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 o
