrition Scleroderma Systemic lupus erythematosus

AIDS, acquired immunodeficiency syndrome; HIV, human immunodeficiency virus.

biliary tract, and stomach (Table 101-1). Whereas in some of these diseases, malabsorption may be the presenting feature, in others malabsorption may be only a minor clinical problem or may be detected only as a laboratory abnormality. This chapter provides an overview of basic pathophy­ siologic mechanisms leading to symptoms or complications of maldigestion or malabsorption, reviews the clinical mani­ festations and complications of malabsorption, describes tests that can be used clinically to evaluate digestive and absorptive function, provides a rational diagnostic approach to the individual patient, and discusses malabsorptive diseases and general measures in the treatment of malab­ sorption syndrome that are not covered in other chapters of this book.

ETIOLOGY AND PATHOPHYSIOLOGY From a pathophysiologic point of view, mechanisms causing malabsorption can be divided into premucosal (luminal)

factors, mucosal factors, and postmucosal factors (vascular and lymphatic). For clinical purposes, this approach is of limited value, because the various clinical pictures caused by malabsorption syndromes are determined mainly by the nature of the malabsorbed substrates. We therefore discuss the mechanisms causing malabsorption on the basis of the malabsorbed substrate. A separate section is devoted to the role of mechanisms compensating for the consequences of malabsorption. Normal uptake of nutrients, vitamins, and minerals by the gastrointestinal tract requires several steps, each of which can be compromised in disease. (Normal digestion and absorption are discussed in Chapter 100.) Solubilization is a prerequisite for the absorption of such nutrients as fat or calcium. Fat and fat-soluble vitamins are solubilized by the formation of micelles, and calcium is solubilized through acidification of the gastrointestinal lumen. Alternatively, increased solubilization of the components of intestinal chyme can contribute to the mani­ festations of gastrointestinal diseases, such as increased absorption of oxalate, which can result in the development of kidney stones in patients with short bowel syndrome.

Chapter 101  Maldigestion and Malabsorption Table 101-2  Mechanisms of Malabsorption, Malabsorbed Substrates, and Representative Causes MECHANISM

MALABSORBED SUBSTRATE(S)

representative causes

Fat Fat-soluble vitamins Calcium Magnesium

Hepatic parenchymal disease Biliary obstruction Bacterial overgrowth with bile acid deconjugation Ileal bile acid malabsorption CCK deficiency

Pancreatic insufficiency

Fat Protein Carbohydrate Fat-soluble vitamins Vitamin B12 (cobalamin)

Congenital defects Chronic pancreatitis Pancreatic tumors Inactivation of pancreatic enzymes (e.g., Zollinger-Ellison syndrome)

Reduced mucosal digestion

Carbohydrate Protein

Congenital defects (see Table 101-14) Acquired lactase deficiency Generalized mucosal disease (e.g., celiac disease, Crohn’s disease)

Intraluminal consumption of nutrients

Vitamin B12 (cobalamin)

Small intestinal bacterial overgrowth Helminthic infections (e.g., Diphyllobothrium latum infection)

Fat Protein Carbohydrate Vitamins Minerals

Congenital transport defects (see Table 101-14) Generalized mucosal diseases (e.g., celiac disease, Crohn’s disease) Previous intestinal resection or bypass Infections Intestinal lymphoma

Fat Protein

Intestinal lymphangiectasia Primary Secondary (e.g., solid tumors, Whipple’s disease, lymphomas) Venous stasis (e.g., from congestive heart failure)

Vitamin B12

Pernicious anemia Atrophic gastritis Previous gastric resection

Decreased gastric mixing and/or rapid gastric emptying

Fat Calcium Protein

Previous gastric resection Autonomic neuropathy

Rapid intestinal transit

Fat

Autonomic neuropathy Hyperthyroidism

Maldigestion Conjugated bile acid deficiency

Malabsorption Reduced mucosal absorption

Decreased transport from the intestine

Other Mechanisms Decreased gastric acid and/or intrinsic factor secretion

CCK, cholecystokinin.

Digestion of macromolecular compounds, such as poly­ saccharides, triglycerides, and proteins, to their molecular components—monosaccharides, fatty acids, and amino acids, respectively—is achieved by soluble or membranebound digestive enzymes. Absorption of undigested or par­ tially digested macromolecular compounds occurs to a very minor degree in health and may be increased slightly in various intestinal diseases. Although such absorption does not play a nutritive role, it may be important for the normal function of the immune system and for the pathogenesis of diseases such as food allergy (see Chapter 9). Liberation of substrate, such as vitamin B12, from binding sites in food or, conversely, binding to factors such as intrin­ sic factor allows absorption to take place. Chemical changes to nutrients may be required for absorp­ tion, such as changing the charge of iron. Mucosal absorption can occur by active or passive carriermediated transport or by simple or facilitated diffusion. Postmucosal transport of absorbed substrates also is important.

Intestinal sensory and motor function permits detection of the presence of nutrients, facilitates adequate mixing of nutrients with intestinal secretions and delivery to absorp­ tive sites, and provides adequate time for nutrient absorp­ tion (see Chapter 97). Neural and hormonal functions are required to stimulate and coordinate digestive secretions, mucosal absorption, and intestinal motility. An overview of pathophysiologic mechanisms of mal­ digestion and malabsorption is provided in Table 101-2. This table also shows the ingested substrates primarily affected by the individual pathophysiologic mechanisms and lists examples of etiologic disorders for these mechanisms.

FATS DEFECTIVE MIXING

For sufficient digestion and absorption of lipids, dietary fat must adequately mix with digestive secretions. Gastric

1737

1738

Section X  Small and Large Intestine Table 101-3  Pathophysiologic Mechanisms That Result in Deficiency of Luminal Conjugated Bile Acids PATHOPHYSIOLOGIC MECHANISM

causes

Decreased synthesis and/or secretion of conjugated bile acids

Parenchymal liver diseases (e.g., cirrhosis, primary biliary cirrhosis) Biliary obstruction (e.g., tumors) Biliary fistulas Inborn errors of bile acid synthesis CCK deficiency

Intestinal loss of conjugated bile acids

Ileal resection Severe ileal mucosal disease Congenital defects of the ileal sodium–bile acid cotransporter

Luminal deconjugation of bile acids

Small intestinal bacterial overgrowth

Binding of bile salts or insolubilization of bile salts due to low luminal pH

Cholestyramine (binding) Zollinger-Ellison syndrome (low pH) Exocrine pancreatic insufficiency (low pH)

CCK, cholecystokinin.

resections or gastrointestinal motility disorders that result in rapid gastric emptying or rapid intestinal transit, such as autonomic neuropathy resulting from diabetes mellitus or amyloidosis, can cause fat malabsorption consequent to impaired gastrointestinal mixing of dietary fat.1

REDUCED SOLUBILIZATION OF FAT

Fat malabsorption due to decreased formation of micelles occurs if the luminal concentrations of conjugated bile acids are lower than the critical concentration required for forming micelles.2,3 Table 101-31,4 details the pathophysiologic mechanisms and representative diseases that cause luminal bile acid deficiency.

DECREASED LIPOLYSIS

If exocrine pancreatic function is severely reduced, impair­ ment of pancreatic lipase and colipase secretion results in decreased luminal hydrolysis of dietary fat.5 Chronic pan­ creatitis, cystic fibrosis, pancreatic duct obstruction by pan­ creatic and ampullary tumors, and pancreatic resection are the most common causes of pancreatic insufficiency.1 Even when pancreatic enzyme concentrations are normal, reduced pancreatic lipase activity due to a low luminal pH,6 exces­ sive calcium ingestion,7 or ingestion of the specific lipase inhibitor orlistat8 can cause pancreatic steatorrhea. Selec­ tive congenital lipase or colipase deficiency is a rare cause of pancreatic fat malabsorption.9

DECREASED MUCOSAL ABSORPTION AND CHYLOMICRON FORMATION

Generalized mucosal diseases, such as celiac disease or tropical sprue, often are associated with fat malabsorption. Defective uptake of free fatty acids and monoglycerides results from reduced mucosal surface area because of villus shortening, reduced enterocyte function, and mucosal inflammation.1 Intestinal fat absorption also is impaired in diseases that result in disturbance of intracellular formation of chylomicrons and accumulation of lipids within the enterocytes, including abetalipoproteinemia, hypobetalipo­ proteinemia, and chylomicron retention disease.10

DEFECTIVE LYMPHATIC TRANSPORT OF CHYLOMICRONS

Impairment of lymphatic transport of chylomicrons is a cause for postmucosal malabsorption of dietary fat.

Decreased lymphatic transport can result from congenital diseases such as primary intestinal lymphangiectasia or from obstruction of lymphatic vessels due to metastatic solid tumors, lymphoma, Whipple’s disease, retroperitoneal fibrosis, or trauma6 (see Chapter 28). Usually, lymphatic vessels in the mucosa become dilated (lymphangiectasia), and chylomicrons are lost into the intestinal lumen post­ prandially and also in the fasting state11; steatorrhea in these situations usually is only mild to moderate.10

PROTEINS AND AMINO ACIDS Defective digestion or absorption of dietary proteins has to be differentiated from excessive loss of serum proteins into the gastrointestinal tract, which is termed protein-losing enteropathy (see Chapter 28).

DEFECTIVE INTRALUMINAL PROTEOLYSIS

Protein digestion may be impaired in patients who have undergone partial or total gastric resection, presumably as a result of poor mixing with digestive secretions, although gastric pepsin deficiency could be contributory. Defective proteolysis also occurs with exocrine pancreatic insuffi­ ciency.1,12,13 In congenital diseases, pancreatic proteolysis can be impaired by inborn errors in the synthesis of proteo­ lytic enzymes (trypsinogen deficiency)13 or by defective acti­ vation of pancreatic proenzymes resulting from congenital deficiency of intestinal enterokinase (see later).14

DEFECTIVE MUCOSAL HYDROLYSIS OF PEPTIDES AND DECREASED ABSORPTION OF OLIGOPEPTIDES AND AMINO ACIDS

Generalized mucosal diseases, such as celiac disease and tropical sprue, result in global malabsorption, which includes malabsorption of oligopeptides and amino acids due to lack of mucosal hydrolysis of oligopeptides and defective mucosal absorption.13 Reduction of intestinal absorptive surface, as in short bowel syndrome or jejuno­ ileal bypass, also results in protein and amino acid malab­ sorption.13,15 Congenital defects of amino acid transporters on the enterocytes, such as Hartnup’s disease and lysinuric protein intolerance, can lead to selective malabsorption of a subgroup of amino acids (see later on).

Chapter 101  Maldigestion and Malabsorption CARBOHYDRATES DEFECTIVE INTRALUMINAL HYDROLYSIS OF CARBOHYDRATES

Pancreatic α-amylase normally is secreted in excess into the intestinal lumen. In mild forms of pancreatic insufficiency, carbohydrate digestion usually is at least partially pre­ served,16 but severe pancreatic insufficiency results in clini­ cally apparent carbohydrate malabsorption and diarrhea due to decreased luminal hydrolysis of ingested starch.17

MUCOSAL DEFECTS OF CARBOHYDRATE DIGESTION AND ABSORPTION

The most common cause of carbohydrate malabsorption is late-onset lactose malabsorption due to decreased levels of the intestinal brush border enzyme lactase (adult-type hypo­ lactasia, acquired primary lactase deficiency). Depending on ethnic background, lactase is present in less than 5% to more than 90% of the adult population; its deficiency results in a selective malabsorption of lactose. Acquired malabsorption of carbohydrates occurs commonly after extensive intestinal resections, in diffuse mucosal diseases such as celiac disease or Crohn’s disease, or temporarily after self-limited gastrointestinal infections (postinfection carbohydrate malabsorption).16,17 The pathophysiologic mechanisms of carbohydrate malabsorption are reduction of the intestinal mucosal surface area and a reduced activity or expression of intestinal oligo- and disaccharidases or transport proteins for monosaccharides.16 Congenital disaccharidase deficiencies (lactase, sucrase-isomaltase, and trehalase)18 and congenital deficiency or malfunction of transport molecules as in congenital glucose-galactose mal­ absorption19 can cause early onset of malabsorption of mono- or disaccharides (see later on). Intolerance of fructose is discussed in a subsequent section.

VITAMINS FAT-SOLUBLE VITAMINS

Diseases causing malabsorption of dietary fat commonly cause malabsorption of fat-soluble vitamins, because they require similar absorptive mechanisms. This is especially important in diseases that result in impaired micelle forma­ tion from bile salt deficiency.20 Fat-soluble vitamins also are malabsorbed in diffuse diseases of the mucosal surface area, in diseases affecting chylomicron formation and transport,21 and in exocrine pancreatic insufficiency.22 Some authors have suggested that absorption of fat-soluble vitamins is less affected by exocrine pancreatic insufficiency than by small intestinal diseases resulting in steatorrhea.23

WATER-SOLUBLE VITAMINS Vitamin B 12 (Cobalamin)

Decreased release of dietary vitamin B12 from food sources because of impaired pepsin and acid secretion, as in atro­ phic gastritis24 or use of acid inhibitory drugs such as proton pump inhibitors,25 usually results in only mild cobalamin malabsorption without clinical consequences. By contrast, deficiency of gastric intrinsic factor secretion, as occurs in pernicious anemia or after gastric resections, or secretion of an abnormal intrinsic factor, as in some congenital diseases, results in severe vitamin B12 malabsorption with clinical consequences.24 Autoimmune gastritis of pernicious anemia is the most common cause of vitamin B12 malabsorption.26 Cobalamin

malabsorption in pernicious anemia is caused by decreased intrinsic factor secretion resulting from parietal cell destruc­ tion and by blocking autoantibodies that inhibit intrinsic factor binding to vitamin B12.26 Mild cobalamin malabsorp­ tion may be found in patients with pancreatic insufficiency and in patients with Zollinger-Ellison syndrome, owing to decreased proteolytic release of vitamin B12 from its complex with R-binding protein24,27 (see Chapters 32, 59, and 100). In bacterial overgrowth syndrome (see Chapter 102) or helminthic infection with Diphyllobothrium latum (see Chapter 110), dietary cobalamin is made unavailable to the host or is consumed by the microorganisms or parasites in the intestinal lumen and, therefore, is not available for intes­ tinal absorption.26 Diseases and conditions affecting the ileal mucosa, such as Crohn’s disease or ileal resection, lead to a reduction of specific absorptive sites for the intrinsic factor-vitamin B12 complex.24 Ileal resections of more than 60 cm usually result in clinically significant vitamin B12 malabsorption.28 Imerslund-Gräsbeck syndrome, a disease of autosomal recessive inheritance due to malfunction of the cubilinamnionless (AMN) complex, is characterized by selective ileal malabsorption of the intrinsic factor–vitamin B12 complex despite normal ileal morphology.24,29 Congenital diseases affecting transcobalamin II also result in malab­ sorption of cobalamin.24,30 In previously healthy persons it usually takes several years of vitamin B12 malabsorption before cobalamin defi­ ciency develops, because the body stores contain large amounts of cobalamin and the daily requirement is rela­ tively small.

Folate Folate malabsorption occurs with mucosal diseases affect­ ing the proximal small intestine, such as celiac disease, Whipple’s disease, and tropical sprue.31 Folate deficiency is common in chronic alcoholism, in which it is postulated to be caused by decreased dietary intake as well as decreased intestinal absorption of folate.32 As discussed later, several drugs result in impaired intestinal uptake of folate, and an inherited form of selective folate malabsorption has been described. In contrast with cobalamin, body stores of folate are small relative to the daily requirements; therefore, folate deficiency develops faster than cobalamin deficiency in the setting of malabsorption. Increased serum folate levels resulting from bacterial formation of tetrahydrofolate have been reported in small intestinal bacterial overgrowth states.33

Other Water-Soluble Vitamins Other water-soluble vitamins, such as ascorbic acid and the B-complex vitamins, are absorbed in the small intestine either by carrier-mediated transport or by passive diffusion. Generalized malabsorption syndromes from intestinal causes impair the absorption of these vitamins, thereby leading to deficiency states.34,35 Deficiency of these watersoluble vitamins also occurs in chronic alcoholism, proba­ bly owing to decreased oral intake and reduced intestinal absorption.32

MINERALS CALCIUM

Severe calcium malabsorption can occur in diseases that affect the small intestinal mucosa, such as celiac disease. In these disease states, calcium absorption is impaired directly because of the reduction of the intestinal surface area and

1739

1740

Section X  Small and Large Intestine indirectly because of formation of insoluble calcium soaps with malabsorbed long-chain fatty acids. Therefore, dis­ eases causing malabsorption of long-chain fatty acids by other mechanisms, such as bile acid deficiency, also can result in calcium malabsorption.21 In many of these dis­ eases, malabsorption and deficiency of vitamin D further contribute to intestinal calcium malabsorption.21 Selective intestinal malabsorption of calcium—that is, without fat malabsorption—can occur in renal disease, hypopara­ thyroidism, and inborn defects in formation of 1α,25dihydroxyvitamin D or in the intestinal vitamin D receptor.21 Calcium malabsorption also occurs commonly after gastric resection (see subsequent section, “Malabsorption after Gastric Resection”).

MAGNESIUM

In many generalized malabsorptive disorders, magnesium malabsorption can result in magnesium deficiency.36 Mal­ absorption is due to the reduction in mucosal absorptive surface area and to luminal binding of magnesium by mal­ absorbed fatty acids; a congenital form of selective intestinal magnesium malabsorption also has been reported.37

IRON

Iron deficiency is common in patients with gastric resection or with celiac disease. Reduction in the mucosal surface area of the small intestine as a result of diffuse mucosal disease, intestinal resection, or intestinal bypass also can result in impaired iron absorption, potentially leading to iron deficiency38; a congenital form of iron malabsorption also has been described (see Table 101-14).39 Intestinal loss of iron from chronic gastrointestinal bleeding is, however, the most common gastrointestinal cause of iron deficiency.40 Worldwide, hookworm infection is a common cause of iron deficiency.

ZINC

Zinc, like other minerals, is malabsorbed in generalized mucosal diseases of the small intestine.41 A congenital selective defect of zinc absorption, acrodermatitis entero­ pathica, is caused by a defect in the zinc transport protein hZIP4.42

Colonic Salvage of Incompletely Absorbed Carbohydrates

In healthy people, between 2% and 20% of ingested starch escapes absorption in the small intestine47; pancreatic insuf­ ficiency or severe intestinal disorders further increase this amount.17 Carbohydrates that reach the colon cannot be absorbed by the colonic mucosa, but they can be metabo­ lized by the colonic bacterial flora. Metabolism by anaerobic bacteria results in the breakdown of oligosaccharides and polysaccharides to mono- and disaccharides, which are metabolized further to lactic acid; short-chain (C2 to C4) fatty acids (SCFAs) such as acetate, propionate, and butyr­ ate; and to odorless gases, including hydrogen, methane, and carbon dioxide.48 Studies in normal subjects have suggested that the bacte­ rial metabolism of starch to small carbohydrate moieties is a rapid process in the normal colon. The rate-limiting step in the overall conversion of polysaccharides to SCFAs appears to be the conversion of monosaccharides to SCFAs.17 Colonic absorption of SCFAs results in a reduction of the osmotic load and, as a result, in mitigation of osmotic diar­ rhea.49 In normal subjects, more than 45 g of carbohydrates must reach the colon to cause diarrhea, and up to 80 g of carbohydrates per day can be metabolized by bacteria to SCFAs; approximately 90% of these SCFAs are absorbed by colonic mucosa50 (Fig. 101-1). Chronic carbohydrate malab­ sorption causes adaptive changes in bacterial metabolic activity that result in an even higher efficiency of the bacte­ rial flora to digest carbohydrates,51 although at the expense of increased flatus production (see later).

Diarrhea CHO

Generalized malabsorption can cause deficiency of copper and selenium.43,44 In Menkes disease (kinky hair disease), an inherited disorder of cellular copper transport, selective intestinal copper malabsorption results (see later on). It is uncertain whether malabsorptive diseases result in deficien­ cies of chromium and manganese.41

MECHANISMS THAT COMPENSATE FOR MALABSORPTION

OA

OA H2, CO2

OTHERS

Gas

90% 20-90%

Caloric salvage

(CH4)

CHO

CHO

ROLE OF THE COLON

The colon has the capacity to absorb a limited number but a wide variety of substances and nutrients including sodium, chloride, water, oxalate, short chain fatty acids, calcium, and vitamin K. Although colonic nutrient absorption does not play a major role in health, the nutritive role of the colon in patients with severe malabsorption is clinically rele­ vant.45 Colonic preservation of malabsorbed nutrients also can result in symptoms and complications of malabsorp­ tion,46 such as colonic hyperabsorption of oxalate, which contributes to formation of renal stones (see later on).

Figure 101-1.  Carbohydrate metabolism and absorption of metabolic products in the colon. Up to 80 g of carbohydrate that reaches the colon can be metabolized by colonic bacteria to organic acids—lactic acid and the short-chain fatty acids acetate, proprionate, and butyrate—and to hydrogen, carbon dioxide, and methane. Approximately 90% of the organic acid produced is absorbed by colonic mucosa, which permits salvage of calories. Osmotic diarrhea results when organic acids that escape absorption and carbohydrate that escapes bacterial metabolism accumulate in the colon. Between 20% and 90% of gases produced in the colon is absorbed by the colonic mucosa; the remainder is excreted as flatus. CHO, carbohydrate; OA, organic acid.

Chapter 101  Maldigestion and Malabsorption Long-chain fatty acids bind calcium in the colon, thereby increasing the amount of sodium oxalate that is absorbed.60 Fatty acids with chain lengths longer than 12 carbons can cause diarrhea, because they increase mucosal permeability and inhibit colonic absorption of fluid and electrolytes.61 An increase in colonic permeability due to long-chain fatty acids also may be a contributing factor for the increased colonic oxalate absorption seen in patients with steatorrhea and hyperoxaluria.62 Patients with short bowel syndrome can gain caloric energy from colonic absorption of medium-chain fatty acids, coming from medium-chain triglyceride supplemen­ tation, if they have at least part of the colon in continuity with the remaining small intestine.63 In the rat colon, absorp­ tion of octanoate is not affected by the simultaneous pres­ ence of other luminal substrates.64

100

% Hydrogen in breath

80 r = –0.94 60

40

20

0 0

400

800

1200

Total flatus volume (mL/6 hr) Figure 101-2.  Relationship between flatus volume and colonic hydrogen absorption during fasting (open circles) and after ingestion of 12.5 grams of lactose (closed circles). At high flatus volumes, the fraction of hydrogen that is excreted in breath decreases to approximately 20% of total hydrogen excretion. The remaining 80% is excreted in flatus. (From Hammer HF. Colonic hydrogen absorption: Quantification of its effect on hydrogen accumulation caused by bacterial fermentation of carbohydrates. Gut 1993;34:818.)

Because SCFAs have caloric values between 3.4 and 5.95 kcal/g,52 their colonic absorption can contribute posi­ tively to overall calorie balance. In patients with short bowel syndrome, colonic salvage of malabsorbed carbohy­ drates can save up to 700 to 950 kcal/day, provided that a substantial part of the colon remains in continuity with the small intestine.53 Not all SCFAs are absorbed by the colon, and those that are not absorbed contribute to osmotic diarrhea. The beneficial effects of colonic bacterial carbohydrate metabolism may be accompanied by side effects due to gas production (see Chapter 16). Up to 10-fold differences in the volume of gas produced in the colon have been observed in normal persons.54 The colon also can absorb gas. If intra­ colonic gas volumes are low, up to 90% of the volume of intracolonic gas can be absorbed; if gas volumes are high, however, this proportion can decrease to 20%54 (Fig. 101-2). Therefore, persons who have the disadvantage of producing more gas in their colons have an additional disadvantage of absorbing a smaller fraction of the gas. Gas produced from bacterial carbohydrate metabolism is odorless. The odor of flatus is due to volatile sulfur-containing substrates that result from bacterial metabolism of protein.55 Impaired colonic salvage of carbohydrates has been suggested to contribute to the diarrhea in Crohn’s disease56 and ulcerative colitis.57 Bacterial carbohydrate metabolism may be lessened by antibiotic treatment.58 In some patients, antibiotic-associated diarrhea may be the result of impaired colonic salvage of carbohydrates that normally are not absorbed or the result of dietary fiber that can accumulate in stool because of decreased bacterial fermentation.59

Role of the Colon in Fat Malabsorption

Long-chain triglycerides or fatty acids, which constitute most dietary fat, cannot be absorbed by the human colon.

Colonic Salvage of Calcium

Although most unabsorbed calcium is insoluble when it reaches the terminal ileum,65 preservation of at least half of the colon in patients with extensive small bowel resection improves calcium absorption by about 40%, compared with calcium absorption in patients who have an ileostomy.66 Absorption of calcium requires solubilization of calcium salts. Bacterial metabolism of dietary fiber or incompletely absorbed carbohydrates can help solubilize calcium by causing a decrease in the pH of luminal contents in the colon. Once calcium is solubilized, it can contact the cecal mucosa, which in the rat has been demonstrated to be the site with the highest calcium absorption rate per surface area in the entire intestine.65 Calcium solubilization in the colon from bacterial fermentation of malabsorbed lactose also can occur in patients with lactose malabsorption, because in this condition, the bioavailability of calcium from milk is greater than that from mineral water.67 In addi­ tion to their effect on luminal pH, the SCFAs acetate and propionate, which are products of bacterial metabolism of lactose, have been shown to enhance calcium absorption directly in the human colon.68

ROLE OF INTESTINAL TRANSIT IN THE SALVAGE OF MALABSORBED NUTRIENTS

The lower parts of the gastrointestinal tract do not normally contact nutrients, and when they do, intestinal transit time is prolonged.69,70 This delay in transit could contribute to the compensation mechanisms in malabsorptive diseases; however, nutritional salvage by this mechanism has not been quantitated.

CLINICAL FEATURES AND EVALUATION Diagnosis of malabsorption requires suspecting its presence, confirming its existence, and demonstrating its cause. Malabsorption usually is suspected on the basis of the patient’s history, signs and symptoms, or findings on routine laboratory evaluations. Malabsorption of an ingested nutri­ ent or substrate can be confirmed by measuring its increased stool concentration or its decreased serum concentration or urinary excretion. Finding the cause of malabsorption often requires tests such as endoscopy with small intestinal biopsy; under certain clinical circumstances, noninvasive tests or radiologic imaging are helpful in providing a spe­ cific diagnosis.

1741

1742

Section X  Small and Large Intestine Table 101-4  Symptoms and Signs of Malabsorption and Relevant Pathophysiology SYMPTOM OR SIGN Gastrointestinal Diarrhea

Abdominal distention, flatulence Foul-smelling flatulence or stool Pain Ascites Musculoskeletal Tetany, muscle weakness, paresthesias Bone pain, osteomalacia, fractures Cutoneous and Mucosal Easy bruisability, ecchymoses, petechiae Glossitis, cheilosis, stomatitis Edema Acrodermatitis, scaly dermatitis Follicular hyperkeratosis Hyperpigmented dermatitis Thin nails with spoon-shaped deformity Perifollicular hemorrhage Spiral or curly hair Other Weight loss, hyperphagia Growth and weight retardation, infantilism Anemia Kidney stones Amenorrhea, impotence, infertility Night blindness, xerophthalmia Peripheral neuropathy Fatigue, weakness Neurologic symptoms, ataxia

PATHOPHYSIOLOGIC EXPLANATION Osmotic activity of carbohydrates or short-chain fatty acids Secretory effect of bile acids and fatty acids Decreased absorptive surface Intestinal loss of conjugated bile acids Ileal resection Severe ileal mucosal disease Congenital defects of the ileal sodium–bile acid cotransporter Bacterial gas production from carbohydrates in colon, small intestinal bacterial overgrowth Malabsorption of proteins or intestinal protein loss Gaseous distention of intestine Protein loss or malabsorption Malabsorption of vitamin D, calcium, magnesium, and phosphate Protein, calcium, or vitamin D deficiency; secondary hyperparathyroidism Vitamin K deficiency and vitamin C deficiency (scurvy) Vitamin B complex, vitamin B12, folate, or iron deficiency Protein loss or malabsorption Zinc and essential fatty acid deficiency Vitamin A deficiency Niacin deficiency (pellagra) Iron deficiency Malabsorption of vitamin C Malabsorption of vitamin C Nutrient malabsorption Nutrient malabsorption in childhood and adolescence Iron, folate, or vitamin B12 deficiency Increased colonic oxalate absorption Multifactorial (including protein malabsorption, secondary hypopituitarism, anemia) Vitamin A deficiency Vitamin B12 or thiamine deficiency Calorie depletion, iron and folate deficiency, anemia Vitamin B12, vitamin E, or folate deficiency

SUSPECTING AND CONFIRMING THE PRESENCE OF MALABSORPTION History and Physical Examination

Table 101-4 lists symptoms and signs suggestive of malab­ sorption, although virtually all can have causes other than malabsorption. For example, greasy stools might indicate malabsorption, but a greasy appearance also can be due to mucus in stool. Floating of stool in the toilet water can be due to a high stool fat content, but it also can be caused by high gas content. Nevertheless, such symptoms and signs are helpful in raising the clinician’s index of suspicion and in guiding the physician as to which specific laboratory tests, structural evaluations, or function tests should be ordered. The current obesity epidemic has led to a changing picture of malabsorption; for example, few patients today with celiac disease are underweight at diagnosis and some are even overweight. These patients have been reported to be less likely to present with classic features, such as diarrhea or anemia. In these patients, a further increase in weight after dietary gluten exclusion may be a cause of morbidity.71

Laboratory Findings

Certain blood tests might yield abnormal results in malab­ sorption, but with rare exceptions they are not specific for malabsorptive diseases. Blood tests also can be used as a screening tool to help the physician decide how vigorously

to evaluate malabsorption. Table 101-5 lists blood tests in which abnormal results should raise the suspicion of mal­ absorption and stool tests that should be used to confirm the suspicion of malabsorption. Quantitative fecal fat measurement followed by measure­ ment of fecal chymotrypsin or elastase concentration may be helpful, both in establishing malabsorption and in dif­ ferentiating between pancreatic and intestinal causes of malabsorption. Low levels of serum β-carotene, cholesterol, triglycerides, and calcium and a prolonged prothrombin time suggest malabsorption of fat and fat-soluble vitamins. Low levels of vitamin B12, folate, iron, and albumin suggest malabsorption of water-soluble substances and, therefore, indicate intestinal disease rather than pancreatic or biliary disease. Severe deficiency of fat-soluble vitamins might indicate intestinal or biliary disorders. Low levels of plasma citrulline are associated with destructive small intestinal disease, such as celiac disease, or can follow intestinal resection,72 although fasting plasma citrulline tests are poor predictors of enterocyte dysfunction in clinical practice; an oral citrulline generation test has been proposed to improve its predictive value.73

DIAGNOSTIC APPROACH Clinical Clues to the Presence of Specific Diseases Clinical clues (Table 101-6) or results of laboratory tests (Table 101-7)74 can indicate the presence of a specific

Chapter 101  Maldigestion and Malabsorption Table 101-5  Laboratory Tests That Are Useful in Patients with Suspected Malabsorption and for Establishing Possible Nutrient Deficiencies TEST Blood Cell Count Hematocrit, hemoglobin Mean corpuscular hemoglobin or mean corpuscular volume White blood cells, differential Biochemical Tests (Serum) Triglycerides Cholesterol Albumin Alkaline phosphatase Calcium, phosphorus, magnesium Zinc Iron, ferritin Other Serum Tests Prothrombin time β-Carotene Immunoglobulins Folic acid Vitamin B12 Methylmalonic acid Homocysteine Citrulline Stool Tests Fat Elastase, chymotrypsin pH

COMMENT(S) Decreased in iron, vitamin B12, and folate malabsorption or with blood loss Decreased in iron malabsorption; increased in folate and vitamin B12 malabsorption Decreased in vitamin B12 and folate malabsorption; low lymphocyte count in lymphangiectasia Decreased in severe fat malabsorption Decreased in bile acid malabsorption or severe fat malabsorption Decreased in severe malnutrition, lymphangiectasia, protein-losing enteropathy Increased in calcium and vitamin D malabsorption (severe steatorrhea); decreased in zinc deficiency Decreased in extensive small intestinal mucosal disease, after extensive intestinal resection, or in vitamin D deficiency Decreased in extensive small intestinal mucosal disease or intestinal resection Decreased in celiac disease, in other extensive small intestinal mucosal diseases, and with chronic blood loss Prolonged in vitamin K malabsorption Decreased in fat malabsorption from hepatobiliary or intestinal diseases Decreased in lymphangiectasia, diffuse lymphoma Decreased in extensive small intestinal mucosal diseases, with anticonvulsant use, in pregnancy; may be increased in small intestinal bacterial overgrowth Decreased after gastrectomy, in pernicious anemia, terminal ileal disease, and small intestinal bacterial overgrowth Markedly elevated in vitamin B12 deficiency Markedly elevated in vitamin B12 or folate deficiency May be decreased in destructive small intestinal mucosal disease or intestinal resection Qualitative or quantitative increase in fat malabsorption Decreased concentration and output in exocrine pancreatic insufficiency Less than 5.5 in carbohydrate malabsorption

Table 101-6  Cardinal Clinical Features of Specific Malabsorptive Disorders DISorder

CARDINAL CLINICAL FEATURES

Adrenal insufficiency Amyloidosis

Skin darkening, hyponatremia, hyperkalemia Renal disease, nephrotic syndrome, cardiomyopathy, neuropathy, carpal tunnel syndrome, macroglossia, hepatosplenomegaly Ileal resection or disease, liver disease Flushing, cardiac murmur Variable symptoms: dermatitis herpetiformis, alopecia, aphthous mouth ulcers, arthropathy, neurologic symptoms, and (life-threatening) malnutrition; abnormal liver biochemical test results, mild iron deficiency Arthritis, aphthous mouth ulcers, episcleritis, uveitis, pyoderma gangrenosum, erythema nodosum, abdominal mass, fistulas, primary sclerosing cholangitis (PSC), laboratory signs of inflammation Chronic sinopulmonary disease, meconium ileus, distal intestinal obstruction syndrome (DIOS), elevated sweat chloride Kidney stones, dermatosis Long history of diabetes and diabetic complications Bloating and cramping, intermittent diarrhea Previous intestinal surgery or trauma, Crohn’s disease Migratory necrolytic erythema Symptoms and signs of thyroid disease Recurrent infections Other ischemic organ manifestations; abdominal pain with eating (chronic mesenteric ischemia) Enlarged mesenteric or retroperitoneal lymph nodes, abdominal mass, abdominal pain, fever Urticaria pigmentosum, peptic ulcer Acquired immunodeficiency syndrome History of pancreatitis, abdominal pain; or alcoholism; large-volume fatty, oily stools History of travel to endemic areas Jaundice, itching Dysphagia, Raynaud’s phenomenon, skin tightening Previous intestinal surgery, motility disorder (scleroderma, pseudo-obstruction), small intestinal diverticula, strictures History of travel to endemic area Specific history of exposure, living in or travel to endemic area, immunosuppression, abdominal mass or intestinal obstruction, ascites Lymphadenopathy, fever, arthritis, cerebral symptoms, heart murmur Peptic ulcers, diarrhea

Bile acid deficiency Carcinoid syndrome Celiac disease Crohn’s disease Cystic fibrosis Cystinuria, Hartnup’s disease Diabetes mellitus Disaccharidase deficiency Gastrointestinal fistulas Glucagonoma Hyperthyroidism, hypothyroidism Hypogammaglobulinemia Intestinal ischemia Lymphoma Mastocytosis Mycobacterium-avium complex infection Pancreatic insufficiency Parasitic infection Primary biliary cirrhosis Scleroderma Small intestinal bacterial overgrowth Tropical sprue Tuberculosis Whipple’s disease Zollinger-Ellison syndrome

1743

1744

Section X  Small and Large Intestine Table 101-7  Laboratory Tests That Are Useful in the Differential Diagnosis of Malabsorption TEST Blood Cell Count Acanthocytes Nuclear remnants in erythrocytes (Howell-Jolly bodies) White blood cells, differential Platelets Other Tests ESR, C-reactive protein Ferritin Iron Liver biochemical tests Immunologic Markers Immunoglobulins Allergen-specific IgE Autoantibodies (e.g., ANA) HLA-DQ2 or HLA-DQ8 Antimitochondrial autoantibodies HIV-ELISA/Western blot Neuroendocrine Markers ACTH, cortisol Chromogranin A 5-Hydroxyindoleacetic acid in urine Gastrin* Glucagon* Serum TSH Somatostatin* Tissue transglutaminase antibodies, EMA Stool Tests Occult blood test Ova and parasites Leukocytes

COMMENT Abetalipoproteinemia Splenic atrophy in celiac disease, inflammatory bowel disease, radiation enteritis, amyloidosis Eosinophilia in eosinophilic gastroenteritis and parasitic disease Low lymphocyte count in lymphangiectasia, tuberculosis, protein-losing enteropathy Low CD4+ count in AIDS Increased in inflammatory diseases Increased in Crohn’s disease, Whipple’s disease, lymphoma Increased in inflammatory diseases, lymphoma; decreased in iron deficiency Decreased in celiac disease, chronic occult intestinal bleeding, chronic inflammatory diseases Increased in primary biliary cirrhosis and other liver diseases, celiac disease IgA deficiency, immunodeficiency syndromes IgE-mediated hypersensitivity Connective tissue diseases Celiac disease, refractory sprue Primary biliary cirrhosis AIDS Abnormal values in Addison’s disease Elevated in neuroendocrine tumors Elevated in carcinoid syndrome Elevated in Zollinger-Ellison syndrome Elevated in glucagonoma Decreased in hyperthyroidism; increased in hypothyroidism Elevated in somatostatinoma (normal in duodenal somatostatinoma) Celiac disease Erosive or ulcerative intestinal disease or tumor Repeated samples may be needed to detect Giardia lamblia Present in some inflammatory diseases of the intestine

*Perform this test if there is a strong suspicion of an underlying neuroendocrine tumor.77 ACTH, adrenocorticotropic hormone; AIDS, acquired immunodeficiency syndrome; ANA, antinuclear antibodies; EMA, endomysial antibodies; ESR, erythrocyte sedimentation rate; HIV, human immunodeficiency virus; HLA, human leukocyte antigen; Ig, immunoglobulin; TSH, thyroid-stimulating hormone.

underlying disease or can help in the differential diagnosis. In addition, the following questions may be helpful and should be asked as part of the history before physical examination: Has the patient undergone previous surgery, such as gastric or small bowel resection or a gastrointestinal bypass operation? Is there a family or childhood history of celiac disease? Is there a history of travel to endemic areas of tropical sprue, giardiasis, or other gastrointestinal infections? Is there excessive alcohol consumption? Does the patient have a history of chronic pancreatitis or symptoms suggesting a pancreatic tumor? Does the patient have clinical features of thyrotoxicosis, Addison’s disease, Whipple’s disease, biliary or liver disease, or diabetic neuropathy? Does the patient eat a diet high in poorly absorbable carbohydrates (sweeteners such as sorbitol or fructose) or fat substitutes or an unbalanced diet that could result in malnutrition? Is there a likelihood of human immunodeficiency virus infection? Is the patient receiving treatment with a drug that can cause malabsorption? Does the patient have a history of stem cell or organ trans­ plantation or abdominal radiation? Does the patient have a history of extraintestinal manifesta­ tions of inflammatory bowel disease, celiac disease, or Whipple’s disease

A rational approach to establishing the cause of malab­ sorption can require several diagnostic steps. Depending on the clinician’s background, the availability of different tests, and the patient’s preferences, different diagnostic approaches may be used. If time constraints are not a consideration, a stepwise approach may be used, starting with noninvasive evaluations that can guide further invasive procedures or even provide a diagnosis. In other instances, the physician may choose a more invasive test in the hope of reaching a diagnosis with the fewest possible tests and in the shortest possible time. Diagnostic approaches differ depending on the epidemiologic or ethnic background of an individual patient. For example, if parasitic infections are a likely possibility, stool examination can provide a rapid diagnosis by noninvasive testing. In populations with a very low prevalence of lactose intolerance, a secondary cause of lactose malabsorption is more likely than it would be in populations with a high prevalence of acquired primary lactase deficiency, and therefore, additional tests are appropriate. The sequence of tests thus depends on the affected per­ son’s symptoms and history, as well as results of previous testing (Table 101-8). Tests that can detect the most common causes of malabsorption or are noninvasive or inexpensive usually should be performed initially (first-line tests). In some patients, testing for rarer causes of malabsorption and use of more invasive or more expensive tests may be necessary to establish the diagnosis (second-line tests). For unusually difficult cases, additional tests may be

Chapter 101  Maldigestion and Malabsorption Table 101-8  Tests to Establish the Cause of Malabsorption Based on Main Symptoms Weight Loss, Osteomalacia or Osteopenia, Diarrhea, Suspected Steatorrhea, or Deficiency of Fat-Soluble Vitamins First-line Tests Abdominal ultrasonography Chymotrypsin and/or elastase concentration in stool EGD with small intestinal biopsies Endomysial and tissue transglutaminase antibodies Laboratory tests (complete blood cell count, white blood cell differential, cholesterol, triglycerides, electrolytes, calcium, magnesium, serum ALT, AST, AP, bilirubin levels, prothrombin time, serum albumin level, erythrocyte sedimentation rate and C-reactive protein, TSH) Ova, parasites, and leukocytes in stool Second-line Tests Abdominal CT, MRI Endoscopic examination of the terminal ileum, including ileal biopsies ERCP/MRCP More extensive laboratory investigation (immunoglobulins, human immunodeficiency virus ELISA, antinuclear antibodies, ferritin, food allergen– specific IgE, adrenocorticotropic hormone, cortisol, chromogranin A, gastrin, urinary 5-HIAA) Quantitative fecal fat Quantitative small intestinal culture or breath tests for bacterial overgrowth Small bowel series/small bowel MRI Special staining of small intestinal biopsies (e.g., Congo red for amyloid, PAS for Whipple’s disease) Therapeutic trial of pancreatic enzymes, antibiotics (tetracycline, metronidazole), or a gluten-free diet Video capsule endoscopy Tests in Unusually Difficult Cases (Third-line Tests) Abdominal angiography Antienterocyte antibodies Endoscopic ultrasonography Enteroscopy, including biopsies Magnetic resonance angiography Positron emission tomography Serum or plasma glucagon, somatostatin Somatostatin (octreotide) scan Special tests of intestinal biopsies (e.g., flow cytometry of intraepithelial lymphocytes for lymphoma and refractory celiac disease, PCR for Tropheryma whipplei or other infective organisms, chromogranin A stain for enteroendocrine cells) Spiral CT of the pancreas Tests for bile acid malabsorption Tube test for exocrine pancreatic secretion (secretin, cholecystokinin, or Lundh test) Bloating, with or without Diarrhea First-line Tests Fructose H2 breath test Lactose H2 breath test Lactose tolerance test Second-line Tests Chymotrypsin and/or elastase concentration in stool EGD with duodenal biopsies Endomysial and tissue transglutaminase antibodies Genetic testing for hypolactasia Quantitative small intestinal culture or breath tests for bacterial overgrowth Stool pH (in patients with diarrhea) Anemia and Suspected Malabsorption Microcytic or hypochromic anemia (low MCV, MCH) EGD with duodenal biopsies Endomysial and tissue transglutaminase antibodies Iron, ferritin, and transferrin in serum Ova and parasites in stool Video capsule endoscopy Macrocytic anemia (high MCV, MCH) First-line Tests Folic acid in serum or red blood cells Vitamin B12 in serum Second-line Tests in Cases of Vitamin B12 Deficiency CT, small bowel series, enteroclysis, video capsule endoscopy EGD with gastric and duodenal biopsies Endomysial and tissue transglutaminase antibodies Evaluation of ileum (e.g., double balloon enteroscopy with biopsy, colonoscopy to ileum with biopsy) Ova and parasites in stool Quantitative small intestinal culture or breath tests for bacterial overgrowth Schilling test (with and without intrinsic factor) Second-line Tests in Cases of Folate Deficiency EGD with duodenal biopsies Endomysial and tissue transglutaminase antibodies ALT, alanine aminotransferase; AP, alkaline phosphatase; AST, aspartate aminotransferase; CT, computed tomography; EGD, esophagogastroduodenoscopy; ERCP, endoscopic retrograde cholangiopancreatography; 5-HIAA, 5-hydroxyindoleacetic acid; IgE, immunoglobulin E; MCH, mean corpuscular hemoglobin; MCV, mean corpuscular volume; MRCP, magnetic resonance cholangiopancreatography; MRI, magnetic resonance imaging; PAS, periodic acid–Schiff; PCR, polymerase chain reaction; TSH, thyroid-stimulating hormone.

1745

1746

Section X  Small and Large Intestine required that may be available only in specialized centers (third-line tests). For some disorders, such as bile acid malabsorption, lactose malabsorption, and bacterial overgrowth, it may be difficult to establish a causal link between symptoms and the malabsorbed substrate. In these conditions, observation of the response to therapy may be critical in proving or disproving a causal relationship.

ANATOMIC INVESTIGATIONS

Endoscopic examination of the stomach, duodenum, or ileum and histologic examination of mucosal biopsy speci­ mens can establish a diagnosis of some conditions causing malabsorption. The role of radiologic imaging examinations is limited mostly to answering questions about abdominal regions not easily accessible to endoscopy, such as parts of the small intestine, parenchymatous organs, the peri­ toneal cavity, the mesentery, or the retroperitoneum. Capsule endoscopy and magnetic resonance imaging (MRI) are contributing to making these areas more accessible to diagnostic evaluation. Radiologic studies of the small intes­ tine can show evidence of stasis, blind loops, diverticula, fistulas, rapid transit, and other abnormalities that can assist in diagnosis (see later on).

ENDOSCOPY, BIOPSY, AND DUODENAL ASPIRATION Endoscopy

Endoscopic inspection of the duodenal mucosa can provide clues to some causes of malabsorption. Aphthae suggest Crohn’s disease, and small, diffuse, white, punctate lesions can be seen in primary or secondary lymphangiectasia. Mosaic-like scalloping of duodenal folds (Fig. 101-3) and reduction in the number of duodenal folds are highly suggestive of villus atrophy in celiac disease, although these abnormalities may be seen in other diseases (see Chapter 104).75 Villus atrophy may be seen endoscopically using magnification endoscopy76 and chromoendoscopy with indigocarmine staining; however, a normal duodenal fold pattern should not deter the endoscopist from taking

Figure 101-3.  Endoscopic image showing scalloping of the duodenal folds in a patient with celiac disease.

mucosal biopsy specimens. Endocrine tumors causing mal­ absorption, such as duodenal gastrinomas or somatostatino­ mas or ampullary tumors obstructing the pancreatic duct, also can be detected during endoscopy. If ileal disease is the suspected cause of malabsorption, visual examination and biopsy of the ileal mucosa may be required to establish a diagnosis; this can be accomplished by retrograde intuba­ tion of the ileum at colonoscopy or by double-balloon endoscopy.

Biopsy

Examination of endoscopic biopsy specimens from the duo­ denum may be diagnostic or highly suggestive of a variety of small bowel disorders resulting in malabsorption (Table 101-9); follow-up small intestinal biopsy can be used to assess treatment effects. Duodenal biopsy specimens should be obtained from patients with atypical or nonspecific gas­ trointestinal symptoms, including abdominal pain, bloat­ ing, and weight loss, and should not be limited only to patients with diarrhea.77,78 Endoscopic biopsy is an adequate substitute for jejunal suction biopsy,79 and its advantages over capsule biopsy are that multiple specimens are more easily obtained and focal or patchy lesions can be identified and targeted for sampling.80 Compared with duodenal biop­ sies, endoscopically obtained biopsies from the jejunum to find changes of celiac disease are helpful in only very few patients.81 The adequacy of mucosal biopsy specimens is a function of their size and the number obtained.82 If large specimens can be obtained using jumbo biopsy forceps, they can be oriented on a piece of filter paper before they are put into a fixing solution83; two or three jumbo biopsy specimens usually are sufficient to allow histologic section­ ing parallel to the villi and crypts. Specimens also may be obtained with smaller forceps, although the number of specimens obtained must then be increased to four to six. Specimens can be inspected with a low-power dissecting microscope or by magnification endoscopy to obtain an initial impression of the villus architecture and to ensure proper orientation before they are placed in formalin. The diagnostic yield of biopsy is influenced by the distri­ bution of histologic abnormalities, which in some diseases is diffuse but in other diseases is patchy. Tropical diarrhea malabsorption syndrome (tropical sprue; see Chapter 105), abetalipoproteinemia, and immunodeficiency usually result in a diffuse alteration of small intestinal mucosa. Thus, a completely normal appearance of a duodenal biopsy speci­ men rules out these disorders. Primary lymphangiectasia has a patchy distribution, so that a single mucosal biopsy might not rule out the disorder (see Chapter 28). Patchy distribution also has been described for the histologic changes in some patients with celiac disease, especially when symptoms are subtle, although this disorder usually affects the small intestine diffusely.84 Other possible sources of error and misdiagnosis include poorly oriented speci­ mens and those obtained proximally, where peptic injury can be the cause of mucosal alterations. Additional biopsy specimens from the stomach and duodenal bulb can help the pathologist to establish the extent of peptic injuries in the upper gastrointestinal tract and to interpret inflamma­ tory changes in the duodenum in relation to these lesions. Distortion of villus architecture over Brunner’s glands or lymphoid aggregates, common in the duodenum, should be interpreted with caution. Specific histologic features may be diagnostic for some rare causes of malabsorption (see Table 101-9)85 such as Whipple’s disease (Fig. 101-4), abetalipoproteinemia or hypobetalipoproteinemia, intestinal lymphangiectasia, giar­ diasis (Fig. 101-5), lymphoma, or collagenous sprue. In most

Chapter 101  Maldigestion and Malabsorption Table 101-9  Causes of Malabsorption That Can Be Diagnosed by Small Bowel Biopsy CAUSE OF MALABSORPTION Generalized Histologic Abnormalities Abetalipoproteinemia, hypobetalipoproteinemia Collagenous sprue (Chapter 104) Mycobacterium-avium complex infection (Chapter 33) Whipple’s disease (Chapter 106) Patchy Histologic Abnormalities Amyloidosis (Chapter 35) Crohn’s disease (Chapter 111) Eosinophilic gastroenteritis (Chapter 27) Lymphangiectasia (Chapter 28) Lymphoma (Chapter 29) Mastocytosis (Chapter 35) Parasites (Giardia lamblia, Strongyloides stercoralis, coccidia) (Chapters 109, 110)

MAIN HISTOLOGIC FEATURES Lipid accumulation and vacuolization of enterocytes Collagenous band below atrophic epithelium Acid-fast bacilli, foam cells Foamy macrophages with PAS-positive inclusion bodies Congo red–stained deposits with apple-green birefringence in polarized light Epithelioid granulomas and characteristic focal inflammation Eosinophilic infiltration Ectatic lymph vessels Clonal expansion of lymphocytes Diffuse infiltration with mast cells Some parasites may be seen on histologic examination

PAS, periodic acid–Schiff. Modified from Riddell RH. Small intestinal biopsy: Who? how? what are the findings? In Barkin JS, Rogers AI, editors. Difficult Decisions in Digestive Diseases. Chicago: Year Book; 1989. p 326.

Figure 101-5.  Small bowel biopsy specimen from an immunocompetent patient with giardiasis. A normal-appearing villus and adjacent pearshaped organisms with red-staining nuclei are evident. (Courtesy of Cord Langner, MD.)

A

B Figure 101-4.  Duodenal biopsy specimen from a patient with Whipple’s disease. A, Hematoxylin and eosin staining shows villus blunting. The lamina propria is infiltrated with pale-staining foamy macrophages. (Courtesy of Cord Langner, MD.) B, High-power view demonstrates purple-red macrophages. (Periodic acid–Schiff stain.) (Courtesy of Günter J. Krejs, MD, Graz, Austria.)

patients with small intestinal disorders, however, histologic examination is not diagnostic85 (Table 101-10) and reveals a spectrum of mucosal responses ranging from infiltration by lymphocytic cells to a flat mucosa with villus atrophy and crypt hyperplasia (Fig. 101-6). In many parts of the world, celiac disease is by far the most common cause of this type of histologic alteration, but a definite diagnosis of celiac disease cannot be established by mucosal biopsy alone (see Chapter 104). Some disease states can be identified only with use of special histologic stains, such as Congo red (intestinal amy­ loidosis), periodic acid–Schiff (PAS) (Whipple’s disease), or immunohistochemical techniques for detecting refractory celiac disease, small intestinal lymphoma, or enteroendo­ crine insufficiency (see later on). Polymerase chain reaction analysis of intestinal biopsy specimens for Tropheryma whipplei may be helpful in evaluating patients in whom Whipple’s disease is suspected (see Chapter 106).86 In cases in which these diseases are a possibility, the clinician has to request these specific tests. Measurement of mucosal enzyme activities in a jejunal biopsy can be used to confirm disaccharidase deficiency, although this is not recom­ mended for routine clinical use.

1747

1748

Section X  Small and Large Intestine Table 101-10  Malabsorptive Diseases with Abnormal but Not Diagnostic Small Intestinal Histologic Findings Increased Lymphocyte Infiltration with or without Crypt Hyperplasia AIDS enteropathy (Chapter 33) Celiac disease (Chapter 104) Infection (due to Giardia lamblia, Cryptosporidium [Chapter 109]; viral enteritis [Chapter 107]) Tropical sprue (Chapter 105) Flat Lesion with or without Mucosal Inflammation Celiac disease (Chapter 104) Drug-induced enteropathy (NSAIDs, colchicine, neomycin) Food protein hypersensitivity (rye, barley, egg, fish, rice, poultry) (Chapters 9, 27) Immunodeficiency (hypogammaglobulinemia) (Chapter 2) Immunoproliferative small intestinal disease (IPSID) (Chapters 29, 104, and 115) Infection (due to Giardia lamblia, Cryptosporidium) (Chapter 109) Intestinal transplantation (Chapter 34) Lymphoma (Chapter 29) Nongranulomatous chronic idiopathic enterocolitis Autoimmune enteropathy Prolonged folate or cobalamin deficiency Protein-calorie malnutrition Traumatic injury Tropical sprue (Chapter 105) Atrophic Lesion Chronic radiation damage (Chapter 39) Cicatrizing Crohn’s disease (Chapter 111) Diffuse lymphoma (Chapter 29) Idiopathic diarrhea of infancy (microvillus inclusion disease) (Chapter 96) Unresponsive gluten sensitivity (lymphoma or ulcerative jejunitis) (Chapter 104)

A

AIDS, acquired immunodeficiency syndrome; NSAIDs, nonsteroidal antiinflammatory drugs. Modified from Riddell RH. Small intestinal biopsy: Who? how? what are the findings? In Barkin JS, Rogers AI, editors. Difficult Decisions in Digestive Diseases. Chicago: Year Book; 1989. p 326.

Aspiration

Fluid aspirated from the descending part of the duodenum may be examined microscopically for Giardia lamblia (see Chapter 109) or cultured to detect bacterial overgrowth in patients with diffuse small intestinal motility disorders (see Chapters 97 and 102).

Video Capsule Endoscopy

Video capsule endoscopy (VCE) is an increasingly popular technique for diagnosing diseases of the small intestine. VCE was initially introduced for evaluating suspected bleeding in the small intestine, but subsequently it has been used to diagnose a wider range of diseases such as Crohn’s disease, celiac disease, and other malabsorptive disorders. In several studies, lesions suggesting Crohn’s disease were detected by VCE when they had been missed by conven­ tional diagnostic procedures.87 These reports need to be interpreted carefully, because no biopsy specimens were obtained, and long-term evaluations to confirm the diagno­ sis are lacking. VCE appears to be superior to conventional radiologic imaging of the small intestine and to computed tomogra­ phy (CT) with small bowel enteroclysis to detect subtle mucosal changes, such as aphthous or erosive lesions of the small intestine.87 In celiac disease, the detection of villus atrophy by VCE has a good correlation to villus atrophy seen in duodenal biopsy specimens,88,89 but it is

B Figure 101-6.  Duodenal biopsy specimen from a patient with untreated celiac disease. A, Subtotal villus atrophy, crypt elongation, and lymphoplasmacytic infiltration of the lamina propria can be seen. B, High-power view demonstrates villus blunting with increased intraepithelial lymphocytes. (Hematoxylin and eosin stain.) (Courtesy of Cord Langner, MD.)

questionable if this procedure can detect subtle changes, such as Marsh 1 and 2 lesions. Changes on VCE that suggest villus atrophy are scalloping, mosaic pattern, and fissuring. In a recent study of VCE in patients with celiac sprue, villus atrophy was seen in the duodenum and jejunum in 59% of cases, in the duodenum only in 32%, and in the jejunum only in 3%.89 In refractory celiac disease, VCE can detect changes such as ulcerations and strictures that suggest T-cell lymphoma but that are missed by conventional techniques.90 This test may be used in patients with established malabsorption in whom no diagnosis has been established despite extensive diagnostic workup.

Chapter 101  Maldigestion and Malabsorption ABDOMINAL IMAGING Small Bowel Follow-through and Small Bowel Enteroclysis

The principal role of small bowel radiologic series in evalu­ ating malabsorption is to identify focal or diffuse abnormali­ ties and alterations that predispose to bacterial overgrowth, including diverticula, stagnant loops of intestine, general­ ized intestinal hypomotility or dilatation, intestinal fistulas, and tumors.91 Small bowel enteroclysis is preferred to small bowel follow-through examinations, because distention of the lumen results in better demonstration of the small bowel contour.92 Double-contrast enteroclysis, in which intubation of the upper jejunum is used to instill contrast material directly into the upper jejunum, has a higher sensitivity than small bowel series for detecting mucosal changes, although it is less acceptable to the patient and can miss focal changes in the duodenum, such as diverticula. Use of an intravenous agent such as glucagon to reduce motility enables overlapping loops of small intestine to be separated and imaged more distinctly. Alterations associated with diffuse, localized, or distal mucosal changes that might have been missed by proximal mucosal biopsy also may be identified. Normal findings on small bowel series do not rule out intestinal causes of mal­ absorption and should not dissuade the clinician from per­ forming biopsy of the small intestine. Ulcerations and strictures may be seen in various malab­ sorptive disorders, including Crohn’s disease, radiation enteritis, celiac disease, intestinal lymphoma, and tubercu­ losis. Aphthous ulcers and cobblestoning of the mucosa, either alone or with thickened and distorted folds, are fea­ tures of Crohn’s disease but also can be present in other conditions. Reduced numbers of jejunal folds and an increased number of and thickening of ileal folds can suggest celiac disease.91 Mass lesions can be found with intestinal lymphoma or, rarely, with hormone-producing tumors. The disadvantage of conventional enteroclysis is that direct imaging of the bowel wall and surrounding structures is not possible, and overlapping bowel loops potentially impair complete visualization of the whole small bowel— hence the rationale for combining enteroclysis with CT or MRI scanning.93

Abdominal Computed Tomography

Abdominal CT for small bowel investigation is performed after administration of oral or intravenous contrast agents.94 Small intestinal CT scanning is useful to detect focal intes­ tinal lesions, such as thickening of the small bowel wall in Crohn’s disease or small intestinal lymphoma, intestinal fistula, and dilated bowel loops; however, mild mucosal changes such as aphthae in Crohn’s disease or villus atrophy of various causes are missed by this technique. Diffuse thickening of the small bowel may be seen in Whipple’s disease and in graft-versus-host disease.94 In some cases of celiac disease, reversal of the jejunoileal fold pattern is observed.95 CT is a sensitive test for detecting enlarged abdominal lymph nodes, which can be present in disorders such as Whipple’s disease, small bowel lymphoma, or small intestinal inflammatory diseases such as Crohn’s disease. Evidence for pancreatic disease that may be detected on CT includes calcifications of the pancreas, dilatation of the pan­ creatic duct, and pancreatic atrophy. Tumors obstructing the pancreatic duct or hormone-secreting neuroendocrine tumors also can be located by CT.

Magnetic Resonance Imaging of the Small Intestine

MRI may be used to image the small intestine either with administration of oral contrast solutions or by enteroclysis. Segmental bowel wall thickening with inflammatory involvement of the mesentery, cobblestoning, and ulcer­ ations may be seen in Crohn’s disease; this method is very sensitive for demonstrating complications of Crohn’s disease, such as intestinal fistula formation. In celiac dis­ ease, small bowel MRI with oral contrast can demonstrate small intestinal dilatation, mucosal thickening, and an increased number of folds in the ileum (ileal jejunization) with flattening of the jejunal folds.96 Most of these signs also are found in other inflammatory diseases of the intestine, but the fold pattern abnormalities are most specific for celiac disease.96 This method also is very useful to detect changes suggesting complications, such as lymphoma or carcinoma. With MRI enteroclysis, subtle mucosal changes might be missed and be more evident on conventional small bowel enteroclysis97 or capsule endoscopy.87 Because MRI or CT imaging of the small intestine requires no tube place­ ment, these techniques have largely replaced classic small bowel enteroclysis.

Other Radiologic Studies

A plain film of the abdomen may be helpful to detect pan­ creatic calcifications if exocrine pancreatic insufficiency is suspected. However, morphologic signs of chronic pancre­ atitis alone do not prove a pancreatic cause of malabsorp­ tion, because the function of the exocrine pancreas must be severely impaired before malabsorption becomes evident. A plain film of the abdomen also can document dilated loops of intestine; dilated loops predispose to small bowel bacterial overgrowth or suggest the presence of an obstruction. Endoscopic retrograde cholangiopancreatography (ERCP) can help establish the cause of pancreatic insufficiency (see Chapter 59). It can help distinguish chronic pancreatitis from pancreatic tumor and can document pancreatic duct stones. ERCP and endoscopic ultrasound (EUS) are the methods of choice for documenting various causes of biliary obstruction. Magnetic resonance pancreatography (MRCP) is increasingly being used to replace diagnostic ERCP. If a neuroendocrine tumor (e.g., gastrinoma, somatostatinoma) is the suspected cause of malabsorption, an indium-111 octreotide scintigraphic scan, positron emission tomogra­ phy (PET), or an EUS examination of the pancreas may be helpful in establishing the diagnosis or demonstrating the extent of disease (see Chapters 31 and 32). Transabdominal ultrasound examination has the advan­ tage of no radiation exposure and therefore can be used in pregnant patients. Ultrasonography often is used to investi­ gate the pancreas, although the sensitivity for detecting pan­ creatic tumors is lower than that of ERCP or CT. Nevertheless, obstruction of the biliary tract, pancreatic calcifications, dilatation of the pancreatic duct, or stones within the pan­ creatic duct may be demonstrated. Ultrasound examination also may be used to document thickening of the bowel wall, abscesses, and fistula in Crohn’s disease.

NONINVASIVE EVALUATION OF GASTROINTESTINAL DIGESTIVE AND ABSORPTIVE FUNCTION

Some conditions causing malabsorption can be diagnosed with noninvasive tests, although, as pointed out in Table 101-11, diagnostic accuracy may be limited, and further tests may be necessary to identify underlying diseases or to

1749

Section X  Small and Large Intestine Table 101-11  Malabsorptive Diseases or Conditions in Which Noninvasive Tests Can Establish Malabsorption or Provide a Diagnosis DISEASE OR CONDITION

DIAGNOSTIC TEST(S)*

COMMENT(S)

Lactose malabsorption

Lactose hydrogen breath test Lactose tolerance test Fructose hydrogen breath test 14 C-d-xylose breath test Glucose hydrogen breath test Schilling test with and without antibiotics SeHCAT test, 14C-TCA test

Tests do not differentiate between primary and secondary lactose malabsorption Questionable clinical relevance A predisposing factor should be sought if the result of any of the tests is positive

Fructose malabsorption Small intestinal bacterial overgrowth (see Chapter 102) Bile acid malabsorption Exocrine pancreatic insufficiency Vitamin B12 malabsorption

Does not differentiate between primary and secondary causes To establish malabsorption in chronic pancreatitis Variable sensitivity and specificity, depending on type of test and stage of the disease Test is performed without intrinsic factor and, depending on result with intrinsic factor, with antibiotics or with pancreatic enzymes (see text). Further tests are needed if small intestinal bacterial overgrowth, terminal ileal disease, or pancreatic disease is suspected

Quantitative fecal fat determination Fecal elastase or chymotrypsin, tubeless tests (see Chapters 56 and 59) Schilling test

*See text for diagnostic accuracy of the different tests listed. SeHCAT, selenium-75-homotaurocholic acid test; TCA, taurocholic acid.

differentiate primary and secondary causes. Apart from pro­ viding a diagnosis, tests evaluating gastrointestinal absorp­ tive and digestive function may be helpful in evaluating complex disease presentations. For most or all of the fol­ lowing tests, the potential benefits with regard to the costs of workup or to patient acceptability have not been estab­ lished. Because test procedures and analytical methods can vary among laboratories,98 each laboratory should establish its own reference values for these tests.

Fat Malabsorption

Quantitative Fecal Fat Analysis The van de Kamer method is the quantitative titration of fatty acid equivalents in which the results are expressed as fecal output of fat in grams per 24 hours. This method is considered the gold standard for fecal fat analysis.99 Modi­ fications in which the extracted fats are weighed rather than titrated100 have an excellent correlation with the results of the van de Kamer method. Near-infrared reflectance analy­ sis may be a less-cumbersome method to quantify fecal fat output in stool collections101 because it requires less han­ dling of stool by laboratory personnel, but it still requires a 48- to 72-hour collection to exclude the influence of day-today variability; the stool must be mixed before a sample is obtained for analysis. Fecal fat excretion of less than 7 g per day with a fat intake of 100 g per day usually is considered normal. However, the volume effect of diarrhea by itself increases fecal fat output to levels of up to 14 g per day (secondary fat malabsorption)102 (Fig 101-7); this latter value could be used as the upper limit of normal in patients with diarrhea. Diet is important in considering causes of steatorrhea; for example, elevated fecal fat values can be observed in patients consuming a diet rich in the fat substitute olestra.100 Quantitative fecal fat analysis is available routinely now only in a few centers. Reasons for the limited clinical use of quantitative fecal fat measurements are as follows: (1) If the main symptom of malabsorption is chronic diar­ rhea, measurement of fecal fat might not influence the sub­ sequent workup, because the diagnostic tests performed to establish the etiology of diarrhea are similar to the tests for the workup of steatorrhea. (2) An elevated fecal fat level usually cannot differentiate among biliary, pancreatic, and enteric causes of malabsorption. (3) In many patients with

20

15 Fecal fat (g/day)

1750

10

5

0 0

400

800

1200

1600

2000

Fecal weight (g/day) Figure 101-7.  Graph showing fecal fat output (average of a three-day stool collection) plotted as a function of fecal weight from normal subjects (�) and from subjects with induced diarrhea (). The washout effect of diarrhea increases fecal excretion of fat to levels above the upper limit of normal (7 g/day). With significant diarrhea, therefore, a fecal fat excretion of 14 g/day should be used as the upper limit of normal. (From Fine KD, Fordtran JS. The effect of diarrhea on fecal fat excretion. Gastroenterology 1936;12;102.)

severe steatorrhea, the stools have the characteristic porridge-like appearance, and quantitative studies are not necessary to establish fat malabsorption. (4) Fat absorption may be normal despite malabsorption of other nutrients, so a normal fat balance does not imply normal absorptive function of the gastrointestinal tract. (5) Finally, accuracy depends on quantitative stool collections for 48 to 72 hours, adherence to a diet comprising 80 to 100  g of fat daily, and a diet diary to determine fat intake. Science aside, quantitative fecal fat analysis has never been popular among patients, physicians, or laboratory personnel per­ forming the test. Despite the limitations of quantitative fecal fat analysis, it nevertheless is still useful in several clinical circum­ stances: to establish malabsorption and avoid nutritional

Chapter 101  Maldigestion and Malabsorption deterioration103 when overt features of intestinal or pancre­ atic disorders are lacking, as in some cases of osteoporosis, osteomalacia, anemia, or weight loss; to monitor treatment in patients with established malabsorptive disorders, such as exocrine pancreatic insufficiency or short bowel syndrome; to estimate fecal calorie loss in patients with severe malabsorption syndromes; and to quantitate fecal fat excretion in patients with diarrhea who have undergone ileal resection, thereby distinguishing steatorrhea due to bile acid deficiency from secretory diarrhea caused by bile acid loss.104 Semiquantitative Fat Analysis For the acid steatocrit test,105 a sample of stool is diluted 1 : 3 with distilled water in a test tube. The diluted stool is homogenized, and a 500-µL aliquot is pipetted into a tube. Then 100 mL of 5M HClO4 is added to allow better fat extraction and separation of the lipid layer. An aliquot of the diluted stool-HClO4 mixture is put into a nonheparin­ ized microcapillary tube and sealed on one end. After cen­ trifugation of this aliquot at 13,000 rpm for 15 minutes, the fatty layer (FL) and the solid layer (SL) are measured, and the acid steatocrit (AS) is determined according to the following equation: AS (%) = [FL (FL + SL)] × 100 An acid steatocrit of less than 31% is normal. In a small study, the acid steatocrit for random spot stool samples had a high sensitivity and specificity for detection of steator­ rhea, compared with the van de Kamer method, which is performed on a 72-hour stool collection. A linear correla­ tion also was found between results obtained with the acid steatocrit and those of the van de Kamer method, although results were quite divergent in some patients.105 Because quantitative fecal fat measurements are based on 48- to 72-hour stool collections (to minimize the effect of day-to-day variability in fecal fat excretion), however, the acid steatocrit cannot be expected to replace quanti­ tative measurement of fat output in borderline cases or in cases in which exact measurement of fecal fat loss is required. Qualitative Fecal Fat Analysis Fat analysis by microscopic examination of random stool samples might provide a clue to the presence of steatorrhea, although it cannot be used to exclude steatorrhea; its sole advantage is its ease of performance. A sample of stool is placed on a glass slide to which several drops of glacial acetic acid and Sudan III stain are added. Acidification of stool samples improves fat extraction and separation of the lipid layer.105 The slide is held over a flame-burner and the acidified mixture is heated to boiling and then examined while still warm for presence of orange fat globules. A count of up to 100 globules with a diameter less than 4 mm per high-power field is normal.6 Results of qualitative fat analysis by this method and of quantitative fat analysis do not correlate very well.106 In a small study, Sudan staining of spot stool samples had a sensitivity of 78% and a specificity of 70% for the detection of steatorrhea.105 A quantitative microscopic method of counting and measuring fat globules using the Sudan stain has been shown to correlate well with chemically measured fecal fat output.107 Breath Tests for Fat Malabsorption The principle of the 14C-triolein breath test is to measure 14 CO2 in the breath after ingestion of a triglyceride that has

been radiolabeled with 14C. Fat malabsorption results in decreased pulmonary excretion of 14CO2.108 Because of erroneous results in a variety of metabolic and pulmonary diseases, lack of sensitivity in mild malabsorption, radiation exposure to the patient, cost of the substrate, and the need for expensive equipment, this test has not found widespread acceptance for clinical use; the nonradioactive isotope 13C is used to label triglycerides instead (see later on). Serum Tests for Fat Malabsorption The photometric measurement of β-carotene at 456 nm109 has been suggested as a useful screening test for steatorrhea, although experience with this technique is limited; values less than 100 mg per 100 mL suggest the presence of steator­ rhea, and values less than 47 mg per 100 mL strongly indi­ cate steatorrhea. Concentrations in excess of 100 mg per 100 mL do not exclude mild steatorrhea, although they make steatorrhea with fat losses in excess of 16 g per day very unlikely. Normal values also have been established in the pediatric population.110 β-Carotene can be falsely low in patients with liver disease or in alcoholics who consume a diet deficient in β-carotene. Disorders in lipoproteins or intake of carotene-containing food additives also can influence the results.

Carbohydrate Malabsorption

The hydrogen breath test is a noninvasive test that takes advantage of the fact that in most people, bacterial metabo­ lism of carbohydrate results in accumulation of hydrogen, which then is absorbed by the colonic mucosa and excreted in the breath. Using different carbohydrates, such as lactose or fructose, the hydrogen breath test can be used to detect malabsorption of these carbohydrates. Measurement of breath hydrogen excretion after ingestion of lactulose has been used to assess orocecal transit time, and glucose has been used as a substrate to detect small bowel bacterial overgrowth, although sensitivity and specificity are poor.111 Unfortunately, up to 18% of people are hydrogen nonexcre­ tors,112 and in these persons, hydrogen breath test results may be falsely negative because hydrogen is metabolized by bacteria to methane. Such limitations and pitfalls of breath hydrogen testing have to be taken into account when test results are interpreted.113 The diagnosis of lactose malabsorption is established if an increase in breath hydrogen concentration of greater than 20 parts per million over baseline occurs after ingestion of 20 to 50 grams of lactose. An increase within the first 30 minutes after ingestion of lactose has to be disregarded, because it may be due to bacterial degradation of lactose in the oral cavity. Up to four hours may be required for the increase in breath hydrogen concentration to occur. Breath hydrogen measurements obtained before and at 30, 60, 90, 180, and 240 minutes after ingestion of 50 grams of lactose provide the best diagnostic yield with the fewest possible measurements.112 The lactose hydrogen breath test still is performed by most clinicians for evaluating lactose malabsorption, but this test can miss the disorder in hydrogen nonexcretors. In these patients, a lactose tolerance test—measurement of blood glucose levels before and 30 minutes after ingestion of 50 grams of lactose—can be used. An increase in glucose concentration of less than 20 mg/dL over baseline within 30 minutes of ingestion of 50 grams of lactose indicates lactose malabsorption. The lactose tolerance test has a lower sensitivity than the lactose hydrogen breath test for diagnos­ ing lactose malabsorption.112

1751

1752

Section X  Small and Large Intestine Lactase deficiency in acquired primary lactase deficiency (adult-type hypolactasia) is not caused by mutations in the gene coding for intestinal lactase (LPH gene). It has been shown, however, that a single-nucleotide polymorphism (SNP), either the C or T nucleotide −13910 upstream of the LPH gene, is involved in the regulation of intestinal lactase expression.114 A CC genotype at −13910 C/T is associated with acquired primary lactase deficiency (adult-type hypo­ lactasia), whereas TC and TT genotypes are linked with lactase persistence.115,116 This polymorphism can be used as a diagnostic test for adult-type hypolactasia.116 This SNP is only associated with adult-type hypolactasia in whites; other SNPs are linked to adult-type hypolactasia or lactase persistence in Africans.117 In patients with diarrhea, a stool test to detect a fecal pH lower than 5.5 can serve as a quali­ tative indicator of carbohydrate malabsorption.118 In the research setting, fecal carbohydrates can be determined by the anthrone method, which measures carbohydrates on a weight basis.119 By contrast, the reducing sugar method gives results on a molar basis and, therefore, provides infor­ mation about the osmotic activity of malabsorbed carbohy­ drates.17 Total SCFAs and lactic acid, which are the products of bacterial carbohydrate metabolism, can be measured in stool by titration.120 Individual SCFAs can be determined by gas chromatography.121

Protein Malabsorption

The classic test to quantify protein malabsorption, measure­ ment of fecal nitrogen content in a quantitatively collected stool specimen,12 rarely is used today. For research pur­ poses, a combined 14C–octanoic acid–13C–egg white breath test, accompanied by measurement of the urinary output of phenol and p-cresol, has been used to assess the effect of gastric acid on protein digestion.122 In this method, labeling of the 13C–egg protein test meal with 14C–octanoic acid allows the simultaneous measurement of protein assimila­ tion and gastric emptying rate. Phenol and p-cresol are the quantitatively most important phenolic compounds in the feces and urine and are specific metabolites of tyrosine, produced by bacterial fermentation in the colon. They result from protein that has escaped digestion and absorption in the small intestine and are rapidly absorbed in the colon, detoxified, and excreted in urine. Recovery of higher amounts of urinary phenols observed after omeprazole treatment in the study of this test indicated an increased availability of protein in the colon.

Vitamin B12 (Cobalamin) Malabsorption

The Schilling Test The Schilling test can be used clinically to distinguish between gastric and ileal causes of vitamin B12 deficiency. Because both intrinsic factor and hydrochloric acid are pro­ duced by parietal cells in humans, alternative approaches to diagnosing pernicious anemia are to document atrophic gastritis by endoscopy and biopsy, to confirm achlorhydria by acid secretion analysis and increased serum gastrin levels, and to look for antibodies in the serum directed against parietal cells or intrinsic factor.26,123 Because the intrinsic factor used in the Schilling test is of bovine origin, the test is not commercially available in some countries. The Schilling test is performed by administering a small oral dose of radiolabeled vitamin B12 and, simultaneously or within one or two hours, a large intramuscular flushing dose of nonradiolabeled vitamin B12. The unlabeled B12 saturates vitamin B12 carriers; thus, any radioactive vitamin B12 absorbed by the intestine is excreted in the urine. If less than 7% to 10% of the administered dose is recovered in urine within 24 hours, vitamin B12 malabsorption is con­

firmed. To specify the site of vitamin B12 malabsorption, a second phase of the Schilling test is performed subsequently with oral administration of intrinsic factor. In patients with pernicious anemia, the results of the Schilling test normal­ ize after oral administration of intrinsic factor.24,123 Patients with pancreatic exocrine insufficiency might have an abnormal result on the Schilling test, with or without added intrinsic factor, but results normalize with addition of pancreatic enzymes (Chapter 59). In small bowel bacterial overgrowth the results of the Schilling test can improve after antibiotic therapy (Chapter 102). In ileal disease or following ileal resection, abnormal results of the Schilling test persist despite intrinsic factor. Schilling test results are normal in patients with dietary vitamin B12 defi­ ciency, in protein-bound (food-bound) vitamin B12 malab­ sorption,24 and sometimes in congenital transcobalamin II deficiency.124 False-positive results on the Schilling test can result from renal dysfunction or inadequate urine collec­ tion.123 The value of this test is diminished by the need for accurately timed urine collections. Results in the 5% to 10% excretion range often are difficult to interpret. A varia­ tion of the standard Schilling test is the dual-isotope or single-stage Schilling test, in which two different cobalamin isotopes are given simultaneously, one of them bound to intrinsic factor. This makes it possible to perform the first two phases of the Schilling test in one day; however, the results of this test are not as accurate as those obtained with the standard protocol.24,123 Serum Test for Vitamin B12 and Folate Deficiency Measurements of serum cobalamin and folate concentra­ tions commonly are used to detect deficiency states of these vitamins. The sensitivity and specificity of these tests are unknown because no gold standard test has been estab­ lished and because serum levels do not always correlate with body stores.24,125 Furthermore, results of vitamin B12 levels vary with different commercial tests.126 Several causes of misleading serum cobalamin levels have been established. Serum vitamin B12 levels can be normal despite depleted body stores in small intestinal bacterial overgrowth (as a result of production of inactive cobalamin analogs by the bacteria), liver disease, myeloproliferative disorders, congenital transcobalamin II deficiency, and with high levels of intrinsic factor antibodies. In contrast, oral contraceptives, pregnancy, and folate deficiency can cause low serum cobalamin levels despite normal body stores.123 Therefore, if there is a high suspicion, especially for cobala­ min deficiency, parenteral replacement with monitoring of the clinical response is recommended.126 Measurement of methylmalonic acid, homocysteine, and holotranscobala­ min are of limited clinical use in establishing vitamin B12 deficiency.126 Serum folate concentrations decrease within a few days of dietary folate restriction, even if tissue stores are normal. Feeding also influences serum folate levels; therefore, deter­ mination of folate in the fasting state is recommended. Mea­ surement of red blood cell folate concentration has been considered a better estimate of folate tissue stores than serum folate levels by some authors.123

Small Intestinal Bacterial Overgrowth

Tests for the diagnosis of bacterial overgrowth are covered in more detail in Chapter 102. Briefly, tests used to diagnose bacterial overgrowth are the quantitative culture of a small intestinal aspirate (which is considered to be the gold stan­ dard diagnostic test), measurement of deconjugated bile acids or vitamin B12 analogs in intestinal aspirates, measure­ ment of serum folate, and several breath tests, including the

Chapter 101  Maldigestion and Malabsorption 14

C-glycocholate breath test, the 14C-d-xylose breath test, the lactulose hydrogen breath test, and the glucose hydrogen breath test. The rationale for the breath tests is the produc­ tion by intraluminal bacteria of volatile metabolites (i.e., 14 CO2 or H2), from the administered substances, which can be measured in the exhaled breath.

Exocrine Pancreatic Insufficiency

Pancreatic function tests are discussed in detail in Chapters 56 and 59. Invasive pancreatic function tests require duo­ denal intubation and measurement of pancreatic enzyme, volume, and bicarbonate output after pancreatic stimulation by a liquid test meal (the Lundh test) or by injection of cholecystokinin (CCK) or secretin. Noninvasive tests include measurement of fecal chymotrypsin or elastase concentra­ tion, the fluorescein dilaurate test, and the N-benzoyl-ltyrosyl para-aminobenzoic acid (NBT-PABA) test. Elastase has a higher sensitivity for the detection of exocrine pan­ creatic insufficiency compared with chymotrypsin,127 but the specificity of elastase is low.128 Measurement of pancreatic enzymes and components of pancreatic fluid in duodenal aspirates obtained during endoscopy and after intravenous stimulation with secretin and CCK can have an excellent correlation with the more classic intubation tests for secretory function.129

Bile Salt Malabsorption

In patients with steatorrhea due to ileal disease or resection, bile salt malabsorption usually is present, but measurement of bile acid malabsorption is of limited clinical value. In patients with diarrhea without steatorrhea, bile salt malab­ sorption may be present in the absence of overt ileal disease, and in such cases, measurement of bile salt absorption is helpful. Measurement of Fecal Bile Acid Output Elevated fecal bile acid concentrations or output can indi­ cate intestinal bile acid malabsorption.130 Under steady-state conditions, the increased fecal bile acid output reflects increased hepatic synthesis of bile acids.131 In severe bile acid malabsorption, fecal bile acid output can be reduced if hepatic synthesis of bile acids is impaired. The measure­ ment can be performed by enzymatic methods or by gas chromatography. This test requires a quantitative stool col­ lection, and the analytic techniques are time consuming and require considerable expertise. Enzymatic methods may be unreliable in severe steatorrhea.132 Carbon-14–Taurocholate Bile Acid Absorption Test The 14C-taurocholate bile acid absorption test requires a 72-hour stool collection after ingestion of radioactively labeled bile acid. The rate of intestinal bile acid absorption is calculated from the fecal recovery of 14C-labeled tauro­ cholic acid (14C-TCA). Normal values for this test have been established in normal persons with laxative-induced diar­ rhea, because diarrhea by itself can increase fecal losses of bile acids,131 presumably because of accelerated intestinal transit.133 Clinical limitations of this test are that it requires substantial analytical work, access to a gamma camera, and a time-consuming stool collection. Therapeutic Trial of Bile Acid–Binding Resins (Cholestyramine) A therapeutic trial of cholestyramine or other bile acid– binding resins can be used to diagnose bile acid malabsorp­ tion as a cause of diarrhea. It is, however, controversial to what extent a clinical response to cholestyramine correlates with the presence of bile acid malabsorption, because cho­

lestyramine may have a nonspecific constipating effect in patients with diarrhea from other causes. Failure of diarrhea to remit within three days of initiation of cholestyramine makes bile acid malabsorption an unlikely cause of diar­ rhea; however, some patients respond only to large doses of cholestyramine. In patients with established bile acid malabsorption in whom no improvement is obtained with bile acid–binding resins, it is very unlikely that bile acid malabsorption is the cause of diarrhea. In these patients, bile acid malabsorption is considered a secondary phenomenon due to a washout effect.131 In patients with severe bile acid malabsorption resulting in steatorrhea, cholestyramine might even aggra­ vate fat malabsorption and diarrhea.3 Therefore, without further testing for bile acid malabsorption, neither a positive nor a negative result of a therapeutic trial of cholestyramine constitutes proof of the presence or absence, respectively, of bile acid malabsorption. Selenium-75–Labeled Homotaurocholic Acid Test The radioactive taurocholic acid analog used for this test is resistant to bacterial deconjugation. After it has been admin­ istered orally, the patient undergoes serial gamma scintig­ raphy to measure whole-body bile acid retention or, as suggested by some authors, bile acid retention in the gall­ bladder.134 A limitation of this test is that normal values for bile acid retention, which are used to compare normal and abnormal bile acid absorption, were obtained only in healthy persons without diarrhea.135 However, secondary bile acid malabsorption can be induced by diarrhea itself and is proportional to the stool weight, as demonstrated with the 14C-TCA test.131,133 For this test to be clinically useful, normal values need to be established for patients with diarrhea. Finally, this test is very time consuming, because bile acid retention needs to be measured either four or seven days (depending on the protocol) after the bile acid administration. d-Xylose

Test

Absorption of the pentose d-xylose is facilitated by passive diffusion. Approximately 50% of the absorbed d-xylose is metabolized, and the remainder is excreted in urine. After an overnight fast, a 25-g dose of d-xylose is swallowed, and the patient is encouraged to drink sufficient volumes of fluid to maintain good urine output. Urine is collected for the next five hours. As an alternative, one hour after inges­ tion of d-xylose, a venous sample may be obtained.136 Less than 4 grams (16% excretion) of d-xylose in the urine col­ lection or a serum xylose concentration below 20 mg/dL indicates abnormal intestinal absorption. The traditional urine test appears to be more reliable than the one-hour blood test. False-positive results occur if the duration of urine col­ lection is too short or if the patient is dehydrated or has renal dysfunction, significant ascites, delayed gastric emp­ tying, or portal hypertension. d-Xylose absorption may be normal in patients with only mild impairment of mucosal function or with predominantly distal small intestinal disease. Because d-xylose is susceptible to bacterial metabo­ lism, absorption is diminished in patients with bacterial overgrowth, although the test has a poor sensitivity for detecting this.137 The test is of limited clinical value today and mostly has been replaced by small intestinal biopsy.

Intestinal Permeability Tests

Intestinal permeability tests mostly are used in studies of the pathophysiology of intestinal disorders; they do not provide a specific diagnosis.138

1753

1754

Section X  Small and Large Intestine Most current permeability tests are based on the diffe­ rential absorption of mono- and disaccharides. Damage to the mucosa can result in an increased permeability for disaccharides and oligosaccharides consequent to epithelial injury, and it can result in a decreased permeability of monosaccharides due to reduction of mucosal surface area.139 Absorption is measured by urinary excretion. Expression of results as the absorption ratio of the mono- and disaccharide minimizes the influences of gastric emptying, intestinal transit, renal and hepatic function, and variations in time of urine collections.140 Increased intestinal permeability has been shown to predict the development of Crohn’s disease or relapse in patients with this disease.141 In celiac disease, the finding of considerably increased permeability is a sensitive marker for advanced disease; permeability tests also have been used to judge response to a gluten-free diet142 or to screen firstdegree relatives for celiac disease. Elevated serum amino­ transferase levels in patients with celiac disease correlate with increased intestinal permeability.143 Disturbances of intestinal permeability have been documented in users of nonsteroidal anti-inflammatory drugs,144 in inflammatory joint disease,145 and in diabetic diarrhea.146

Carbon-13 Breath Tests

The increasing availability of stable isotopes has raised interest in replacing radioactive 14C with nonradioactive 13C for breath tests.111,147 In malabsorptive diseases, 13C-labeled substrates have been evaluated in the diagnosis of steator­ rhea,148 small bowel bacterial overgrowth, and exocrine pancreatic insufficiency149 and in the evaluation of the digestibility of egg protein. Because of concerns about diag­ nostic accuracy, cost, and limited availability, these tests have not gained widespread acceptance.

MALABSORPTION IN SPECIFIC SITUATIONS AND DISEASE STATES LACTOSE MALABSORPTION AND INTOLERANCE

Deficiency of the intestinal brush border enzyme lactase can lead to lactose malabsorption, which can result in lactose intolerance. Unlike other intestinal disaccharidases, which develop early in fetal life, lactase levels remain low until the 34th week of gestation.150 Transient lactase deficiency in premature infants can lead to symptoms of lactose malabsorption, such as diarrhea, until normal intestinal lactase activity develops. In rare cases, enzyme deficiency is manifest at the time of birth and is permanent, and con­ genital lactase deficiency (OMIM #223000)* is diagnosed. Reversible lactase deficiency can occur at all ages as a result of transient small intestinal injury associated with acute diarrheal illnesses. Acquired primary lactase deficiency (adult-type hypolac­ tasia, OMIM #223100) is the most common form of lactase deficiency worldwide. Most populations lose considerable lactase activity in adulthood.151 The decline in lactase activ­ ity is a multifactorial process that is regulated at the gene transcription level152 and leads to decreased biosynthesis or retardation of intracellular transport or maturation of the enzyme lactase-phlorizin hydrolase.153 In whites, a single-

*The Online Mendelian Inheritance in Man (OMIM) system assigns numbers to specific diseases according to a continuously updated catalog of human genes and genetic disorders (http://www.ncbi.nlm.nih.gov/omim/).

Table 101-12  Ethnic Groups with High and Low Prevalence Rates of Acquired Primary Lactase Deficiency (Adult-Type Hypolactasia) Lactase Deficiency–Predominant Ethnic Groups (60%-100% of Population is Lactase Deficient) Middle East and Mediterranean: Arabs, Israeli Jews, Greek Cypriots, southern Italians Asia: Thais, Indonesians, Chinese, Koreans Africa: South Nigerian, Hausa, Bantu North and South America: Alaska Natives, Canadian and U.S. Native Americans, Chami Indians Lactase Persistence–Predominant Ethnic Groups (2%-30% of Population is Lactase Deficient) Northern Europeans Africa: Hima, Tutsi, Nomadic Fulani India: Indians from Punjab and New Delhi areas Data from Johnson JD. The regional and ethnic distribution of lactose malabsorption. In: Paige DM, Bayless TM, editors. Lactose Digestion. Clinical and Nutritional Implications, 1st ed. Baltimore: Johns Hopkins University Press; 1981, p 11.

nucleotide polymorphism (SNP) −13910 T/C upstream of the gene coding for the enzyme lactase-phlorizin hydrolase (LPH gene) has been found to be involved in the regulation of lactase-phlorizin hydrolase.114 The CC genotype of the SNP −13910 T/C upstream of the LPH gene is associated with adult-type hypolactasia; TC and TT genotypes are linked with lactase persistence.115 In other populations, such as some African and sub-Saharan African populations, the −13910*T polymorphism is not associated with lactase persistence.117,154 Because it is present in most of the adult human population, this form of lactase deficiency has to be considered normal, rather than abnormal. Lactase deficiency usually produces symptoms only in adulthood, although lactase levels in affected persons start to decline during childhood.155 Lactase activity persists in most adults of Western European heritage156 (Table 101-12). Even in this group, the activity of lactase is only approxi­ mately half the activity of sucrase and less than 20% of the activity of maltase.155 Accordingly, in these persons, lactase activity is much more susceptible to a reduction in function with acute or chronic gastrointestinal illnesses. In lactose malabsorbers, it may be unclear whether lactose malabsorption is from acquired primary lactase deficiency or is the consequence of another small intestinal disorder. Therefore, in the individual lactose malabsorber, especially with an ethnic background associated with a low prevalence of acquired primary lactase deficiency, it may be necessary to exclude other malabsorptive disorders, such as celiac disease. The main symptoms of lactose intolerance are bloating, abdominal cramps, increased flatus, and diarrhea. The development of bloating and abdominal cramps pre­ sumably is associated with increased perception of luminal distention by gas,157 because no clear relation has been observed between the amount of lactose ingested and the severity of symptoms.158 Ingestion of as little as 3 g of lactose to as much as 96 g of lactose may be required to induce symptoms in persons with lactose malabsorption.159 Gastro­ intestinal symptoms, including diarrhea, have been shown to be more severe in adults with shorter small intestinal transit time,160 but no such relation between intestinal transit and symptoms is observed in children.161 Also, in pregnant women and in thyrotoxic patients with Graves’ disease, changes in intestinal motility play a role in the clinical manifestation of lactose malabsorption.162,163 To make a diagnosis of lactose intolerance, and in view of the poor correlation between lactose malabsorption and lactose intolerance, it is very important to monitor symptoms

Chapter 101  Maldigestion and Malabsorption

H2 concentration (ppm × 10), symptoms

10

INCOMPLETE ABSORPTION AND INTOLERANCE OF FRUCTOSE

Lactose malabsorption with intolerance

8

H2 concentration 6 4 Symptoms

2 0 0

60

A 7 H2 concentration (ppm × 10), symptoms

120

180

240

Minutes Lactose malabsorption without intolerance

6 5

H2 concentration

4 3 2

Symptoms

1 0 0

B

60

120

180

240

Minutes

Figure 101-8.  Illustration of the role of symptoms in determining the clinical importance of lactose malabsorption. Assessment of the clinical relevance of an abnormal lactose hydrogen breath test is made by monitoring abdominal symptoms (bloating, cramps, pain) during the test. Breath hydrogen concentration in parts per million (ppm) and gastrointestinal symptoms using an arbitrary scoring system for two different patients are plotted on the graphs. A, The patient has symptoms associated with an increase in breath hydrogen concentration and therefore can be con­sidered to have lactose intolerance. B, The patient has no increase in symptoms, although the breath hydrogen concentration increases con­siderably; therefore, the patient has lactose malabsorption without lactose intolerance.

during a lactose hydrogen breath test and to confirm that any symptoms experienced by the patient during the test are truly those that the patient complains of and that they are associated with a significant increase in breath hydrogen levels (Fig. 101-8). Adult-type hypolactasia also may be a risk factor for developing osteoporosis and bone fractures, owing to patients’ avoidance of dairy products164 or interference with calcium absorption.165 Patients in whom a clear association can be established between symptoms and lactose malabsorption should be educated about a lactose-reduced or lactose-free diet. Yogurt may be tolerated by such patients166 and provides a good source of calcium. Consuming whole milk or chocolate milk, rather than skim milk, and drinking milk with meals can reduce symptoms of lactose intolerance, presumably as a result of prolongation of gastric emptying. Alternatively, supplementation of dairy products with lactase of microbio­ logic origin may be suggested.167 Furthermore, because many carbohydrates other than lactose are incompletely absorbed by the normal small intestine,47 and because dietary fiber also may be metabolized by colonic bacteria, persistence of some symptoms while the patient is on a lactose-free diet is not uncommon. It also must be kept in mind that symptoms arising after ingestion of dairy prod­ ucts may be due to milk protein allergy or to fat intolerance rather than lactose intolerance.

Fructose is found in modern diets either as a constituent of the disaccharide sucrose or as the monosaccharide and it is used as a sweetener in a variety of food items. The average daily intake of fructose varies from 11 to 54 g around the world.168 Fructose as a constituent of sucrose is absorbed by a well-characterized absorptive system that integrates enzy­ matic hydrolysis of the disaccharide sucrose by sucrase and transfer of the resulting two monosaccharides, glucose and fructose, through the apical membrane of the epithelial cell; the absorptive capacity for fructose that is not accompanied by glucose, however, is relatively small.169 The normal absorption capacity of fructose depends on other nutrients as well and is poorly understood. Healthy subjects have the capacity to absorb up to 25 g of fructose, but many have malabsorption and intolerance with intake of 50 g of fructose.170 Ingesting food that contains fructose in excess of glucose can result in symptoms such as abdominal bloating or diar­ rhea171 and especially can provoke symptoms in patients with irritable bowel syndrome172; it has been suggested that as little as 3 g of fructose can precipitate symptoms in patients with functional bowel disorders. Women might complain more often of symptoms and exhibit more fruc­ tose malabsorption than do men; there is no adaptation to regular consumption of fructose.173 Fructose malabsorption usually is identified by a positive result on a hydrogen breath test after ingestion of 25 or 50 g of fructose. Because the fructose content in fruit and in soft drinks usually is less than 8 g per 100 g of fruit or drink, the amounts of fructose used in the hydrogen breath test are not physiologic, and no data are available on how many asymptomatic people would have a positive test result; fructose contents of 30 to 40 g per 100 g can be present in chocolate, caramel, and pralines.174 In a group of patients with isolated fructose malabsorp­ tion, no defect could be detected of the gene encoding the luminal fructose transporter (GLUT5).175 It is therefore unlikely that patients who present with gastrointestinal symptoms have a true defect of intestinal fructose absorp­ tion; it is more likely that they belong to a subset of people in whom ingestion of foods rich in fructose provokes symp­ toms related to other disorders, such as irritable bowel syndrome, or resulting from unique, but not necessarily abnormal, colonic bacterial activity. The latter is suggested by a study in asymptomatic and symptomatic persons with fructose malabsorption in which it was demonstrated that the disappearance rate of fructose in anaerobic, but not aerobic, stool cultures was significantly elevated in the symptomatic group compared with the asymptomatic group.176 Testing for fructose malabsorption by the hydrogen breath test may be useful in identifying patients in whom dietary restriction of foods with excessive fructose content may be beneficial in treating bloating and diarrhea. Symptoms in these persons probably are the result of ingestion of unphys­ iologic amounts of fructose and not the consequence of a defect in fructose absorption.

ILEAL BILE ACID MALABSORPTION

Bile acid malabsorption usually is present in patients who have undergone ileal resection or bypass operations or who have severe disease of the ileum, where specific bile acid transport proteins are located. The clinical consequences of bile acid malabsorption depend on whether bile acid loss can be compensated by increased hepatic synthesis.177 Ileal

1755

1756

Section X  Small and Large Intestine resection of more than 100 cm usually results in severe bile acid malabsorption that cannot be compensated by increased hepatic synthesis; in such cases, steatorrhea results from impaired micelle formation due to decreased luminal con­ centrations of conjugated bile acids.3,177 In ileal resections of less than 100 cm, bile acid malabsorption usually can be compensated by increased hepatic synthesis, and malab­ sorbed bile acids cause secretory diarrhea rather than steat­ orrhea.3,177 Secretory diarrhea caused by or associated with bile acid malabsorption is discussed in detail in Chapter 15. Knowledge of the differing pathophysiology of steator­ rhea and of secretory diarrhea from bile acid malabsorption is important not only for understanding the clinical presen­ tation but also for choosing the appropriate therapy. In patients with compensated bile acid malabsorption, binding of bile salts in the lumen of the intestine by cholestyramine reduces diarrhea. By contrast, in decompensated bile acid malabsorption, cholestyramine further depletes the bile acid pool, thereby worsening steatorrhea. In several cases of decompensated bile acid malabsorption after extensive ileal resections, intestinal fat absorption was improved markedly by oral administration of conjugated bile acids.104,178,179 Cholylsarcosine in a dose of 2 to 3 g per meal has been reported to enhance fat absorption and nutritional status in patients with short bowel syndrome who have residual colon.104,179,180 Natural conjugated bile acids lessen severity of steatorrhea in such patients to a smaller extent.180 Improved fat absorption also was associated with decreased urinary oxalate excretion.179 A syndrome of primary bile acid malabsorption with normal ileal morphology has been reported in children who, at birth, develop severe diarrhea, severe steatorrhea, and failure to thrive and who have reduced plasma choles­ terol levels.181 In an index case, this type of bile acid mal­ absorption was shown to be caused by mutations in the ileal sodium–bile acid cotransporter gene (SLC10A2).4 Adultonset bile acid malabsorption is not caused by SLC10A2 mutations,182 and although its exact pathophysiology is unknown, accelerated intestinal transit may be a causative factor.183

AMYLOIDOSIS

Malabsorption has been reported in AL-type amyloidosis, AA-type amyloidosis, and hereditary amyloidosis (see Chapter 35). Malabsorption occurs in 5% to 13% of patients with AL or AA amyloidosis,184,185 whereas it was present in 58% of Swedish patients with familial amyloidosis.186 Fecal fat excretion can reach levels up to 60 g/day.186 Gastro­intestinal absorption of d-xylose and vitamin B12 can be reduced,186,187 and protein-losing enteropathy can develop.188 Amyloid deposits are found in the muscle layers, the stroma of the lamina propria and the submucosa, the wall of mucosal and submucosal blood vessels in the gastro­intestinal tract, and in enteric and extraenteric nerves.189,190 In many patients with amyloidosis who have diarrhea or malabsorption, or both, symptoms suggesting autonomic neuropathy are present.133,186 Autonomic neuropathy can cause rapid intestinal transit, which in turn can lead to severe diarrhea and malabsorption despite normal transport capacity of the intestinal mucosa.133 Other suggested mecha­ nisms of malabsorption in amyloidosis are decreased absorption from a physical barrier effect of amyloid depos­ its190 and small intestinal bacterial overgrowth, which also might be a consequence of autonomic neuropathy.187 Bile acid malabsorption is found in many patients with amyloidosis associated with autonomic neuropathy191 and

is caused by rapid intestinal transit rather than impaired absorptive transport in the terminal ileum.133 Diarrhea in these patients usually fails to respond to bile acid–binding agents.133 Barium studies in patients with amyloidosis usually are normal but might show thickened folds, nodular lesions, filling defects, dilatation of bowel segments, or altered transit.192 The endoscopic appearance of the gastrointestinal mucosa can show a fine granular appearance, polypoid protrusions, erosions, ulcerations, atrophic changes, and mucosal friability, but in many patients, no macroscopic changes are evident.190,192 Histologic examination demonstrates amyloid deposits in 72% of esophageal, 75% to 95% of gastric, 83% to 100% of small intestinal, and 75% to 95% of colorectal biopsy speci­ mens.185,190 Subcutaneous fat pad aspiration or biopsy can make the diagnosis more safely without having to resort to endoscopic biopsy and the potential risk of intestinal bleed­ ing. Amyloid deposits might not be seen with routine his­ tologic stains but become more evident with Congo red staining. Therapy of diarrhea in patients with amyloidosis includes attempts to prolong intestinal transit time with opioids or octreotide and to avoid further amyloid deposition in the tissue by treating the underlying disorder in AA amyloidosis, the plasma-cell dyscrasia in AL amyloidosis, and by administering colchicine in familial Mediterranean fever.

MALABSORPTION CAUSED BY DRUGS AND FOOD SUPPLEMENTS

Table 101-1321,24,25,32,102,123,193-212 lists drugs and food supple­ ments reported to induce malabsorption of vitamins, miner­ als, or nutrients, as well as the suggested pathophysiologic mechanisms by which this occurs.

MALABSORPTION AFTER GASTRIC RESECTION OR BARIATRIC SURGERY Gastric Resection

Severe steatorrhea after total and partial gastric resections has been a long-observed complication of these operations. Fecal fat excretion rates usually are between 15 and 20 g/ day,1 but values of up to 60 g/day have been reported.208 Suggested mechanisms for steatorrhea include defective mixing of nutrients with digestive secretions, lack of gastric acid and gastric lipase secretion, decreased small bowel transit time, small intestinal bacterial overgrowth, and pan­ creatic insufficiency.1,213 Studies have shown that pancreatic enzyme supplements214 and antibiotic treatments213 neither improve fat absorption nor relieve symptoms after gastric resection. Total and partial gastric resections also can result in significant protein malabsorption, whereas absorption of carbohydrates seems not to be significantly impaired. Nutrient malabsorption in these patients also can result in gastrointestinal symptoms, such as diarrhea and severe weight loss.215 Vitamin E deficiency can occur if food does not pass through the duodenum. The differential diagnosis of neuro­ logic symptoms in postgastrectomy patients should include hypovitaminosis E.216 Loss of parietal cells after total gastric resection results in diminished intrinsic factor secretion, which in turn leads to malabsorption of vitamin B12 and, in approximately 30% of patients, vitamin B12 deficiency; bacterial overgrowth and lack of release of food-bound cobalamin due to diminished

Chapter 101  Maldigestion and Malabsorption Table 101-13  Drugs and Dietary Products That Cause Malabsorption SUBSTANCE

SUBSTRATE MALABSORBED

SUGGESTED MECHANISM

Acarbose Antacids Azathioprine Biguanide (metformin)

Carbohydrate Phosphate, iron, vitamin A Generalized malabsorption Cobalamin, folate, glucose

Carbamazepine Cholestyramine Colchicine

Folate Fat, fat-soluble vitamins, bile acids Fat, xylose, nitrogen, cobalamin, carotene

Contraceptives, oral*

Folate

Ethanol

Xylose, fat, glucose, nitrogen, thiamine, cobalamin, folate

Fiber, phytates Glucocorticoids Histamine H2 receptor antagonists*

Iron, calcium, magnesium, zinc Calcium Cobalamin

Laxatives, irritant type (phenolphthalein, bisacodyl, anthraquinones) Methotrexate

Fat, glucose, xylose

Inhibition of α-glucosidase Luminal binding of substrates Villus atrophy Reduced ileal absorption of intrinsic-factorcobalamin complex; inhibition of intestinal glucose or folate absorption Inhibition of intestinal folate absorption Binding of conjugated bile salts Mucosal damage and villus atrophy at high doses (impaired processing of IF-cobalamin receptor [the cubilin-amnionless complex]) Inhibition of pteroylpolyglutamate hydrolase (folate conjugase) Mucosal damage; decreased disaccharidase activity; decreased pancreatic exocrine function and bile secretion Chelation Inhibition of calcium absorption Impaired release of food-bound B12 owing to reduced gastric acid and pepsin secretion (and reduced IF secretion) Washout effect; toxic effect on the mucosa

†

Methyldopa Neomycin

Folate, fat, cobalamin, xylose

Olestra* Orlistat Para-aminosalicylate Phenytoin

Generalized malabsorption Fat, nitrogen, fat-soluble vitamins, cobalamin, mono- and disaccharides, iron Fat-soluble vitamins Fat, fat-soluble vitamins Fat, cobalamin, folate Folate, calcium

Proton pump inhibitors*

Cobalamin

Pyrimethamine

Folate

Somatostatin analogs (e.g., octreotide)

Fat

Sulfonamides and sulfasalazine Tetracycline Thiazides Triamterene*

Folate Calcium Calcium Folate

Mucosal damage; inhibition of intestinal folate transport Mucosal damage Mucosal damage; disruption of micelle formation Binding of fat-soluble vitamins Inhibition of pancreatic lipase Unknown Inhibition of folate and calcium absorption due to luminal alkalinization; impaired vitamin D metabolism Impaired release of food-bound cobalamin by pepsin owing to reduced gastric acid secretion; small intestinal bacterial overgrowth Competitive inhibition of intestinal folate absorption Inhibition of hepatobiliary bile acid secretion; inhibition of pancreatic enzyme secretion; inhibition of cholecystokinin release Inhibition of pteroylpolyglutamate hydrolase and folate transport Precipitation of luminal calcium Decreased 1,25 dihydroxyvitamin D synthesis Competitive inhibition of intestinal folate absorption

REFERENCE(S) 193 194 195 194, 196, 197 1101 194 24, 194, 199 194 32, 194 200 21 201 102, 194 194, 200 202 194, 200 203, 212 193 24, 194 21, 200, 204 25 205 206, 207 123, 200 209 210 205, 211

*Malabsorption usually does not result in deficiency states. † Findings in case reports.

gastric acid and pepsin secretion have been implicated as additional pathogenetic factors. Iron malabsorption resulting in iron deficiency anemia also is commonly present in patients who have undergone gastric resection, although the mechanisms for iron malab­ sorption are not fully established; lack of acid secretion with resultant decreased solubilization of iron salts has been sug­ gested as a possible cause. Calcium absorption can be severely impaired in patients with gastric resections, resulting in reduced bone density.217 The mechanisms for calcium malabsorption probably are several, including decreased solubilization of calcium salts owing to the loss of gastric acid secretion, rapid

intestinal transit, low calcium intake secondary to milk intolerance, and malabsorption of vitamin D. Studies in gastrectomized rats have suggested that diminished calcium absorption after gastric resections is due mainly, if not entirely, to decreased calcium solubilization.218 By contrast, studies in humans have shown that calcium absorption is normal in patients with atrophic gastritis and in persons in whom acid secretion was inhibited by acid-inhibiting drugs.219 Treatment for patients who have undergone gastric resec­ tion should include the adequate supplementation of mal­ absorbed vitamins and minerals, to prevent serious long-term complications.220

1757

1758

Section X  Small and Large Intestine Bariatric Surgery

The number of patients undergoing bariatric surgery is increasing; indications for and procedural details of the various procedures are described in Chapter 7. These patients need to be monitored for long-term problems, such as changes in bone metabolism. Risk can even increase over time due to poor compliance with supplementation, inad­ equate intake, or ongoing malabsorption. Gastrointestinal long-term problems of bariatric surgery depend on the type of surgical procedure performed. Primarily restrictive procedures such as Roux-en-Y gastric bypass have only a mild component of noncaloric malab­ sorption. Other procedures, like the biliopancreatic diver­ sion, which has been used more extensively in the past, can result in severe malnutrition.221 Roux-en-Y gastric bypass can result in deficiency of proteins, iron, calcium, folate, vitamin B12 and vitamin D. Deficiencies in vitamin B1 are rare but potentially serious.222,223 Iron deficiency after gastric bypass can develop for several reasons, such as intolerance to red meat, diminished gastric acid secretion, and exclusion of the duodenum. Menstruat­ ing or pregnant women may be particularly predisposed to developing iron deficiency after gastric bypass surgery. Postoperatively, oral iron and vitamin C supplementation should be prescribed, because once iron deficiency has developed it may be refractory to oral treatment.224 In Roux-en-Y gastric bypass, colonization of both gastric chambers with aerobic and anaerobic bacteria has been demonstrated, resulting in a positive hydrogen breath test in 41% of subjects; no clinical symptoms such as diarrhea, malabsorption, or pneumonia could be attributed to this bacterial overgrowth.225 It has been suggested that after bariatric surgery patients should have yearly measurements of a basic metabolic panel, magnesium, complete blood count, iron studies, vitamin D, parathyroid hormone, and bone density.221 The routine and lifelong use of multivitamins is considered necessary.226

MALABSORPTION IN THE ELDERLY

Malabsorption in elderly persons should not be ascribed to the aging process; it should be evaluated just like malab­ sorption occurring in younger patients. In healthy elderly persons, small bowel histologic features are normal despite a decline in cell turnover and continual cell renewal.227,228 Malabsorption of fat has been described in chronic conges­ tive heart failure229 and in chronic intestinal ischemia (see Chapter 114), but this is not due to aging per se. Elderly persons may be more susceptible to gastrointestinal insult and subsequent decompensation of gastrointestinal func­ tion.230 Changes in pancreatic anatomy and secretion occur, but only in a minority do they result in overt pancreatic insufficiency.231 Deficiencies of some nutrients, presumably caused by malabsorption, however, may be present in elderly persons with no overt gastrointestinal disease. An increased risk of folate and vitamin B12 deficiency, despite adequate intake of these vitamins, has been reported in the elderly.232 Mal­ nutrition in the elderly can contribute considerably to mor­ bidity and mortality, although it may be difficult to ascertain whether weight loss results from altered appetite, increased catabolism, or malabsorption. Small bowel bacterial overgrowth in elderly persons with gastric hypochlorhydria from atrophic gastritis or treatment with a proton pump inhibitor usually is not associated with clinically significant malabsorption,233 but an improvement in nutritional status after antibiotic treatment has been described in some elderly patients.234 An increased preva­

lence of lactose malabsorption in the elderly may be the result of clinically unapparent small bowel bacterial overgrowth.235

CONNECTIVE TISSUE DISEASES Systemic Sclerosis

The gastrointestinal tract is involved to a variable degree in most patients with systemic sclerosis. Early pathologic changes are characterized by vasculopathy, which results in ischemia and progressive organ dysfunction.236 Typical histologic findings include atrophy of the muscle layers with increased deposition of elastin and collagen in the submucosa and serosa and between smooth muscle bundles of the muscularis externa.237 Small bowel biopsy specimens might reveal an increased number of plasma cells within the lamina propria and collagen deposits around and between lobules of Brunner’s glands in the submucosa of the duodenum.238 Malabsorption in scleroderma usually results from bacte­ rial overgrowth secondary to ineffective motility in the small intestine,239 but other factors, such as decreased mucosal blood flow,240 also can contribute. Malabsorption and bacterial overgrowth are not limited to patients with diffuse disease; they also can occur in patients with longstanding limited cutaneous systemic sclerosis.241 Elevated serum concentrations of motilin and CCK have been described in patients with systemic sclerosis and fat malab­ sorption, but they are thought to result from myogenic or neurogenic disturbances of intestinal or gallbladder con­ traction.242 In addition to antibiotic treatment of bacterial overgrowth, low doses of octreotide (50 µg subcutaneously every evening for three weeks) have been shown to induce intestinal migrating motor complexes, reduce bacterial over­ growth, and relieve abdominal symptoms.239

Lupus Erythematosus and Other Connective Tissue Diseases

Excessive fecal fat excretion associated with abnormalities of d-xylose breath testing may be found in some patients with lupus erythematosus; these findings may be accom­ panied by flattened and deformed villi with an inflamma­ tory infiltrate seen on duodenal biopsy.243 Malabsorption that resolved after treatment with prednisolone also has been described in association with the hypereosinophilic syndrome in lupus erythematosus.244 Malabsorption is an uncommon feature of mixed connective tissue disease and polymyositis.245,246

CONGENITAL DEFECTS THAT CAUSE MALABSORPTION

Table 101-14 lists congenital intestinal diseases that result in malabsorption of specific substrates or in a generalized malabsorption syndrome.*

Amino Acid Transport Defects

Amino acids are absorbed by the enterocyte as oligopep­ tides, dipeptides, and free amino acids. In several inborn diseases, transport defects for different groups of amino acids have been identified in the intestine and kidney (see Table 101-14). In iminoglycinuria, Hartnup’s disorder, and cystinuria, the intestinal transport defect seems to be of no or only Text continued on p. 1763 *See references 4, 10, 14, 24, 29, 37, 39, 42, 124, 247-271.

Malabsorption of Amino Acids Hartnup’s disorder

SLC7A7

?

?

?

?

OCRL1

Lysinuric protein intolerance

Isolated lysinuria*

Iminoglycinuria

Blue diaper syndrome*

Methionine malabsorption syndrome* (Oasthouse syndrome)

Lowe oculocerebral syndrome

Type A: SLC3A1 Type B: SLC7A9

SLC6A19

DISORDER

Cystinuria (types A, B, AB)

CAUSATIVE GENE

XR

AR

AR

AR

?

AR

AR (type A) and incomplete AR (type B)

AR

SUGGESTED MODE OF INHERITANCE

Lysine, arginine

Methionine

Tryptophan

l-Proline

Lysine

Dibasic amino acids (lysine, ornithine, arginine)

Neutral amino acids (tryptophan, leucine, methionine, phenylalanine, tyrosine, valine, ?histidine, ?lysine) Cystine and/or dibasic amino acids (lysine, ornithine, arginine)

MALABSORBED SUBSTRATES

Impaired intestinal lysine and arginine absorption

Intestinal methionine absorption defect

Decreased intestinal absorption of specific free amino acids owing to a defective amino acid transporter at the brush border membrane. Type A: no transport of cystine, lysine, or arginine Type B: reduced or normal cystine transport and reduced or no lysine and arginine transport Defect of the basolateral transporter (y+LAT-1) for dibasic amino acids (also malabsorption of di- and oligopeptides) Decreased intestinal absorption of lysine Impaired intestinal absorption of l-proline in a subgroup of subjects Intestinal tryptophan absorption defect

Decreased intestinal absorption of free neutral amino acids

SUGGESTED MECHANISM OF MALABSORPTION

Table 101-14  Congenital Disorders of the Gastrointestinal Mucosa That Result in Malabsorption271

Blue discoloration of diapers, failure to thrive, hypercalcemia, nephrocalcinosis Mental retardation, convulsions, diarrhea, white hair, hyperpnea; urine has characteristic sweet smell of dried celery Aminoaciduria, mental retardation, cataracts, rickets, choreoathetosis, renal disease

Sparse hair, hyperammonemia, nausea, vomiting, diarrhea, protein malnutrition, failure to thrive, aversion to protein-rich food Mental retardation, malnutrition, failure to thrive Aminoaciduria; benign disorder

Aminoaciduria, cystine stones in the urinary tract

Most patients are asymptomatic; some patients have photosensitive skin rash, intermittent ataxia, psychotic behavior, mental retardation, diarrhea

CLINICAL FEATURES

253

252

251

257

249

249

Continued

248, 250

247

REFERENCE(S)

Chapter 101  Maldigestion and Malabsorption 1759

APOB

SAR1B

LIPA

Familial hypobetalipoproteinemia

Chylomicron retention disease Anderson’s disease

Wolman’s disease, cholesteryl ester storage disease

MTP

AR

AR

Incomplete AD

AR

AR

SLC5A1

Glucose-galactose malabsorption

Malabsorption of Fat Abetalipoproteinemia

AR

?

Trehalase deficiency

AR

Sucroseisomaltose gene

AR

SUGGESTED MODE OF INHERITANCE

Sucrase-isomaltase deficiency

Malabsorption of Carbohydrates Congenital lactase LCT deficiency

DISORDER

CAUSATIVE GENE

Fat

Fat

Fat, fat-soluble vitamins

Fat, fat-soluble vitamins

Glucose, galactose

Trehalose

Sucrose, starch

Lactose

MALABSORBED SUBSTRATES

Deficient activity of hLAL, cholesterol ester hydrolase, causing accumulation of cholesteryl esters and triglycerides in various body tissues; infiltration of intestinal mucosa with foamy cells, intestinal damage

Defective chylomicron formation and accumulation in the enterocyte

Defective lipoprotein assembly owing to a lack of MTP, resulting in triglyceride accumulation in the enterocyte and no chylomicron formation Triglyceride accumulation in the enterocyte in homozygotes owing to formation of a truncated apolipoprotein B

Sucrase activity is absent; isomaltase activity is absent or reduced; reduced maltase activity Lack of intestinal trehalase activity Defect of the brush border sodium-glucose cotransporter (SGLT1)

Permanent very low lactase activity

SUGGESTED MECHANISM OF MALABSORPTION

Table 101-14  Congenital Disorders of the Gastrointestinal Mucosa That Result in Malabsorption—cont’d

Steatorrhea, hepatosplenomegaly, abdominal distention; failure to thrive, adrenal calcifications

Steatorrhea, diarrhea, neurologic symptoms, retinitis pigmentosa, failure to thrive, absence of chylomicrons and VLDL in the blood, acanthocytosis Homozygotes: clinical manifestations as for abetalipoproteinemia Heterozygotes: fat absorption is probably normal; hypolipidemia, neurologic manifestations Steatorrhea, failure to thrive, absence of chylomicrons and reduced LDL levels in the blood; neurologic symptoms in some patients

Diarrhea and/or vomiting after ingesting mushrooms Neonatal onset of osmotic diarrhea, dehydration, intermittent or constant glycosuria

Diarrhea, bloating, and dehydration in the first days of life Osmotic diarrhea after starch or sucrose ingestion; failure to thrive

CLINICAL FEATURES

259, 260

10, 258

10

10

256

255

255

255

REFERENCE(S)

1760 Section X  Small and Large Intestine

Occipital horn syndrome (X-linked cutis laxa)

Isolated magnesium malabsorption (hypomagnesemia with secondary hypocalcemia [HOMG]) Menkes disease

Hereditary folate malabsorption Malabsorption of Minerals Acrodermatitis enteropathica

Imerslund-Gräsbeck syndrome (ileal B12 malabsorption, megaloblastic anemia type I) Transcobalamin II deficiency

Malabsorption of Vitamins Congenital IF deficiency (congenital pernicious anemia)

DISORDER

XR

XR

ATP7A

ATP7A

AR

AR

SLC39A4

TRPM6

AR

AR

TCN2

?

AR

AR

SUGGESTED MODE OF INHERITANCE

CUBN or AMN

GIF

CAUSATIVE GENE

Copper

Copper

Magnesium

Zinc

Folate

Cobalamin (vitamin B12)

Cobalamin (vitamin B12)

Cobalamin (vitamin B12)

MALABSORBED SUBSTRATES

General copper transport disorder; intestinal copper malabsorption with copper accumulation in the intestinal mucosa owing to a defective transmembrane coppertransporting ATPase (MNK) Milder form of same defect as in Menkes disease; low levels of functional MNK

Defective zinc absorption in the small intestine owing to a defect in the zinc transport protein (hZIP4) Selective defect in intestinal magnesium absorption

Defective synthesis of IF or synthesis of an abnormal IF with either reduced affinity for cobalamin or for the ileal IF receptor, or increased susceptibility to proteolysis Impaired ileal absorption of IF-cobalamin complex owing to defects in the cubilin-AMN complex (IF-cobalamin receptor) Defective transport of cobalamin out of enterocytes into the portal blood due to absence or malfunction of transcobalamin II Defective folate transport across the intestinal mucosa

SUGGESTED MECHANISM OF MALABSORPTION

Inguinal hernias, bladder and ureteral diverticula, skin and joint laxity, chronic diarrhea, bone changes

Cerebral degeneration, diarrhea, abnormal hair, hypopigmentation, arterial rupture, thrombosis, hypothermia, bone changes

Diarrhea, scaling erythematous dermatitis, alopecia, neuropsychiatric symptoms; onset after weaning Tetany, convulsion, diarrhea, hypomagnesemia with secondary hypocalcemia

Megaloblastic anemia, diarrhea, neurologic symptoms

Vomiting, diarrhea, failure to thrive, anemia, immunodeficiency, neurologic symptoms

Megaloblastic anemia, neurologic symptoms, proteinuria

Megaloblastic anemia, neurologic symptoms, delayed development

CLINICAL FEATURES

262

262

37

42

267

Continued

30, 124

24, 29, 124

30, 124, 261

REFERENCE(S)

Chapter 101  Maldigestion and Malabsorption 1761

AR

Generalized malabsorption Generalized malabsorption

AR

NEUROG3

EpCAM

Generalized malabsorption

Carbohydrates, fat, cobalamin, electrolytes, water Generalized malabsorption

XR

AR

AR

MYO5B

Bile acids, fat

USH1C, ABCC8, and KCNJ11 FOXP3

AR

SLC10A2

Protein, fat

Calcium

Calcium

Iron

MALABSORBED SUBSTRATES

Intestinal epithelial cell dysplasia and villus atrophy

Lack of enteroendocrine cells

Villus atrophy

Villus atrophy with microvillus inclusions in enterocytes, absent or shortened brush border microvilli Enteropathy with villus atrophy and inflammation

Defective activation of pancreatic proenzymes owing to lack of intestinal enterokinase Defect of the ileal ASBT

Defective 25(OH)D 1αhydroxylase, resulting in 1α,25(OH)2D deficiency and reduced intestinal calcium absorption Malfunction of the vitamin D receptor owing to defective hormone binding, defective receptor translocation to nucleus, or defective receptor binding to DNA, resulting in malabsorption of calcium

Intestinal iron transport disorder

SUGGESTED MECHANISM OF MALABSORPTION

Severe diarrhea , failure to thrive, type I diabetes mellitus Severe diarrhea, failure to thrive

Hyperinsulinism, profound congenital sensorineural deafness, enteropathy, renal tubular dysfunction Polyendocrinopathies, severe diarrhea, hemolytic anemia

Steatorrhea, diarrhea, failure to thrive Severe watery diarrhea and steatorrhea requiring total parenteral nutrition

Diarrhea, failure to thrive, hypoproteinemia, edema, anemia

Bone pain, deformities and fractures, muscle weakness, alopecia

Iron deficient anemia that is unresponsive to oral iron supplementation Bone pain, deformities and fractures, muscle weakness

CLINICAL FEATURES

269

268

270

266

265

4

14, 264

263

263

39

REFERENCE(S)

*Reported in only a few case reports. AD, autosomal dominant; AMN, amnionless; AR, autosomal recessive; ASBT, sodium bile acid cotransporter; CUBN, cubulin; hLAL, human lysosomal acid lipase; IF, intrinsic factor; LDL, low-density lipoprotein; MTP, microsomal triglyceride transfer protein; 1α,25(OH)2D, 1α,25-dihydroxyvitamin D; 25(OH)D, 25-hydroxyvitamin D; VLDL, very-low-density lipoprotein; XR, X-linked recessive.

Congenital tufting enteropathy

Immune dysregulation polyendocrinopathy and enteropathy, X-linked (IPEX) Enteric anendocrinosis*

Hyperinsulinism, with enteropathy and deafness

Congenital bile acid malabsorption Microvillus inclusion disease

AR

AR

VDR

PRSS7

AR

CYP27B1

Hereditary selective deficiency of 1α,25(OH)2D (pseudo– vitamin D deficiency rickets) Hereditary generalized resistance to 1α,25(OH)2D (vitamin D–resistant rickets)

Other Defects Enterokinase deficiency

AR

TMPRSS6

SUGGESTED MODE OF INHERITANCE

Iron-refractory iron deficient anemia

DISORDER

CAUSATIVE GENE

Table 101-14  Congenital Disorders of the Gastrointestinal Mucosa That Result in Malabsorption—cont’d

1762 Section X  Small and Large Intestine

Chapter 101  Maldigestion and Malabsorption minor clinical significance, because the amino acids affected by the transporter defects still can be absorbed as oligopep­ tides and dipeptides, and protein malnutrition can be avoided.247,248,257 Manifestations in these diseases, therefore, are mainly due to amino acid transport defects in the kidney. In Hartnup’s disorder, oral administration of nicotinamide and a high-protein diet have been shown to relieve symp­ toms to some extent.247 In lysinuric protein intolerance, however, the transport defect is located on the basolateral membrane of the entero­ cyte, leading to malabsorption of cationic amino acids in both their monopeptide and dipeptide forms. Patients with lysinuric protein intolerance therefore cannot tolerate highprotein foods, with consequent development of protein mal­ nutrition. Malabsorption of lysine with resultant deficiency of this essential amino acid is thought to be an important factor in the development of several disease manifestations in these patients.249 Treatment consists of protein restriction and supplementation with oral citrulline.

Disaccharidase Deficiency and Transport Defects for Monosaccharides

In sucrase-isomaltase deficiency, affected infants usually become symptomatic after weaning, when starch and sucrose are introduced to the diet. Symptoms and signs include osmotic diarrhea, failure to thrive, excess flatus, and occasional vomiting. Diagnosis can be established by an oral sucrose absorption test. Treatment includes avoidance of dietary starch and sucrose.255 In patients with this disease, symptoms tend to resolve spontaneously with age. Patients with glucose-galactose malabsorption suffer from severe diarrhea, leading to dehydration in the first days of life. The diarrhea stops only if glucose and galactose are eliminated from the diet. Older children and adults tolerate the offending carbohydrates better, but the transport defect is lifelong. Diagnosis can be established with an oral glucose tolerance test or by in vitro glucose absorption tests per­ formed on intestinal biopsy specimens. Therapy consists of a fructose-based diet free of glucose and galactose. After the age of three months, addition of foods containing low quan­ tities of glucose or galactose such as vegetables, fruits, and cheese is considered safe.256

Congenital Disorders of Lipid Absorption

Abetalipoproteinemia is a disorder of autosomal recessive inheritance characterized by triglyceride accumulation in the enterocyte. This disease seems to be caused by muta­ tions in the gene for microsomal triglyceride transfer protein (MTP), resulting in defective assembly of triglyceride-rich lipoproteins.10 Familial hypobetalipoproteinemia, a disorder of auto­ somal dominant inheritance, has clinical manifestations similar to those of abetalipoproteinemia when in the homo­ zygous state. This disease seems to be caused by mutations of the apolipoprotein B gene in most cases.10 Chylomicron retention disease and Anderson’s disease are caused by defective release of chylomicrons by entero­ cytes. The distinction between the two conditions derives from differences in the partitioning of lipid between mem­ brane and cytoplasmic compartments; however, both disease are due to a defect in the same gene (SAR1B). General treatment measures in abetalipoproteinemia, hypobetalipoproteinemia, chylomicron retention disease, and Anderson’s disease include the replacement of triglycerides containing long-chain fatty acids with medium-chain triglycerides and dietary supplementation with tocopherol.10

Wolman’s disease and the milder, late-onset cholesteryl ester storage disease are seemingly caused by mutations in different parts of the LIPA gene, resulting in infiltration of intestinal mucosa with foam cells and intestinal damage.

Congenital Disorders of Cobalamin Absorption

Several congenital diseases can result in vitamin B12 malab­ sorption. Absence of intrinsic factor (IF) synthesis is the most common cause of congenital cobalamin deficiency; abnormal results on Schilling tests normalize with the coadministration of IF.24,124 In some patients, an abnor­ mal (nonfunctional) IF is secreted that has a decreased affinity for cobalamin, a decreased affinity for the ileal IFcobalamin receptor (cubilin-amnionless [AMN] complex), or an increased susceptibility to proteolysis.24,124 Imerslund-Gräsbeck syndrome is a congenital disease characterized by malabsorption of the cobalamin-IF complex despite normal ileal morphology. This syndrome can be caused by mutations in two genes that code for the cubilin and AMN proteins, which are colocalized in the ileal mucosa and form the IF-cobalamin receptor.29 In transcobalamin II deficiency, serum levels of cobala­ min commonly are normal, although in most patients intes­ tinal cobalamin absorption is abnormal.124 Diagnosis is established by demonstrating the absence of transcobalamin II in the plasma.124 All congenital disorders of cobalamin absorption are treated by the parenteral administration of cobalamin, although high-dose oral cobalamin also might be effective.

Intestinal Enterokinase Deficiency

Enterokinase is an enzyme that is secreted by the intestinal mucosa and that initiates the activation of pancreatic proen­ zymes. Several patients have been reported to have an inborn deficiency of this enzyme, with resultant diarrhea, failure to thrive, and hypoproteinemia mainly from protein malab­ sorption. These patients respond well to pancreatic enzyme replacement, and some patients show a tendency to improve with age. Secondary enterokinase deficiency also has been reported in patients with villus atrophy, although patients with celiac disease seem not to be affected.14

PRIMARY IMMUNODEFICIENCY DISEASES

Malabsorption commonly occurs in entities that are charac­ terized by deficiencies in humoral or cellular immunity272 (see Chapter 2). The immunodeficiency syndromes most commonly associated with malabsorption are selective IgA deficiency, common variable immunodeficiency (CVID), and severe combined immunodeficiency. The etiology of the malabsorption varies for the different syndromes.

Selective Immunoglobulin A Deficiency

Selective IgA deficiency (OMIM #609529) is the most common primary immunodeficiency disorder and is char­ acterized by a selective near-absence of secretory and serum IgA, leading to susceptibility to respiratory, urogenital, and gastrointestinal infections. Autoimmune and allergic dis­ eases also commonly develop in patients with this disorder. A 10- to 16-fold increased incidence of gluten-sensitive enteropathy has been reported in patients with IgA defi­ ciency.273 However, at least a subgroup of patients have sprue-like small intestinal lesions, leading to severe diar­ rhea and malabsorption, but are unresponsive to a glutenfree diet.274 Improvement with immunosuppressive therapy has been described in one case report.275 Pernicious anemia, giardiasis, and secondary disaccharidase deficiencies also are seen with increased frequency in patients with selective IgA deficiency.274,276

1763

1764

Section X  Small and Large Intestine Common Variable Immunodeficiency

Common variable immunodeficiency (CVID; OMIM #240500), or CVID-acquired hypogammaglobulinemia, is an immunodeficiency disorder characterized by decreased serum immunoglobulin (Ig)G levels and variably decreased serum levels of other immunoglobulin subclasses with T-cell defects. Familial and sporadic forms are caused by mutations in the TNFRSF13B gene.277 Onset of the disease usually is in adulthood, and it manifests with recurrent respiratory and gastrointestinal infections. Affected patients also are at increased risk for autoimmune and neoplastic diseases. Malabsorption and diarrhea occur in 9% to 40% of patients with CVID274; malabsorption involves dietary fat, carbohydrates, vitamin B12, and folate.272,278 Small intestinal biopsy specimens show either sprue-like features, including villus shortening with increased numbers of lymphocytes in the epithelium and in the lamina propria, or a pattern similar to that in graft-versus-host disease (see Chapter 34).274,279 Some specific histologic features, including a near-absence of plasma cells, are observed. The disease does not respond to a gluten-free diet, and it appears that the sprue-like syndrome in CVID is a distinct entity,278 sometimes referred to as “hypogammaglobulin­ emic sprue.”280 In some patients with CVID, foamy macro­ phages are present, as in Whipple’s disease, but in contrast with Whipple’s disease, the macrophages do not contain material that stains with periodic acid–Schiff.279 In addition, nodular lymphoid hyperplasia can be detected in the gas­ trointestinal tract in a high proportion of CVID patients; the presence of nodular lymphoid hyperplasia does not corre­ late with the presence of malabsorption. The incidence of small bowel lymphoma is increased in CVID and both dis­ orders have to be considered as potential causes of malab­ sorption in these patients. Giardia organisms often are isolated from patients with CVID, and small bowel bacterial overgrowth has been documented in a number of these patients. Unfortunately, only some patients with malabsorption associated with CVID respond to antimicrobial treatment.279 Some patients with sprue-like intestinal changes have ben­ efited from glucocorticoids276 or immunoglobulins. Patients with CVID have a higher prevalence of atrophic gastritis with cobalamin malabsorption, although antibodies against parietal cells and intrinsic factor are absent.274,278

X-Linked Infantile Agammaglobulinemia (Bruton’s Agammaglobulinemia)

X-linked infantile agammaglobulinemia (Bruton’s agamma­ globulinemia; OMIM #300300) is an immunodeficiency disease characterized by lack of mature B lymphocytes and failure of Ig heavy chain rearrangement; it is caused by mutations in the gene for Bruton tyrosine kinase.281 This disease usually manifests after the first six months of life and is characterized by recurrent severe bacterial infections. Severe gastrointestinal problems, such as malabsorption and chronic diarrhea, seem to be less common than in CVID279; the prevalence of chronic gastroenteritis was 10% in one large series.282 In affected patients, the possi­ bility of giardiasis and bacterial overgrowth needs to be evaluated.279,282

Immune Dysregulation-PolyendocrinopathyEnteropathy–X-Linked Syndrome

Immune dysregulation-polyendocrinopathy-enteropathy– X-linked syndrome (IPEX) (OMIM #304790) is a disorder of early childhood characterized by protracted diarrhea,

dermatitis, insulin-dependent diabetes mellitus, thyroiditis, thrombocytopenia, and hemolytic anemia. It is a disorder of X-linked recessive inheritance caused by mutations in the FOXP3 gene.270 Diarrhea and malabsorption are secondary to severe villus atrophy with inflammation. Antienterocyte antibodies commonly are present. The enteropathy usually does not respond to a gluten-free diet, but immunosuppres­ sive therapy has been shown to be of some benefit. IPEX usually is fatal in childhood. Successful bone marrow trans­ plantation with amelioration of enteropathy has been reported in some cases.283

Other Congenital Immunodeficiency Syndromes

In severe combined immunodeficiency (OMIM #300400), diarrhea and malabsorption are common. Symptoms are associated with stunting of intestinal villi or their complete absence. The pathophysiology of malabsorption is unknown, and the syndrome usually fails to respond to antimicrobial treatment.274,276 Malabsorption also has been reported in DiGeorge’s syndrome (thymic hypoplasia, OMIM #188400) and chronic granulomatous disease of childhood (OMIM #306400), but little is known about its cause in these disorders.274

NEUROFIBROMATOSIS TYPE 1 (VON RECKLINGHAUSEN’S DISEASE)

Malabsorption can be an intestinal complication of neuro­ fibromatosis type 1 (OMIM #162200). Mechanisms of malabsorption include periampullary duodenal tumors, which are mainly somatostatin-containing neuroendocrine tumors, and pancreatic carcinomas with resultant pan­ creatic duct obstruction; tumors can cause exocrine pancre­ atic insufficiency and biliary obstruction.284 Duodenal somatostatinomas in von Recklinghausen’s disease usually do not increase plasma somatostatin levels, although one case of somatostatinoma syndrome has been reported.285 Infiltrating mesenteric plexiform neurofibromas and vascu­ lar damage caused by proliferation of nerves can cause lymphatic or vascular obstruction (or both), resulting in abdominal pain, protein-losing enteropathy, diarrhea, steatorrhea, and bowel ischemia.286,287 In patients with von Recklinghausen’s disease, an increased incidence of neuro­ endocrine tumors in other locations has been observed; gastrinomas with Zollinger-Ellison syndrome also have been reported in some of these patients.288

NONGRANULOMATOUS CHRONIC IDIOPATHIC ENTEROCOLITIS AND AUTOIMMUNE ENTEROPATHY

Nongranulomatous chronic idiopathic enterocolitis is an entity that is distinct from refractory celiac disease and inflammatory bowel disease.289 The etiology of this disease is unknown, although chronic infection and an autoimmune cause have been suggested. Severe diarrhea and malabsorp­ tion occur as a result of diffuse villus atrophy, and ulcer­ ations may be present in the small and large intestine. Small intestinal villus atrophy and neutrophilic inflamma­ tion of the mucosa with crypt abscesses may be seen in biopsy specimens from the small intestine and colon (Fig. 101-9); the number of intraepithelial lymphocytes is not increased.289,290 Patients respond dramatically to glucocorticoids, and most require long-term low-dose maintenance therapy.289,290 Improvement with cyclosporine and long-term antibiotic therapy has been reported in one patient each.291 The condition is associated with a high mortality rate.289,290

Chapter 101  Maldigestion and Malabsorption terocyte antibodies and anti–goblet cell antibodies. Absence of antibodies, however, does not exclude the diagnosis.292 Therapy of autoimmune enteropathy is challenging, and some patients have been treated successfully with glucocor­ ticoids and immunosuppressive drugs.

ENDOCRINE AND METABOLIC DISORDERS Adrenal Insufficiency (Addison’s Disease)

Some patients with adrenal insufficiency, independent of its etiology, have fat malabsorption, and fecal fat excretion of up to 30 g/day has been documented.293 Fat malabsorp­ tion also is observed in rats after adrenalectomy.294 The pathophysiologic mechanism of malabsorption in this disease is unknown, but fat absorption normalizes after glu­ cocorticoid replacement. Isolated autoimmune Addison’s disease has been asso­ ciated with pernicious anemia295 and celiac disease.296 An increased incidence of celiac disease and pernicious anemia also is found in autoimmune polyglandular syn­ drome (APS) type 2 (Schmidt’s syndrome), which is char­ acterized by the association of autoimmune Addison’s disease and other autoimmune endocrine disorders except hypoparathyroidism.297

A

B Figure 101-9.  Duodenal biopsy specimen from a patient with nongranu­ lomatous chronic idiopathic enterocolitis. A, Histopathologic features include villus atrophy, diffuse infiltration of the lamina propria with inflammatory cells, and crypt abscesses (arrow). B, High-power view demonstrates crypt infiltration by neutrophils (arrow). (Hematoxylin and eosin stain.) (Courtesy of Cord Langner, MD.)

Nongranulomatous chronic idiopathic enterocolitis shares several clinical and histologic features with adult autoimmune enteropathy.290,292 In many patients with adult autoimmune enteropathy, antienterocyte antibodies, anti– goblet cell antibodies, and other autoimmune disorders are present. Symptoms are chronic severe high-output diarrhea and malabsorption.292 Diagnosis relies on a combination of clinical and histologic findings. Proposed diagnostic criteria require the presence of chronic diarrhea and malabsorption; exclusion of other small intestinal diseases, such as celiac disease; histologic changes on intestinal biopsies such as partial or complete villous blunting, deep crypt lympho­ cytosis, increased crypt apoptotic bodies, and minimal intraepithelial lymphocytosis; and the presence of antien­

Enteroendocrine Deficiency (Autoimmune Polyglandular Syndrome Type 1 and Enteric Anendocrinosis)

Autoimmune polyglandular syndrome type 1 (APS 1) (OMIM #240300) is characterized by multiple endocrine organ failure (especially hypoparathyroidism and adrenal insufficiency) due to autoimmune destruction, with ecto­ dermal dystrophy and susceptibility to chronic Candida infections.297 APS type 1 is inherited as an autosomal reces­ sive disorder and is caused by mutations in the AIRE gene.298 Severe malabsorption, which tends to be recurrent, devel­ ops in approximately 20% of patients with APS type 1. In two patients, malabsorption was caused by a transient and selective destruction of small intestinal enteroendocrine cells, leading to a temporary deficiency of enteroendocrine hormones (especially CCK),299,300 These patients have auto­ antibodies to tryptophan hydroxylase, which are directed against enteroendocrine cells (including CCK-producing cells).301 The long-known association between hypoparathy­ roidism and steatorrhea may be caused by the same mecha­ nism, because in most reports of this association, patients fulfill the diagnostic criteria for APS type 1.302,303 Selective absence of small intestinal enteroendocrine cells can be diagnosed by special immunohistochemical stains for these cells, such as immunohistochemical stains for chromogranin A or CCK (Fig. 101-10) or by measure­ ments of postprandial serum levels of the affected hor­ mones. Patients with APS type 1 also have an increased incidence of vitamin B12 malabsorption as a result of auto­ immune gastritis.297 Lack of enteroendocrine cells also results in congenital malabsorption in a newly described disease resulting from a mutation in the NEUROG3 gene (diarrhea 4; enteric anendocrinosis OMIM #610370) (see Table 101-14).268

Hyperthyroidism and Autoimmune Thyroid Disease

Some reports suggest that up to 25% of hyperthyroid patients have at least some degree of fat malabsorption, but data from large series of patients are lacking. Fecal fat values in hyperthyroid patients can reach 35 g/day.304 The mecha­ nism of steatorrhea in hyperthyroidism has not been estab­ lished. Motility studies in hyperthyroid patients (including

1765

1766

Section X  Small and Large Intestine

A

B

Figure 101-10.  Chromogranin A immunohistochemical staining of enteroendocrine cells in duodenal biopsy specimens obtained from a normal subject (A) and from a patient with malabsorption associated with autoimmune polyglandular syndrome type 1 (B). In B, enteroendocrine cells are absent. See text for details.

patients with and without diarrhea) have demonstrated accelerated small intestinal and whole-gut transit times305; however, fecal fat values were not reported in these patients. It can be hypothesized that more-pronounced disturbances of intestinal transit can lead to decreased mixing of food and digestive secretions and reduced intestinal absorption of nutrients. Some of the steatorrhea in hyperthyroid patients might be due to hyperphagia with increased dietary intake of fat.306 An increased number of lymphocytes and plasma cells and some degree of edema in small intestinal biopsy speci­ mens have been found in patients with steatorrhea and hyperthyroidism; villus architecture is normal.304 Absorp­ tion of glucose and d-xylose is normal in hyperthyroid patients with and without malabsorption.306 Fat malabsorp­ tion tends to normalize when patients attain a euthyroid state.304,306 In patients with autoimmune thyroid diseases, an increased prevalence of celiac disease296 and primary biliary cirrhosis,295 both of which can result in fat malabsorption, has been recognized. The prevalence of celiac disease in patients with autoimmune thyroid disease is approximately 2% to 4%.296 Cobalamin malabsorption resulting from auto­ immune gastritis is found in a considerable number of patients with autoimmune thyroid disease.26,295

bacterial overgrowth,310 and pancreatic insufficiency,313 can be associated with diabetes mellitus. In patients with dia­ betes mellitus type 1, a high prevalence (3% to 8%) of celiac disease has been reported from screening studies; however, most patients identified were asymptomatic.314 Markedly reduced pancreatic exocrine function, as determined by fecal elastase measurement, has been reported in 30% of patients with type 1 diabetes and 17% with type 2 diabetes, compared with 5% of control subjects. In 40% of diabetic patients with reduced fecal elastase levels, fat malabsorp­ tion with fecal fat output of more than 10 g/day was detected.315 Gastrointestinal symptoms and steatorrhea in these patients, however, did not correlate with fecal elastase levels.313,315 In addition, the unresolved specificity of elas­ tase raises the possibility that not all of these patients truly had pancreatic insufficiency.316 The presence of cobalamin malabsorption caused by autoimmune atrophic gastritis is increased three- to five-fold in patients with diabetes mel­ litus type 1 compared with the nondiabetic population.317 Ingested carbohydrates are malabsorbed in patients receiving acarbose as part of their diabetes treatment, which in turn can lead to symptoms of diarrhea and malabsorp­ tion. Foods recommended to diabetics because they contain poorly absorbable carbohydrates such as fructose or sorbitol also can result in bloating and diarrhea.

Diabetes Mellitus

Metabolic Bone Disease

Chronic diarrhea is common in patients with diabetes mel­ litus, especially in those with long-standing diabetes mel­ litus type 1.307 Mild steatorrhea often is present in patients with diabetic diarrhea and in diabetic patients who do not complain of diarrhea.308 Although the pathophysiologic mechanism of malabsorp­ tion and diarrhea in patients with diabetes mellitus is unknown, poor glycemic control is an important cofactor.309 Most of these patients have signs of autonomic neuropathy, such as orthostatic hypotension, impotence, bladder dys­ function, incontinence, inappropriate heart rate variability, and abnormal sweating.310 Therefore, in some patients, the cause of diarrhea and malabsorption has been attributed to rapid gastric emptying and rapid intestinal transit, causing impaired mixing of nutrients with digestive secretions and decreased contact time between nutrients and the intestinal mucosa. The clinician has to be aware, however, that certain treat­ able diseases, such as celiac disease,311,312 small intestinal

Special consideration has to be given to osteoporosis and osteomalacia in malabsorptive diseases. Patients with these metabolic bone diseases usually do not present with sugges­ tive symptoms or abnormalities either on physical examina­ tion or on routine laboratory examinations. Reduced bone mineral density is a common finding in patients with gastric resection,318 celiac disease,319 and lactose malabsorption.164 Osteoporosis has been suggested to result from calcium mal­ absorption or reduced calcium intake, which leads to sec­ ondary hyperparathyroidism, which in turn increases bone turnover and cortical bone loss. Vitamin D malabsorption probably is of lesser importance. Although up to one half of patients on a gluten-free diet have osteoporosis,320 some studies have shown significant improvement in bone mineral density one year after starting a gluten-free diet.321 In inflammatory bowel diseases such as Crohn’s disease, which may be accompanied by malabsorption, other factors such as glucocorticoid use or testosterone deficiency322 may contribute to decreased bone mass.

Chapter 101  Maldigestion and Malabsorption In addition to treating the underlying cause of malabsorp­ tion, calcium supplementation is needed to ensure a daily intake of 1500 mg of calcium. Vitamin D deficiency must be corrected. If osteoporosis is present, bisphosphonate treat­ ment is suggested.319 Nutritional management is discussed in more detail in Chapters 4 and 5.

GENERAL APPROACH TO MANAGEMENT Treatment of malabsorptive diseases must be directed against the underlying condition, if possible. In addition, nutritional deficits must be corrected. The reader is referred to the relevant chapters of this book for discussion about the treatment of specific diseases and their nutritional man­ agement. In patients with abdominal bloating and gasrelated complaints as a result of sugar malabsorption, a diet with reduced content of poorly absorbable carbohydrates such as fructose, sobitol, or fermentable dietary fibers is a long-term effective therapy.323 In pancreatic insufficiency, in disorders of intestinal fat absorption, and in short bowel syndrome, medium-chain triglycerides can be used as a source of dietary calories. In patients with short bowel syndrome and some remaining colon, colonic salvage capacity can be used to regain calo­ ries from carbohydrates324; these patients, therefore, should consume a diet rich in carbohydrates and medium-chain triglycerides. In bile acid malabsorption after extensive ileal resections, intestinal fat absorption can be improved markedly by oral administration of natural conjugated bile acids178 or of syn­ thetic cholylsarcosine.104,179,180 Replacement of conjugated bile acids also reduces urinary oxalate excretion and there­ fore should protect against development of kidney stones.179 Patients with cystic fibrosis or short bowel syndrome who are unable to absorb vitamin D from their diet may benefit from treatment with an ultraviolet lamp, which emits ultra­ violet radiation similar to sunlight.325 In patients with malabsorption and an intact colon, fluid depletion must be avoided to prevent kidney stones associ­ ated with hyperoxaluria.326 In patients with malabsorption syndrome, special care should be given to the replacement of vitamins, iron, calcium, and trace elements to avoid defi­ ciency syndromes (see Chapters 4 and 5).

In patients with diarrhea, symptomatic treatment with opiates or loperamide can increase the time available for absorption of nutrients.

KEY REFERENCES

Fine KD, Fordtran JS. The effect of diarrhea on fecal fat excretion. Gas­ troenterology 1992; 102:1936-9. (Ref 102.) Fine KD, Schiller LR. AGA technical review on the evaluation and management of chronic diarrhea. Gastroenterology 1999; 116:146486. (Ref 98.) Gibson PR, Newnham E, Barrett JS, et al. Review article: Fructose mal­ absorption and the bigger picture. Aliment Pharmacol Ther 2007; 25:349-63. (Ref 168.) Hofmann AF, Poley JR. Role of bile acid malabsorption in pathogenesis of diarrhea and steatorrhea in patients with ileal resection. Gastroen­ terology 1972; 62:918-34. (Ref 3.) Holt PR. Intestinal malabsorption in the elderly. Dig Dis 2007; 25:14450. (Ref 231.) Horton KM, Corl FM, Fishman EK. CT of nonneoplastic diseases of the small bowel: Spectrum of disease. J Comput Assist Tomogr 1999; 23:417-28. (Ref 94.) Longstreth GF, Newcomer AD. Drug-induced malabsorption. Mayo Clin Proc 1975; 50:284-93. (Ref 194.) Online Mendelian Inheritance in Man. McKusick-Nathans Institute of Genetic Medicine, Baltimore: Johns Hopkins University; and Bethesda (Md): National Center for Biotechnology Information, National Library of Medicine [2008 Sep 30]. Available from: http://www.ncbi. nlm.nih.gov/omim/ (Ref 271.) Poitou Bernert C, Ciangura C, Coupaye M, et al. Nutritional deficiency after gastric bypass: Diagnosis, prevention and treatment. Diabetes Metab 2007; 33:13-24. (Ref 222.) Riddell RH. Small intestinal biopsy: Who? How? What are the findings? In: Barkin JS, Rogers AI, editors. Difficult Decisions in Digestive Diseases. Chicago: Year Book Medical Publishers; 1989. p 326-31. (Ref 85.) Romagnuolo J, Schiller D, Bailey RJ. Using breath tests wisely in a gastroenterology practice: An evidence-based review of indications and pitfalls in interpretation. Am J Gastroenterol 2002; 97:1113-26. (Ref 111.) Ryan ER, Heaslip IS. Magnetic resonance enteroclysis compared with conventional enteroclysis and computed tomography enteroclysis: A critically appraised topic. Abdom Imaging 2008; 33:34-7. (Ref 97.) Ryan ME, Olsen WA. A diagnostic approach to malabsorption syn­ dromes: A pathophysiological approach. Clin Gastroenterol 1983; 12:533-50. (Ref 6.) Seetharam B. Gastrointestinal absorption and transport of cobalamin (vitamin B 12). In: Johnson LR, editor. Physiology of the Gastrointes­ tinal Tract. New York: Raven Press; 1994. p 1997-2026. (Ref 24.) Wilson FA, Dietschy JM. Differential diagnostic approach to clinical problems of malabsorption. Gastroenterology 1971; 61:911-31. (Ref 1.) Full references for this chapter can be found on www.expertconsult.com.

1767

CHAPTE R

102 Enteric Microbiota and Small Intestinal Bacterial Overgrowth Seamus O’Mahony and Fergus Shanahan

CHAPTER OUTLINE Composition and Molecular Analysis of the Enteric Microbiota  1769 Interactions between Host and Microbes  1771 Metabolic Activity of the Microbiota  1772 Small Intestinal Bacterial Overgrowth  1772 Etiology and Predisposing Factors  1772

Increasing recognition of the role of the intestinal micro­ biota (flora) in the development and function of the gastro­ intestinal tract has led to a resurgence of scientific and clinical interest in the enteric ecosystem. The contribution of the microbiota to mucosal homeostasis is so profound that it is no longer acceptable to study intestinal patho­ physiology without considering the activities of the indig­ enous bacteria. The lesson of Helicobacter pylori in the genesis of peptic ulcer disease and gastric cancer has been a sobering reminder of the potential pathogenic role of luminal bacteria. Contributions of the intestinal microbiota to health and disease are complex and influenced by a variety of environmental factors, bacterial factors, and host-related factors, including antibiotic exposure, diet, and immunologic and genetic status. As discussed in this chapter, the enteric bacteria are critical to health and host defense, but under conditions such as those favoring bacte­ rial overgrowth, components of the microbiota can become a liability.

COMPOSITION AND MOLECULAR ANALYSIS OF THE ENTERIC MICROBIOTA Most human enteric bacteria cannot be cultured, because of a lack of truly selective growth media. Nonetheless, mole­c­ ular profiling has shown that whereas the microbiota appear distinct in different persons, the composition of each per­ son’s microbiota is relatively stable after infant weaning and throughout adulthood. Evidence from studies of twins sug­ gests that the individuality of human microflora may be genetically controlled,1 but environmental variables includ­ ing diet and sanitation appear to have profound effects on early intestinal colonization with bacteria.2,3 In adulthood, dietary fluctuations appear to induce changes in bacterial enzymes and metabolic activity rather than changes in the relative populations of the microflora.2,4,5

Mechanisms of Malabsorption  1774 Clinical Features  1774 Diagnosis  1775 Treatment  1777

The composition of the microflora varies quantitatively and qualitatively over the longitudinal and the crosssectional axes of the alimentary tract. Beyond the oral cavity, which harbors approximately 200 different bacterial species, the size and diversity of the microflora increase distally along the digestive tract (Fig. 102-1). Gastric acid restricts bacterial numbers within the stomach to fewer than 103 colony-forming units (CFU)/mL. The gradient in bacte­ rial density is greatest across the ileocecal valve, with approximately 108 bacteria per gram of ileal contents and up to 1012 bacteria per gram of colonic contents, comprising more than 1000 different bacterial species.2,4-6 More than 99% of the culturable bacteria in the ileum and the colon are obligate anaerobes, but the composition of the flora at the mucosal surface differs from that within the lumen; ratios of anaerobes to aerobes are lower at mucosal surfaces. Culture-independent methods, such as the various mole­c­ ular approaches described later, suggest that mucosaassociated bacteria differ from those recovered from feces, thus supporting the idea that host-related factors have a role in determining the enteric microflora7 and implying that bacterial aspirates from the lumen may be an incomplete reflection of mucosa-associated bacteria. The microbiota of the proximal small intestine consist predominantly of Gram-positive facultative bacteria— bacteria that can survive under aerobic or anaerobic con­ ditions—although enterobacteria and Bacteroides species also may be present. Peristalsis is the principal factor restricting bacterial numbers in the small intestine. In the distal small intestine, the composition of the microflora resembles that of the colon, with a preponderance of Gramnegative anaerobes. The most prominently represented genera in the distal bowel include Bacteroides, Clostridium, Lactobacillus, Fusobacterium, Bifidobacterium, Eubacterium, Peptococcus, and Escherichia species.2,6 The impact of diet on the composition of the enteric microflora has been studied by several investigators, with varying results. The consensus is that breast-fed babies have

1769

1770

Section X  Small and Large Intestine

Oral cavity 200 species

Stomach Helicobacter pylori

Duodenum and proximal jejunum 2 10 –103 bacteria/mL

Ileum 108 bacteria/mL

Figure 102-1.  Composition and distribution of bacterial flora within the human alimentary tract.

a greater proportion of bifidobacteria in their fecal flora than that observed in infants who are formula fed; highcarbohydrate and high-fiber diets are associated with increased bifidobacteria, whereas a high-fat diet is linked with an increase in Bacteroides species. It is the metabolic activity of the microbiota, however, rather than their composition, that exhibits the greatest response to dietary changes in a given person. Detailed analysis of the enteric microflora has been con­ founded by the limitations of traditional culture-dependent microbiology.8 First, obtaining representative material from different niches within the intestine is problematic; because most of the indigenous microflora are obligate anaerobes, major methodologic difficulties are encountered with sam­ pling, contamination, transport, and storage. Second, the lack of truly selective growth media precludes culture of most components of the microflora. In this respect, it is noteworthy that culture of H. pylori and C. difficile was accomplished only within the past two to three decades. This difficulty has led to a shift in emphasis from conven­ tional bacterial phenotyping toward genotyping and mo­ lecular approaches to study the unculturables.4,9,10 Bacterial nucleic acid extracted from feces or mucosal biopsy samples can provide a profile of the composition of the indigenous microflora. The small ribosomal subunit RNA (16S rRNA in bacteria) contains highly conserved regions of base sequences that reflect an absence of evolu­ tionary change. These conserved sequences are interspersed with hypervariable regions, which contain mutational changes reflecting the evolutionary divergence of different species. Sequencing of 16S rRNA, therefore, represents a method for identification and phylogenetic classification of intestinal bacteria.

Colon 1010 – 1011 bacteria/g 400-500 species including Bacteroides Eubacterium Peptostreptococcus Bifidobacterium Ruminococcus Bacillus Fusobacterium Clostridium Lactobacillus Enterococcus Enterobacter

For rapid profiling of the dominant culturable and non­ culturable organisms within a complex ecosystem such as that in the intestinal tract, 16S rRNA can be amplified by polymerase chain reaction (PCR) with universal primers spanning conserved and variable regions. The mixture of hypervariable RNA fragments can then be separated by a chemical denaturing gradient or a temperature gradient gel electrophoresis (DGGE and TGGE). Complete denaturation of the RNA fragments is prevented by incorporating a GC-rich 5′ end to one of the primers (a GC clamp).11 Varia­ tions in migration distance through the denaturing gradients reflect the diversity of 16S species in the sample (Fig. 1022). Theoretically, the technique is semiquantitative, because the more dominant the organism, the more abundant the specific PCR product. The specific PCR product can be cut from the gel and further amplified, cloned, and sequenced to identify individual bacterial strains without requiring a conventional culture step. Further refinements of the technique can be achieved by using species-specific PCR primers. Other molecular techniques for analysis of specific bacte­ rial species now are possible because of the increasing avail­ ability of genomic sequence data for the major components of the bacterial flora. These techniques include fluorescence in situ hybridization (FISH) flow cytometry (FISH-flow) and bacterial DNA microarrays. In disorders such as inflamma­ tory bowel disease, immunologic reactivity against compo­ nents of the microflora has been used to identify microbes that may be etiologic in the disease. Marker antibodies gen­ erated by hybridoma or phage-display technology have been used as reagents to identify microbial antigens. For example, antineutrophil cytoplasmic antibody (pANCA), which is associated with ulcerative colitis, has been used to identify

Chapter 102  Enteric Microbiota and Small Intestinal Bacterial Overgrowth Patient 1 0

4

8

Patient 2 12

0

4

8

12

Figure 102-2.  Representative examples of molecular fingerprinting profiles. Denaturing gradient gel electrophoresis analysis of the intestinal microbiota in two patients tested over time, at 0, 4, 8, and 12 months. Although the profile for each patient is unique, it also is stable.

colonic bacteria expressing a pANCA-related epitope.12,13 In addition, candidate microbes can be identified by the presence of unique bacterial nucleic acid sequences asso­ ciated with a particular lesion or disease location, by sub­ tractive cloning using genomic representational difference analysis. With this technique, a sequence representing a bacterial transcription factor from an apparent commensal organism, Pseudomonas fluorescens,14-16 was found in lesions of Crohn’s disease but not in adjacent nonlesional mucosa. Serologic expression cloning also has been used to identify bacterial flagellin as a dominant antigen in Crohn’s disease.17 The new science of metagenomics—the sequencing of genes from whole microbial environments at once—prom­ ises to address many of the unresolved questions about the microbiota. The microbiota comprise a repository of genetic information (microbiome) that greatly exceeds that of the host genome. By combining metagenomics with bioinfor­ matics, biochemistry, and traditional bioassays, new insights into the metabolic capacity of the human intestinal micro­ biota can be achieved.4 Major consortia (including the human microbiome project4) around the globe are under way using metagenomics as a tool for bioprospecting the intestinal microbiota in health and disease.

INTERACTIONS BETWEEN HOST AND MICROBES The microbiota exert both positive and negative regulatory effects on the development and function of the intestine. These complex influences first were shown in comparative studies of germ-free and conventionally colonized animals. A sterile intestine is associated with reductions in mucosal cell turnover, digestive enzyme activity, cytokine produc­ tion, lymphoid tissue, lamina propria cellularity, vascular­ ity, muscle wall thickness, and motility, but with an increase in enterochromaffin cell area.18 The molecular events under­ pinning this regulatory signaling from the lumen currently are being explored using modern techniques such as laser capture microdissection and gene array analysis; such

studies promise to reveal new molecular targets to be exploited for the design of novel therapeutics.19,20 Thus, for example, when applied to animals colonized with only a single bacterial strain, Bacteroides thetaiotaomicron, this combined approach has illustrated the impact of bacteria-derived signaling on the expression of host genes controlling mucosal barrier function, nutrient absorption, angiogenesis, and development of the enteric nervous system. Incoming bacterial signals include secreted chemoattrac­ tants, such as the formylated peptide f-Met-Leu-Phe, cellu­ lar constituents such as lipopolysaccharide (LPS) and peptidoglycans, flagellin, and bacterial nucleic acids (i.e., CpG DNA). Detection of bacterial stimuli by the host and discrimination of pathogens from commensals are mediated in part by pattern recognition receptors such as Toll-like receptors (TLRs) that are present on epithelial and immune (dendritic) cells. In health, engagement of TLRs by ligands from the commensal microbiota appears to be required for mucosal homeostasis.21,22 Thus, not only are bacterial signals required for optimal mucosal and immune development, but they also actually are required to maintain and condi­ tion the mucosa for responses to injury.22 The immune system mediates the sense of microbial danger and responses to injury. Although the primary lym­ phoid organs are developed at birth, mucosal immune func­ tions require continual education and fine-tuning of cytokine balances and T-cell responses; this process is achieved by microbial colonization and sporadic mucosal infections. Without the microbiota, mucosal lymphoid tissue is rudi­ mentary, and induction of mucosal immune responses and tolerance is suboptimal.23,24 With a surface area similar to that of a tennis court (approximately 400 m2) and only one cell layer separating the internal milieu from the lumen, the enteric mucosa is well adapted to immunologic sampling of the intraluminal microbial community. Sampling of the microbiota across the epithelial barrier is mediated by M cells, which deliver particulate and microbial antigens to underlying immune cells, and by mucosal dendritic cells, which appear to extend processes into the lumen between the surface entero­ cytes without disrupting tight junctions.25 It appears that intestinal dendritic cells can ingest and retain intact live commensal bacteria and then transit to the mesenteric lymph node, where immune responses to commensals are induced locally.26 Thus, the mesenteric lymph node acts as a gatekeeper, preventing access of commensal bacteria to the internal milieu and protecting the host from harmful sys­ temic immune reactivity. The immunosensory function of dendritic cells is facilitated by their plasticity and versatility of responses,27 depending on whether they are presented with commensals or pathogens; moreover, they appear to exhibit tissue-specific specialization in the intestine.23,24 In addition to specific immune responses to enteric bac­ teria, the surface epithelial cells serve a sensory function to detect microbial danger by producing chemokines that acti­ vate the host immune response and recruit it to any breach in the mucosal barrier caused by pathogenic infection.28 Transduction of bacterial signals into host immune responses after engagement of TLRs may proceed along more than one molecular pathway. The transcription factor nuclear factor-κB (NF-κB) is the pivotal regulator of epithe­ lial responses to invasive pathogens, but nonpathogenic bacteria can attenuate inflammatory responses by delaying the degradation of IκB, which is counter-regulatory to NF-κB.29 Other signal transduction pathways are likely to emerge to account for the anti-inflammatory effects of probiotics and other commensal organisms such as

1771

1772

Section X  Small and Large Intestine Bacteroides thetaiotaomicron. This anaerobic commensal can antagonize the proinflammatory effects of NF-κB within the epithelial cell by enhancing the nuclear export of its transcriptionally active subunit (RelA) in a peroxisome proliferator-activated receptor-γ (PPAR-γ)– dependent manner.30

METABOLIC ACTIVITY OF THE MICROBIOTA The enteric microbiota are tantamount to a hidden meta­ bolic organ (Table 102-1). Although our understanding of indigenous bacterial metabolites is still superficial, coevolu­ tion with this living inner mass of bacteria has several apparent benefits for the host. In addition to the production of regulatory signals for mucosal homeostasis as discussed earlier, the microbiota exhibit important metabolic proper­ ties not possessed by the host. These include biotransforma­ tion of bile acids; degradation of oxalate; breakdown of otherwise indigestible dietary components, such as plant polysaccharides; and production of short-chain fatty acids, a major energy source for colonic epithelium, from ferment­ able carbohydrates. Other activities include synthesis of biotin, folate, and vitamin K.2,20 Clinicians also have exploited enteric bacterial enzymes such as azoreductase to convert prodrugs such as sulfasalazine to active drug metab­ olites (e.g., aminosalicylate). Other examples of bacterial action on drug bioavailability include the metabolism of l-dopa to dopamine and degradation of digoxin. Not all of the metabolic changes induced by the enteric microbiota are beneficial to the host, however, and although bacteria prob­ ably degrade some carcinogens, they also might promote the production of carcinogens from dietary procarcinogens.31 A striking example of the importance of bacterial metabo­ lism is exemplified by the regulatory effect that the enteric microbiota exert on fat storage.32 It has long been known that germ-free animals need a significantly greater caloric intake to sustain a body weight similar to that of normal colonized animals. Thus, the normal host-microbiota rela­ tionship has nutritional benefit, in contrast with the nega­ tive nutritional effect associated with bacterial overgrowth syndromes. Elegant studies with germ-free mice have shown that upon colonization, body weight increases despite a reduced caloric intake. The bacteria in the micro­ biota colonizing the intestine promote storage of dietary calories in fat by increasing absorption of monosaccharides and suppressing epithelial-derived fasting-induced adipo­ cyte factor (FIAF).32 Thus, the composition and activity of the intestinal microbiota should be considered as a dietinfluenced variable that can influence susceptibility to obesity.

Table 102-1  Examples of Metabolic Activities of Intestinal Microbiota Biotransformation of bile acids Breakdown of dietary oxalate Conversion of prodrugs to active metabolites Degradation of polysaccharides of plant origin Production of folate, B vitamins, and vitamin K Production of nutrient short-chain fatty acids Production of regulatory signals for mucosal and immune homeostasis Regulation of fat storage

One of the outcomes of bacterial metabolic activity is gas production. Of the five gases—N2, O2, CO2, H2, CH4—that constitute 99% of flatus, the latter three are produced by the enteric bacteria, and bacteria are the sole source of hydrogen and methane in the intestine. Hydrogen production by bacterial action on carbohydrates, and to lesser extent on protein, normally occurs in the colon. In patients with small intestinal bacterial overgrowth, however, the small intestine also becomes a site of H2 production. Bacterial methanogens occur in the colon and produce methane from H2 and CO2, with significantly detectable excretion in approximately 30% of humans.33-37 The principal gases produced are odor­ less, but bacterial metabolism also is responsible for produc­ ing various trace and odiferous gases in flatus such as hydrogen disulfide.38,39 Qualitative and quantitative vari­ ability in gas production with diet illustrates the fluctua­ tions in bacterial metabolic activity despite the apparent stability of the microbiota in adulthood.

SMALL INTESTINAL BACTERIAL OVERGROWTH Small intestinal bacterial overgrowth (SIBO) is character­ ized by malabsorption and overgrowth of bacteria in the small intestine. The syndrome often is referred to as blind loop syndrome because of recognition of the disorder in patients with predisposing anatomic abnormalities. Other terms that have been used to describe the disorder include stagnant loop syndrome, contaminated small bowel, small intestinal colonization, and small bowel stasis. In 1939, Barker and Hummel40 reported macrocytic anemia in association with intestinal strictures and anastomoses and postulated that the anemia was secondary to bacterial overgrowth, or “putrefaction.” SIBO is not confined to humans and is well recognized in dogs.41 The syndrome is associated with a variety of anatomic disturbances, such as blind loops,42 and motility disorders, such as scleroderma,43 but it can occur in the absence of any specific predisposing factor. It is likely that the condition is underdiagnosed, particularly in the elderly.44 Patients with SIBO do not necessarily present with a florid malabsorption syndrome, and symptoms may be minor and nonspecific. Considerable debate has concerned the relationship between irritable bowel syndrome (IBS) and SIBO (see Chapter 118).45 Bacterial overgrowth has been documented in asymptomatic elderly persons in the com­ munity,46,47 in whom it is debatable whether the phenome­ non is of any significance48; in the absence of malabsorption or related symptoms, such overgrowth probably should not be considered to represent true SIBO. Asymptomatic SIBO may be termed simple colonization and probably results from achlorhydria and abnormal fasting intestinal motility (see later). The diagnosis of SIBO usually is made by noninvasive breath testing,49 even though studies on the accuracy of these tests report very variable results. For this reason, culture of a small intestinal aspirate must be regarded as the diagnostic gold standard. Unfortunately, much of the pub­ lished literature on SIBO is based on breath tests, rather than on culture, and the findings must be interpreted with caution.

ETIOLOGY AND PREDISPOSING FACTORS

The upper small intestine is an environment of relatively low bacterial counts because of the combined effects of gastric acid and peristalsis. Bacterial counts in aspirates

Chapter 102  Enteric Microbiota and Small Intestinal Bacterial Overgrowth Table 102-2  Pathophysiology and Some Conditions Associated with Small Intestinal Bacterial Overgrowth PATHOPHYSIOLOGY

CONDITION

Anatomic abnormalities

Blind loop (Billroth II gastrectomy, end-to-side anastomosis) Small intestinal diverticulosis Small intestinal stricture (Crohn’s disease, radiation enteritis, focal segmental ischemia) Diabetes mellitus Idiopathic intestinal pseudo-obstruction Scleroderma Acid-lowering medication Atrophic gastritis Previous vagotomy Gastrocolic or enterocolic fistula Resection of ileocecal valve

Motility disorders Reduced gastric acid secretion Abnormal connection between colon and proximal bowel Various mechanisms

Celiac disease Cirrhosis Chronic pancreatitis Chronic kidney disease Radiation enteritis Rheumatoid arthritis

from the normal upper small intestine generally are less than 1000/mL. Pathophysiology and predisposing condi­ tions are listed in Table 102-2.

Anatomic Abnormalities

A variety of anatomic abnormalities, including iatrogenic and disease-related abnormalities, lead to stagnation of small intestinal contents, resulting in bacterial overgrowth. The bacteria in SIBO are similar to those found in the normal colon, and certain organisms are common. Common aerobic organisms include Escherichia coli and Streptococcus, Staphylococcus, Micrococcus, Klebsiella, and Proteus species. Common anaerobic species include Lactobacillus, Bacteroides, Clostridium, Veillonella, Fusobacterium, and Peptostreptococcus.50 The classic anatomic cause of SIBO is a blind loop resulting from abdominal surgery, such as Billroth II partial gastrectomy; other anatomic abnormalities that can result in SIBO include intestinal strictures and small bowel diverticulosis.

Motility Disorders

Disorders affecting small intestinal peristalsis, such as scleroderma,43 diabetes mellitus,51 and chronic idiopathic intestinal pseudo-obstruction52 constitute the next most common cause of SIBO after anatomic abnormalities. SIBO is well recognized in scleroderma and occurs mainly in patients with small intestinal involvement72 who have limited cutaneous systemic sclerosis. Diarrhea is the most important symptom. The somatostatin analog octreotide is effective in the management of SIBO associated with scleroderma.73 Although small intestinal dysmotility is thought to be the main predisposing factor in diabetes, SIBO in diabe­ tics is not especially associated with autonomic neuro­ pathy.74 Treatment of SIBO in diabetics improves orocecal transit time.75

Fistula or Ileocecal Valve Resection

The ileocecal valve prevents reflux of colonic bacteria into the small intestine, and resection of the valve or develop­ ment of fistulas between the colon and upper gastrointesti­ nal tract can lead to reflux of colonic contents into the small intestine, with ensuing bacterial overgrowth.53,54

Reduced Gastric Acid Secretion

Achlorhydria is known to be a predisposing factor for SIBO, and SIBO has been described in patients after vagotomy,55 in those with atrophic gastritis, and in those taking acid suppressants.56-59 SIBO occurs more often in patients taking proton-pump inhibitors (PPIs) than in those taking hista­ mine H2 receptor antagonists,58 but clinical malabsorption does not appear to occur in this situation.59

Aging

Advancing age seems to be an independent risk factor for SIBO, but it is not clear if overgrowth results from the aging process itself and age-related changes in intestinal motility or if it is a consequence of achlorhydria. Early studies in this area found that SIBO was a common (and commonly unrecognized) cause of malabsorption in the elderly44,60 and that many such patients did not have an obvious predispos­ ing factor, such as a blind loop. More-recent studies have reported SIBO in asymptomatic elderly persons residing in the community. These patients, although asymptomatic, had lower weights and body mass indices (BMI) than expected, and treatment with antibiotics increased both weight and BMI.46,47 In contrast, a Japanese study reported SIBO (diagnosed by glucose hydrogen breath test) in 25.6% of disabled older adults but in none of the healthy older adults.61

Chronic Liver Disease

SIBO appears to be common in patients with chronic liver disease,62,63 is more common in patients with advanced (Child class C) liver disease,63 and may be an independent risk factor for spontaneous bacterial peritonitis,64 although this association is controversial.65 No association with any particular cause of chronic liver disease has been found,66 but SIBO does not occur in cirrhotic patients who do not have portal hypertension.67 The etiology of SIBO in patients with chronic liver disease is likely to be related to distur­ bances in gastrointestinal motility67 and possibly to the use of antacids,65 both of which can foster proliferation of bacteria. Small intestinal dysmotility is more severe in cirrhotic patients with a history of spontaneous bacterial peritonitis,64 and treatment of SIBO improves motility.68 Liver transplantation improves small bowel dysmotility in cirrhotic patients.68 Antibiotics and prokinetic agents are effective in reducing the SIBO associated with cirrho­ sis.69 SIBO in cirrhosis is associated with systemic endo­ toxemia.70 Oral conjugated bile acids reduce bacterial overgrowth and endotoxemia in cirrhotic rats, suggesting a contributory role for cholestasis in cirrhotic patients with SIBO.71

Other Causes

SIBO is present in many patients with celiac disease who have persistent symptoms despite their adherence to a gluten-free diet.76 It is not clear why this is so, but a motility disturbance seems the most likely explanation. SIBO is common in Crohn’s disease, particularly in patients who have had previous intestinal resection, and orocecal transit time has been reported to be prolonged in Crohn’s patients with SIBO.77 Positive results on glucose

1773

1774

Section X  Small and Large Intestine hydrogen breath tests are particularly associated with the presence of a small bowel stricture.78 SIBO is common in chronic pancreatitis.79 SIBO in this setting may be caused by small bowel dysmotility resulting from chronic opioid use and achlorhydria. Furthermore, pancreatic juice may have an antibacterial effect, so its absence might allow enteric bacteria to proliferate more freely.80 SIBO in rheumatoid arthritis is associated with high disease activity and does not appear to be related to achlorhydria.81 SIBO occurs in late radiation enteritis and appears to be related to intestinal dysmotility.82 SIBO is common in chronic kidney disease, which is associated with neuropathic-type abnormalities of small intestinal motility.83 SIBO occurs commonly in cystic fibrosis (CF). Use of azithromycin is paradoxically associated with an increased risk of a positive breath test for SIBO.84 In a murine model of CF, eradication of SIBO decreased intestinal mucus secretion.85 SIBO has been reported in patient populations with inter­ stitial cystitis,86 acne rosacea,87 morbid obesity,88 fibromyal­ gia,89 acromegaly,90 and focal segmental ischemia (see Chapter 114.) Many patients with SIBO have more than one predispos­ ing factor. For example, SIBO can develop in patients with scleroderma who receive PPI therapy for gastroesophageal reflux.

MECHANISMS OF MALABSORPTION

SIBO classically causes a combination of megaloblastic anemia (due to vitamin B12 deficiency) and steatorrhea (due to fat malabsorption). Megaloblastic anemia was described in association with intestinal strictures as long ago as 1897.91 Vitamin B12 deficiency is caused by bacterial utilization of the vitamin within the intestinal lumen before it can be absorbed across the mucosa.92 Anaerobic organisms mainly are responsible for the vitamin B12 deficiency, and in animal studies, only therapy directed against anaerobes reversed the deficiency.93 Unlike aerobic bacteria, anaerobes can use vitamin B12 both in its free form and complexed with intrin­ sic factor.94 Anaerobic bacteria deprive the host of ingested vitamin B12 and exacerbate B12 deficiency by using the vitamin to produce inactive cobamides, which then can compete with dietary B12 for ileal binding sites, thereby decreasing absorption of the vitamin.95 Deficiencies of thia­ mine96 and nicotinamide also have been reported in SIBO.97 Folate levels tend to be high in SIBO, because the bacteria synthesize folate,98 which then is absorbed and used by the host.99 Much of our knowledge on the mechanisms of malabsorp­ tion in SIBO is derived from animal models of blind loops.100,101 Malabsorption of fat and fat-soluble vitamins results mainly from deconjugation of bile acids,102 and administration of conjugated bile acids has been reported to reverse steatorrhea in human and animal studies.103 Defi­ ciencies of vitamins A, D,104 and E105 have been reported, but vitamin K deficiency is uncommon because production of vitamin K by luminal bacteria offsets any deficiencies attributed to fecal fat loss. SIBO leads to carbohydrate malabsorption by reducing brush border disaccharidase levels.106-108 In animal studies, bacterial extracts of cultures from experimentally created blind loops contain proteases that can remove components of the intestinal surface membrane.109 These proteases appear to have elastase-like substrate specificity and may be etiologic in disaccharidase deficiency. Lactose intolerance

is common and contributes to the diarrhea that typifies SIBO. Bacterial fermentation of carbohydrates contributes to abdominal discomfort and bloating in SIBO and is the basis for the various breath tests used to diagnose the condition. Protein malabsorption in SIBO is caused by a number of factors: decreased absorption of amino acid and peptides, which has been described in animal models and can result from mucosal damage110; low levels of enterokinase, which can impair the activation of pancreatic proteases111; and protein-losing enteropathy.112 Although hypoproteinemia is common in SIBO, manifestations of severe hypoprotein­ emia, such as edema, are rare. Small intestinal histologic findings generally are normal in patients with SIBO, and in one study, morphometric findings in the small intestine also were described as normal.113 Abnormalities of small intestinal mucosa (e.g., villus atrophy, cellular infiltration of the lamina propria, intraepithelial lymphocytosis) have been described in some patients with SIBO, and these changes revert to normal fol­ lowing treatment with antibiotics.114 Electron microscopy studies of experimental animals with SIBO have described enterocyte abnormalities, such as vacuolization of micro­ villus membranes and mitochondrial swelling.115

CLINICAL FEATURES

SIBO may be difficult to diagnose because symptoms associ­ ated with the predisposing disorder can predominate. The classic clinical presentation of SIBO is that of a malabsorp­ tive state characterized by steatorrhea and vitamin B12 defi­ ciency that is not reversible with intrinsic factor. Patients with vitamin B12 deficiency can present with neurologic symptoms, central or peripheral neuropathy, and symptoms of anemia, such as fatigue, breathlessness, and chest pain. Patients with steatorrhea can report weight loss, diarrhea, and abdominal bloating and discomfort. Associated fatsoluble vitamin deficiency can occur, leading to night blind­ ness (in vitamin A deficiency) and metabolic bone disease (in vitamin D deficiency). Osteoporosis is another wellrecognized complication of SIBO.116,117 The clinical presentation of SIBO appears to be changing. Older references to clinical features of SIBO emphasized steatorrhea, megaloblastic anemia, and a history of surgery leading to blind loop syndrome. In a somewhat more modern series, Toskes and Kumar reported data for 100 consecutive, albeit highly selected, symptomatic patients referred for 14C-xylose breath testing118 and found a history of gastrointestinal surgery in only 15%. The three most common associated conditions, which accounted for more than 90% of the positive results on 14C-xylose breath tests for the patients in their referral center, were gastroparesis, chronic pancreatitis, and irritable bowel syndrome. Diar­ rhea, bloating, and flatulence were the most common symp­ toms. More recent studies demonstrate that the clinical presentation of SIBO may be less dramatic than the classic descriptions of SIBO and have milder symptoms. The wide use of breath tests is one reason we see this newer type of SIBO patient.

Small Intestinal Bacterial Overgrowth and Irritable Bowel Syndrome

Many patients with SIBO fulfill the Rome Criteria for IBS, and considerable debate on the relationship between IBS and SIBO has ensued since a study in 2000 reported that 78% of patients with Rome criteria-positive IBS tested posi­ tive for SIBO by lactulose hydrogen breath testing.119 Anti­ biotic therapy led to clinical improvement and normalization of the breath test. This study also provoked criticism of its

Chapter 102  Enteric Microbiota and Small Intestinal Bacterial Overgrowth selection criteria and study design and raised concerns regarding the accuracy of the lactulose hydrogen breath test.120 Two subsequent studies found no difference in lactu­ lose hydrogen breath test positivity between IBS patients and controls.121,122 A Swedish study, using cultures of intes­ tinal aspirate to diagnose SIBO, reported a prevalence of SIBO of 4% in both IBS patients and controls.123 The hy­ pothesis that SIBO is a significant etiological factor in IBS remains unproved. Nonetheless, the putative relationship between SIBO and IBS has stimulated a number of clinical trials of antibiotic therapy in IBS. Three relatively small controlled trials reported a modest benefit of antibiotic therapy in IBS patients,124-126 although the design and statistical analysis of these trials have been criticized.45 The major benefit appeared to be in patients with abdominal bloating, suggest­ ing that the predominant effect of antibiotic therapy is to reduce gas-forming bacteria. Symptom improvement in these studies did not consistently correlate with normaliza­ tion of the lactulose hydrogen breath test, however, suggesting that the benefit of antibiotic therapy might be reduction of colonic, rather than small intestinal, gasforming bacteria. These topics are discussed more fully in Chapter 118.

Small Intestinal Bacterial Overgrowth and Nonalcoholic Steatohepatitis

Considerable interest has arisen in the putative association between SIBO and nonalcoholic steatohepatitis (NASH). It has been postulated that SIBO might play a role in the pathogenesis of NASH,127 because NASH is a common com­ plication of jejunoileal bypass surgery for morbid obesity and can be reversed with metronidazole treatment. Anti­ biotic treatment prevents hepatic and bile duct injury in genetically susceptible rats with surgically created blind loops and SIBO,128,129 although the pattern of hepatic and biliary injury in this experimental situation was histologi­ cally and radiologically more compatible with primary scle­ rosing cholangitis than with NASH. Wigg and colleagues130 postulated that SIBO might lead to NASH by altering small intestinal permeability and thereby increasing absorption of endotoxin. They studied 22 patients with NASH and found SIBO (by lactulose hydrogen breath testing) in 50% of sub­ jects; serum endotoxin levels and small intestinal permea­ bility, measured by the lactulose-rhamnose test, however, were normal in the patients with SIBO.

DIAGNOSIS

The diagnosis of SIBO should be considered in any patient with malabsorption and a predisposing condition. As mentioned earlier, most patients today do not have pre­ disposing surgically induced anatomic abnormalities. It is likely that SIBO is commonly overlooked in patients without known predisposing factors and in patients who have nonspecific symptoms. Blood tests in patients with SIBO typically reveal a macrocytic anemia: Vitamin B12 levels are low, and folate levels may be high. Steatorrhea may be confirmed by three-day quantitative fecal fat collection; this test has understandably fallen from favor with patients and laboratory staff, and qualitative microscopic examina­ tion of fresh stool for fat globules usually is performed.131,132 If an anatomic defect is suspected as the cause of SIBO, appropriate barium studies may be used to define the anatomy.

Aspiration

The gold standard test for the diagnosis of SIBO is aspiration of small intestinal fluid with culture and bacterial counts of

the aspirate; presence of more than 105 CFU/mL of duodenal aspirate is considered diagnostic. Unfortunately, such aspi­ ration is invasive and time-consuming. Moreover, although it still is recommended by most experts, some investigators have raised concerns that the test might miss bacterial overgrowth occurring more distally in the small intestine. Corazza and colleagues,133 however, collected intestinal juice at two different levels of the proximal jejunum and reported a highly significant correlation between the bacte­ rial counts at these sites. Other potential problems with aspiration of small intestinal fluid include contamination of the aspirate with bacteria from the mouth and technical difficulties with transport and culture of the aspirate. Contamination with oropharyngeal bacteria may be con­ trolled for by simultaneous culture of saliva and jejunal aspirate.134 Several techniques for collecting small intestinal contents have been described, including duodenal intubation with fluoroscopic guidance and endoscopic collection of fluid,135 and brushing of the duodenal mucosa with a cytology brush.136 Culture of unwashed small intestinal mucosal biopsy specimens is an alternative to culture of a small intestinal aspirate, although the former method appears to have a lower sensitivity compared with culture of aspi­ rates.137 Aspirate can be collected easily during routine endoscopy, and this is probably the easiest method in routine clinical practice. Small intestinal aspirate is col­ lected by placing a sterile suction catheter inside a sterile overtube, which is passed through the suction channel of the endoscope. The aspirate should be placed immediately in aerobic and anaerobic transport vials, and the aspirate should be plated for aerobic and anaerobic organisms as soon as possible. High levels of jejunal fluid volatile fatty acids, such as acetate and propionate, have been reported in SIBO.138 These acids may be measured by gas-liquid chromatogra­ phy; although the technique is highly specific, the sensitiv­ ity is low,133 and this test is rarely used.

Breath Tests

A variety of noninvasive tests have been developed for diagnosing SIBO. The 14C-glycocholic acid breath test was one of the first breath tests used for this purpose and is based on the ability of bacteria to deconjugate bile salts. 14 C-glycine is produced and metabolized, resulting in a peak of 14CO2 in the expired air. The test has a low sensitivity, because not all bacteria are capable of such deconjugation, and the test has a low specificity, because increased colonic deconjugation of bile salts can occur with ileal disease or following ileal resection.139 The test therefore cannot distin­ guish between SIBO and ileal malabsorption and has largely fallen out of favor. The currently used breath tests are based on the ability of bacteria to produce hydrogen or radiolabeled carbon dioxide after metabolizing a substrate such as glucose, lactulose, or xylose. Breath tests are simple and noninvasive and there­ fore are more attractive than is duodenal intubation or endoscopy for collecting intestinal aspirates. These breath tests, however, do have several potential problems49: About 15% of the population are methane producers (in persons who are colonized with Methanobrevibacter smithii, hydro­ gen reacts with carbon dioxide to form methane, so less hydrogen is produced than in non-methane producers). Both slow and rapid small intestinal transit can affect the accuracy of these tests. An acidic environment in the colon, such as occurs with ingestion of nonabsorbable carbohy­ drates (e.g., lactulose), inhibits bacterial carbohydrate metabolism.140 Several patient-related factors, such as recent

1775

1776

Section X  Small and Large Intestine Table 102-3  Sensitivity and Specificity of Breath Testing in the Diagnosis of Small Intestinal Bacterial Overgrowth: Summary of Clinical Studies NUMBER OF PATIENTS REFERENCE Glucose Hydrogen 133 144 145 146 147 148 Lactulose Hydrogen 133 148 149 150 14 C-D-Xylose 147 152 153 154 155 156 157

Total

With Overgrowth

SENSITIVITY (%)

SPECIFICITY (%)

77 45 30 40 46 81

44 27 20 29 24 32

62 93 75 27-52 58 44

83 78 30 36-80 86 80

77 81 27 28

44 32 9 18

68 31 89 17

44 86 100 70

46 12 60 30 10 20 47

24 12 23 20 6 10 14

42 100 65 95 60 60 79

86 — 59 100 40 90 85

diet, smoking, and exercise, can influence baseline levels of breath hydrogen (see later). The literature on breath tests in SIBO is confusing, with wide variations in sensitivity and specificity (Table 102-3).

colleagues,133 reported sensitivity of 62% and specificity of 83%. Very rapid intestinal transit can lead to a false-positive test result, because glucose can reach the colon before it can be absorbed.

Glucose Hydrogen The glucose hydrogen breath test probably is the most widely used breath test in clinical practice: The substrate is inexpensive, and the hydrogen meter is economical, por­ table, and easy to use. The glucose hydrogen breath test first was reported as a diagnostic test for SIBO in 1972 by Bond and Levitt.141 Normally, glucose is absorbed completely in the upper small intestine; with bacterial overgrowth, however, the glucose is cleaved by bacteria into carbon dioxide and hydrogen. The hydrogen is measured in the exhaled breath (at baseline and then every 30 minutes for 2 hours); a rise of 20 parts per million (ppm) above the base­ line is regarded as diagnostic of SIBO. Fasting breath hydro­ gen levels of more than 20 ppm also are considered positive. High baseline hydrogen levels also are common in untreated celiac disease and normalize after gluten withdrawal for as-yet-unknown reasons.142 Patient preparation is important for the glucose hydrogen test143: Patients must avoid smoking and ingestion of non­ fermentable carbohydrates, such as pasta and bread, the night before the test, because these factors can raise baseline breath hydrogen values. Exercise can induce hyperventila­ tion, thereby reducing baseline breath hydrogen values, and should be avoided for two hours before the test. Some authors recommend an antibacterial mouth rinse before testing to prevent premature hydrogen or carbon dioxide production from the action of the oral flora on the glucose substrate. A number of studies have compared the glucose hydrogen breath test against the gold standard of culture of intestinal aspirate (see Table 102-3). Sensitivity levels from 27% to 52% have been reported, with specificity rates between 30% and 83%133,144-148; the largest study, by Corazza and

Lactulose Hydrogen The lactulose hydrogen breath test is based on a principle similar to that of the glucose hydrogen breath test: Lactulose is a disaccharide that is not absorbed in the small intestine but is metabolized by bacteria in the proximal colon, pro­ ducing a late peak in exhaled hydrogen. In the presence of bacterial overgrowth, an early hydrogen peak is observed. Results of this test may be difficult to interpret with either slow or fast intestinal transit, and sensitivity and specificity have been disappointing133,149,150 (see Table 102-3); Corazza and associates reported sensitivity and specificity rates of 68% and 44%, respectively.133 Sensitivity of the test may be increased by the addition of scintigraphy to correct for abnormalities of intestinal transit,150 but the lactulose hydrogen breath test cannot be recommended for routine clinical use. Xylose The 14C-xylose and 13C-xylose breath tests measure labeled carbon dioxide that is produced by breakdown of labeled substrates by bacteria. The isotope may be radioactive (14C) or stable (13C); the stable isotope has been used in chil­ dren.151 d-Xylose is the most widely used substrate and is a good substrate for breath testing for SIBO because it is absorbed completely in the small intestine, is metabolized minimally, and is catabolized by Gram-negative bacteria. The 14C-d-xylose breath test appears to perform better than the glucose or lactulose hydrogen breath test (see Table 102-3) but, as with these other breath tests, widely differing levels of accuracy have been reported, with sensitivity rates ranging from 42% to 95% and specificity rates between 40% and 100%.147,152-157 The 14C-d-xylose breath test result is con­ sidered positive when the “cumulated dose at four hours

Chapter 102  Enteric Microbiota and Small Intestinal Bacterial Overgrowth exceeds 4.5% of the administered radioactivity.”49 Distur­ bances in intestinal transit particularly affect the perfor­ mance of this test, and accuracy may be improved by the addition of a transit marker (such as barium or diatrizoate meglumine–diatrizoate sodium [Gastrografin]) and radio­ logic imaging.156

Other Tests

Other noninvasive tests described for SIBO include mea­ surement of urinary cholyl-para-aminobenzoic acid (PABA) and serum bile acids. Cholyl-PABA is a synthetic substrate made by conjugating cholic acid with PABA, which is hydrolyzed by the bacterial enzyme cholyl hydrolase to release PABA158; this PABA-based test, however, does not accurately distinguish between SIBO and other causes of malabsorption.159 Elevated free serum bile acids have been reported in SIBO, but the test depends on the presence of bacteria that deconjugate bile salts, such as Bacteroides.160

Choice of Test

It is our view that duodenal intubation with collection and culture of small intestinal aspirate remains the gold stan­ dard test for diagnosing SIBO. Small intestinal juice can be collected easily during standard endoscopy, and we recom­ mend endoscopy with duodenal biopsy and collection of small intestinal juice for culture as a standard diagnostic approach in patients with malabsorption. It is common practice, however, to provide empirical antibiotic treatment for patients with suspected SIBO, without either breath testing or culture of small intestinal aspirate.

TREATMENT

Attention should be given to the patient’s nutritional state, and any vitamin deficiency should be corrected (see Chap­ ters 4, 5, and 100). A lactose-free diet can ameliorate the diarrhea. If possible, any predisposing anatomic or func­ tional abnormality should be corrected, but in practice, this is unlikely to be an option. Acid-lowering medication should be discontinued, if possible. A variety of antibiotics have been reported to be effective in SIBO, but little evidence exists to favor one agent over another. Antibiotics that have been reported to be effective include metronidazole, amoxicillin, amoxicillinclavulanate, ciprofloxacin, tetracycline, and cotrimoxazole. One randomized crossover trial reported that norfloxacin and amoxicillin-clavulanate were effective in SIBO.161 In another study, rifaximin and chlortetracycline normalized results on glucose hydrogen breath testing in 70% and 27%, respectively, of patients with SIBO.162 Both ciprofloxacin and metronidazole were found to be highly effective in SIBO associated with Crohn’s disease, and although these antibiotics have been used for primary therapy in Crohn’s disease, normalization of breath tests occurred in most of the patients in this study.163 There have been several reports on the use of the nonab­ sorbable antibiotic rifaximin in SIBO. Rifaximin at a dose of 1.2 g/day and 1.6 g/day leads to normalization of the glucose hydrogen breath test in 58% and 80% of patients, respectively.164 Metronidazole is more effective than rifa­ ximin, at least in patients with SIBO associated with the blind-loop syndrome.165 Recurrence of SIBO after rifaximin treatment is common.166 Therapy usually is given initially for two weeks, and then clinical response is assessed; it may be useful to repeat a breath test or culture of small intestinal aspirate. Many patients with an underlying anatomic or motility disorder require permanent antibiotic treatment; in such patients, it

is usual to rotate antibiotic treatment every two weeks or, alternatively, to give antibiotics for two of every four weeks. Continuous treatment with a single agent can lead to antibiotic resistance or to side effects associated with long-term use, such as peripheral neuropathy in patients given metronidazole. The somatostatin analog octreotide stimulates intestinal motor activity when administered in low dosage. Given subcutaneously at 50 µg once daily for three weeks, it has been reported to be effective in SIBO associated with scleroderma.73 At higher doses (200 µg three times daily), octreotide paradoxically can cause SIBO by inducing hypo­ motility.167 The prokinetic agent cisapride has been reported to be effective in SIBO associated with cirrhosis,168 but the drug is no longer available in the United States, and its use in several countries is strictly controlled because of risk of drug interactions and cardiac arrhythmias. Probiotic therapy is a logical and attractive approach to the management of SIBO, but it has been examined in only a few studies. Saccharomyces boulardii does not appear to be effective, and in one double-blind crossover study, Lactobacillus fermentum KLD showed no advantage over placebo.169 A small uncontrolled trial showed that Lactobacillus plantarum 299V and Lactobacillus GG bene­ fited children who had SIBO associated with short bowel syndrome.170 In developed countries, SIBO is probably second only to celiac disease as the most common cause of malabsorption. The condition no longer manifests commonly with the classic features of steatorrhea and megaloblastic anemia, and most patients do not have a blind loop or other predis­ posing anatomic abnormalities. Many patients have nonspe­ cific symptoms similar to those of IBS. Although the glucose hydrogen and 14C-xylose breath tests are simple and nonin­ vasive, the gold standard test for diagnosis is culture of small intestinal aspirate. The aspirate can be easily col­ lected at endoscopy, which usually is performed to obtain biopsy specimens of the small intestine during evaluation of malabsorption. Treatment with one of several broadspectrum antibiotics is simple and effective.

KEY REFERENCES

Artis D. Epithelial-cell recognition of commensal bacteria and mainte­ nance of immune homeostasis in the gut. Nat Rev Immunol 2008; 8:411-20. (Ref 28.) Backhed F, Ding H, Wang T, et al. The gut microbiota as an environ­ mental factor that regulates fat storage. Proc Natl Acad Sci U S A 2004; 101:15718-23. (Ref 32.) Bratten J, Spanier J, Jones MP. Lactulose hydrogen breath testing does not discriminate patients with irritable bowel syndrome from healthy controls. Am J Gastroenterol 2008; 103:958-63. (Ref 122.) Castiglione F, Rispo A, Di Girolamo E, et al. Antibiotic treatment of small bowel bacterial overgrowth in patients with Crohn’s disease. Aliment Pharmacol Ther 2003; 18:1107. (Ref 163.) Di Sefano M, Miceli E, Missanelli M, et al. Absorbable vs non-absorbable antibiotics in the treatment of small intestine bacterial overgrowth in patients with blind-loop syndrome. Aliment Pharmacol Ther 2005; 21:985. (Ref 165.) Marchesi J, Shanahan F. The normal intestinal microbiota. Curr Opin Infectious Dis 2007; 20:508-13. (Ref 8.) O’Hara AM, Shanahan F. The gut flora as a forgotten organ. EMBO Rep 2006; 7:688-93. (Ref 2.) Lauritano EC, Gabrielli M, Scarpellini E et al. Small intestinal bacterial overgrowth recurrence after antibiotic therapy. Am J Gastroenterol 2008; 103:2031-5. (Ref 166.) Parlesak A, Klein B, Schecher K, et al. Prevalence of small bowel bacterial overgrowth and its association with nutrition intake in nonhospitalised older adults. J Am Geriatr Soc 2003; 51:768-73. (Ref 47.) Pimentel M, Park S, Mirocha J, et al. The effect of a non-absorbed oral antibiotic (rifaximin) on the symptoms of the irritable bowel

1777

1778

Section X  Small and Large Intestine syndrome: A randomized trial. Ann Intern Med 2006; 145:557-63. (Ref 125.) Rakoff-Nahoum S, Paglino J, Eslami-Varzaneh F, et al. Recognition of commensal microflora by Toll-like receptors is required for intestinal homeostasis. Cell 2004; 118:229-41. (Ref 21.) Riordan SM, McIver CJ, Wakefield D, et al. Small intestinal mucosal immunity and morphometry in luminal growth of indigenous gut flora. Am J Gastroenterol 2001; 96:494-500. (Ref 113.) Romagnuolo J, Schiller D, Bailey RJ. Using breath tests wisely in a gastroenterology practice: An evidence-based review of indications and pitfalls in interpretation. Am J Gastroenterol 2002; 97:2113-26. (Ref 49.) Scarpellini E, Gabrielli M, Lauritano CE et al. High dose rifaximin for the treatment of small intestinal bacterial overgrowth. Aliment Pharmacol Ther 2007; 25:781-6. (Ref 164.) Sharara AI, Aoun E, Abdul-Baki H, et al. A randomized double-blind placebo-controlled trial of rifaximin in patients with abdominal

bloating and flatulence. Am J Gastroenterol 2006; 101:326-33. (Ref 126.) Turnbaugh PJ, Ley RE, Hamady M, et al. The human microbiome project. Nature 2007; 449:804-10. (Ref 4.) Vanner, S. The lactulose breath test for diagnosing SIBO in IBS patients: Another nail in the coffin. Am J Gastroenterol 2008; 103:964-65. (Ref 120.) Vanner S. The small intestinal bacterial overgrowth/irritable bowel syn­ drome hypothesis: Implications for treatment. Gut 2008; 57:1315-21. (Ref 45.) Wigg AJ, Roberts-Thomson IC, Dymock RB: The role of small intestinal bacterial overgrowth, intestinal permeability, endotoxaemia, and tumour necrosis factor alpha in the pathogenesis of non-alcoholic steatohepatitis. Gut 2001; 48:206-11. (Ref 130.) Full references for this chapter can be found on www.expertconsult.com.

CHAPTE R

103 Short Bowel Syndrome Alan L. Buchman

CHAPTER OUTLINE Etiology  1779 Incidence and Prevalence  1779 Pathophysiology  1779 Loss of Absorptive Surface Area  1780 Loss of Site-Specific Transport Processes  1782 Loss of Site-Specific Endocrine Cells and Gastrointestinal Hormones  1782 Loss of the Ileocecal Valve  1783 Intestinal Adaptation to Resection  1783 Medical Management  1783 Limited Ileal Resection  1783 Extensive Small Intestinal Resection and Partial Colectomy  1784

Short bowel syndrome (SBS) is characterized by malabsorption due to insufficient intestinal surface area, such that the affected person is unable to absorb sufficient fluid, energy, or nutrients to sustain life in the absence of specialized nutritional support. This syndrome, also known as intestinal failure, occurs in adults in whom less than 200 cm of small intestine is present. The spectrum of SBS, however, ranges from limited ileocolonic resections with moderate nutritional compromise to extensive small intestinal and colonic resections, with duodenostomy, proximal jejunostomy, or jejunocolonic anastomosis and severe nutritional consequences. SBS also may be a congenital condition or can result from a variety of congenital causes.

ETIOLOGY The major causes of SBS in adults are Crohn’s disease for which multiple intestinal resections have been performed; mesenteric infarction from venous or arterial thrombosis, arterial embolism, or midgut volvulus; massive enterectomy performed to manage traumatic injuries or tumor resection, and radiation injury (Table 103-1). The causes of SBS in the pediatric population are congenital abnormalities (see Chapter 96), including gastroschisis, intestinal atresia, malrotation, aganglionosis, and necrotizing enterocolitis. More than 90% of infants now survive the extensive intestinal resections required for these conditions, and these patients need careful follow-up for their SBS as they mature to adulthood. Intestinal failure also can result from chronic intestinal pseudo-obstruction syndrome in both adults and children (see Chapter 120), as well as from unclassified sprue in adults (see Chapter 104) and congenital villus atrophy in children.

Home Parenteral Nutrition  1786 Complications  1788 Gallstones  1788 Liver Disease  1790 Calcium Oxalate Kidney Stones  1790 d-Lactic Acidosis  1790 Other Complications  1791 Surgical Management  1791 Intestinal Lengthening Procedures  1791 Intestinal Transplantation  1791 Pharmacologic Enhancement of Bowel Adaptation  1793 Survival and Quality of Life  1795

INCIDENCE AND PREVALENCE The incidence of SBS is difficult to assess in the United States, because of a lack of a national registry for affected persons and a lack of prospective studies in defined populations of patients who have undergone extensive intestinal resections. The incidence of severe SBS necessitating longterm parenteral nutrition is estimated to be 2 to 4 cases per 1 million persons per year, based on multinational European data.1 It is estimated that between 10,000 and 20,000 patients in the United States follow a home parenteral nutrition regimen for SBS. Approximately 50% to 70% of patients with SBS who initially require parenteral nutrition can be weaned from this therapy and therefore might not be reflected in the prevalence estimates.2,3 Such patients often still require aggressive nutritional monitoring. The incidence and prevalence of SBS associated with Crohn’s disease are decreasing now that infliximab and strictureplasty have become commonplace.

PATHOPHYSIOLOGY The major consequence of extensive intestinal resection is loss of absorptive surface area, which results in malabsorption of macronutrients, micronutrients, electrolytes, and water.4 The degree of malabsorption is determined by the length of the remnant intestine; the specific portions of small and large intestine resected, along with their sitespecific transport processes and endocrine cells; and the adequacy of adaptive processes in the residual intestine over time. Three types of intestinal resections typically are encountered: limited ileal resection for Crohn’s disease, often with cecectomy or right hemicolectomy; extensive

1779

1780

Section X  Small and Large Intestine Proximal small intestine Fat Sugars Peptides and amino acids Iron Folate Calcium Water Electrolytes

Ileocolonic Jejunocolonic End-jejunostomy anastomosis anastomosis Figure 103-1.  The three common types of intestinal resection and anastomosis observed in patients with short bowel syndrome: ileocolonic anastomosis, jejunocolonic anastomosis, and end-jejunostomy.

Table 103-1  Causes of Short Bowel Syndrome and Intestinal Failure In Adults Catastrophic vascular accidents Superior mesenteric arterial embolism Superior mesenteric arterial thrombosis Superior mesenteric venous thrombosis Chronic intestinal pseudo-obstruction* Intestinal resection for tumor Midgut volvulus Multiple intestinal resections for Crohn’s disease Radiation enteritis* Refractory sprue* Scleroderma and mixed connective tissue disease* Trauma In Children Congenital villus atrophy* Extensive aganglionosis* Gastroschisis Jejunal or ileal atresia Necrotizing enterocolitis *Functional short bowel syndrome also can occur in conditions associated with severe malabsorption, in which the bowel length often is intact.

ileal resection with or without partial colectomy and with jejunocolonic anastomosis; and extensive small intestinal resection and total colectomy resulting in proximal jejuno­ stomy (Fig. 103-1). Patients in the two latter groups commonly suffer from Crohn’s disease or had mesenteric infarction.

LOSS OF ABSORPTIVE SURFACE AREA Nutrient Malabsorption

The length of the small intestine is estimated at 3 to 8 meters, and nutrient absorption is preserved until more than one half of the small intestine is resected.5-9 Most macronutrients (carbohydrate, fat, and nitrogen) are absorbed in the proximal 100 to 150 cm of intestine.10 Specific areas of absorption in the small intestine of nutrients, minerals, vitamins, electrolytes, and trace elements are discussed in Chapters 99 to 101 and are illustrated in Figure 103-2. Enterocytes lining the small intestine appear uniform from the duodenum to the ileocecal valve, but a distinct proximal-to-distal gradient exists in both morphology and function.11 Villi are taller and crypts are deeper in the

Colon Water Electrolytes MCTs Amino acids

Middle small intestine Sugars Peptides and amino acids Calcium Water Electrolytes Distal small intestine Bile salts Vitamin B12 Water Electrolytes

Figure 103-2.  Specific areas of absorption of dietary constituents and secretions in the small intestine and colon. Macronutrients and micronutrients are absorbed predominantly in the proximal jejunum. Bile acids and vitamin B12 (cobalamin) are absorbed only in the ileum. Electrolytes and water are absorbed in both the small and the large intestine. Medium-chain triglycerides (MCTs), calcium, and some amino acids can be absorbed in the colon.

jejunum than in the ileum, and the activity of microvillus enzymes and nutrient absorptive capacity per unit length of intestine are several-fold higher in the proximal than in the distal small intestine; loss of part of the jejunum initially compromises nutrient absorption more than does loss of an ileal segment of similar length, because of these morphologic and functional differences. The ileum, however, eventually is able to compensate for jejunal loss, whereas the jejunum is unable to compensate for ileal absorption of bile salts and vitamin B12. Normal digestion and absorption depend on the gradual gastric emptying of partially digested nutrients, mixing of these nutrients with bile and pancreatic enzymes in the duodenum, and rapid digestion and absorption of the digestive products in the proximal small intestine. Patients with a proximal jejunostomy have rapid gastric emptying of liquids and rapid intestinal transit, which can compromise the gastric phase of digestion and result in inadequate mixing with biliary and pancreatic secretions, insufficient enzymatic digestion, and nutrient maldigestion. Rapid intestinal transit decreases nutrient-enterocyte contact time, and therefore, segmental absorption is decreased. Patients with a high jejunostomy are net secretors of salt and fluid, because jejunal fluid secretion is stimulated by oral intake and subsequent gastric emptying of nutrients; these patients excrete more fluid than they ingest, and accordingly, their fluid management may be challenging.12 Most patients whose jejunum is shorter than 100 cm and who have no colon require long-term parenteral nutrition. Preservation of even some colon at surgery is highly beneficial for nutrient absorption. The ileocecal valve acts as a brake to slow intestinal transit, thereby increasing nutrient-enterocyte contact time and enhancing absorption.

Chapter 103  Short Bowel Syndrome 50-g bread meal

Table 103-2  Daily Stomal or Fecal Losses of Electrolytes, Minerals, and Trace Elements in Severe Short Bowel Syndrome*

24 g of malabsorbed CHO Colonic bacterial fermentation

240 mmol of SCFA (60% acetate), hydrogen, methane, carbon dioxide, and sulfides

72 kcal Figure 103-3.  Colonic absorption of malabsorbed carbohydrate (CHO) in a hypothetical patient with short bowel syndrome following ingestion of a 50-g bread meal. Unabsorbed carbohydrates (∼24 g), nonstarch polysaccharides, and soluble fiber are fermented by colonic bacterial flora to hydrogen, methane, carbon dioxide, sulfides, and ∼240 mmol short-chain fatty acids (SCFAs), including acetate, butyrate, and propionate to generate 72 kcal. By comparison, normal persons absorb 220-720 mmol SCFA from fermentation of 30-60 g nonstarch polysaccharides. 10

100% colon remaining 1 tsp/day) Pretzels Soybean curd (tofu)

1787

1788

Section X  Small and Large Intestine Fever (may occur only during infusion); chills when catheter is flushed or during infusion

Erythema of SQ tunnel tract

Blood cultures from the catheter and peripherally for bacteria and fungi CBC with peripheral blood smear (examine for bacteria, yeast)

Tunnel infection

Remove catheter

Exit site infection

Empiric treatment for MRSA with vancomycin or linezolid; adjust antibiotics based on C&S

Rx for coag. neg. Staph and Gram neg. organisms empirically; adjust antibiotics based on C&S + hold TPN for 24 hrs

Empiric treatment for MSSA and MRSA with vancomycin or linezolid

IV antibiotics × 1–3 wks or antibiotic lock technique × 1 wk

Catheter may be replaced in a different site

Bacteria

Fever resolves in 72 hrs

No

Severe sepsis despite antimicrobial therapy

Remove catheter

Infection present

Fungus

Continue appropriate antibiotic(s)

Continue antibiotic(s)

Tenderness, erythema or purulence of exit site (or on dressing)

Remove catheter Ampho B 100–250 mg total

Rx for an additional 2 wks with IV antibiotics

Replace catheter once patient is afebrile × 48–72 hrs or blood culture is negative in the absence of fever

Infection cured

No further Rx Remove catheter

Replace catheter at a different site

Recurrence within 6 wks of antibiotic discontinuation (same organism)

Retreat Unsuccessful

Remove catheter

Figure 103-6.  Algorithm for the diagnosis and management of catheter-related infection. ampho B, amphotericin B; coag., coagulase; CBC, complete blood count; C&S, culture and sensitivity testing; MSSA, methicillin-sensitive Staph aureus; MRSA, methicillin-resistant Staph aureus; neg., negative; Rx, treatment; SQ, subcutaneous; Staph, Staphylococcus; S. aureus, Staphylococcus aureus; TPN, total parenteral nutrition.

Patients in whom the frequency of TPN infusions can be reduced to fewer than five nights per week should have their micronutrient status monitored two to three times yearly to detect deficiencies. At the clinical visit, particular attention should be paid to the catheter exit site for evidence of erythema, purulent discharge, warmth, or tenderness. A catheter may remain in place indefinitely if it is properly maintained.88

COMPLICATIONS GALLSTONES

Interruption of the enterohepatic circulation of bile acids by ileal resection results in decreased hepatic bile acid secretion and altered composition of hepatic bile in terms of its organic components: bile acid, cholesterol, and phospholipids (see Chapters 64 and 72). Hepatic bile

Chapter 103  Short Bowel Syndrome Slowed rate of TPN and/or lipid emulsion infusion, resistance to flushing of catheter, leakage or swelling of the exit site

Empirically treat for possible thrombosis with tPA (2 mg in 2 mL)

Radiologic contrast study of the catheter to verify or exclude occlusion

Attempt aspiration through catheter in 60 min

Malposition

Attempt to adjust position with a pig-tail catheter (sterile technique by radiologist)

Successful

Unsuccessful

Consider minidose warfarin (1–2 mg by mouth daily)

Radiologic contrast study of the catheter

Successful

Unsuccessful

Thrombosis No further treatment

Replace catheter

tPA (2 mg in 2 mL)

Unsuccessful

Occlusion, but no thrombosis

Remove and replace catheter

HCI (0.1N) or ethanol (70%)

NaOH (0.1N)

Attempt aspiration through catheter after 1 hr

Attempt aspiration through catheter after 6–7 hrs

Successful

Unsuccessful

Review TPN formula for incompatibilities

Replace catheter

Successful

Unsuccessful

Figure 103-7.  Algorithm for the diagnosis and management of thrombotic or nonthrombotic catheter-related occlusion. tPA, tissue plasminogen activator; TPN, total parenteral nutrition.

becomes supersaturated with cholesterol, with subsequent formation of cholesterol crystals and gallstones in gallbladder bile (see Chapter 65). Most gallstones in patients with SBS, however, are composed of calcium bilirubinate; the pathophysiology is unclear. A prevalence of 44% of asymptomatic gallstones was documented in a study of 84 patients with severe SBS who required TPN.19 Formation of biliary sludge and gallbladder hypomotility probably

contribute to the high prevalence of these stones, because many of these patients are on long-term parenteral nutrition.89 Postprandial CCK concentration is decreased in some patients with SBS,90 and injections of CCK have been used experimentally to induce gallbladder contraction, although this therapy is not always successful and results in nausea, vomiting, and abdominal pain in some patients.91,92

1789

1790

Section X  Small and Large Intestine LIVER DISEASE

Liver disease often develops in patients who require longterm TPN. After five years of TPN, more than 50% of these patients will be found to have severe liver disease defined as grade 2 fibrosis, cirrhosis, or one of the following: total serum bilirubin greater than 3.5 mg/dL for longer than one month, ascites, portal hypertension, hepatic encephalopathy, or liver failure with a factor V concentration less than 50% of normal.93 Liver failure develops in approximately 15% of all TPN-dependent patients.94 The incidence, prevalence, and severity of liver disease in young children and infants, in particular, are much greater than in adults.95 The incidence and prevalence of liver disease and liver failure specifically in patients with SBS requiring TPN are unknown. Although these disorders often are referred to as TPN-associated liver disease, the pathogenesis probably is related to malabsorption of nutrients such as choline96 and to the route of nutrient assimilation, namely, through the central axis, rather than the portal circulation.97 Patients with the least amount of residual intestine are at greatest risk for developing liver disease.98,99 Diagnosis of liver disease related to intestinal failure in the patient with SBS requires the exclusion of other potential causative disorders. SBS-related liver disease can manifest as cholestasis, steatosis, or steatohepatitis; cholestasis is more common in infants. Studies have suggested benefit from oral lecithin, although it is poorly absorbed, and from intravenous choline (investigational) and, to a lesser extent, ursodeoxycholic acid.100-105 A recent uncontrolled case series of 18 infants described the substitution of a fish oil-based lipid emulsion for the conventional long-chain triglyceride-based emulsion. Compared with historic controls from a different era, reversal of cholestasis occurred sooner (9.4 vs. 44.1 weeks in the fish oil-supplemented group.106 Dextrose overfeeding (greater than 40 kcal/kg/ day) and excessive fat emulsion infusion (2.5 g/kg/day, possibly only 1.0 g/kg/day) should be avoided.93 A minimum of 2% to 4% of total calories, however, should be provided as linoleic fatty acid (50% of most lipid emulsions), to prevent essential fatty acid deficiency. Carnitine supplementation is not useful.107

CALCIUM OXALATE KIDNEY STONES

Fat malabsorption secondary to bile acid deficiency in patients with extensive ileal resection is associated with an increased risk of oxalate kidney stones if the colon is preserved. Oxalate in food usually precipitates as calcium oxalate in the intestinal lumen and is lost in the stool. Lipolysis in patients with SBS and fat malabsorption is normal, and unabsorbed long-chain fatty acids compete with oxalate for available luminal calcium. Consequently, a larger amount of free oxalate is lost to the colon, where it is absorbed and ultimately excreted by the kidney (Fig. 103-8), manifesting as just hyperoxaluria or with calcium oxalate stone formation. Patients with SBS who do not have a colon in continuity are not at increased risk. In one study, symptomatic kidney stones developed within two years of enterectomy in 9 of 38 patients (24%) with SBS and an intact colon.19 Urinary oxalate excretion should be monitored regularly in these patients. Treatment of hyperoxaluria consists of restriction of oxalate-containing food products (see Table 103-6). If hyperoxaluria persists, then oral administration of calcium citrate should be tried; the extra calcium precipitates dietary oxalate, and the citrate prevents stone growth in the urine. A single case report describes the use of conjugated bile acid supplementation to reduce hyperoxaluria.108 Hyperoxaluria also may be

Normal

Steatorrhea +2 LCFA Ca

Ca+2 Oxalate

Insoluble

Oxalate

Insoluble

Excreted Excreted Figure 103-8.  Mechanism of oxalate hyperabsorption in patients with steatorrhea. Normally, oxalate in food is precipitated as calcium oxalate in the intestinal lumen and lost in the stool (left). Lipolysis is normal in patients with short bowel syndrome with fat malabsorption (right), and unabsorbed long-chain fatty acids (LCFA) compete with oxalate for the available calcium. Consequently, a larger amount of free oxalate passes from the small intestine to the colon, where it is absorbed and ultimately excreted by the kidney, often resulting in hyperoxaluria.

related to the metabolism of the vitamin C in TPN solution in the presence of light.109 D -LACTIC d-Lactic

ACIDOSIS

acidosis is a rare complication of SBS and in this setting is observed only in patients with a preserved colon. The episodes of acidosis usually are precipitated by increased oral intake of refined carbohydrates and can be induced in the patient with SBS by carbohydrate overfeeding.110 Malabsorbed carbohydrate is metabolized by colonic bacteria to SCFAs and lactate, which lower the intracolonic pH. A lower pH inhibits the growth of the predominant Bacteroides species and promotes the growth of acidresistant, Gram-positive anaerobes (Bifidobacterium, Lactobacillus, and Eubacterium), which have the capacity to produce d-lactate. d-Lactate is absorbed from the colon and is metabolized to only a limited extent in humans because of our lack of d-lactate dehydrogenase. The main excretory route for d-lactate is the kidney.111 Absorbed d-lactate results in the development of a metabolic acidosis and characteristic neurologic signs and symptoms of nystagmus, ophthalmoplegia, ataxia, confusion, and inappropriate behavior. Patients with d-lactate acidosis often are suspected of being inebriated, but their blood alcohol levels are normal. The constellation of specific neurologic symptoms and metabolic acidosis in a patient with SBS should raise the suspicion of possible d-lactic acidosis. Blood tests will confirm a metabolic acidosis and a normal lactate level; however, the clinical laboratory should be notified to quantify the d-lactic acid rather than the l-lactic acid concentration. The diagnosis is confirmed by measurement of whole-blood d-lactate concentration, which will be elevated significantly (to greater than 3 mmol/L, compared with the normal level of less than 0.5 mmol/L). Treatment consists of correcting the acidosis with sodium bicarbonate and stopping oral intake, which usually results in rapid abatement of the neurologic symptoms. The poten-

Chapter 103  Short Bowel Syndrome tial benefit of antibiotic treatment to change the colonic flora is debated. Substitution of refined carbohydrates for starch has prevented recurrent d-lactic acidosis in a few patients.112 The mediator of the neurologic symptoms still is unknown, and infusion of d-lactic acid in normal subjects to achieve blood levels commonly observed in patients with d-lactic acidosis does not cause any neurologic symptoms. The neurologic symptoms have a striking resemblance to those of Wernicke’s encephalopathy, and in one patient with SBS, recurrent d-lactic acidosis was prevented by thiamine supplementation.113

intestinal transplantation, and it should be performed only in centers with significant experience in this area. To date, no studies have been conducted to compare medical and surgical therapies. A less complex procedure, the serial transverse enteroplasty (STEP), developed by Kim, is a novel technique during which a linear surgical stapler is applied from alternating and opposite directions along the intestine’s mesenteric border to incompletely staple and divide the dilated intestine (Fig. 103-11).119 This procedure leads to the tapering of the intestine in a zig-zag pattern, which results in nutrients being channeled along a narrower but longer intestine. Rather than an intestinal lengthening procedure, this technique is better described as an intestinal tapering procedure. Results reported from an international registry comprising 38 patients from 19 centers have indicated the procedure increases intestinal length by almost 50%, and it has resulted in close to a 100% increase in nutrient absorption.120,121 These improvements in some patients, however, may have been due in part to increased segmental absorption, observed as part of the natural postenterectomy adap­tation process. Nevertheless, the tapering of a dilated, essentially nonfunctional loop of bowel might decrease bacterial overgrowth and improve nutrient absorption. Furthermore, the STEP might have an advantage over the Bianchi procedure in addition to being less technically demanding, in that by avoiding intestinal transection, it may be easier to preserve the blood supply of the intestine.

OTHER COMPLICATIONS

Renal dysfunction,114 metabolic bone disease,115 memory deficits,116 and neurologic abnormalities117 all have been described in patients with SBS who require long-term TPN.

SURGICAL MANAGEMENT INTESTINAL LENGTHENING PROCEDURES

The most important surgical procedure is reanastomosis of the residual small bowel to the residual colon. This procedure carries relatively low mortality and morbidity rates and allows enhanced energy absorption from SCFAs produced from the bacterial fermentation of unabsorbed carbohydrate. A number of other surgical procedures, such as tapering enteroplasty, construction of intestinal valves, creation of recirculating loops, reversal of a short intestinal segment, or colonic interposition, have been attempted to increase intestinal transit time. These procedures are considered experimental, the experience with each is limited, and outcomes generally are not optimal.1 Longitudinal intestinal lengthening and tailoring (Bianchi procedure) (Fig. 103-9) may be useful in patients who have segmental dilation and nonfunctional intestine due to dysmotility and bacterial overgrowth. In this procedure, the surgeon divides the dilated bowel, creates two hemiloops, and anastomoses the hemiloops in an endto-end fashion, thereby doubling the bowel length (Fig. 103-10).118 Although the surface area is not truly increased, bowel function can improve, allowing reduction or elimination of TPN. Nearly all of the approximately 100 operations reported have been undertaken in children. This procedure should be attempted only as a last resort before

INTESTINAL TRANSPLANTATION

Intestinal transplantation is being performed in an increasing number of centers worldwide. The main indication for transplantation in children and adults is TPN-dependent SBS complicated by progressive liver disease. Combined intestine-liver transplantation is the only alternative for patients in whom end-stage liver disease has developed. Isolated intestinal transplantation may be considered for patients with clinically significant liver disease that has not yet progressed to cirrhosis.122 Patients who have significant fluid losses and who experience frequent episodes of severe dehydration despite appropriate medical management also may be candidates for isolated intestinal transplantation. Medicare has approved other indications, including twomajor-vessel thrombosis, a single episode of fungemia, a single episode of bacterial sepsis with shock, and two lifetime episodes of catheter sepsis, although the preponder-

Y

X

Y

X

A

B

C

Figure 103-9.  The Bianchi procedure for intestinal lengthening depicted in a schematic diagram. A, The bowel is split lengthwise. B, Two hemiloops (X and Y) are created. C, The hemiloops (X and Y) are anastomosed end to end.

1791

1792

Section X  Small and Large Intestine

B A

D

C

Figure 103-10.  The Bianchi procedure shown in intraoperative photographs. A, The tips of the forceps are within the dilated loop of intestine, which has been opened; the beginning of each hemiloop is evident (right side). B, A blood vessel can be seen going to the left hemiloop. C, Completed anastomosis. D, The suture shows the gain in the length of intestine. The first hemiloop extends from the tip of the forceps to the first perpendicular suture line. The distance from that point to the end of the thread represents the gain in intestinal length (approximately 26 cm in this infant). (Photographs kindly provided by Kishore R. Iyer, MD, New York, NY.)

Table 103-7  Patient and Graft Survival Rates (%) for Transplants Performed for Short Bowel Syndrome from January 1988 to March 30, 2007, in the United States 1 Year TRANSPLANT TYPE Isolated intestine (n = 510) Intestine/liver (n = 405) Multivisceral (n = 454)

3 Years

5 Years

10 Years

Patient

Graft

Patient

Graft

Patient

Graft

Patient

Graft

83.2 63.6 71.1

78.4 61.3 67.8

65.8 52.4 54.3

55.7 49.8 54.3

55.0 47.4 58.2

43.2 44.4 47.9

43.6 39.2 52.7

24.1 36.0 NA

Based on Organ Procurement and Transplantation Network (http://optn.transplant.hrsa.gov/) data as of June 18, 2008. This work was supported in part by Health Resources and Services Administration contract 231-00-0115. NA, not available.

ance of evidence does not support these as appropriate indications for transplantation. Survival has improved considerably since intestinal transplantation was initiated, with reported survival and nutritional autonomy of up to 18 years.123 As of May 31, 2007 (most recent data from the International Small Bowel Transplant Registry), 1720 transplantation procedures had been performed worldwide in 1608 patients, 909 of whom were still alive. This experience included 746 isolated intestine, 594 intestine-liver, and 380 multivisceral transplants. Patients who have undergone transplantation more recently, generally have had better survival because of improved technique and optimized immunosuppressive regimens. Mean hospitalization was 56 days for isolated intestine, 76 days for intestine-liver, and 62 days for multivisceral transplant recipients. Additional information can be found on the International Intestinal Transplant Registry website,

http://intestinaltransplant.org, which is updated every 2 years (last time in 2007). Current (2008) patient and graft survival data for the United States are presented in Table 103-7. The mortality rate for patients waiting for an intestinalliver transplant is significantly greater than for those waiting for an isolated liver transplant.1 Therefore, early referral to an intestinal transplantation center at the first sign of liver disease is recommended, even if a transplant does not ultimately become necessary. Intestinal and multiorgan transplantations are expensive and generally cost between $250,000 and $3 million per case. Post-transplantation complications, and the most common causes of death afterward, include acute rejection, chronic rejection, cytomegalovirus infection, sepsis (often complicating rejection), and post-transplantation lymphoproliferative disease (PTLD).124 Antirejection medications amount to another $10,000

Chapter 103  Short Bowel Syndrome 12%, and MELD scores were significantly elevated up to 180 days before death, although they were less reliable when obtained 90 days or more before death in a group of 133 patients with intestinal failure; mortality was 50% at 328 days patients with a MELD score between 15 and 25. Increased CRP was also an independent predictor of death within 90 days before death, and a CRP of four or greater universally predicted mortality.126 One unit increase in CRP was associated with a 20% increased risk of immediate death. Further evaluation of such predictors of poor outcome will be necessary, however, before they can be used reliably to support early intestinal transplantation.

PHARMACOLOGIC ENHANCEMENT OF BOWEL ADAPTATION

A

a

b

B

c

Figure 103-11.  The serial transverse enteroplasty (STEP) procedure shown intraoperatively (A) and in line drawings (B). STEP is a technique in which a linear surgical stapler is applied from alternating and opposite directions along the intestine’s mesenteric border to incompletely staple and divide the dilated intestine. The arrows in a and b point to an invagination produced by the staples. The configuration after recovery is shown in c. (Photograph kindly provided by Kishore R. Iyer, MD, New York, NY.)

yearly, in addition to repeated hospitalizations for infection and rejection. For patients who do well, however, nearly all are successfully weaned from TPN, although a few require some maintenance intravenous fluids. This expense compares with a charge of $100,000 to $150,000 per year for home TPN, in addition to the costs of hospitalization for complications. The actual costs of TPN (including pharmacists’ time) however, are closer to $20 to $30 per day. Intestinal transplantation has reached a stage at which it is a feasible, but not yet practical, alternative to conservative treatment of the patient with SBS. One of the greatest dilemmas facing intestinal transplantation is balancing the avoidance of premature transplantation with late referral for transplantation; the latter often requires addition of a liver graft and often results in a lessoptimal outcome.125 High-risk patients likely to develop complications on home TPN need to be identified early, and every attempt must be made to enhance nutrient and fluid absorption and decrease the need for TPN. Both the Model for End-stage Liver Disease (MELD) score and C-reactive protein (CRP) can be used to predict mortality in patients who require long-term TPN. In a study by Putchakayala and colleagues,126 each point of increase in the MELD (to more than 15) was associated with an increased death risk of

The growing knowledge about growth factors has stimulated several clinical studies in patients with SBS. The promising results with the use of growth hormone and dietary lglutamine in a large uncontrolled study of TPN-dependent patients with SBS127 raised hopes that intestinal mucosal growth could be enhanced beyond the adaptive period.6 Two placebo-controlled studies of identical growth hormone and l-glutamine supplementation failed to show any beneficial effect on absorption,46,47 however, and two other studies showed only marginal improvements in fluid and nutrient retention.128,129 GLP-2 is an intestinotrophic enteric hormone that initially was evaluated in a small uncontrolled study of eight patients with SBS, who received native GLP-2 400 mg subcutaneously twice a day for 35 days.130 The treatment resulted in an increase in several absorptive parameters, body weight, and mucosal growth. Use of a synthetic analog of GLP-2 (teduglutide) was associated with increased villus height and increased fluid absorption, with more modest improvements in energy and nitrogen absorption, that regressed once the medication was discontinued.131 A double-blind, randomized, multicenter study indicated administration of the GLP-II analog resulted in a significant decrease in the requirement for parenteral nutrition, although only a few subjects were able to be fully weaned.132 Increased fluid retention was associated with less chronic dehydration,133 a primary factor in the development of nephropathy in patients who require long-term TPN.134 The rapid advance in our knowledge of epithelial growth factors undoubtedly will lead to discovery of still other growth factors that can stimulate intestinal epithelial growth and thus benefit these patients. A double-blind, randomized, controlled trial of growth hormone (0.1 mg/kg/day for 4 weeks) in 41 TPN-dependent patients showed that TPN requirements in treated patients could be reduced by an additional 2 L per week (or one night weekly) over the reduction with standard therapy described earlier in this chapter.135 It is unclear whether these effects were related to improved absorption or appetite stimulation. This study led to the FDA approval of growth hormone injections for treating TPN-dependent SBS. The benefit from this therapy lasted nearly four months following completion of three weeks of daily growth hormone injections; it is unclear whether booster injections will be required. The benefits of this therapy must be weighed against the potential side effects, which include fluid retention, edema, arthralgias, and carpal tunnel syndrome; it also is unknown whether any of the potential growth factor therapies would be more effective if administered during the adaptive phase following enterectomy.

1793

1794

Section X  Small and Large Intestine Jejunoileal anastomosis, no colon present

Jejunoileal anastomosis

End jejunostomy

Re-anastomosis of colon, if possible

≥100 cm residual small intestine

100 cm small intestine + colon

4 stools/day > normal None Streaks of blood Obvious blood Mostly blood Normal Mild friability Moderate friability Exudation, spontaneous bleeding Normal Mild Moderate Severe

*Sutherland index: Range, 0-12. (From Sutherland LR, Martin F, Greer S, et al. 5-Aminosalicylic acid enema in the treatment of distal ulcerative colitis, proctosigmoiditis, and proctitis. Gastroenterology 1987; 92:1894.)

1991

1992

Section X  Small and Large Intestine Table 112-8  Endoscopic and Histologic Assessment of Disease Activity in Ulcerative Colitis

Table 112-9  Induction Therapy for Ulcerative Colitis Based on Disease Severity

SCORE

Mild Disease 5-Aminosalicylates Topical (distal colitis) Oral (distal/extensive colitis) Combination Moderate Disease 5-Aminosalicylates Topical (distal colitis) Oral (distal/extensive colitis) Combination Glucocorticoids Topical (distal colitis) Oral (distal/extensive colitis) Combination Azathioprine or 6-mercaptopurine Severe Disease IV glucocorticoids IV cyclosporine IV infliximab

CRITERIA

Endoscopic Assessment 0 Normal mucosa 1 Loss of vascular pattern 2 Granular, nonfriable mucosa 3 Friability on rubbing 4 Spontaneous bleeding, ulceration Histologic Assessment 0 Normal 1 No significant inflammation: Possibly architectural changes of chronic disease and small foci of lymphocytes but no acute inflammation, crypt abscesses, or epithelial destruction 2 Mild to moderate inflammation: Edema, vascularity, increased acute and chronic inflammatory cells but intact epithelium 3 Severe inflammation: Heavy infiltrate of acute and chronic inflammatory cells, crypt abscesses, ulceration of surface epithelium, purulent exudate

severe disease if the score is greater than 10. Clinical response generally is accepted to be reflected by a decrease of three points from the patient’s initial baseline score. An index very similar to the UCDAI that has been used extensively in recent randomized, controlled trials (RCTs) is the Mayo score, which incorporates the same four components as the UCDAI.126 Other scales also have been developed, many of which are modifications of the Truelove and Witts classification and the UCDAI. None of these disease activity instruments has ever been formally validated. There also exist many endoscopic and histologic scales for grading the severity of colitis (Table 112-8).127,128 Endoscopic findings do not always correlate with clinical symptoms, and such correlations generally are more consistent within individuals. Thus, although therapeutic decisions are based primarily on clinical status, it may be useful to follow the sigmoidoscopic mucosal appearance over time in an individual patient, if the clinical response to treatment is uncertain. In addition to the typical categorization of disease activity into mild, moderate, and severe, an important subgroup is fulminant colitis. Patients with severe colitis who appear toxic, with fever higher than 38.3°C (101°F), tachycardia, abdominal distention, signs of localized or generalized peritonitis, and leukocytosis, are considered to have fulminant colitis. Toxic megacolon is said to occur when there is radiologic evidence of transverse colon dilatation to greater than 6 cm in an acutely ill patient. Fulminant colitis and toxic megacolon are clinical diagnoses, and complete co­ lonoscopic examination should be avoided in patients with severe or fulminant colitis because of the risk of inducing megacolon or perforation. In this patient population, a limited flexible sigmoidoscopy is appropriate to ensure that the etiology of the symptoms is UC itself and not other conditions.

TREATMENT MEDICAL

The goals of therapy of UC are to induce remission, to maintain remission, to maintain adequate nutrition, to minimize disease- and treatment-related complications, and

IV, intravenous.

Table 112-10  Maintenance Therapy for Ulcerative Colitis 5-Aminosalicylates Topical (distal colitis) Oral (distal/extensive colitis) Azathioprine or 6-mercaptopurine Infliximab

Disease distribution Distal Extensive Prior therapy

Severity of disease Mild Moderate Severe

Medication choice

Response Side effects Compliance

Figure 112-12.  Factors that should be considered in the choice of medical therapies for ulcerative colitis.

to improve the patient’s quality of life. Current management strategy focuses on using appropriate medical therapy and optimizing timing of surgery. Several factors should be considered in determining optimal therapy for patients with UC (Fig. 112-12). Current therapeutic strategies can be classified broadly, based on disease activity, into those that treat active disease (induction therapy) (Table 112-9) and those that prevent recurrence of disease once remission is achieved (maintenance therapy) (Table 112-10). This concept of induction and maintenance of remission forms the basis of our evaluation of the efficacy of a specific therapy. The extent of disease is an important consideration that helps determine the route of administration of medication. Thus, for example, proctitis may be treated with suppositories or foam preparations as well as with oral therapy, and enema preparations may be used alone or in combination with systemic therapy for patients with

Chapter 112  Ulcerative Colitis left-sided disease. Other important factors to consider are a patient’s prior response to or side effects from a specific medication and compliance with medication. These factors might favor or preclude the use of a specific agent. Given the chronic nature of UC, medications need to be efficacious and well accepted by patients from the standpoints of safety and ease of administration. The mainstay of medical therapy focuses on regimens that alter host response to decrease mucosal inflammation. Therapies that target other aspects of the systemic inflammatory process or manipulate the enteric flora also have been developed to treat UC. One important consideration when evaluating the efficacy of a particular medication (e.g., in a RCT that compares a novel therapy to placebo) is the placebo response rate. Even though placebos often are thought of as inert agents, they have been noted to lead to improvement in a variety of both subjective and objective outcome measures in a number of different medical conditions, such as anxiety, depression, insomnia, pain, asthma, obesity, hypertension, and even myocardial infarction.129 In some of these disorders, placebo response rates of up to 40% have been reported. With respect to placebo response rates in UC, a meta-analysis of 40 randomized, placebo-controlled trials, in which the most commonly used activity indices were the Mayo Score or the UCDAI, reported a pooled placebo response rate of 28% and remission rate of 13%.130 Studies using less-stringent outcome definitions were noted to have higher placebo response and remission rates. Univariate analysis suggested that longer follow-up duration, higher number of follow-up visits, longer disease duration, and lower disease severity at study entry were positively associated with the placebo remission rate.

Aminosalicylates

Oral Sulfasalazine consists of an antibacterial component, sulfapyridine, bonded by an azo bond to a salicylate, 5-aminosalicylic acid (5-ASA, mesalamine) (Fig. 112-13).131 The drug was synthesized by Nana Svartz in 1938-1939 and its benefit for the treatment of IBD was discovered serendipitously in 1941-1942 by her when patients with UC receiving this medication for a presumed diagnosis of rheumatoid arthritis noted improvement in colitis symptoms132; in retrospect, these patients had peripheral arthropathy associated with their IBD. Research subsequently established that 5-ASA is the principal therapeutic moiety of sulfasalazine in IBD and that the sulfapyridine component of the parent drug serves as an inactive carrier, largely preventing absorption of 5-ASA in the small intestine and allowing it to be released in the colon.133,134 Approximately 90% of sulfasalazine reaches the colon, and only a small amount is absorbed in the small intestine. On reaching the colon, the enzyme azoreductase, which is elaborated by colonic bacteria, cleaves the azo bond to release the active constituent moiety, 5-ASA. After 5-ASA is absorbed from the colon, 20% of the compound undergoes hepatic acetylation, forming N-acetyl 5-ASA, and is excreted in the urine. Sulfasalazine is one of several agents in the class of 5-ASA compounds that is considered to be the first line of therapy for inducing remission in patients with mild to moderate UC.131,135 Mesalamine derivatives have not been evaluated in a randomized, controlled fashion in patients with severely active disease. At a dose of 3 to 6 g/day, sulfasalazine induces remission in 39% to 62% of patients with mild to moderate UC, about twice the remission rate of placebotreated patients.136,137 Various formulations and controlled-release systems (Table 112-11; see Fig. 112-13) have been developed to

Sulfasalazine HOOC

O N

HO

S

N

NH

O 5-Aminosalicylate

N

Sulfapyridine

Olsalazine 5-ASA

N

N

5-ASA

Mesalamine (Pentasa brand) HOOC NH2

HO

Ethylcellulose

Mesalamine (Asacol brand) HOOC Eudragit-S NH2

HO

Balsalazide HOOC HO

O N

N

Figure 112-13.  Molecular preparations.

structures

CH2

COOH

5-aminosalicylate

(5-ASA)

NH

C of

CH2

deliver 5-ASA to specific sites of the gastrointestinal tract without the sulfapyridine moiety, which is thought to be responsible for most of the side effects. Olsalazine (Dipentum) is a 5-ASA dimer linked by an azo bond and is formulated in gelatin capsules. Balsalazide (Colazal) consists of a 5-ASA monomer linked to a biologically inactive carrier molecule, 4-aminobenzoyl-β-alanine. Similar to sulfasalazine, 5-ASA is released from olsalazine and balsalazide in the colon upon cleavage of the azo bond via the bacterial enzyme azoreductase. Approximately 99% of the drug is delivered intact to the colon, and its metabolites are cleared rapidly in the urine. Three commonly used mesalamine preparations allow delivery of 5-ASA before the drug reaches the colon: Pentasa, Asacol, and Lialda. Pentasa uses ethyl cellulose-coated microgranules that release mesalamine from the duodenum throughout the small bowel and the colon; about 50% of 5-ASA is released in the small intestine, and the remainder is released in the colon. Asacol is a Eudragit-S-100–coated mesalamine tablet that is released at a pH greater than 7, usually in the distal ileum and the colon. With Asacol, about 15% to 30% of mesalamine is released in the small intestine. Lialda (MMx mesalamine) is a novel mesalamine formulation that uses a multimatrix structure composed of an inner lipophilic matrix and an outer hydrophilic matrix. It is coated with a pH-dependent polymethacrylate film to

1993

1994

Section X  Small and Large Intestine Table 112-11  Oral 5-Aminosalicylic Acid Preparations and Sites of Delivery in the Gastrointestinal Tract

Table 112-12  Side Effects of Sulfasalazine and 5-Aminosalicylates

DRUG

Dose-Related Alopecia Anorexia Back pain Folate malabsorption (with sulfasalazine) Headache Nausea, vomiting, dyspepsia Non–Dose-Related Agranulocytosis, aplastic anemia Arthralgia Colitis Fever Fibrosing alveolitis, pulmonary eosinophilia Hemolytic anemia (Heinz bodies) Hepatitis Hypersensitivity skin rashes (occasionally with photosensitivity) Male infertility (with sulfasalazine) Pancreatitis Pericarditis, myocarditis

Prodrugs Sulfasalazine Olsalazine Balsalazide

FORMULATION

Sulfapyridine + 5-ASA 5-ASA dimer 4-aminobenzoyl β-alanine + 5-ASA Mesalamine Preparations Asacol, Claversal, pH sensitive, Salofalk resin-coated; delayed release Rowasa Enema Canasa Suppository Pentasa Ethylcellulose-coated microgranules; controlled release Lialda pH sensitive, multi-matrix and polymethacrylate coated; delayed and slow release

SITE OF DELIVERY Colon Colon Colon Distal ileum, colon Distal colon Rectum Duodenum to colon Distal ileum, colon

5-ASA, 5-aminosalicylate.

allow the delayed release of mesalamine in the terminal ileum and colon at a pH greater than 7. This technology also allows mesalamine to be released slowly and in close proximity to the colonic mucosa. These oral 5-ASA derivatives (mesalamines) have been shown to be superior to placebo for mildly to moderately active UC.137-142 Meta-analyses have demonstrated that the mesalamines are as efficacious as sulfasalazine, and the various mesalamine preparations appear to be comparable in efficacy.137,140 Balsalazide has been shown to have superior efficacy and a more rapid response compared with traditional mesalamine agents.143,144 In a RCT, balsalazide 6.75 g/day, a dose equivalent to mesalamine 2.4 g daily, achieved higher rates of remission and had better tolerance compared with pH-dependent mesalamine 2.4 g/day.143 It has been suggested that the greatest benefit of balsalazide is in patients with newly diagnosed left-sided UC.144 More important than the specific 5-ASA preparation is the dose-dependent response when 5-ASA is used as an induction therapy for active UC.137,140 For this indication, mesalamine is not effective at doses lower than 2 g daily, and there is an increased response at doses of 4 to 4.8 g daily. The ASCEND I and II trials showed that mesalamine at doses of 2.4 and 4.8 g/day have similar efficacy for patients with mildly active disease, but the higher dose (4.8 g/day) was more efficacious in patients with moderately active disease.145,146 This dose of mesalamine is comparable to 12 g/day of sulfasalazine, which is impractical in clinical practice because of the high probability of intolerance. No RCT has evaluated the use of aminosalicylates for severely active UC, but these agents are generally thought not to be effective in severely active disease. Once remission is achieved, sulfasalazine and other 5-aminosalicylates are effective in maintaining it.147-150 This benefit appears to be dose dependent for sulfasalazine, with a dose of 2 g/day often used to balance efficacy and adverse side effects.147 Such a dose-dependent response, however, has not been found with the other 5-ASA preparations, and at doses of 1.5 to 4.8 g/day, remission can be maintained in more than 50% of patients.151 One meta-analysis has suggested that sulfasalazine might have a slight but statistically

significant therapeutic superiority relative to the newer 5-ASAs in maintaining remission when considering trials of six months’ duration; however, when these trials were combined with those of 12 months’ duration, this statistically significant benefit was lost.151 A double-blind RCT comparing two doses of balsalazide (1.5 g twice daily and 3 g twice daily) with mesalamine 0.5 g three times daily for six months reported a remission rate of 77.5% with the higher dose of balsalazide compared with remission rates of 56.8% and 43.8% with mesalamine and the lower dose of balsalazide, respectively.152 In general, the same dose of 5-ASA derivative that induces remission is recommended for maintenance therapy, although this recommendation has not been formally tested in a randomized, placebocontrolled fashion. Common side effects of sulfasalazine include fever, rash, nausea, vomiting, and headache (Table 112-12). Other, lesscommon but important side effects of sulfasalazine include hypersensitivity reactions, reversible sperm abnormalities, and impairment of folate absorption. Approximately 15% of patients taking sulfasalazine develop significant side effects that require discontinuing the medication. Up to 90% of patients who are intolerant to sulfasalazine, however, can tolerate mesalamine.153 In clinical trials, the newer 5-ASA preparations and balsalazide have been shown to be better tolerated than sulfasalazine,140,154,155 although the adverse event profiles during maintenance therapy appear to be similar for 5-ASA preparations and sulfasalazine.151 Sulfasalazine can impair folate absorption (by competitively inhibiting the jejunal enzyme, folate conjugase) thereby contributing to anemia, and folate supplementation should be prescribed to patients receiving sulfasalazine. Olsalazine is associated with drug-induced diarrhea in up to 10% of patients, which often limits its use. It has been noted that if olsalazine is ingested with meals and is continued despite the diarrhea, the incidence of this side effect can be lessened substantially to 3%. A systematic review of oral 5-ASA for maintenance of remission in UC found olsalazine to be significantly inferior to sulfasalazine, and this reduced efficacy was related mostly to a significantly higher rate of withdrawals because of adverse events.151 Oral mesalamine preparations do not appear to have significant dosedependent toxicity.

Chapter 112  Ulcerative Colitis Topical Topical aminosalicylates can be administered in the form of 5-ASA enemas, 5-ASA suppositories, and, in Europe, 5-ASA foam. The use of enemas allows the medication to be delivered up to the level of the splenic flexure in about 95% of patients, and suppositories can be used to treat disease up to 15 to 20 cm from the anal verge. Topical mesalamine derivatives may be used as an alternative monotherapy or as an adjunctive therapy to oral agents in patients with left-sided colitis or pancolitis. They are effective for inducing remission in patients with mildly to moderately active distal UC, without a clear dose-response effect156,157 in nonrefractory patients. The standard dosing regimens used to induce remission are 1 to 4 g of 5-ASA in the form of an enema nightly, or mesalamine suppositories 1 to 1.5 g either nightly or in divided doses throughout the day. Mesalamine enemas have been shown to be comparable to oral sulfasalazine in the treatment of active distal UC, with fewer side effects.155 Similar efficacies have been demonstrated for mesalamine enemas regardless of whether the 1-, 2-, or 4-g formulation is used for inducing remission in patients with mild to moderate left-sided UC not requiring concurrent glucocorticoids or immunomodulators. In fact, mesalamine enemas are perceived to be even more effective than topical glucocorticoid enemas in this setting.157,159 A combination of topical and oral mesalamine also may be more effective than either agent alone in patients with left-sided colitis or pancolitis, suggesting a dose-response effect.160,161 In patients with proctitis, mesalamine suppositories, 500 mg administered twice daily, have been shown to be beneficial for treating active disease.162 Mesalamine foam has a more uniform distribution and longer persistence in the distal colon compared with mesalamine enemas. The foam preparation has been shown to have better patient acceptance than the enema preparation,163 but mesalamine foams currently are not available in the United States. Topical mesalamine preparations also are effective for maintaining remission in left-sided UC or proctitis.156,157 The effective maintenance dosing interval ranges from nightly to every three days. Topical mesalamine is as effective as oral mesalamine,164 and the combination of topical and oral mesalamine may be more effective than oral mesalamine alone as a maintenance regimen.165

Glucocorticoids

Systemic At doses equivalent to 40 to 60 mg/day of oral prednisone, glucocorticoids are effective first-line therapy for moderate or severe flares of UC.124,166-170 The use of doses higher than 60 mg/day is associated with increased side effects without appreciable clinical benefit and thus should be avoided. The addition of sulfasalazine to corticosteroids in moderately to severely active UC does not offer any incremental benefit. Although no study has directly compared the efficacy of oral and parenteral glucocorticoids, the latter commonly are used in severe disease. No adequately designed controlled study has been performed to confirm the clinical impression that continuous infusion of parenteral glucocorticoids is superior to pulse therapy. The use of adrenocorticotropin (ACTH) has been suggested as an alternative to conventional glucocorticoid therapy of active UC in small studies.171 One double-blind RCT suggested that intravenous ACTH was more effective than intravenous hydrocortisone for the treatment of severely active UC only in steroid-naïve patients172; this observation has not been confirmed. Because most patients with severely active flares have been treated previously with glucocorticoids, ACTH rarely is used in clinical practice. A

noteworthy complication of ACTH therapy is bilateral adrenal hemorrhage. Glucocorticoids have no maintenance benefits in patients with UC. Steroid-dependent patients, or patients who are unable to taper off glucocorticoids without experiencing disease exacerbation, benefit from the addition of steroidsparing agents. There has been no trial to date assessing mesalamine therapy and its efficacy in maintaining remission induced with glucocorticoids. The long-term remission rate in patients who require parenteral glucocorticoids for severe UC is approximately 50%.173 Immunomodulatory agents, as discussed, should be considered in patients who are dependent on steroids, who require two courses of glucocorticoids for induction of clinical response or remission within one year, or who require parenteral glucocorticoids to induce remission. In addition to the use of immunomodulatory agents, one should consider using infliximab for steroid-dependent patients. Glucocorticoids are associated with many mild and serious side effects in patients with IBD (Table 112-13). These side effects occur commonly and involve nearly every organ system. Every effort should be made to minimize glucocorticoid use and exposure. Budesonide is a glucocorticoid preparation that is structurally different from prednisone. The presence of 16α,17αacetyl side chains allows enhanced topical anti-inflammatory activity and affinity for glucocorticoid receptors compared with prednisone.174 In addition, budesonide has an approximately 90% first-pass metabolism in the liver and erythrocytes and is converted to metabolites that have little or no biological activity. The resultant low systemic bioavailability translates to significantly less toxicity compared with

Table 112-13  Side Effects of Glucocorticoids Cutaneous Acne Impaired wound healing Purpura, ecchymoses, petechiae Striae Endocrine Adrenal insufficiency Cushingoid appearance Gastrointestinal Dyspepsia Dysphagia/odynophagia (candidiasis) Infectious complications Numerous pathogens Metabolic Electrolyte imbalance, hypokalemia Fluid retention Growth retardation Hyperglycemia, secondary diabetes mellitus Hyperlipidemia, altered fat distribution Hypertension Musculoskeletal Myopathy Osteonecrosis Osteoporosis Neuropsychiatric Anxiety, mood swings Depression Insomnia Psychosis Ocular Cataracts Glaucoma

1995

1996

Section X  Small and Large Intestine traditional glucocorticoids. Entocort is a controlled-ilealrelease oral budesonide preparation consisting of EudragitL-100–coated microgranules with an internal ethyl cellulose component; it releases budesonide at pH greater than 5.5, and about 50% to 80% of budesonide is absorbed in the ileocecal region. There currently is no oral formulation of budesonide that provides optimal release characteristics for the entire length of the colon. A small uncontrolled study has suggested that Budenofalk, which is not available in the United States, may be effective for prednisone-dependent UC.175 Controlled studies have not shown the benefit of oral budesonide for the treatment of active UC.176 Topical Topical glucocorticoids in liquid and foam formulations are effective short-term therapy for active UC distal to the splenic flexure.177,178 Foam preparations often are tolerated better by patients and may be easier to retain than liquid preparations. Topical glucocorticoids have been found to be less effective than topical mesalamine for inducing remission of distal UC159; however, the combination of topical corticosteroids and topical mesalamine has been more efficacious than either alone in the short-term treatment of distal UC.179 Whereas systemic absorption of glucocorticoids with topical therapy is significantly less than that with oral administration, prolonged treatment with topical glucocorticoids still may be associated with steroid-related side effects and should be avoided. As mentioned previously, budesonide is a potent corticosteroid with a rapid first-pass metabolism. Budesonide enemas, which currently are neither available nor approved in the United States, have been shown to be effective for the treatment of active distal UC in several controlled trials. In a double-blind RCT of patients with active distal UC, budesonide, 2 mg/100 mL for six weeks, resulted in a remission rate of 19% compared with 4% in patients receiving placebo therapy (P < 0.05).180 Subsequent trials have shown budesonide enema to be as efficacious as or even superior to prednisolone enema without resultant depression of endogenous cortisol levels.181-183 Budesonide enema perhaps is inferior in efficacy to me­ salamine enema,184 but it clearly presents an alternative topical glucocorticoid for treatment of distal UC. The optimal dose for budesonide enema consistently has been shown to be 2 mg/100 mL once daily.180,181,185 Budesonide in foam preparation also has been shown to have comparable efficacy with traditional hydrocortisone foam for the treatment of active proctosigmoiditis.186 Additional studies are needed to determine the effect of longer-term topical budesonide use. As with other glucocorticoid preparations, budesonide enema is not effective for maintaining remission in UC.185

Immunomodulators

Azathioprine and 6-Mercaptopurine Of the various immunomodulatory agents, the most widely used are azathioprine and 6-MP. These two agents are purine analogs that interfere with nucleic acid metabolism and cell growth and exert cytotoxic effects on lymphoid cells. They are inactive prodrugs with subtle structural differences. Azathioprine is nonenzymatically converted to 6-MP, which is then metabolized through a series of enzymatic pathways to active and inactive metabolites (see Fig. 111-8). The two primary metabolites of 6-MP are 6thioguanine nucleotides (6-TGNs) and 6-methylmercaptopurine (6-MMP). The 6-TGN metabolites are thought be responsible for the immunomodulatory action of azathio-

prine and 6-MP and their bone marrow suppression property, whereas hepatotoxicity is thought to be related to 6-MMP. One key enzyme involved in the biotransformation of 6-MP is thiopurine methyl­transferase (TPMT), which converts 6-MP to its inactive metabolites, 6-MMP and 6-methylmercaptopurine ribonucleotides. There is a population polymorphism in the TPMT gene: 89% of the population have homozygous wild-type TPMT, and 11% and 0.3% of the population have heterozygous and homozygous mutations, respectively.187 Persons with heterozygous and homozygous TPMT mutations have decreased to absent enzyme activity. The clinical significance of this genetic polymorphism is that inherited differences in TPMT may be responsible for most of the variability in drug response observed among individual patients. The efficacy of azathioprine in the treatment of UC is a matter of debate. Four RCTs have evaluated azathioprine for inducting remission in active UC (Table 112-14).188-191 These four studies were small, heterogeneous in design, used different outcome definitions for response, and reached different conclusions. Two of the studies involved steroiddependent patients,190,191 one other study used steroids for induction,188 and two studies used 5-ASAs as a comparator group rather than placebo.189,191 Only one study showed a significant benefit with azathioprine compared with 5-ASA for induction therapy in steroid-dependent disease.191 With respect to the use of azathioprine for maintenance of remission in UC, four RCTs have been performed (see Table 112-14).188,192-194 Just as with studies of induction therapy, these four studies also had small sample sizes, used heterogeneous designs with different outcome definitions of response, allowed for various cotherapies, and again reached different conclusions. One of the studies was in steroiddependent disease,192 another allowed the use of steroids for relapse,188 one study used 5-ASA as a comparator group rather than placebo,194 and another included patients who were mostly taking 5-ASAs and was actually a study of azathioprine withdrawal.193 Only this withdrawal study showed a benefit with continued azathioprine. Thus, for the purpose of induction or maintenance therapy for UC, our use of azathioprine is largely based on its established efficacy in Crohn’s disease rather than any proven benefit in UC. One subset of patients, however, has been shown to obtain benefit with the use of azathioprine, specifically patients who have severely active UC and who are able to attain induction of remission with intravenous followed by oral cyclosporine. In these patients, maintenance therapy with azathioprine has been reported to decrease colectomy rates (see later). The optimal dose of azathioprine or 6-MP for treating UC is unclear, and no formal dose-ranging study has been reported in the literature. The effective doses for 6-MP and azathioprine generally are 1 to 1.5 mg/kg/day and 2 to 3 mg/ kg/day, respectively.195 At these doses, however, there still may be nonresponders and, for them, higher doses may be necessary. Induction of leukopenia had been advocated for dose optimization,196 but this practice was not supported by subsequent studies.197-199 Monitoring metabolite levels may be beneficial in determining the optimal dose of azathioprine or 6-MP. To date, at least 13 studies examining response in IBD with respect to 6-TGN level have been published. A metaanalysis of the first 12 of these studies found that the studies were similar in that they were retrospective and the majority of patients were adults with Crohn’s disease, but they were heterogeneous with respect to sample size, the proportion of patients in remission, and the activity indices used to

Chapter 112  Ulcerative Colitis Table 112-14  Randomized, Controlled Trials of Azathioprine for Ulcerative Colitis REFERENCE Induction 188 189 190 191 Maintenance 188 192 193 194

N

DOSE (MG/KG/D)

DURATION OF THERAPY (months)

RESPONSE (AZA)

RESPONSE (control)

P-VALUE

1 3 6 6

78% 60% NR 53%

68% 80% NR 19%

NS NS NS .006

Glucocorticoids in all None; control = 5-ASA None None; control = 5-ASA Glucocorticoids for relapse None 5-ASA in most AZA withdrawal Glucocorticoid induction Control = 5-ASA

80 20 44* 72*

2.5 2.5 2-2.5 2

80 30* 67

1.5-2.5 1.5 NR

11 6 12

40% NR 64%

23% NR 41%

NS NS .039

25

2.5

18

42%

62%

NS

CO-THERAPY

*All patients in this study were glucocorticoid-dependent. 5-ASA, 5-aminosalicylates; AZA, azathioprine; mo, months; N, number of patients; NR, not reported; NS, not significant.

assess response.200 Of the seven studies that reported data on 6-TGN threshold levels, a pooled analysis of the first six studies showed a three-fold significantly higher rate of remission among patients with a 6-TGN level of greater than 230 to 260 pmol/8 × 108 red blood cells. Incorporation of 6-TGN metabolite measurement into the management regimen of patients receiving azathioprine or 6-MP therapy for IBD is not mandatory and it is a subject of continuing controversy. Currently, 6-TGN measurement appears to be most useful for identifying reasons for nonresponse to therapy and for suspected noncompliance. If used, metabolite levels should be determined at least two weeks following any dose adjustment to allow sufficient time for the metabolites to reach steady-state. Currently, it is recommended in the package insert and by the U.S. Food and Drug Administration (FDA) to determine TPMT genotype or phenotype before initiating therapy. The active metabolites, 6-TGNs, also are responsible for myelosuppression with therapy, and patients with TPMT mutation or decreased TPMT enzyme activity are more likely to experience this toxicity because of preferential shunting of 6-MP metabolism toward the excessive production of 6-TGN.201 Thus, identifying TPMT polymorphism before initiating azathioprine or 6-MP therapy can decrease the risk of myelotoxicity. Patients with homozygous wildtype TPMT or normal (to high) TPMT enzyme activity level may receive these agents starting at the weight-based optimal dose of 2.5 mg/kg/day for azathioprine or 1.5 mg/kg/day for 6-MP. It has been suggested by some investigators that in patients with heterozygous TPMT mutation or intermediate enzyme activity level, 6-MP or azathioprine should be started at 50% of the weight-based optimal dose. Alternative therapy should be considered in patients with homozygous mutations for TPMT. Regardless of whether a patient’s TMPT genotype or phenotype is known, continued frequent monitoring of complete blood counts remains necessary, because only 27% of all patients with leukopenia have TPMT mutations.202 In addition, two studies have reported that TPMT testing may be cost effective.203,204 Azathioprine and 6-MP therapy have a delayed onset of action. The mean time to clinical response with azathioprine or 6-MP therapy in patients with UC has been reported to be three to four months in uncontrolled studies,205,206 a

Table 112-15  Side Effects of Azathioprine and 6-Mercaptopurine Abnormal liver biochemical test results Bone marrow suppression Hypersensitivity reactions (fever, rash, arthralgia) Infections Lymphoma Nausea, abdominal pain, diarrhea Pancreatitis

figure that is similar to the 17 weeks’ response time to clinical benefit in placebo-controlled trials of azathioprine or 6-MP therapy for active Crohn’s disease.207 Intravenous loading of azathioprine at 40 mg/kg for 36 hours does not shorten the time required for a therapeutic response in patients with Crohn’s disease.202 Such practice presumably would have the same results if attempted in patients with UC. Because azathioprine or 6-MP therapy is associated with a number of potentially significant toxicities, its duration of therapy should be determined by weighing clinical benefit against these potential toxicities. The optimal duration of maintenance therapy with azathioprine or 6-MP currently is unknown in patients with UC. In patients with Crohn’s disease, the maintenance benefit of azathioprine or 6-MP can be observed for at least five years.208,209 Based on these data in Crohn’s disease and the paucity of alternative maintenance therapies, in patients with UC in whom remission is maintained with azathioprine or 6-MP, treatment generally is continued indefinitely as long as there is no significant adverse side effect. Common side effects of azathioprine and 6-MP therapy include nausea, vomiting, bone marrow suppression, pancreatitis, allergic reactions, and infections (Table 11215).210,211 Bone marrow suppression occurs in 2% to 5% of patients.210,212 It is dose dependent and manifests primarily as leukopenia, although all three cell lines may be affected. This hematologic toxicity can increase with concurrent use of sulfasalazine or mesalamine compounds.198,213-215 It is

1997

1998

Section X  Small and Large Intestine known that mesalamine can interact with the enzyme TPMT, leading to increased levels of 6-TGN, and that this interaction has been associated with leukopenia. Bone marrow suppression is managed by reducing the dosage of immunomodulator or withdrawing the medication. Routine monitoring of complete blood count with differentials is necessary for patients receiving azathioprine or 6-MP and should be continued for the entire duration of therapy. Allergic reactions to azathioprine or 6-MP usually manifest as fever, rash, and arthralgia and resolve following discontinuation of these medications.212,216 Recurrence of similar reactions occurs with medication challenge, although patients who develop allergic reactions to one agent may be able to tolerate subsequent challenge with the other.217 Pancreatitis also is idiosyncratic and independent of dosage.212,218,219 It usually occurs during the first month of therapy and is reversible upon withdrawal of the drug. Patients using azathioprine or 6-MP therapy can have abnormal liver biochemical tests, but these usually resolve following drug withdrawal.220 Because liver biopsy is not performed routinely in these patients, their pattern of hepatic injury, if any, is unknown. Cholestasis with inflammation, nodular regenerative hyperplasia, and peliosis hepatis have been reported with azathioprine and 6-MP therapy.212,220 As is the case for complete blood counts, routine monitoring of liver biochemical tests is recommended. An increased risk of malignancy, primarily lymphoma, has been reported, but not consistently. A metaanalysis of six studies examining this risk reported a fourfold elevated risk of lymphoma with 6-MP/azathioprine.221 The lymphoma that develops in patients who have IBD and receive these immunomodulatory agents appears to be associated with Epstein-Barr virus.222 Cyclosporine Cyclosporine A is a potent inhibitor of cell-mediated immunity. Its use in UC is primarily in patients with severe, steroid-refractory disease. There has only been one randomized, placebo-controlled trial evaluating the efficacy of intravenous cyclosporine in severe UC. In this study of 20 patients who did not respond to at least seven days of intravenous hydrocortisone, nine (82%) of the 11 patients receiving continuous intravenous infusion of cyclosporine at 4 mg/kg/day responded, compared with none of the nine patients receiving placebo therapy.223 The time to clinical response was rapid, at a mean of seven days. After the intravenous route of therapy was converted to oral cyclosporine, 44% of those patients who responded initially required colectomy during the sixmonth follow-up period. Intravenous cyclosporine monotherapy may be as effective as intravenous glucocorticoids in patients with severely active UC; its use thus potentially minimizes the toxicities of combination therapy.224 The addition of azathioprine or 6-MP in patients who have responded to intravenous cyclosporine has been shown in other studies to reduce the rate of relapse or colectomy.225,226 Thus, cyclosporine can be considered a bridge therapy to control active disease in patients with steroid-refractory UC while waiting for elective surgery or the onset of action of azathioprine or 6-MP. With the addition of azathioprine, long-term remission at one year may be more likely in patients who initially respond to intravenous cyclosporine monotherapy than in those who respond to intravenous glucocorticoids. A European retrospective cohort study of 142 patients who were treated with cyclosporine, 118 of whom responded initially, reported the probability of avoiding colectomy to be 63% at one year, 41% at four years, and 12% at seven years; overall,

54% of patients required colectomy at some point.227 Patients who were already taking 6-MP or azathioprine at the time cyclosporine was initiated continued taking their current dose, and those who were naïve to 6-MP or azathioprine were started at target doses at the time of response to cyclosporine during their hospitalization. The authors found that 59% of patients previously taking 6-MP or azathioprine required eventual colectomy, compared with 31% for patients naïve to these drugs (P < 0.05). Because most of the serious adverse effects associated with the use of cyclosporine are dose-dependent, intravenous doses lower than 4 mg/kg that still can achieve efficacy are desirable. One RCT has shown that a dose of 2 mg/kg is as effective as 4 mg/kg given intravenously in patients with severely active UC, judged by clinical response rates, time to response, and short-term colectomy rates.228 The mean plasma cyclosporine levels were 237 ng/mL in patients receiving the 2 mg/kg dose and 332 ng/mL in patients receiving the 4 mg/kg dose. Thus, initiating therapy at 2 mg/kg may be reasonable, but regardless of the dose used, careful monitoring of plasma cyclosporine trough levels is necessary. Cyclosporine has been associated with many adverse effects, including paresthesias, tremors, headache, hyper­ trichosis, and gingival hyperplasia (Table 112-16). Other potentially serious toxicities include hypertension, seizures, electrolyte and liver biochemistry abnormalities, nephrotoxicity, anaphylaxis, and opportunistic infections. These complications are mostly dose-dependent. Severe complications have been reported with cyclosporine in up to 12% of patients with UC,229 and two large series have reported death rates of 1.8% to 2.8% with cyclosporine, more than half of which were due to infections acquired while taking the drug.227,229 Careful monitoring for adverse effects is critical during cyclosporine therapy. Baseline serum electrolytes, creatinine, cholesterol, and liver biochemical values should be measured. Cyclosporine therapy should be avoided in patients with an impaired creatinine clearance to minimize the risk of severe nephrotoxicity. Patients with serum cholesterol lower than 120 mg/dL should receive nutritional support to improve the level before initiating cyclosporine therapy, because a low cholesterol level is associated with an increased risk of seizures. During intravenous therapy, cyclosporine levels should be monitored daily, and the dose should be adjusted to achieve a trough concentration (measured one hour before dosing) between 200 and 400 ng/mL, determined by high-pressure liquid chromatography. Serum electrolytes and serum creatinine levels should be monitored daily or every other day. The dose of cyclosporine also

Table 112-16  Side Effects of Cyclosporine Anaphylaxis Diarrhea Electrolyte abnormalities Gingival hyperplasia Headache Hepatotoxicity Hirsutism Hypertension Infections Nausea, vomiting Opportunistic infections Paresthesia Renal insufficiency Seizure Tremor

Chapter 112  Ulcerative Colitis should be decreased when the serum creatinine increases by 20% to 30% over baseline. If patients respond to intravenous cyclosporine, the route of administration can be changed to oral therapy with 2 mg of oral agent for each 1 mg of intravenous cyclosporine. The drug can be administered in two divided doses daily. Drug monitoring during oral cyclosporine therapy includes weekly trough cyclosporine levels and weekly to biweekly electrolyte and creatinine levels. Oral cyclosporine should be continued for three to six months, while waiting for surgery or for azathioprine or 6-MP to take effect. Patients on long-term cyclosporine therapy should receive Pneumo­ cystis carinii pneumonia prophylaxis with trimethoprimsulfamethoxazole. Methotrexate Methotrexate is a folic acid antagonist and has antime­ tabolite and anti-inflammatory properties. Although early reports suggested potential benefit of methotrexate administered intramuscularly or orally in UC, the only randomized, placebo-controlled trial failed to demonstrate its efficacy for the treatment of active UC.230 In this study of 67 patients with chronic active UC, oral methotrexate at 12.5 mg/wk for nine months was comparable to placebo therapy in the rate of achieving first remission, time to first remission, relapse following remission, and the mean glucocorticoid dose. It is unknown if methotrexate at higher doses administered intramuscularly or subcutaneously may be beneficial in inducing or maintaining remission in UC. Given the absence of data supporting its efficacy, metho­ trexate cannot at this time be considered a standard therapy for UC. Other Immunomodulators Alternative immunomodulators have been explored for patients who do not tolerate or have not responded to the previously mentioned immunosuppressants. Mycophenolate mofetil has pharmacodynamic properties similar to those of azathioprine and 6-MP but a more rapid onset of action. A pilot study of patients with chronic active UC receiving concomitant prednisolone found azathioprine to be superior to mycophenolate mofetil throughout the oneyear study period, with remission rates at one year of 100% and 88%, respectively.231 Uncontrolled studies reported less than 50% remission rates with mycophenolate mofetil therapy in patients with steroid-dependent UC232,233 and the intolerance rate was high.232 A substantial number of patients developed adverse effects necessitating drug withdrawal, including recurrent upper respiratory tract infection, bacterial meningitis, depression, and migraine headache.231,233 Tacrolimus is another immunosuppressant with actions similar to those of cyclosporine. In contrast to cyclo­ sporine, it has a 100-fold greater potency and a more-rapid onset of action. A number of small uncontrolled studies have suggested benefit of oral or intravenous tacrolimus for the treatment of patients with refractory UC. The only randomized, placebo-controlled trial of tacrolimus in UC involved 63 Japanese patients with either steroiddependent or steroid-refractory disease who were randomized to receive either initial oral tacrolimus at 0.05  mg/kg or placebo twice daily.234 Patients in the high-trough concentration (10 to 15  ng/mL) tacrolimus group had a significantly higher rate of response and nonsignificantly higher rate of remission than those in the placebo group at week two, and a number of patients demonstrated response or remission (or both) after an additional 10 weeks of openlabel therapy. As with cyclosporine, tacrolimus can result in a number of toxicities including nephrotoxicity, electro-

lyte abnormalities, nausea, diarrhea, headache, tremors, paresthesias, insomnia, alopecia, hirsutism, and gingival hyperplasia.235,236 Thus, given the limited data and potential for harmful adverse events, the use of these alternative immunomodulators currently is not incorporated into standard practice.

Antibiotics

Antibiotics have a limited role in the management of UC, and most controlled studies have not demonstrated their benefit either in active disease or maintenance of remission.237-241 The most commonly used antibiotics in this setting are metronidazole and ciprofloxacin. One RCT found oral tobramycin to be superior to placebo as a short-term adjunctive therapy to glucocorticoids for active UC.242 Another RCT reported a modest benefit for the addition of ciprofloxacin for six months in patients with UC refractory to mesalamine and corticosteroids.243 At present, the data showing efficacy of antibiotics for treatment of patients with UC are not as convincing as are the data for antibiotic treatment of Crohn’s disease. Thus, at present the primary role of antibiotics in the treatment of UC is in the management of its suppurative complications.

Probiotics, Prebiotics, and Synbiotics

Probiotics are living organisms in foods and dietary supplements that might beneficially affect the host in a number of ways, including improving its intestinal microbial balance, blocking adhesion sites on colonocytes (which might improve mucosal barrier function), and enhancing local immune response.39,244 A probiotic can be a specific nonpathogenic strain of a bacterial species or a mixture of multiple species and strains, most commonly including Lactobacillus or Bifidobacterium species; sometimes they contain fungal antigens as well. An example of a common probiotic is VSL#3, which contains four strains of Lactoba­ cillus (Lactobacillus acidophilus, Lactobacillus delbrueckii subspecies bulgaricus, Lactobacillus plantarium, and Lactobacillus casei), three strains of Bifidobacterium (Bifidobacterium infantis, Bifidobacterium longum, Bifido­ bacterium breve), and one strain of Streptococcus (Strepto­ coccus salivarius subspecies thermophilus). Prebiotics are nondigestible food ingredients that selectively stimulate the growth or activity of one or more organisms of the intestinal microbiota, such as Lacto­ bacillus or Bifidobacterium species, thereby potentially conferring beneficial effects to the host.245,246 The majority of prebiotics are nondigestible oligosaccharides, with galacto-oligosaccharide, fructo-oligosaccharide, lactulose, and inulin being the most commonly used agents. Because probiotics have the challenge of competing with indigenous microbiota for nutrients, scientists have developed synbiotics, which are combinations of probiotics and prebiotics, in the hope of facilitating the survival of probiotics in the intestines. With respect to the use of these agents for inducing remission in mildly to moderately active UC, four RCTs have been performed using different agents.247-250 Two of three studies that measured rates of remission found no benefit of probiotics (VSL#3 in one study, fermented milk in the other) added to 5-aminosalicylates247,248; the third study found that E. coli Nissle 1917 combined with glucocorticoids had efficacy similar to that of mesalazine combined with glucocorticoids.249 The fourth study, which used a synbiotic, reported a nonsignificant improvement in disease activity when the synbiotic was combined with standard therapy.250 With respect to the use of these agents for the maintenance of remission in mildly to moderately

1999

2000

Section X  Small and Large Intestine active UC, six RCTs have been published.249,251-255 Two of these studies reported significantly lower rates of relapse for patients receiving a probiotic (Bifidobacterium in one study, fermented milk in the other) after medically induced remission compared with those receiving placebo,251,252 and the other four studies (using E. coli Nissle in three studies and Lactobacillus rhamnosus strain GG in the fourth) found no difference in rates of relapse.250,253-255 Nontraditional probiotic therapies that also have been evaluated include Saccharomyces boulardii and Trichuris suis.256,257 A small, uncontrolled study of 24 patients with mild to moderate active UC suggested a potential benefit of Saccharomyces boulardii when used in addition to mesalamine.256 The use of helminths in active UC was investigated by Weinstock and colleagues, who randomized 54 patients with active disease to receive 2500 T. suis ova or placebo orally every 2 weeks for 12 weeks and reported that rates of improvement were significantly higher in the active treatment group at week 12 (43% vs. 17%, P = 0.04); significant improvement was seen as early as week six.257 In summary, at present, there is no convincing evidence to support the use of probiotics, prebiotics, or synbiotics for the treatment of UC. However, future large well-designed RCTs are necessary to address this issue more definitively.

Nutritional Therapy

Short-chain fatty acids, especially butyrate, have been shown to be the main energy substrate for colonocytes. Butyrate metabolism accounts for approximately 70% of colonocyte oxygen use. The suggestion that there is an impairment of colonocyte oxidation of short-chain fatty acids in UC led to therapeutic investigations of this form of nutritional therapy. Indeed, placebo-controlled studies have found butyrate enemas to be beneficial in treating mildly active, left-sided colitis.258-260 Fish oil containing eicosapentaenoic acid has been found to attenuate colitis in animal models of colitis, probably by protecting the integrity of colonic mucosa, suppressing the inflammatory response, or both.261-263 In a small, placebocontrolled, crossover study of patients with mild to moderate UC, treatment with fish oil resulted in a 56% reduction in disease activity compared with a 4% reduction in controls (P < 0.05).264 This benefit has not been confirmed in other studies, and a benefit in maintaining remission has not been observed.265-267 Furthermore, compliance is limited because of side effects and the odor of the fish oil preparation. In contrast to Crohn’s disease, where bowel rest and total parenteral nutrition can improve disease, multiple studies have not found total parenteral nutrition with or without bowel rest to have any therapeutic advantage in patients with UC.268,269 Parenteral nutrition, however, can offer nutritional benefit in these patients. In general it is important to provide adequate nutrition to patients with UC who are about to undergo surgery. Nutrition is no more effective than placebo, however, for use as primary therapy of active UC.

Nicotine

Based on the observation that smoking is associated with a decreased risk of developing UC and that a former smoker with active colitis may gain clinical benefit on resuming smoking, nicotine has been used to treat patients with this disease. RCTs have shown some benefit of transdermal nicotine in the treatment of active UC.270-273 When administered at the highest tolerated dosage of 22 mg/day or less for four weeks in patients with mildly to moderately active UC, transdermal nicotine resulted in clinical improvement in 39% of patients compared with 9% of patients who received

placebo therapy (P = 0.007).270 As a single therapy, however, transdermal nicotine was not as effective as low-dose prednisolone.274 Common side effects included nausea, lightheadedness, itching, and tremor. Topical nicotine therapy has fewer side effects and may be an alternative. Pilot studies have shown topical nicotine to be beneficial in patients with distal UC, but no large RCT has been performed,275,276 and transdermal nicotine has not been found to be effective as a maintenance therapy.277 Thus, based on available data on clinical efficacy and the overall poor patient tolerability, nicotine cannot be considered part of the standard treatment for patients with UC.

Heparin

Heparin, a group of sulfated glycosaminoglycans, has antiinflammatory and immunomodulatory properties in addition to its well-known anticoagulant activity. The exact mechanism whereby heparin might ameliorate UC remains uncertain. An anticoagulant benefit, however, might not be responsible, because similar efficacy has not been observed in patients with IBD when treated with warfarin. Because of their negative charge, the glycosaminoglycans that constitute heparin have varied biological effects, including significant anti-inflammatory actions and augmentation of the peptide growth factors involved in intestinal mucosal repair and regeneration. Based on reports of fortuitous improvement in patients with UC receiving heparin for treatment of deep venous thromboses, pilot studies have suggested that unfractionated heparin may be effective for inducing remission in patients with severe, refractory UC.278,279 Compared with glucocorticoids as a first-line therapy, however, small RCTs have reported conflicting results.280,281 Intravenous heparin therapy has been associated with substantial bleeding complications. Low-molecular-weight heparin (LMWH) offers advantages over unfractionated heparin in its subcutaneous route of administration, and preliminary studies suggested a benefit of LMWH in the treatment of active UC.282,283 Unfortunately, this finding was not confirmed in a large, placebo-controlled trial of patients with mildly to moderately active UC receiving LMWH for six weeks.284 At this time, the use of either unfractionated heparin or LMWH cannot be advocated as primary therapy for patients with active UC.

Biological Therapy

Recent advances in our understanding of the pathogenesis of IBD have resulted in the development of therapies targeted at specific molecules or mediators involved in the inflammatory processes of these diseases. Most studies evaluating the efficacy of these agents have been performed in patients with Crohn’s disease, and only limited data are available for patients with UC. Anti-Tumor Necrosis Factor Antibodies TNF is a key proinflammatory cytokine that has been demonstrated to play a role in several disease states, including IBD. Elevated TNF concentrations have been found in inflamed intestine in patients with Crohn’s disease and UC, and stool and mucosal concentrations of TNF in patients with IBD have been shown to correlate with clinical disease activity. Infliximab (Remicade) is a chimeric monoclonal antibody of IgG1 subclass directed against human TNF-α. It consists of 75% human and 25% murine components (Fig. 112-14). The efficacy of infliximab in Crohn’s disease is well established, and it is approved by the FDA to treat Crohn’s disease and UC. Infliximab is thought to operate in Crohn’s disease via a multitude of mechanisms, including

Chapter 112  Ulcerative Colitis Human IgG antibody

Chimeric human/mouse antibody

Variable

Constant

Fc

Fc portion

infliximab Mouse component Figure 112-14.  Structural diagram of anti–tumor necrosis factor antibody. Normal human immunoglobulin G (IgG) antibody is shown on the left and infliximab is shown on the right. Infliximab is 75% human and 25% murine.

antagonizing the activity of TNF-α,285,286 initiating cytotoxicity on immune cells,287 and inducing T-cell apoptosis.288,289 Results from two large, multicenter, randomized, doubleblind, placebo-controlled trials (ACT 1 and 2) showed efficacy of infliximab therapy in UC.290 In these two similarly designed trials, 728 patients with moderately to severely active UC who failed conventional therapy with glucocorticoids alone or in combination with thiopurines (ACT 1) or glucocorticoids alone or in combination with thiopurines and 5-aminosalicylates (ACT 2) were randomized to placebo, infliximab 5 mg/kg, or infliximab 10 mg/kg at weeks 0 and 2 and then every eight weeks through week 46 (ACT 1) or week 22 (ACT 2). With respect to clinical response at week 8, in ACT 1 69% and 61% of patients receiving infliximab at 5 and 10 mg/kg, respectively, had a clinical response, compared with 37% of patients receiving placebo (P < 0.001 for both comparisons). In ACT 2 at week 8, 64% and 69% of patients receiving infliximab at 5 mg/kg and 10 mg/kg, respectively, had a clinical response, compared with 29% of patients receiving placebo (P < 0.001 for both comparisons). With respect to clinical remission at week 8 in ACT 1, 39% and 32% of patients receiving infliximab at 5 mg/kg and 10 mg/kg, respectively, attained remission, compared with 15% of patients receiving placebo (P < 0.003 for both compar