Inflammatory Bowel Disease: From Bench to Bedside

Inflammatory Bowel Disease: From Bench to Bedside 2nd Edition

Inflammatory Bowel Disease: From Bench to Bedside 2"^ Edition

Edited by Stephan R. Targan Cedars-Sinai IBD Center, Los Angeles, CA, USA

Fergus Shanahan Cork University Hospital, Cork, Ireland

Loren C. Karp Cedars-Sinai IBD Center, Los Angeles, CA, USA

Springe]

Library of Congress Cataloging-in-Publication Data Inflammatory bowel disease : from bench to bedside / edited by Stephan R. Targan, Fergus Shanahan, and Loren C. Karp-2nd ed. p. ; cm. Includes bibliographical references and index. 1.Inflammatory bowel diseases. I. Targan, Stephan R. II. Shanahan, Fergus. III. Karp, Loren C. [DNLM: 1. Inflammatory Bowel Diseases. WI420 14237 2002] ISBN 1-4020-0713-2 (HC)

ISBN - 10 0-387-25807-8 (SC) ISBN-13 978-0387-25807-2

E-ISBN 0-387-25808-6 E-ISBN 978-0387-25808-9

Printed on acid-free paper. First softcover printing, 2005 Copyright © 2003 by Springer Science+Business Media. Inc. All rights reserved. TTiis work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science-hBusiness Media, Inc., 233 Spring Street, New York. NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now know or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks and similar terms, even if the are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the dale of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed in the United States of America. 9 8 7 6 5 4 3 2 springeronline.com

1

SPIN 11423430

Preface

Why A Second Edition? The pace of research in inflammatory bowel diseases has accelerated over the last decade, with a particularly rapid sprint occurring as we approached the new millennium. Advances in basic and technologic research have enabled scientists to examine the inflammatory process at the cellular and molecular levels. The powerful research tools of the current biotech and genotech era are now being applied successfully to inflammatory bowel disease research. Although there are many unanswered questions, today as before, the pathogenesis of inflammatory bowel disease can now be discussed at an increasingly fundamental level. Today, new therapeutic strategies are based on understanding of pathophysiology and are no longer introduced merely on an empiric basis. Many hitherto unexplained features of inflammatory bowel diseases can be accounted for because of improvements in understanding of the immune and inflammatory responses in the gut. Like the first edition of Inflammatory Bowel Diseases: From Bench to Bedside, this book is intended to be more than a comprehensive compilation of reviews by individual authors on different aspects of these disorders. It is intended that the individual chapters be components of a coherent, albeit detailed, story of the local and systemic pathophysiology of intestinal inflammation, with a well-reasoned series of management strategies. Our goal was not only to produce a standard reference text, but also to present research advances and current concepts of etiopathogenesis in the context of what is already known of the clinicopathologic features of these disorders. Our vision was to have a book that would blend recent advances in the basic and clinical sciences as they relate to inflammatory bowel disease. It is our hope that that the book will illustrate the effectiveness of a team approach of basic scientists and clinician investigators in the field of inflammatory bowel disease. The book will give the reader a glimpse of where the field is moving and an idea of

likely research directions in the future. Of course, it is our personal wish that this book will stimulate ideas for future research. In early 1994, when the first edition of Inflammatory Bowel Diseases: From Bench to Bedside, was released, our introductory pages included predictions and considerations for the future of management of these disorders, as follows: (1) heterogeneity and the 'reagent grade' patient; (2) combination drug therapy rather than a stepwise progression; (3) emphasis on intestinal immunophysiology; (4) mucosa-specific rather than systemic immunomodulation; and (5) focus on factors promoting healing and remission rather than relapse. Each of these concepts has been validated and extended in the ensuing years, and are now the very foundation of current laboratory and clinical research on the inflammatory bowel diseases. In addition, we posed three questions related to the predictions that we felt fundamental to research both at the bench and at the bedside. Perhaps the single most important technological advance toward answering these questions has been the application of molecular technology to the creation of a new society of colitic animal models, from which to learn about human disease. With the use of such models much progress has been made on first question, "Are Crohn's disease and ulcerative colitis diff*erent expressions of the same disease or are they discrete entities?" Based on numerous factors, including genetic associations, marker antibodies, and environmental agents, it has become increasingly clear that the varying clinical manifestations of inflammatory bowel disease reflect unique pathogenic processes in the mucosa. Indeed, there are numerous discrete entities that can be stratified by a variety of subclinical and clinical markers that may well identify which patients are likely to respond to any particular therapeutic intervention. The second question, "Do infectious agents have a role in the etiology or pathogenesis of inflammatory

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, v-vi. © 2003 Kluwer Academic Publishers. Printed in Great Britain

VI

bowel disease?" is a major focus for researchers at the present time. Mounting evidence, including the identification of specific bacterial products, implicate such agents in the pathogenic process. A key objective for investigators is to understand the interplay between bacteria and the altered immune response leading to mucosal inflammation. Evidence suggests that these immune responses are to normal commensal bacteria rather than any specific pathogen. In 2001, a Crohn's disease associated gene mutation was discovered in NOD2, a protein that is responsible for regulating appropriate responses between bacteria and host. This finding corroborates that an abnormality in this interaction is fundamental to the disease process in at least some forms of Crohn's disease, and further confirms that is unlikely that infectious agents will be directly correlated with the disease responses that are characteristic of the inflammatory bowel diseases. More likely, the process is indirect, requiring any number of combinations of genetics, immune responses, and environmental triggers, to manifest as disease. Specific manipulation of bacterial expression and the corresponding immune response may be the basis of very effective therapies in the near future. Finally, we posed the questions, ''Where is research taking us? How will it change the management of inflammatory bowel disease?" The effect of the research has already become apparent in the example of therapeutic anti-tumor necrosis factor-a

Preface monoclonal antibodies (anti-TNF-ot). This treatment is targeted at a very specific point in the immune response, over-production of TNF-oc, commonly found in patients with Crohn's disease. That this treatment is only effective in a portion of patients with Crohn's disease points to the fact that there are unique immune mechanisms that underly the intestinal inflammation in diff'erent subpopulations of patients. An even more recently developed therapeutic approach for those not responding to anti-TNF-oc - or even those that do - involves using an antibody to oc4 to prevent the recirculation of cells into the mucosa, a method which has been shown to decrease inflammation. In the second edition, chapters across the spectrum of Inflammatory Bowel Disease: From Bench to Bedside will reflect the advances delineated above and lay the groundwork for ongoing research and treatment of these disorders. In the near future, the molecular basis of the interaction between host genetics and the environment will become more clear, and the linkage with the immune system will reveal not only more effective therapy but the ability to predict responders and non-responders to individual therapies. We would like to express our gratitude to each of the authors for their carefully conceived contributions. In encouraging the authors to include their individual perspective and philosophical approaches, we likely made their job more difficult.

List of Authors

Maria T. Abreu Director, Basic and Translational Research, IBD Center Cedars-Sinai Medical Center Los Angeles, CA USA

Richard S. Blumberg Chief, Gastroenterology Division Brigham and Women's Hospital Harvard Medical School Boston, MA USA

David H. Alpers William B. Kountz Professor of Medicine Washington University School of Medicine St Louis, MO USA

Lawrence J. Brandt Professor of Medicine and Surgery Albert Einstein College of Medicine Chief of Gastroenterology, Montefiore Medical Center and AECOM Bronx, NY USA

Kristen O. Arseneau Instructor of Research Digestive Health Center of Excellence University of Virginia Health System Charlottesville, VA USA Filip Baert Heilig Hartziekenhuis Roeselare Belgium IBD Consultant University Hospital Leuven Belgium Charles N. Bernstein Professor of Medicine, Head Section of Gastroenterology and Director, University of Manitoba IBD Clinical and Research Center University of Manitoba Winnipeg, Manitoba Canada Scott W. Binder Chief, Dermatopathology Associate Professor of Pathology UCLA Medical Center Los Angeles, CA USA

Jonathan Braun Professor and Chair Department of Pathology and Lab. Medicine David Gefifen School of Medicine at UCLA Los Angeles, CA USA Arlene Caplan Clinical Psychologist, Division of Gastroenterology, Hepatology, Hopital Sainte Justine Associated Professor, Department of Psychology University of Montreal, Quebec Canada Roger Chapman John Radcliffe Hospital Oxford UK Gregg W. Van Citters Research Fellow Gonda Diabetes Research Center City of Hope National Medical Center/Beckman Research Institute Duarte, CA USA

List of authors

Vlll

Offer Cohavy Post-doctoral Fellow Cedars-Sinai IBD Center Los Angeles, CA USA Stephen M. Collins Professor of Medicine, Chief of Gastroenterology McMaster University Hamilton, Ontario Canada Fabio Cominelli David D. Stone Professor of Internal Medicine Director, Digestive Health Center of Excellence University of Virginia Health System Charlottesville, VA USA William R. Connell St Vincent's Hospital Melbourne Victoria Australia

Themistocles Dassopoulos Assistant Professor of Medicine Department of Medicine, Division of Gastroenterology, Johns Hopkins University Baltimore, MD USA Sander J.H. van Deventer Professor of Gastroenterology Academic Medical Center Amsterdam The Netherlands Laurence J. Egan Assistant Professor of Medicine Mayo Medical School, Mayo Clinic Rochester, MN USA Mark Eggena Post-doctoral Fellow Division of Infectious Diseases University of California, San Francisco San Francisco, CA USA

Galen Cortina Assistant Professor Department of Pathology and Laboratory Medicine University of California at Los Angeles Los Angeles, CA USA

Anders Ekbom Professor, Department of Medicine at Karolinska Hospital, Karolinska Institutet Stockholm Sweden

Kenneth Croitoru Associate Professor of Medicine, Division of Gastroenterology McMaster University Hamilton, Ontario Canada

Charles O. Elson Basil I. Hirschowitz Chair inn Gastroenterology Professor of Medicine and Microbiology University of Alabama at Birmingham Birmingham, AL USA

Cornelius C. Cronin Consultant Physician/Gastroenterologist Mallow General Hospital University College Cork Cork Ireland

Sue C. Eng Senior Fellow in Gastroenterology University of Washington School of Medicine Seattle, WA USA

Sue Cullen John Radcliffe Hospital Oxford UK

Richard J. Farrell Co-Director of the Center for IBD Beth Israel Deaconess Medical Center Assistant Professor of Medicine Harvard Medical School Boston, MA USA

List of authors Michael J.G. Farthing Professor of Medicine Faculty of Medicine University of Glasgow Glasgow Scotland, U K

IX

Geert D'Haens Gastroenterologist, Imelda General Hospital Bonheiden and University Hospital Leuven Leuven Belgium

Brian G. Feagan Professor of Medicine, Epidemiology and Biostatistics University of Western Ontario London, Ontario Canada

Rodger C. Haggitt (deceased) Director of Hospital Pathology, University of Washington Medical Center and Professor of Pathology and Adjunct Professor of Medicine University of Washington Seattle, WA USA

Claudio Fiocchi Division of Gastroenterology Professor of Medicine, Pathology and Pediatrics Case Western Reserve University School of Medicine Cleveland, OH USA

Stephen B. Hanauer Joseph B. Kirsner Professor of Medicine and Clinical Pharmacology Chief, Section of Gastroenterology and Nutrition University of Chicago Chicago, IL USA

Edward Fitzgerald Consultant Radiologist Mercy Hospital Grenville Place Cork Ireland

Robert Hershberg Corixa Corporation Seattle, WA USA

Michael N. Goke Priv.-Doz. Dr. med. Gastroenterology, Hepatology and Endocrinology Medizinische Hochschule Hannover Hannover Germany D. Neil Granger Professor Department of Molecular and Cellular Physiology LSU Health Sciences Center Shreveport, LA USA Matthew B. Grisham Professor Department of Molecular and Cellular Physiology LSU Health Sciences Center Shreveport, LA USA

E. Jan Irvine Professor of Medicine McMaster University Hamilton, Ontario Canada Derek P. Jewell Professor of Gastroenterology University of Oxford Oxford UK Lori Kam San Fernando Valley Gastroenterology Medical Group Tarzana, CA USA Loren C. Karp Researcher/Specialist, IBD Center Cedars-Sinai Medical Center Los Angeles, CA USA

List of authors Ciaran P. Kelly Chief, Blumgart Firm, and Director Gastroenterology Fellowship Training Beth Israel Deaconess Medical Center Boston, MA USA Lorraine Kyne Assistant Professor of Medicine Harvard Medical School, and Gerontology Division Beth Israel Deaconess Medical Center Boston, MA USA F. Stephen Laroux Post-doctoral Fellow Department of Molecular and Cellular Physiology LSU Health Sciences Center Shreveport, LA USa Klaus Lewin Professor, Department of Pathology and Laboratory Medicine University of California at Los Angeles Los Angeles, CA USA Steven N. Lichtman Professor of Pediatrics University of North Carolina at Chapel Hill Chapel Hill, NC USA Henry C. Lin Director, GI Motility Program and Section of Nutrition Cedars-Sinai Medical Center Associate Professor of Medicine, UCLA Los Angeles, CA USA Ian Lindsey Consultant Colorectal Surgeon John Radcliffe Hospital Oxford UK

Uma Mahadevan Clinical Assistant Professor of Medicine University of California at San Francisco San Francisco, CA USA Phillipe Marteau Professor, University of Paris (LHP) Hepato-gastroenterology Unit Laennec Hospital Paris France Robin S. McLeod Professor of Surgery and Health Policy, Management and Evaluation University of Toronto Head, Division of General Surgery Mount Sinai Hospital Toronto, Ontario Canada Stephen J. Meltzer Professor of Medicine University of Maryland School of Medicine Baltimore, MD USA Neil J. McC. Mortensen Professor of Colorectal Surgery John Radcliffe Hospital Oxford UK Diarmuid O'Donoghue Gastroenterology Department Centre for Colorectal Disease St. Vincent's University Hospital Dublin Ireland Timothy R. Orchard Consultant Gastroenterology St Mary's Hospital Praed Street London UK

List of authors Remo Panaccione Assistant Clinical Professor University of Calgary Calgary, Alberta Canada Konstantinos A. Papadakis Assistant Professor of Medicine UCLA School of Medicine, and Research Staff Physician Gastroenterology/IBD Center Cedars-Sinai Medical Center Los Angeles, CA USA Seth Persky Gastroenterologist Long Island Digestive Disease Consultants Setauket, New York Clinical Instructor of Medicine SUNY-Stonybrook School of Medicine Stonybrook, NY USA Theresa T. Pizarro Assistant Professor of Internal Medicine Digestive Health Center of Excellence University of Virginia Health System Charlottesville, VA USA Daniel K. Podolsky Mallinckrodt Professor of Medicine Harvard Medical School Chief, Gastrointestinal Division Massachusetts General Hospital Boston, MA USA

XI

Jerome I Rotter Director, Division of Medical Genetics and Board of Governors' Chair in Medical Genetics Cedars-Sinai Medical Center Professor of Medicine Pediatrics and Human Genetics UCLA School of Medicine Los Angeles, CA USA Paul Rutgeerts Professor of Medicine Department of Gastroenterology University Hospital of Gasthuisberg 3000 Leuven Belgium William J. Sandborn Professor of Medicine Mayo Medical School, Mayo Clinic Rochester, MN USA R. Balfour Sartor Professor of Medicine, Microbiology and Immunology University of North Carohna at Chapel Hill Chapel Hill, NC USA Ernest G. Seidman Professor and Chief, Division of Gastroenterology, Hepatology and Nutrition Sainte Justine Hospital Department of Pediatrics University of Montreal, Quebec Canada

Robert H. Riddell Pathologist, Mount Sinai Hospital and Professor, Department of Laboratory Medicine and Pathobiology University of Toronto Canada

Fergus Shanahan Professor of Medicine Chairman, Department of Medicine Cork University Hospital University College Cork Cork Ireland

Gerhard Rogler PD Dr. med. Dr. phil. Department of Internal Medicine I University Clinic of Regensburg Regensburg Germany

Kieran Sheahan Consultant Histopathologist Pathology Department St Vincent's University Hospital Dublin Ireland

List of authors

Xll

Konrad H. Soergel Professor of Medicine and Physiology Medical College of Wisconsin Milwaukee, WI USA Catherine J. Streutker Pathologist, St Michael's Hospital and Lecturer, Department of Laboratory Medicine and Pathobiology University of Toronto, Ontario Canada Christina M. Surawicz Professor of Medicine Assistant Dean for Facuky Development University of Washington School of Medicine Seattle, WA USA Lloyd R. Sutherland Professor of Medicine University of Calgary Calgary, Alberta Canada Stephan R. Targan Director, Division of Gastroenterology, IBDCenter, and, Immunobiology Institute Cedars-Sinai IBD Center Los Angeles, CA USA Kent D. Taylor Research Scientist Medical Genetics and Birth Defects Center, and Inflammatory Bowel Disease Center Cedars-Sinai Medical Center Assistant Professor

Pediatrics, UCLA School of Medicine Los Angeles, CA USA William J. Tremaine Professor of Medicine Director, IBD Clinic Mayo Clinic Rochester, MN USA Eric A. Vasiliauskas Associate Clinical Director, IBD Center Cedars-Sinai Medical Center Assistant Professor of Medicine and Pediatrics UCLA School of Medicine Los Angeles, CA USA Huiying Yang Director, Genetic Epidemiology Program Medical Genetics Birth Defects Center Cedars-Sinai Health System Associate Professor of Pediatrics and Epidemiology UCLA School Medicine and Public Health Los Angeles, CA USA Casey T. Weaver Professor of Pathology and Microbiology University of Alabama at Birmingham Birmingham, AL USA C. Mel Wilcox Professor of Medicine University of Alabama at Birmingham Birmingham, AL USA

Table of Contents

Preface Stephan R. Targan/Fergus Shanahan/Loren C. Karp

v

SECTION 1: The laboratory bench 1. 2. 3. 4.

5. 6. 7. 8. 9.

Introduction: Inflammatory bowel diseases: from bench to bedside Fergus Shanahan, Loren C Karp, Stephan R. Targan

3

The changing faces of Crohn's disease and ulcerative colitis Anders Ekbom

5

Genetics of inflammatory bowel disease Kent D. Taylor, Jerome I. Rotter, Huiying Yang

21

Experimental mouse models of inflammatory bowel disease: new insights into pathogenic mechanisms Charles O. Elson and Casey T. Weaver

67

The normal intestinal mucosa: a state of 'controlled inflammation' Claudia Fiocchi

101

The lymphocyte-epithelial-bacterial interface Robert Hershberg and Richard S. Blumberg

121

The mucosal inflammatory response. Cytokines and chemokines Fabio Cominelli, Kristen O. Arseneau and Theresa T Pizarro

147

Role of the microcirculation in chronic gut inflammation Matthew B. Grisham, F Stephen LarouxandD. Neil Granger

111

Remission, relapse, intestinal healing and repair Michael N. Goke and Daniel Podolsky

197

10. Antibodies in the exploration of inflammatory bowel disease pathogenesis and disease stratification Jonathan Braun, Offer Cohavy and Mark Eggena

211

11. Pathophysiology of inflammatory bowel disease: the effect of inflammation on intestinal function Stephen M. Collins and Kenneth Croitoru

223

12. Systemic consequences of intestinal inflammation Konstantinos A. Papadakis and Maria T Abreu

235

SECTION II: The bedside 13. Understanding symptoms and signs in inflammatory bowel disease Cornelius C Cronin and Fergus Shanahan

253

xiv

Table of contents

14. Clinical course and complications of ulcerative colitis and ulcerative proctitis Remo Panaccione and Lloyd R. Sutherland

269

15. Clinical features and complications of Crohn's disease William J. Tremaine

291

16. Mechanisms of systemic inflammation associated with intestinal injury R. Balfour Sartor and Steven N. Lichtman

305

17. Pathology of inflammatory bowel diseases: a critical appraisal in diagnosis and management Galen Cortina and Klaus Lewin

337

18. An endoscopic and histologic perspective of diagnosis: when, where, and what to do Charles N. Bernstein and Robert H. Riddell

357

19. Radiologic (radiographic) and imaging features of ulcerative cohtis and Crohn's disease Edward Fitzgerald

371

20. New diagnostic approaches in inflammatory bowel disease Lori Kam and Eric Vasiliauskas

409

21. Differential diagnosis of colitis Sue C Engand Christina M. Surawicz

431

22. 'Disease management' in chronic medical conditions Da vid K A Ipers

457

23. Pharmacoeconomics and inflammatory bowel disease Brian G. Feagan

471

24. Measuring quality of life in inflammatory bowel disease E. Jan Irvine 25. Clinical pharmacology in inflammatory bowel disease: optimizing current medical therapy Laurence J. Egan and William J. Sandborn

481 495

26. Multi-site therapeutic modalities for inflammatory bowel diseases - mechanisms of action Gerhard Rogler

523

27. Targeted therapies for inflammatory bowel disease SanderlH. van Deventer

553

28. Role of antibiotics and probiotics in the management of inflammatory bowel disease Philippe Marteau and Fergus Shanahan

573

29. Nutrition in inflammatory bowel disease Gregg WVan Citters and Henry C Lin

587

30. Medical management of ulcerative colitis William 1 Sandborn

605

31. Surgical management of ulcerative colitis Ian Lindsey and Neil 1 McCMortensen

631

32. Pouchitis: clinical characteristics and management Uma Mahadevan and William J. Sandborn

643

33. Medical therapy for Crohn's disease Stephen B. Hanauerand Themistocles Dassopoulos

659

Table of con ten ts

xv

34. Surgery for Crohn's disease Robin S. McLeod

681

35. Postoperative prevention of recurrence of Crohn's disease Filip Baert, Geert D'Haens and Paul Rutgeerts

697

36. The molecular pathology of inflammatory bowel disease-associated neoplasia and preneoplasia Stephen Meltzer

111

37. Dysplasia in inflammatory bowel disease: clinical pathology and current surveillance methods Catherine! Streutker, Roger C. Haggitt and Robert H. Riddell

719

38. Hepatobiliary disorders Sue Cullen and Roger Chapman

731

39. Articular and ocular complications of inflammatory bowel disease Timothy R. Orchard and Derek R Jewell

141

40. Cutaneous manifestations of inflammatory bowel disease Scott W Binder

757

41. Fertility and pregnancy in inflammatory bowel disease William Connell

763

42. Special considerations in the diagnosis and management of inflammatory bowel disease in the pediatric age group Ernest G. Seidman andArlene Caplan

113

43. Microscopic colitis: collagenous and lymphocytic colitis Diarmuid O'Donoghue and Kieran Sheahan

791

44. Colon ischemia Seth E. Persky and Lawrence J. Brandt

799

45. Diversion colitis KonradH. Soergel

811

46. Pseudomembranous colitis and Clostridium difficile infection Richard J. Farrell, Lorraine Kyne and Ciardn P Kelly

823

47. Infectious colitis Michael J. G Farthing

845

48. Human immunodeficiency virus and inflammatory bowel disease Charles Mel Wilcox

863

49. Bone metabolism and inflammatory bowel disease Maria T. Abreu

875

SECTION III: Back to the laboratory bench 50. Epilogue: Bench to bedside and back to bench Stephan R. Targan, Loren C Karp and Fergus Shanahan Index

887 893

Section I THE LABORATORY BENCH

1

Introduction: Inflammatory bowel disease: from bench to bedside FERGUS SHANAHAN, LOREN C. KARP AND STEPHAN R. TARGAN

The adventurous physician goes on, and substitutes presumption for knowledge. From the scanty field of what is known, he launches into the boundless region of what is unknown (Thomas Jefferson)

The introductory section of the first edition of this book closed with the following upbeat comments: 'inflammatory bowel disease is now an exciting field for the basic researcher, the clinician-investigator, and the clinical practitioner. Much work remains to be done ... Onward!'. In the same year the late Professor Anne Ferguson challenged the medical, political and research communities with a provocative editorial in which Crohn's disease and ulcerative colitis were referred to as 'important and disabling diseases, still under-researched.' Since then there have been remarkable improvements in our understanding of the pathophysiology of these disorders. The molecular mediators and major pathways of tissue injury have been identified. Improvements in molecular genetic technology have ensured that fundamental questions regarding mucosal inflammation can now be asked and answered. The new information promises to be translated into improvements in patient management. This is reflected in a progressive shift from therapeutic empiricism to evidence-based management. Changes in research strategy have perhaps been even more important than technological progress in providing an integrated and coherent overview of disease mechanisms. Complex disorders require research input from a diversity of perspectives, including traditional disciplines, such as biochemistry, microbiology and immunology. It is at the interface of these seemingly disparate disciplines where incisive advances have been made; hence the emergence of hybrid disciplines such as microbial pathophysiology, immunophysiology and psychoneuroimmunology. This is a recurring theme throughout this text with emphasis on transdisciplinary topics such as intercellular crosstalk

within the mucosa, lymphoepithelial dialog, neuroimmune interactions, and epithelial-microbe signalling (Chapters 3-11). This approach is also reflected in chapters explaining the fundamental basis of systemic symptoms and signs and extraintestinal manifestations experienced by patients with Crohn's disease and ulcerative colitis (Chapters 11-13 and 16). Understanding the role of endogenous outcome modifiers such as psychological stress, the brain-gut axis, and even the placebo response of the individual, requires the same transdisciplinary perspective. Perhaps the most intriguing convergence of research avenues is the interaction among the three major ingredients of the pathophysiology of inflammatory bowel disease (genetic predisposition, environmental bacteria and immune dysregulation). Indeed, the interface at the center of this triad appears to have become the basis of a unifying concept for the development of most autoimmune disorders. It is perhaps not surprising that one of the genes associated with increased susceptibility to Crohn's disease (N0D2) is linked to the mechanism of immune p e r c e p t i o n of the b a c t e r i a l microenvironment (Chapter 2). One might predict that additional genes will be identified which regulate how the host immune system handles the microbial flora within the gut. Unraveling the pathogenesis of inflammatory bowel disease is of more than passing interest to the clinician and has several immediate implications for patient management. Genetic studies have already come to the patient bedside with increasing emphasis on phamacogenomic prediction of drug eflicacy and toxicity (Chapters 25 and 26). Immune mechanisms of tissue damage in inflammatory bowel disease have been successfully exploited for diagnostic and therapeutic purposes. Thus, immunologic alterations in patients with Crohn's disease and ulcerative colitis have utility as noninvasive diagnostic tools (Chapter

Stephan R. Targan, Fergus Shanahan and Loren C. Karp (eds.). Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 3-4. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Introduction: Inflammatory bowel disease: from bench to bedside 20), and blockade of specific immune mediators has been one of the most elegant examples of bench to bedside medicine in recent years (Chapter 27). The challenge for clinicians now will be to estabhsh the hierarchy of importance of site-specific therapeutics versus more traditional nonspecific modalities (Chapter 26) for different patients and the place of combination therapy. What of the remaining contributors to the pathogenesis - the enteric bacteria? Here again there is already evidence for translation of research data to the clinical setting. Clinicians have been using the bacterial flora for decades to metabolize the prodrugs sulfasalazine or olsalazine to the active aminosalicylate moiety. The impact of enteral feeding in Crohn's disease is now thought to be mediated in part though effects on the flora and epithelial barrier function. Today, manipulation of the enteric flora with probiotics, prebiotics and synbiotics is an emerging therapeutic strategy as a safe adjunct to immunomodulation (Chapter 28). Tomorrow, perhaps one can predict that molecular fingerprinting of the enteric flora will help explain the timing of disease onset, the subset of bacteria driving the inflammatory response and the influence of the host on the composition of the indigenous flora. Because of the lesson of Helicobacter pylori and peptic disease, basic investigators and clinicians will be reluctant to dismiss the possibility of a specific infectious etiology in some or all patients with inflammatory bowel disease. Although this could account for the changing nature and prevalence of Crohn's disease and ulcerative colitis (Chapter 2), the information from animal models of disease suggests that more than one pathogenic pathway may be involved in leading to the same phenotypic outcome (Chapter 4). The animal models also indicate that a genetically determined defect in immune regulation can indeed lead to chronic inflammation in the presence of the normal bacterial flora, and this does not require a specific pathogenic infection in the traditional sense of a transmissible agent. Thus, these diseases are unlikely to represent a simple cause and eff'ect relationship with a simple struggle between microbe and humans (Chapter 6). Extrapolation of the diversity of animal models to the human condition has the obvious implication

that clinicians need to contend with an extraordinary degree of patient heterogeneity. While this has not been particularly evident from conventional clinical criteria, improvements in genetic, immunologic and microbial molecular diagnostics are poised to facilitate categorization of patient subsets to improve the design and interpretation of clinical trials in the future (Chapter 23). Today the splitters appear to have the advantage over the lumpers. But the learning curve may have reached the point of inflexion, and with more research data the apparent complexity is likely to give way to commonalities and patterns of disease and therapeutic responsiveness. Of course, the management of inflammatory bowel disease today must not only be evidence-based (Chapter 30), but must also be accountable in the context of modern concepts of disease management (Chapter 22) and outcome assessments (Chapter 23). In the midst of relentless scientific progress, clinicians will do well to uphold the traditional principles of caring for patients with chronic disease. Despite all the advances at the research bench, patients seem to have increasing expectations of health care and lower tolerance of illness. Dissatisfaction with modern medicine is reflected in greater litigation and expenditure on alternative medicine. Patient surveys consistently reflect the importance of the doctorpatient relationship. Compassion, time, and a commitment to long-term management are unlikely to become obsolete or superseded by any novel drug therapy. In summary, inflammatory bowel disease is more exciting than ever; it remains a rewarding field for basic researchers and clinicians. Notwithstanding the extraordinary advances brought about by the genotech-biotech era, clinical clues at the bedside continue to pose the correct questions to be tackled in the research laboratory. Likewise, observations at the laboratory bench will continue to be translated into enhanced diagnostic and therapeutic strategies in the clinic. This bench-bedside interface shared by the clinician and the basic investigator needs to be nurtured, as it can pay handsome dividends for patient welfare. Therein lies the essence of what is intended with this book.

The changing faces of Crohn's disease and ulcerative coiitis ANDERS EKBOM

Introduction There are few diseases which have changed faces to such an extent as the inflammatory bowel diseases, i.e. ulcerative colitis (UC) and Crohn's disease (CD). At the beginning of the 20th century inflammatory bowel disease (IBD) was a rarity, and at the end of the same century these disease entities were something gastroenterologists in the Westernized world encounter not once, but repeatedly, on a daily basis. Fifty years ago high socioeconomic status and Jewish ethnicity were two commonly accepted risk factors, associations that today are either questioned or have been refuted in observational studies. Pancolitis, in the case of UC and CD confined to terminal ileum, were the most common clinical features during the first part of the 20th century as opposed to today. Nowadays ulcerative proctitis is the most common clinical presentation among UC patients and CD confined to the terminal ileum constitutes a minority of CD patients. Geographically there was a north-south gradient; a finding reproduced in different settings and continents that does not seem to exist today. Instead, we can see the emergence of an eastwest gradient. A hypothesis of the origin of the diseases included Mycobacterium paratuberculosis as a potential agent, a hypothesis that was refuted early on, then reintroduced, and later refuted again. These different faces of the two diseases are probably in part due to bad methodology or science, but this also illustrates that the two diseases have changed faces over time. It is a rather straightforward endeavor to describe the different faces over time, as we are fortunate to have clinicians and epidemiologists who, during the past 100 years, have taken a great interest in IBD. The difficult part is to interpret these findings and to make sense of them. The goal is to provide benchmarks for other scientists, which can be used to test the different hypotheses, which will emerge. This way we will

eventually understand the etiology of IBD and primary prevention will become a possibility.

Occurrence Temporal trends Although historic figures such as Alfred the Great [1] and Bonnie Prince Charles [2] have been proposed to have suff'ered from CD and UC, respectively by clinicians turned historian, it is obvious that the two disease entities were rare until the 20th century. However, at the end of the 19th century there were quite a few case reports of patients with UC in Great Britain, and a symposium was held at the Royal Society of Medicine, London, as early as 1909, at which 317 patients from diff'erent hospitals in London were presented [3]. Similarly, in 1913 Kenneth Dalziel, a Scottish surgeon, reported nine patients with a new entity described as 'chronic intestinal enteritis and not tuberculosis' [4]. There is a consensus that those nine cases constitute the first case series of what later was called CD. Moreover, a retrospective study from Ireland has described 29 cases of CD treated during the latter half of the 19th century [5]. There are also reports outside the AngloSaxon world of early cases. For instance, the earliest bona-fide case of CD in Sweden is a 13-year-old boy operated on in 1918 due to suspected appendicitis. A bypass procedure was performed because of stenosis of the terminal ileum and, interestingly, the surgeon dismissed the diagnosis of tuberculosis as highly improbable. Nothing was heard from the patient until 1969 when he was admitted for perianal fistulas. As a subsequent barium enema was difficult to interpret, a laparotomy was performed and a resection including the bypassed segment of the ileum was done. The subsequent histopathologic examination revealed changes typical for CD of the terminal ileum. However, it was not until Dr Burrill B. Crohn,

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 5-20. © 2003 Kluwer Academic Publishers. Printed in Great Britain

The changing faces of Crohn's disease and ulcerative colitis in 1932, introduced the term 'regional ileitis' for the disease that was later named after him, when CD became a distinct clinical entity [6]. There are no incidence figures available for either UC or CD until the 1930s. During this decade there are incidence rates from two retrospective studies in two defined populations. The first was done retrospectively in Rochester, Minnesota, where the authors were able to demonstrate an annual incidence of UC of 6.0 per 100 000 for the period 1934-1944 [7] and an annual incidence of CD of 1.9 per 100 000 for the period 1935-1954 [8]. In Europe there is only one estimate during the 1930s for CD, from Cardiff, United Kingdom, of 0.2 per 100 000 for the period 1935-1945 [9]. Thereafter, there have been an increasing number of incidence studies pubHshed either for both UC and CD, or for one of these conditions for different populations and different time periods. Besides cancer and cardiovascular diseases there are probably no other disease entities where so many incidence studies have been undertaken in so many different populations during the later part of the 20th century. However, most of these studies have generally dealt with small populations and/or short time periods; in most instances less than 10 years. Moreover, different case-finding methods have been used, and in many instances no age standardization has been done, making comparisons between different populations and time periods impossible to perform. In spite of these shortcomings there are remarkable similarities in the temporal trends for UC and CD in different populations in the Westernized world. There is a consistent finding from both western Europe and northern America of a substantial increase in incidence of both CD and UC since the Second World War. There is also a strong correlation in the occurrence of the two diseases. Areas or populations with a high incidence or mortality attributed to UC also have a high incidence or mortality due to CD, and viceversa [10]. Although misclassification of either disease could lead to a false correlation, the consistency of these findings in different settings argues convincingly against such a bias. Moreover, although this chapter will not deal with genetics, relatives of patients with UC are at higher risk for UC and also, to a lesser extent, for CD. Relatives of patients with CD are at higher risk for CD and, to a lesser extent, UC [11]. The temporal trends for both diseases and genetics consequently leave us with two different hypotheses: (1) UC and CD represent the

opposite ends of a continuous spectrum of IBD, but with different clinical characteristics; or (2) there are some shared genetic and/or environmental risk factors for UC and CD. Another common feature, with regard to the temporal trends for UC and CD, is that, in those instances where there are incidence figures for the shift from a low-incidence area to a high-incidence area, an increasing incidence of UC precedes an increase in CD with a time lag of around 15-20 years. The time lag is apparent in studies from Sweden [1215], Iceland [16-18], Copenhagen [19-21], the Faroe Islands [22, 23], and the US [7, 8, 24-27]. In Fig. 1 the incidence figures for UC and CD are given for Uppsala County, Sweden, from 1945 to 1983. These figures are the result of four different studies performed in this population [12-15]. These studies are of additional interest as the different researchers have failed to identify more than a handful of prevalent cases, i.e. patients with symptoms or a diagnosis before 1945 for either disease. Different case-finding methods have been utilized in the different studies, but yielding the same results. These findings, like those from Cardiff [9, 28] and Minnesota [7, 8, 24-27], where a constant monitoring has been present, strongly suggest that the low incidence figures in the 1930s and 1940s are not due to flawed case-finding methods, but that this increase in incidence for both UC and CD is real. Since the 1960s and 1970s the excess mortality compared to the background population is only marginal in patients with UC and CD. Thus, the presence of a shift, i.e. transition between low and high incidence of either UC or CD, makes prevalence figures less suitable to compare the occurrence of those diseases in different populations. The prevalence will then be a measure not only of the incidence but also of the duration since this change in incidence. Incidence figures should therefore be used in comparing populations and over time. Another consistent finding in the descriptive epidemiology of IBD up to recently is that the incidence of UC is higher than for CD. In the case of CD there seems to be a leveling-off in most populations with an incidence of up to 6.0 per 100 000 and for UC an incidence between 15 and 20 per 100000 although higher incidence figures have been reported for both diseases [9,29]. This corresponds to a lifetime risk for either one of the two diseases in high-incidence areas between 0.5 and 1.0% which is close to the risk for a disease such as rheumatoid arthritis, which no one disputes constitutes a public health problem. How-

Anders Ekbom

Incidence per 100,000 15

-Ulcerative colitis

Crohn's disease

10

5 •]

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

Year Figure 1. Annual incidence of ulcerative colitis and Crohn's disease in Uppsala County, Sweden, 1945-1983.

ever, as usual with IBD, there are different faces. Recent reports from both Belgium and France [30, 31] play havoc with the notion that UC is more common than CD as the opposite seems to be true in those two adjacent areas.

Age-specific incidence The age-specific incidence rates for UC and CD vary substantially in different populations and over time. Populations with low annual incidence have an almost flat age-specific curve, and during periods of a rising annual incidence this increase will be most pronounced in the age group 20-40 years for both UC [27, 32] and CD [13, 26]. In Fig. 2 this is demonstrated by comparing the number of CD patients during the transient period in Uppsala County, Sweden, 1956-1961 versus 1962-1967. To what extent there is a second peak in high incidence areas in the older age groups (60+) remains controversial. It has been argued that this peak represents a delayed diagnosis made when the disease relapses, as demonstrated in the first bona-

fide case of CD in Sweden and in the case of President Eisenhower [33]. However, in a study from the United Kingdom the authors demonstrated one of the highest incidence figures for CD ever reported, which was mainly due to a second peak [9]. The authors expressed concern that this would be the start of a new trend, but further data from the same population have to some extent challenged this [28]. Another area of recent interest is pediatric IBD. The incidence rates among juveniles have been stable for both diseases in different populations since the Second World War. However, during the 1980s there were reports of a substantial increase in juvenile onset of CD in Scotland [34]. An incidence of 1 per 100 000 inhabitants per year in 1968-1976 has changed to 3 per 100 000 in 1990-1992, with the largest increase in the age group 12-16 years. A similar finding was reported from Stockholm, Sweden, where an annual incidence of 1.1 per 100 000 during late 1980s has changed to an incidence of 5.4 during the late 1990s [35]. No such changes in the incidence of UC could be demonstrated. There was also a change in the pattern of the disease appearance, with a higher frequency of fistulating disease than could

The changing faces of Crohn's disease and ulcerative colitis No. of patients 25

20

1956-1961 1962-1967 15

10

5-9

10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 Age groups

Figure 2. The number of patients with Crohn's disease in different age groups for the years 1956-1961 and 1962-1967 respectively according to the year of first diagnosis. (From Norlen BJ, Krause U, Bergman L. An epidemiological study of Crohn's disease. Scand J Gastroenterol 1970; 5: 385-90; permission to publish given by author BJ Norlen).

previously be demonstrated, as well as the presence of children with CD at very young age (below 5 years). This increase in incidence in juvenile CD seems to be confined to Stockholm, as other areas in Sweden have not been able to demonstrate any changes [36].

Gender The sex distribution of UC and CD has also changed faces over time. In the case of UC there is a male predominance in high-incidence areas, most pronounced in patients with ulcerative proctitis or distal colitis [15]. Fig. 3 illustrates the change in sex distribution documented by 59 different descriptive studies over time [37]. In the case of CD there is an opposite trend compared to UC. Mortality from CD is higher among males in low-incidence areas but higher among females in high-incidence areas [10]. Temporal trends in incidence also show a similar

trend, from an even sex distribution to a female predominance in high-incidence areas [15].

Extent of disease Extent of disease is probably the best example of the changing faces of IBD. In the case of UC, pancolitis constituted the majority of cases in some early epidemiologic studies [38] but not all [27]. In the 1960s and 1970s ulcerative proctitis or distal colitis started to emerge to an extent of disease as common as pancolitis. Patients with ulcerative proctitis have a more pronounced male predominance than patients with more extensive disease and also seem to have an older age at onset. It has therefore been discussed to what extent ulcerative proctitis should be a disease entity of its own [39], but it is quite common that patients with ulcerative proctitis will eventually experience a more extensive disease [40,41], but very little is known about the long-term prognosis with regard to extent of disease in patients diagnosed with

M:Fratio

1990 Figure 3. Male.female (IVI.f) ratio versus time period in 59 reported studies of ucierative colitis from 1930 to 1990. (From Tysk C, Jarnerot G. Has smoking clianged tiie epidemiology of ulcerative colitis? Scand J Gastroenterol 1992; 27: 508-12; permission to publish given by author 0 Tysk).

ulcerative proctitis the past 20 years. This is a concern as patients with ulcerative proctitis or distal colitis presently constitute 50% of all patients w^ith UC diagnosed in a defined catchment area [42]. Is this shift toward a less extensive disease only a sign of better ascertainment of cases? This might be, at least partly, the explanation in some studies, but it is not the underlying reason for the results in studies from Belgium and France. The proportions of ulcerative proctitis or distal colitis do not differ from those of other centers, arguing against the theory that the 'differences francaise and/or belgique' should be due to underascertainment. In the case of CD, terminal ileitis was the original disease [6]. It is therefore of great interest to follow the patient's clinical characteristics at the original center over time even if a denominator is lacking [43]. Terminal ileitis was followed by ileocolitis and finally, during the 1960s, a new disease entity emerged - Crohn colitis. This shift can partly be explained by the understanding of Crohn colitis as an entity in the 1960s [44], but even thereafter there has been a decreased proportion of patients with terminal ileitis and an increased proportion of CD with colonic involvement at Mount Sinai as well as in most population-based studies.

Lately, a new entity has been introduced: indeterminate colitis [45]. Is this a new clinical entity or not? It has been argued that this is not the case. Classical CD of the colon is not always clear-cut in a clinical setting. The introduction of this entity could also be due to methodological problems. Most incidence studies published since the Second World War have been retrospective, i.e. data on cases have been assembled years or even decades after diagnosis. One major advantage of such a study design is that the evaluation process will include a long duration of follow-up of the clinical course. This additional information, compared with what is available at the time of diagnosis, will give an additional edge to retrospective studies compared to prospective ones. It has been shown repeatedly that in any cohort of patients with IBD there will be a continuous change in the diagnoses over time, and up to 10% will have a different final diagnosis to that originally assigned [15]. This is also shown by the frequent reports of patients with a definite UC being operated on with a pouch, who later will have a recurrence typical of that for CD [46]. Thus, the pressure to assign a definite diagnosis when a patient presents with a disease, which unequivocally can be classified as IBD, is probably one major reason for this disease entity.

10 In a prospective study from 20 European centers [47], 5% of all patients with IBD were classified as having indeterminate colitis, and there was a wide variation of that percentage in different populations. In Norway, for instance, there were 93 cases of indeterminate colitis compared to 525 cases of UC [29]. In a reevaluation of the Norwegian material 1 year after diagnosis [48], which included endoscopy and new histopathologic examination, 33% of the patients with indeterminate colitis were reclassified with UC and 17% with CD. There were also patients originally classified with UC or CD who were reclassified as indeterminate colitis. There are very few studies which have tried to assess the specific characteristics in patients with indeterminate colitis, but patients with indeterminate colitis seem to have a younger age at onset, a more extensive disease and a more severe clinical course than patients with UC [48, 49]. However, so far there are no compelling reasons to introduce a third clinical entity - indeterminate colitis.

Geographic differences Observational studies during the 1960s and 1970s, both in Europe [50] and northern America [51], suggested a north-south gradient in IBD. As the existence of such a gradient could give new clues to the etiology of both diseases, a major study was undertaken in Europe to test this hypothesis [47]. Twenty European centers identified patients with either UC or CD prospectively during a 2-year period. There was a uniform diagnostic approach in all centers and close collaboration during the study period. Although there was a wide variation of the annual incidence in the different populations, there was no consistent pattern which would substantiate the presence of a north-south gradient. Even after adjusting for tobacco consumption as well as education no substantial gradient emerged. Another interesting feature of IBD is the uniformity in incidence figures in different countries, and these features seem to follow not natural boundaries, but national borders. For instance, until 1 July 2000, there was 10 miles of water easily accessible by ferry between Copenhagen, Denmark, and Malmo, Sweden. In Fig. 4 the temporal trends for CD in these two adjacent cities are shown [20, 52], and it is obvious that the increase in incidence in CD occurred later in Copenhagen than in Malmo, and a similar time difference is present for UC [21, 49].

The changing faces of Crohn's disease and ulcerative colitis In contrast, the incidence of IBD in the northern part of Sweden [53] and the middle part of Sweden [14, 15, 32, 54-56] are similar to the one in Malmo, and likewise the temporal trends. A uniform incidence of IBD also seems to be present in Italy [57, 58]. Comparisons of the incidence of IBD made in Greece during the 1990s, on the other hand, revealed substantial differences, especially for CD which seems to be almost nonexistent in the northern part of Greece [59] as compared to Crete where the incidence does not differ from that of the rest of Europe [60, 61]. However, there are now strong indications of an east-west gradient evident from case reports [62, 63] and incidence figures [64] from Eastern Europe and the former Soviet Union.

Cotiort effects It is obvious from descriptive epidemiology that the increasing incidence in IBD occurred after the Second World War. However, this does not seem to be due exclusively to a period effect, but there are reports of a birth cohort phenomenon. This was first reported as early as the 1970s in a major incidence study of CD in Stockholm County, Sweden [54]. This finding was confirmed in an adjacent population in the Uppsala Health Care Region where a birth cohort phenomenon could be demonstrated both for UC and CD [15]. In the case of UC the birth a cohort phenomenon was more pronounced for extensive colitis than ulcerative proctitis. The existence of such a birth cohort phenomenon has, however, been questioned, and two studies [20, 55], one from Sweden and one from Denmark, failed to show any such effect. Both these studies, however, had problems with statistical power, and that could be the underlying reason for the inconsistent results. In a slightly different approach, mortality data from England and Switzerland were used in order to analyze temporal trends in both UC and peptic ulcer disease [65]. The authors were able to demonstrate similar patterns for both diseases and the results were consistent with a birth cohort phenomenon. The major shortcoming with this study is that the authors have utilized mortality figures instead of incidence figures, and their finding of a peak in the incidence of UC among those born in the latter half of the 19th century is probably a result of this. However, the similarity in the temporal trends between duodenal ulcer disease and UC implicates early exposures, possibly of infectious origin. Moreover, a birth cohort phenomenon argues against genetic determi-

Anders Ekbom

11

Incidence per 100,000 8

Malmo 1958-1973 Copenhagen 1962-1987

I

'"'1

1958

1960

1962

1964

1966

1968

1970

1972

\

1974

1

1

1976

1

1

1978

1

1

1980

\

[

1982

1

1

1984

!

1—

1986

Year Figure 4. Annual incidences of Crohn's disease in Copenhagen County during the period 1962~-1987 and Malnfio during the period 1958-1973.

nants as the major cause, but strongly indicates environmental factors interacting with genetic susceptibility for the diseases.

Risk factors There is no good hypothesis to explain and identify what sort of exposures are of importance for the emerging epidemic of IBD during the 20th century. This is probably the underlying reason for the abundance of different risk factors which have been proposed. With the exception of smoking, they all have in common that the results from different observational studies are not consistent, and quite often lack any biologic rationale. There has been an uniform pattern during the past 50 years: a hypothesis is proposed, followed by enthusiastic early reports of an association, and then a nullifying of the initial results in better-conducted epidemiologic studies. This includes different dietary habits, food items such as margarine and cornflakes, toothpaste.

chewing gum, etc. There are, however, some exposures which are worth discussing more in detail.

Ethnicity In Crohn's original description all 14 patients were Jewish [6]. Ever since that time there has been a perceived association between Jewish ethnicity and an increased risk of IBD. This has been further substantiated by data from the US Army, where the risk of UC in 1944 was higher among those with a Jewish ethnicity compared to other groups [66]. Analytical studies from Sweden [52, 54], South Africa [67], the United Kingdom [68] and the US [69] have given further credence to such an association. However, there are reasons to question the methodology in most of these studies. In the two Swedish studies [52, 54] the nominator - that is the number of subjects with Jewish ethnicity - was assessed through the family name, but the denominator was assessed in a different manner, an

12 approach that would yield an incidence rate which is bound to be inflated. The study from South Africa [67] has similar drawbacks as the denominator was unknown and therefore ascertained from old census data, probably too low, thus generating inflated incidence rates. It is also of interest that descriptive epidemiologic studies from Israel do not find any incidence or prevalence rates, which diff'ers from other high-incidence areas such as the United Kingdom and Scandinavia [47]. The country of origin seems to be a strong predictor for the risk of IBD, and in subgroup analyses of data from Israel, Jews born either in North America or western Europe seemed to be at highest risk [70]. Similar findings have been reported from Los Angeles, US, where the authors could demonstrate that the Jews originating from Central Europe had a higher risk for CD compared to those originating for Poland or Russia [71]. The authors'conclusion that there is a subset of individuals of Jewish origin with an especially high predisposition for CD might be valid, but the suggestion that all individuals with a Jewish background show an increased risk of either UC or CD still remains to be proven. In areas with high incidence of IBD, certain ethnic minorities, especially those with lower socioeconomic status, often have a lower annual incidence. Such associations have been seen among blacks in Baltimore [69], Indian migrants in the United Kingdom [72], blacks in South Africa [67] and Bedouin Arabs in Israel [73]. However, these observations are not constant over time - changing faces - and there are examples of these differences being diminished over time [72, 74, 75]. In some instances, as in Great Britain, even opposite associations have been reported. In a study using data from the National Representative British Birth Cohort utilizing everyone born in Britain during one week in 1970, the authors could demonstrate a more than 6-fold increased risk for IBD among those with a parent from India, Pakistan or Bangladesh compared to the indigenous British population [76]. Similar results were found in a study from Leicester in the United Kingdom, where the mean incidence in the south Asian community was more than double that of those of European descent [77]. Thus, although ethnicity previously was a good predictor for the risk of IBD, this does not seem to be the case today.

The changing faces of Crohn's disease and ulcerative colitis

Socioeconomic status Another consistent finding in early observational studies of both UC and CD was the association between high socioeconomic status and an increased risk [78, 79]. This finding has been used by some authors to explain the lower incidence of IBD in minority groups. However, this accepted truth is another example of the changing faces of IBD, as recent studies in different populations have failed to find such an association [80, 81]. Socioeconomic status as such is of course not a biologically relevant marker. It stands for differences in diet, crowding, and hygiene, exposures that change over time in various ways in different populations. Thus, it is not surprising that the faces have changed, but the question which remains to be answered is: 'What is the underlying mechanism?'

Diet Because both UC and CD mainly affect the gastrointestinal tract, it is natural that diff'erent dietary exposures have been proposed as the main etiologic factor for the increase in incidence. As early as 1925 it was proposed that UC was due to food allergy [82]. Further credence to this has been lent by the fact that an elimination diet has a beneficial effect on quite a substantial number of patients with UC, where avoidance to cow milk was recognized early to have an eff'ect [2, 83]. Exposure to dairy products therefore seems to be of importance for a clinical course of the disease, and it has been shown that patients with more severe UC have decreased levels of lactase compared to other patients with UC [84]. It is, however, highly unlikely that milk or milk products are of any etiologic relevance for either UC or CD, as there is no consistent finding of such an association in different etiologic studies [85]. Besides milk and milk products, an abundance of other dietary compounds have been implicated in different studies. However, in a review of all such studies conducted up to the mid-1980s for both UC and CD [86], the authors were highly critical of those studies and concluded that both the study design and data analysis made it impossible to infer anything from the results presented that far. Studies since then have not improved our understanding, as they face substantial methodologic problems as in most instances they rely on retrospective assembled data. Any researcher dealing with diet and IBD has to face

Anders Ekbom one major issue: the insidious onset of the disease, i.e. symptoms might have changed the dietary habits, and this is probably one of the reasons why exposure to refined sugar, as well as an inverse association with diet rich in fiber, have been implicated repeatedly, especially for CD [86]. One should also refrain from infering a causal association through ecologic data such as the rise in incidence in IBD and the introduction of margarine in the Westernized world [87], a hypothesis which could be refuted by a well-conducted correlation study [88]. Cornflakes [89], fast food [85], and cola drinks [90] are examples of other food items which have been proposed to be casually linked with IBD, but here also the underlying biologic mechanisms remain elusive [85]. In conclusion, although some interactions with diet cannot be ruled out, it is at present highly unlikely that the increase in incidence of IBD is due to any of those dietary compounds mentioned above.

Smoking During the 1950s an Austrian physician made the observation that smoking was less prevalent among patients with UC than among other patient groups [91]. Unfortunately, he used as a comparison group patients with duodenal ulcer disease, where smoking is an accepted risk factor. A report from Sweden during the 1970s, published only in Swedish, reported the same association [92]. It was not until the 1980s that this observation spread to the AngloSaxon medical literature [93]. Since then there has been an abundance of studies which have shown that smoking is protective against UC, and non-smoking or former smoking is associated with an increased risk [94]. This finding is present in both sexes, all age groups and all extents of disease. It has, however, been argued to what extent smoking cessation is an even stronger risk factor than lifelong non-smoking status [56]. The consistency of these results has even led to the initiation of pharmacologic therapy by nicotine, mostly as patches, which in most instances have shown beneficial effects [95]. In contrast to UC, smoking is a risk factor for CD in almost all analytical studies with a more than doubled risk [74], but even higher risk estimates have been reported [96]. There are clear signs of a doseresponse gradient, which further strengthens the hypothesis of a causal association. However, the risk of former smoking does not differ from the risk of never smoking [97], indicating that smoking is perhaps not an initiator but rather a promoting

13 factor. Further credence to this theory comes from the fact that the clinical course in CD seems to differ among smokers compared to non-smokers where smokers seem to have a more severe clinical course [98]. The contradictory findings in UC and CD have led to speculations that smoking may determine the type of IBD that develops in a predisposed individual. This speculation, as mentioned previously, assumes an otherwise common etiology for the two diseases, which is at least partly contradicted both by the genetics and the temporal trends. In the case of CD there is a reasonable biologic model for such an association if one assumes that part of the disease process is a multifocal gastrointestinal infarction which would be potentiated by cigarette smoke [99], similar to what has been proposed in the process of arteriosclerosis. This is also in line with smoking as a promoting factor more than initiating. In the case of UC, so far no biologic model exists, but there are quite a few speculations [100]. To what extent can smoking and changes in smoking habits explain the time trends in IBD? In the case of UC we should keep in mind that the majority of patients are never smokers, thus the introduction of cigarette smoking as a common feature in a population during the 20th century could not explain the emergence of this disease. It has been proposed that the change in sex ratios over time in UC can be explained by a higher frequency of former smokers in the male section of the population compared to the female population over time [37]. The authors speculate that this male predominance will eventually disappear when the same frequency of ex-smokers is present among females. So far, however, there are no signs of such a trend. There are exceptions to the reported associations between IBD and smoking in studies from Israel [101, 102]. It has been proposed that this could be due to a genetic interaction, i.e. higher genetic predisposition to IBD, especially CD, among Jews. The incidence figures for Israel, as mentioned previously, do not substantiate such a predisposition, and even if that were the case, it is rare that a known risk factor will not show up among the genetically predisposed. It is, however, not unlikely that the results from these two studies illustrate the problem with a reported control group in case-control studies.

14

Oral contraceptives The introduction of oral contraceptives in the 1960s among women in the Westernized world and the emerging epidemics of IBD have led some authors to infer a casual association, especially for CD [103]. It was therefore not surprising that in the early 1980s there were quite a few case reports describing an association between the use of oral contraceptives and the occurrence of CD [104]. Since then there have been many analytic studies dealing with the subject but, in contrast to smoking, the results have not been consistent [105]. The underlying biologic mechanism which has been proposed is similar to that for smoking, i.e. multifocal gastrointestinal infarction mediated by chronic mesenteric vasculitis which is aggravated by oral contraceptives [99]. Oral contraceptives would then, similar to smoking, be a promoter and not an initiator. However, in contrast to smoking, the use of oral contraceptives does not seem to affect the recurrence rate or severity of the disease [106]. There is also an absence of a consistent interaction between smoking and oral contraceptive use, and opposite findings showing the presence of such an interaction [107] and the non-presence [108] have been reported. However, oral contraceptives are still the focus of interest, and in a study from the US [109], the relative risk of 5.5 was presented following oral contraceptive use for more than 6 years. The authors also calculated the attributable fraction and concluded that almost 16% of all CD patients in the American population was due to oral contraceptives. There are, however, reasons to be very skeptic of such numbers, especially as the incidence of CD in some populations such as Sweden [15] and Great Britain [28] have not increased since the introduction of oral contraceptives.

Clusters There have been reports of clusters both in time and space for CD [110-112]. This includes CD occurring in spouses [113-115] and affected families in France [116]. These findings have been interpreted as an indication for a transmittable agent. However, in the case of spouses it has so far not been shown that the number of affected pairs differs from that expected. There are always inherent problems in the analyses of clusters. Any researcher has to be aware of the potential presence of a Texan sharpshooter, i.e. the target is decided after the use of the gun. Some of the

The changing faces of Crohn's disease and ulcerative colitis clusters that have been described can possibly be explained by such a bias, but especially from France the report indicates that some transmittable agent is operating. However, so far the search for this agent has failed [116], but it would be premature to rule out an infectious cause, keeping in mind the history of duodenal ulcer disease. There are also quite a few reports of seasonality of both onset and exacerbation in IBD, especially UC, indicating the existence of an infectious agent triggering the disease [117-121]. To what extent this is of etiologic importance remains unanswered, and the less frequent reports of seasonality for CD could be due to this disease having a more insidious onset, which makes it difficult to establish the start of the symptoms.

Transmittable agents Chronic mycobacterial infection was proposed by Dalziel as early as 1913 [4] to be the cause for what would later be CD. The similarities between Johne's disease in cattle, sheep, and goats [122] and CD have been used as one of the major arguments. Johne's disease is due to an early infection by Mycobacterium paratuherculosis and clusters of patients in England [110] and Wales [123] have been attributed to exposure to this agent [124]. It has been proposed that M. paratuherculosis is spread by clinically or subclinically infected animals around the implicated areas, and this could be the underlying explanation for the higher incidence described in urban compared to rural areas. However, it was pointed out in a recent review [125] that morphologic similarities between Johne's disease and CD are superficial and more suggestive of differing etiopathogenesis. Bacteriologic cultures for M paratuherculosis, immunocytochemistry, use of polymerase chain reaction and trials to transmit CD to animals have all been negative or inconclusive [125]. Moreover, trials with antibiotics directed against M paratuherculosis has also been inconclusive or negative [126, 127]. One should also be aware that in areas such as Iceland, where M. paratuherculosis is a problem, especially in sheep, is not characterized by any remarkably high incidence of CD. On the contrary, the increase in incidence of CD emerged later in Iceland compared with other Scandinavian countries [16-18]. Thus, it presently seems very unlikely that M paratuherculosis is an etiologic factor for CD, although one cannot rule out that its presence in patients with IBD can have a

Anders Ekbom clinical impact. Other transmittable agents which could lead to a persistent infection, such as Listeria [128], Mycoplasma [129] and viruses, especially measles [130], have also been suggested. There are, however, no consistent results so far, and the search for a transmittable agent will continue. As the story of H. pylori demonstrates, one should be cautious before ruling out a persistent infection as a cause.

Psychological factors An association between stressful events and both UC and CD has been proposed repeatedly. It is, however, highly unlikely that psychological stress should be the underlying reason for the emerging epidemic of IBD. An extensive review published 1990 of 138 studies of psychiatric factors and their relation to UC, and to some extent to CD also, seemingly refutes such an association [131]. The authors were able to demonstrate that those studies where a positive association was found were particularly likely to be flawed, and those reporting solely systematic investigations failed to find any association. However, stressful events as an aggravating or even initiating factor cannot be ruled out, especially with the emerging understanding of a brain-gut interaction [132].

Appendectomy As early as the 1980s the first report was published which showed that patients with UC were less likely to have been subjected to appendectomy compared to controls [80]. Further research in this area has shown that this protective effect seems to be most pronounced if the appendectomy was done before the age of 20 [133, 134]. There are two mutually exclusive interpretations of these findings. One is that the removal of the appendix is causally linked to a decreased risk of UC, and animal models have given some credence to this [135]. There have even been trials to utilize appendectomy in a clinical setting as an alternative therapy. The other is that there is an association between appendicitis and UC, i.e. patients who will get appendicitis are less likely to succumb to UC [136]. Data from a Swedish study seem to support this latter alternative, as the authors could find no protective effect against UC following the removal of a non-diseased appendix [137]. In Sweden there has been a dramatic decline in appendectomies. For instance, during the past 10

15 years the decrease was close to 20%, from 13 000 to 10 000, and this decline has been most pronounced in the younger age groups [138]. There are reasons to believe that this decline is not due to better diagnostic procedures, but actually reflects a decrease in the incidence of appendicitis. Appendicitis is another disease where the underlying etiology remains an enigma. There was a rise in appendicitis in the late 19th century, which seemed to peak in the 1950s followed by a decline in the Westernized world. It has been hypothesized that the rise in appendicitis could be due to improvements in sewage disposal and water supplies, leading to an enteric infection in childhood at an older age [139]. It is therefore not unlikely that appendicitis, UC and CD are part of the same disease spectrum, perhaps similar to other disease where hygiene in childhood is of importance for the future risk for disorders such as allergy and asthma [140].

Early exposures The hypothesis of early exposures as of major importance for the risk of IBD was introduced in the 1970s by Whorwell et al [141]. In two subsequent studies from Canada [142,143] the authors were able to demonstrate that both early gastroenteritis and bottle feeding was associated with an increased risk for both UC and CD. Early weaning has also been reported as a risk factor [144], but the results have not been consistent [80]. Hygiene in early childhood with access to hot tap water has been impUcated as a risk factor for CD [133, 134], as has being a single child or a first-born [146], which indicates later exposure to infectious events. As mentioned previously, similar associations have been proposed for the increasing incidence of appendicitis during the early 20th century. There has also been a dramatic decrease in perinatal mortality during this century, and this has been proposed to be the underlying reason for the emergence of IBD [147]. Ecologic data (Fig. 5), where perinatal mortality 20-30 years before and incidence figures for CD are compared, are seductive. The underlying hypothesis is t h a t children who previously should have been most vulnerable due to infectious events or an impaired immunologic defense are those who will succumb in a population with high perinatal mortality, and those survivors will be individuals who today will cause the rising incidence. The reasons for the decrease in perinatal

The changing faces of Crohn's disease and ulcerative colitis

16

14 n

12 H

Netherlands

OS 0)

1 France

>s

so 8 o o

XJ = 0.01 for all IBD families to the chromosome 7 region [286]. A mouse chromosome region containing a susceptibility locus for dextran sulfate-induced colitis, Dssc-1 [238], is syntenic with this human chromosome 7 region. For the chromosome 3 region, negative results have been reported for the G protein G-alpha-i-2 gene (GNAI2) [315]. A modest association between markers near the mutL homologue 1 gene (colon cancer, non-polyposis type I; MLHl) and has been observed in a small sample of 45 CD, 36 UC and 45 controls [316]. For chromosome 7 the mucin-3 (MUC3) gene has been tested for association to IBD; alleles of a 51 bp VNTR polymorphism are increased in UC patients from both Great Britain and Japan [317]. This observation of an association from two populations combined with the observation of a reduction of altered mucin species IV in UC patients [318] make this an attractive finding. Chromosome 1 Two studies have observed evidence for linkage to the D1S2670-D1S2682 region [286, 295]. Families with linkage to this region also show greater linkage to IBDl, suggesting that this susceptibility locus interacts with IBDl [286]. Further narrowing of this region to 130 kb near D1S2697 and D1S3669 has been accomplished by homozygosity mapping in a collection of American Iraqi Chaldean families with IBD [319]. Since both CD and UC are observed in these families, these authors have proposed that this locus determines a generic susceptibility to IBD, with the concomitant inheritance of IBDl then determining CD [319].

49

Candidate gene studies In the candidate gene approach, genetic variants are tested based on a prior hypothesis regarding the role of that gene or gene product in the pathophysiology of the disease. Such a hypothesis may be based on evidence from clinical observation or physiologic studies of affected individuals (in-vivo studies), from studies of known disease-related processes (in-vitro studies), from animal models of disease (transgenic construction or targeted disruption), and from the effects of drugs or chemicals on disease in either humans or animals (pharmacogenetic studies). For IBD, possible candidates could be selected from genes involved in the regulation of the balance between Thl and Th2 cells, the response of humans to gut flora, and the release of cytokines during the immunologic destruction and healing of the intestinal mucosa.

The major histocompatibility complex (MHC) The MHC on chromosome 6p21.3 is the most genedense region of the human genome sequenced thus far, with 224 genes identified within 3.6 megabases (Mb) [320]. Since many of the genes in this cluster function in the regulation of the immune system and in antigen processing and presentation, the MHC is a candidate region for most, if not all, diseases with dysregulation of the immune system as the possible underlying pathophysiology. The MHC has therefore long been considered a candidate region for IBD. However, also among these genes are the most polymorphic human proteins known, the class I and class II molecules; and some of these proteins have over 100 allelic variants. This great variation and the high gene density of the MHC have contributed to the difficulty of assigning the specific role of a specific gene to the pathophysiology of IBD. The MHC class II region In general, the class II molecules are dimeric, with an oc-chain and a p-chain that form a groove for presenting an extracellular antigenic peptide to CD4"^ T cells. The three class II molecules are HLA-DP, HLA-DQ, and HLA-DR. In each case the a- and (3chains are encoded by A and B genes, respectively. For HLA-DR there is an invariant HLA-DRA gene and up to three distinct and highly polymorphic HLA-DRB genes. One of these HLA-DRB genes, HLA-DRBl, is present in all individuals and is the most polymorphic. The study of HLA-DRBl genes has been an important tool in the study of the role of

Genetics of inflammatory bowel disease

50

Table 10. Associations of Crohn's disease and ulcerative colitis with HLA class il alleles Serologic

DR1

OR

Genetic allele

DRBrOI DRBr03 DRBr07 DRBIW DRBr0103

DR2

DRB1*15

DRBri501 and DRBr 1502 DRBri502

15% CD, 9% controls not found in ANCA neg UC 17% CD, 11% controls 18% CD, 10% controls 8.6% UC, 3.2% controls 14% UC (16% extensive UC), 23% extraintestinal manifestations 6% UC, 0.2% control increased with extensive UC, colectomy 7.9% CD, 2.2% controls 8.9% UC, 2.2% controls meta-analysis, association with UC

Reference

1.75

0.003 (corr)

1.58 1.9 RR

0.008 (corr) 0.0001 0.0074 < 0.0001

2.9

27.6 33,84

3.9 4.4

0.0002 < 0.0001 0.004 0.001

325 325 326 326 321

A^ T cell reactivity to the bacterial flora with subsequent colitis unless they sustained a triggering event that activated the pathogenic process. The exact relationship of DNP-KLH to the flora is unclear and other types of antigens were not tested. However, CFA contains mycobacterial antigens as an adjuvant and these mycobacterial antigens were not sufficient to trigger disease on their own, suggesting that some specificity was required for triggering of disease. Little is known about how colitis is triggered in genetically susceptible hosts, and this seems to be an excellent model to explore such questions further.

79 as a growth factor for local, extrathymic T cell development in the intestine. To examine the role of IL-7 in mucosal immune homeostasis, transgenic mice were generated that expressed IL-7 under control of the SV40/HTLV1 LTR viral promoter [92]. In at least one of the SRa/ IL-7 transgenic founder lines, clinical evidence of colitis (diarrhea, weight loss, rectal prolapse, and perianal bleeding) was observed at 6-10 weeks of age, although significant variability in the penetrance and age of onset of colitis was found. Intestinal inflammation was most severe in the rectum, although involvement of the ileum and colon was also found. Histologically, the inflammatory lesions were characterized by dense infiltrates of CDA^ T cells, with lesser numbers of monocytes/macrophages, CD8"^ T cells, and neutrophils. Focal erosions were noted, but there was no ulceration. Goblet cell depletion, crypt abscesses, and Paneth cefl metaplasia were also described. The composition of the infiltrating lymphocyte population was dominated by CD4"^ ap T cells. Ex vivo stimulation of 004"*" T cells isolated from colitic lesions demonstrated increased production of IFN-y and IL-2 but decreased production of IL-4 compared to those isolated from control mice [92]. Curiously, increased IL-7 production in the inflammatory lesions of transgenic mice was due to increased expression by infiltrating lymphocytes, not inflamed epithelial cells. Similarly, IL-7R expression was increased on colonic mucosal lymphocytes. Although the mechanism by which dysregulated IL7 production generates colitis/proctitis in this model is unclear, variable expression of disease suggests that environmental cofactors, such as commensal bacteria, may be contributory.

TNF-a 'knockin' (TNF^"^) mice IL-7 transgenic mice IL-7 is a pleiotrophic cytokine with growthpromoting activity for both immature and mature lymphocytes. IL-7 is produced by stromal cells in the bone marrow and thymus, as well as stroma in other organs. IL-7 mRNA and protein expression have been demonstrated in both mouse and human intestinal epithelium and the IL-7 receptor (IL-7R) is expressed by intestinal lymphocytes in both the lEL and lamina propria (LP) compartments [89, 90]. Indeed, 78 lEL T cells are deficient in IL-7- and IL7R-deficient mice [91], and IL-7 has been implicated

Evidence from a number of animal models of IBD supports a central role for tumor necrosis factor (TNF) in the pathogenesis of chronic intestinal inflammation. Human clinical trials have also implicated TNF, and agents that block TNF activity have recently been used with success in treating patients with Crohn's disease [93]. Mice with impaired regulation of TNF biosynthesis that develop chronic, unremitting inflammation in the intestines and joints were recently described [94]. In this model, deletion of a repeated octanucleotide AU-rich motif in the 3'untranslated region of the Tnfa gene results in enhanced mRNA stability and increased T N F

80 production by macrophages and other hematopoietically derived cells. Mutant mice homozygous -AARE/^ ( T N FrAARE/AAREx ^ ^ ^ ^ ^ ^ ^ ) or heterozygous (TNF^^^^^) for the ARE deletion have elevated circulating levels of TNF-a. Homozygous animals fail to thrive and die between 5 and 12 weeks of age. Both homozygous and heterozygous TNF^^^^ mice develop intestinal inflammation, although the tempo of disease development and progression is accelerated in homozygous mice. Disease is localized primarily to the terminal ileum and, less frequently, the proximal colon. Initial lesions consist of villus blunting and broadening that is associated with mucosal and submucosal infiltration of chronic and acute inflammatory cells, including mononuclear leukocytes, plasma cells, and scattered neutrophils. Severe intestinal inflammation is usually observed by 4 weeks of age in homozygous mice and 8 weeks of age in heterozygous mice, with an increased number of submucosal lymphoid aggregates and follicles becoming evident at this time. With disease progression there is transmural extension of inflammation, and in older heterozygous mice (4-7 months), complete loss of villus structure and poorly organized g r a n u l o m a s . Focal skin lesions and severe symmetrical joint disease develop concomitantly with the intestinal lesions. The roles of the two TNF receptor family members in intestinal disease have been evaluated by crossing the T N F ^ ^ ^ ^ mutation into the T N F R I - and TNFRII-gene backgrounds [94, 95]. X N F ^ ^ ^ ^ ^ ^ ^ mice show complete elimination of gut and joint pathology when backcrossed onto a TNFRI-deficient background. Similarly, backcross of the AARE/AARE mice onto aTNFRII-deficient backTNF ground results in significant attenuation in intestinal inflammation, although there is mild mucosal and submucosal inflammation in these mice. This is in contrast to joint disease in the TNFRII background, which is much more aggressive than in the control. Similarly, introduction of the TNF"^^^ mutation into a RAG-1-deficient background lacking both mature T and B lymphocytes results in marked attenuation of intestinal disease but retention of the severe arthritic phenotype, suggesting distinct chronic inflammatory mechanisms in the intestinal and joint tissues. This model supports a central role for TNF in the pathogenesis of chronic intestinal inflammation. Interestingly, there is clearly a gene dosage effect of the mutated TNF allele because homozygous mice develop the clinical phenotype much more rapidly

Experimental mouse models of inflammatory bowel disease than heterozygotes. Studies with this model suggest that excessive or dysregulated production of TNF-oc in response to otherwise normal inflammatory stimuli in the gut results in disease development. This model further provides evidence that TNF-dependent IBD pathogenesis requires the adaptive immune response and is dependent on the presence of both TNF receptor family members (TNFRI and TNFRII). TNFRII is expressed primarily on hematopoietic cells, suggesting that TNF modulation of cells of hematopoietic origin is important in maintenance of mucosal immune homeostasis.

A20-deficient mice Many of the inflammatory responses induced by TNF and IL-1 are mediated by activation of nuclear factor KB ( N F - K B ) , which regulates the transcription of genes of multiple other proinflammatory cytokines. In its inactive form, N F - K B is sequestered in the cytoplasm, bound by members of the IKB family of inhibitory proteins, including IicBa. Stimuli that activate N F-KB, such as TN F-a or IL-1, result in the phosphorylation of IKB, which is followed by its ubiquitination and subsequent degradation. The release of N F - K B following IKB degradation results in its nuclear translocation where it binds to a consensus sequence present in a number of genes, thus activating their transcription. A20 is a cytoplasmic zinc finger protein that inhibits the TNF-ocstimulated activation of N F - K B at multiple points along the TNF-a receptor signaling cascade. Mice deficient in the A20 protein were recently generated by gene targeting [96]. Mice heterozygous for the A20 gene deficiency {A20'^' ) develop normally and show no evidence of pathology. In contrast, homozygous A20-deficient mice (A20 ^ ) were runted as early as 1 week of age and died shortly thereafter. Gross and histological examination of 3 6-week-old A20~^~ mice demonstrated severe, multiorgan inflammation, including severe intestinal inflammation. The inflammatory lesions were characterized by increased numbers of activated T cells, granulocytes, and macrophages. Interestingly, A20~^~ mice bred into a RAG-1-deficient background demonstrated similar multiorgan inflammation, indicating that the adaptive immune response is not required for disease to develop in this model. Similar multiorgan inflammation has been described in iKBa-deficient mice, indicating that attenuation of TNF-oc-mediated inflammatory effects by regulators of N F K B are essential for maintenance of gut home-

Charles O. Elson and Casey T. Weaver

81

ostasis. The mechanism of A20-mediated attenuation of TNF-induced N F K B activation is not yet defined. The effector cell and cytokines mediating the inflammation are not known.

B7-related protein 1-Fc transgenic B7-related protein-1 (B7RP-1) is a costimulatory molecule expressed on antigen-presenting cells, which is related to but distinct from B7 (CD80, CD86). This molecule interacts with its ligand ICOS (inducible co-stimulator) on the surface of activated T cells. Transgenic mice overexpressing B7RP-1 fused to a human IgGi Fc region have been generated [97]. Transgene expression was driven by the apolipoprotein E promoter in the liver. One-third of the transgenic mice developed intestinal inflammation between 5 and 12 months of age, with the most severe involvement in the proximal and distal colon and milder changes in the smafl intestine. In the colon there was transmural inflammation and Assuring ulceration. The effector cell and cytokines are not yet known, but increased levels of serum IgE were found, suggesting a Th2 pathway.

CD40 iigand transgenic mice CD40 Ugand (CD40L) is a member of the T N F family that is expressed primarily by activated T lymphocytes. Engagement of CD40L on activated T cells by its receptor, CD40, provides an important signal for naive T cell activation and differentiation. Conversely, binding of CD40L expressed on the T cell surface by CD40 expressed on B cells during cognate B-T interactions provides a critical costimulatory signal for B cell prohferation, differentiation, and Ig class switching [98]. CD40 is also expressed on other cell lineages, including monocytes and dendritic cells, where it plays an important role in proinflammatory signaling. Engagement of CD40 on monocytes and dendritic cells during interactions with activated T cells induces expression of proinflammatory cytokines including TNF-oc, IL-1, IL-8 and IL-12, as well as various cell surface molecules such as CD54 and CD86. CD40L-CD40 interactions are required for sustained release of IL-12 which is needed to initiate and sustain Thl responses. CD40- and CD40L-deficient mice generated by gene targeting demonstrate profound defects in both cellular and humoral immunity [99]. In contrast, transgenic mice that overexpress CD40L under control of the LCK proximal promoter, which directs transgene expression to T cells and some B cells, demonstrate multiorgan inflammation and morbid-

Table 3. Models wjth perturbations of the epithelium Bacterial flora driven

Strain variation/ genetic modifiers

IFN-Y, IL-1,IL-6, TNF-oc, IL-12

Probable: antibiotics prevent and treat

FVB

120

Unknown

Undefined

Unknown

Unknown

119

CD4"' Th1

IFN-Y, IL-1, IL-6, TNF-a, IL-12

Unnown

129,C3H > (129xB6)F1 B6

Area involved

Effector cell

Key cytokines

Multi-drug resistance genela-deficient

Colon

CD4"' Th1

NCADA chimera

Small bowel, cecum

Gai2-deficient

Entire colon

Model

Reference

122,166 >

Intestinal trefoil factor-deficient (with DSS)

Colon

Unknown

Undefined

Unknown

BALB/c

124

Keratin 8-deficient

Cecum, colon

Unknown

Undefined

Unknown

FVB/N strain; embryonic lethal in129, B6

125

Experimental mouse models of inflammatory bowel disease

82 Table 4. Models with spontaneous intestinal inflammation

Bacterial flora driven

Strain variation/ genetic modifiers

IFN-yJNF-a, IL-1, IL-6,IL-12

Yes

Yes; multiple loci chromosome 3 > 8,1

CDA"" Th1

IFN-yJNF-a

Yes

AKR

110

Unknown

Undefined

Unknown

NHo colitis in other tamarin strains

167

IVIodel

Area involved

Effector cell

C3H/HeJBir

Cecum > colon

CD4^ Th1

SAMP1/Yit

Ileum > cecum, perianal

Cotton-top tamarin

Colon

Key cytokines

ity associated with IBD [100]. Different founder lines show a transgene dose-dependent alteration in thymic development, such that founders with higher transgene copy number demonstrate progressively profound depletion of thymocytes of all subsets, a loss of thymic cortical epithelium, increased thymic B cells in the medulla, and splenomegaly with myeloid hyperplasia. Mononuclear cell infiltrates are found in the lung, liver, pancreas, and intestinal tract. The degree of tissue inflammation, like the degree of disrupted thymic and peripheral lymphoid tissue organization, shows a direct correlation with transgene copy number. In the highest copy number lines there is severe IBD that is associated with earlyonset wasting and diarrhea beginning at 3-6 weeks of age. Morbidity and mortality in high copy number mice appears to correlate with the severity of intestinal inflammation. Inflammation in the intestine was most marked in the colon but extended into the small bowel. In the colon there was multifocal ulceration, which could be transmural and granulomatous in areas. There were large numbers of CD40L"^ T cells in the inflamed intestine as well as large numbers of B cells and activated T cells in the mesenteric lymph node. In light of studies in several mouse models of IBD, wherein blockade of CD40-CD40L interactions is associated with diminished IL-12 and IFN-y production and inflammation [33, 52, 101], one can speculate that constitutive, high level expression of CD40L by peripheral T cells results in increased proinflammatory cytokine expression by cells of the innate immune system. As yet no characterization of the effector cell or cytokines in the inflammatory lesions of CD40L transgenic mice has been reported.

Reference 102,105

Similarly, it is unknown whether the bacterial flora plays a role in the inflammation seen in this model.

Models of spontaneous IBD Colitis has been reported to occur spontaneously in animals, but this has been sporadic and uncommon. There are now several instances in which colitis has occurred in high frequency in a strain of animals: C3H/HeJBir mice, S A M P l / Y i t mice, and the cotton-top tamarin (Table 4). The cottontop tamarin is an interesting model but little is known about the pathogenesis. The interested reader is referred to an earlier review [1].

C3H/HeJBir mice C3H/HeJBir is a substrain of C3H/HeJ mice which was generated by a program of selective breeding at the Jackson Laboratory. These mice are highly susceptible to the development of colitis, and under certain housing conditions can develop colitis spontaneously. In the original studies, the spontaneous colitis was localized mainly to the cecum and right colon of young mice, with onset in the third to fourth week of life. Colitis was usually mild, resolving by 10-12 weeks of age. The lesions were focal, mainly in the cecum but sometimes extending into the right colon. There was acute and chronic inflammation, ulceration, crypt abscesses, and epithelial regeneration, but no thickening of the mucosal layer or granulomas [102]. A comparison of immunologic cell types and their functions between C3H/HeJBir and the C3H/HeJ strain reveals few differences in the systemic com-

Charles O. Elson and Casey T. Weaver partment. Both have the Lps^ mutation, now known to be a mutation of toll-Hke receptor (TLR) 4, which renders them unresponsive to bacterial endotoxin. The differences that have been found are compatible with a potential defect in mucosal immune regulation including increased secretory IgA, high titer serum antibodies to commensal bacterial antigens, and increased T cell responses to orally delivered antigen in C3H/HeJBir mice [103]. lEL of C3H/ HeJBir mice have been reported to have increased cytotoxicity and homotypic aggregation due to increased expression of Py integrins [104]. C3H/HeJBir mice have no significant B cell or T cell reactivity to food or intestinal epithelial cell antigens, but do have strong B cell and T cell responses to cecal bacterial antigens. Serum IgG antibodies, particularly IgG2a, bind to a set of antigens on Western blots of electrophoresed homogenates of cecal bacteria. This antibody reactivity is highly selective, recognizing a limited but reproducible subset of antigens from the thousands of proteins present [44]. These antibodies probably do not play a major role in disease pathogenesis in that most mice develop them after colitis has resolved. C3H/HeJBir €04"" T cells demonstrate strong proliferative and cytokine responses to cecal bacterial antigens which are detectable as early as 4 weeks of age. This T cell response is directed at protein antigens and is MHC class II restricted. The cytokines produced are mainly IFN-y and IL-2, consistent with a predominantly Thl response. Adoptive transfer of bacterial antigen-activated €04"^ T cells from C3H/HeJBir but not from control C3H/HeJ mice into histocompatible C3H/HeSnJ Scid/Scid recipients induces focal ulcers and colitis mainly in the cecum, similar to what has been found in the C3H/HeJBir donors [105]. This formally demonstrates that CD4^ T cells reactive with conventional antigens of the enteric bacterial flora can mediate chronic IBD. The number of antigens stimulating this T cell response also appears to be limited in that a restricted number of TCRp chain gene famihes is expanded after antigen stimulation in vitro [101]. A number of CD4"^ T cell lines reactive with enteric bacterial antigens has been generated in vitro. Most of these are predominantly Thl and cause colitis in all recipients after adoptive transfer into C3H SCID recipients. High levels of IFN-y and IL12 are found in colon explant cultures of colitic mice. The sustained IL-12 production in the mucosa that is required for disease to occur depends on CD40LCD40 interactions, in that antibodies to CD40L

83 block IL-12 production in vitro and prevent colitis in vivo [101]. A bacterially reactive CT>A^ T cell line with properties similar to Trl cells has also been generated. This cell line is able to block the development of colitis by a pathogenic Thl cell line when both are transferred together into SCID recipients [106]. Again, environmental factors appear to be important in disease expression, in that the expression of the phenotype of spontaneous colitis is reduced or eliminated when the mice are housed in SPF conditions. However, C3H/HeJBir (and C 3 H / H e J ) remain very susceptible to several forms of colitis, including both TNBS-induced and DSS-induced colitis [107]. In addition, C3H/HeJBir carrying the IL-10 null mutation contract severe disease in the first weeks of life [46]. Thus C3H/HeJBir mice have a high susceptibility for colitis, possibly due to defects in mucosal immune regulation, that can express spontaneous colitis when stressed by certain conventional housing conditions.

SAMPI/Yitmice Another model of spontaneous intestinal inflammation is the recently described SAMPl/Yit mouse [108]. The SAMPl/Yit strain is a subline of the SAM (Senescence Accelerated Mice) PI strain, which was derived by extended sibling mating of AKR/J mice. SAMPl mice are so named because of their shortened life span (approximately 9 months) and propensity to develop early senescence associated with spontaneous amyloidosis, alopecia, and osteoporosis [109]. The SAMPl/Yit subline of the SAMPl strain was derived by selective sibling breeding of SAMPl mice with spontaneous skin ulcerations [108]. Unlike the parent SAMPl strain, the SAMPl/Yit strain shows no shortened hfe span or features of early senescence, but does develop a spontaneous enteritis and cecitis under SPF conditions. Unlike most of the IBD models described in this chapter, the intestinal inflammation in SAMPl/Yit mice is unusual in its preferential development in the small intestine; the absence of colonic involvement is characteristic. Extraintestinal manifestations of disease include focal lymphocytic infiltrates in the liver and inflammatory skin lesions on the dorsum and eyelids. Animals maintained under SPF conditions develop discontinuous inflammatory lesions of the distal small intestine as early as 10 weeks of age, and show 100% penetrance by 30 weeks of age [110].

84

Experimental mouse models of inflammatory bowel disease

Grossly, there is thickening of the distal ileum with focal stricture formation. Histologically, the disease is most severe in the terminal ileum with progressively diminished inflammation in the proximal ileum and variable, mild involvement of the cecum. There is no significant involvement of the jejunum, duodenum, stomach, colon, or rectum. Early lesions are characterized by neutrophilic infiltrates of the epithelium, particularly that over Peyer's patches or preexisting lymphoid follicles, an injury that resembles aphthous ulceration in human IBD. Mature lesions are characterized by villus atrophy and crypt hyperplasia with transmural inflammatory cell infiltrates. There is a mixed inflammatory cell infiltrate composed of mononuclear cells (lymphocytes, macrophages, and plasma cells) and neutrophils. Loose aggregates of macrophages consistent with granuloma formation and focal microabscess formation have been described in mature lesions. In more advanced disease there can be development of mucosal fibrosis with budding and branching of glands, as well as muscular and neural hyperplasia. Phenotypic analyses of inflammatory cell infiltrates by immunohistochemistry and flow cytometric analysis have demonstrated increases in activated T cells [110]. There is an increase in the ratio of CDSoc"^ I EL T cells bearing ocp versus y5 TCRs. Disease can be transferred into MHC-matched SCID mice by unfractionated or €04"*" T cell-enriched mesenteric lymph node (MLN) cells from 30-week-old SAMPl / Yit mice in a dose- and time-dependent manner. The cytokine profile of MLN cells from SAMPl/Yit or adoptive transfer recipient mice is characterized by increased production of IFN-y and TNF-a. Treatment of adoptive transfer recipients with neutralizing antibodies to either TNF-a or IL-12 suppressed disease incidence and severity. Thus, Thl-type cytokines play a pathogenic role in this model. Administration of antibodies to adhesion molecules that have been implicated in T cell homing and neutrophil trafficking (E- and P-selectin and oc4-integrin) also attenuates disease development in SAMPl /Yit mice. Development of terminal ileitis in SAMPl/Yit mice requires the endogenous flora. Germfree SAMPl/Yit mice fail to develop intestinal inflammation, whereas germfree SAMPl/Yit mice reconstituted by transfer of the flora from SPF SAMPl/ Yit mice developed disease as early as 15 weeks postconventionalization [108]. These results suggest that T cell reactivity to bacterial antigens localized in the terminal ileum, presumably the commensal flora, may be responsible for disease pathogenesis. It is

unknown whether disease progression represents continued reactivity to antigens provided by the commensal flora or rather represents cross-reactivity to self antigens expressed in the terminal ileum.

Models with perturbations of tlie epitlielium The intestinal epithelium is recognized increasingly as an active partner in the mucosal immune system. Epithelial (lEC) cells produce and respond to a wide variety of cytokines and express molecules able to interact with lymphoid cells. Epithelial cells signal the innate immune system by rapid release of chemokines upon bacterial invasion [111, 112].The epithelial layer is exposed to and interacts also with luminal bacteria. I EC express a number of toll-like receptors (TLR), proteins which bind to and thus recognize certain classes of microbial products based on molecular patterns. There are some 9-10 TLR genes identified to date, and each appears to recognize a diff'erent assortment of microbial molecules, e.g. TLR2 binds CpG nucleotides, TLR4 binds lipopolysaccharides, and TLR5 binds flagellins. Epithelial cells alter the expression of some genes based on the composition of the bacterial flora, and so are able to detect changes in the flora, perhaps via TLR or other pattern-recognition receptors. Thus, one could more accurately view the dynamic interactions at the mucosal surface as a bacterial-epithelial-lymphoid circuit with each component communicating with the others (Fig. 4). It is of some interest that NOD2, the first gene identified as conferring susceptibility to Crohn's disease, appears to be a microbial patternrecognition receptor [113, 114]. The epithelial hypothesis of the pathogenesis of IBD proposes that abnormal epithelial cell function can result in chronic intestinal inflammation even in a host with a normal immune system. The inflammation is viewed as secondary to the epithelial abnormality. Support for this hypothesis comes from the observations that some patients with Crohn's disease have increased intestinal permeability for small molecules [115] and that enterocytes from patients with IBD demonstrate abnormal stimulation of allogenic T cells due to deficient expression of a glycoprotein (gp 180) on their surface [116]. Further support for this hypothesis comes from experimental animal models in which the major abnormality appears to be epithelial. The best example of this is the mJr7a-deficient mouse, as

Charles O. Elson and Casey T. Weaver

85

Bacteria Epithelium Altered gene expression Induction of cytokines Modulation of tight junctions

Lymphoid cells Altered cell types Induction of T cells, B cells Activation of Te/ Tr cells Induction of cytokines Figure 4. Bacterial-epithelial-lymphoid circuit. Enteric bacteria and their products can act on epithelial cells to alter epithelial gene expression and function. The epithelial cells in turn are interacting with lymphoid cells within the epithelial layer itself and in the lamina propria. Mucosal lymphoid cells can influence the bacterial flora through the production of cytokines and immunoglobulins such as IgA. Thus there is a very dynamic dialog or circuit occurring at the mucosal surface. From this viewpoint it is not difficult to see how abnormal epithelial function could perturb this dialog and skew the lymphoid response into an inflammatory one.

discussed below. However, this hypothesis is not incompatible with the immunological hypothesis discussed above, particularly when one views the epithelial layer as an active player in a dynamic bacterial-epithelial-lymphoid circuit at the mucosal surface. From this perspective it is not surprising that mutations that affect or target epithelial cells can result in perturbation of the mucosal immune system and thus eventuate in chronic intestinal inflammation (Table 3). Providing a barrier is one important function of the epithelium, but only one of them [117]. Although increased permeability is generally assumed to equate to increased mucosal immune reactivity and thus chronic intestinal inflammation, direct proof that this is true is lacking. The chronicity of the altered barrier function is likely an important variable. For example, instillation of acetic acid into the colon of mice, which interrupts the mucosal barrier for days, results in inflammation and increased mRNA for IL-1, TNF, and IL-6 but no

up-regulation of T cell cytokine mRNA despite the presence of memory T cells in the lamina propria [118]. The best support for a role of increased permeability leading to IBD comes from the Ncadherin dominant negative mutant chimera, discussed below, and in this model the inflammation takes months to develop and remains histologically mild [119]. Given the advances in understanding of the many roles epithelial cells play in mucosal immune homeostasis, concepts of their potential role in intestinal inflammation need to be expanded beyond alterations in permeability.

Multi-drug resistance gene 1a (/ntfr7a)-deficient mice The murine multiple drug resistance gene, mdrla encodes a 170 kDa transmembrane transporter protein that is expressed by intestinal epithelial cells, as well as a subset of lymphoid cells and hematopoietic

86

Experimental mouse models of inflammatory bowel disease

cells, mdrla is one of a family of transporters known as ATP-binding transporters that are characterized by their ability to transport small amphiphilic and hydrophobic molecules across cell membranes in an ATP-dependent manner. Three mdr genes have been identified in rodents, each with a restricted pattern of tissue expression and presumably distinct functions. Two mdr genes have been identified in humans. Originally defined on the basis of their ability to confer resistance to chemotherapeutic agents in neoplasia, thus the name, the physiologic function of these transporters is not known. Mice with a targeted deletion of the mdrla gene were recently generated on a FVB background; 2-025% of m B6 > NON; multiple genes [107]

127

CD4'' Th1

IFN-y, IL-12,TNF-a

Probable

SJL, C3H, BALB > B6 > DBA/2

168

Colon

CD4"' Th2

IL-4, IL-5,TGF-P

Unknown

SJL

169

Acetic acid

Colon

Unknown

IL-1,IL-6,TNF-a

No

BALB/c

118

Indomethacin (rat)

Small intestine

Unknown

Undefined

Yes

Lewis > Fischer

170 171

Peptidoglycanpolysaccharide (rat)

Injection site: ileum, cecum

T cells

IL-1,IL-6,TNF-a

Yes

Lewis, S-W susceptible; Buffulo resistant

172

Area involved

Effector cell

Key cytokines

Dextran sulfate sodium

Colon

Unknown

TNBS/ethanol

Colon

Oxazalone

Model

Reference

Charles O. Elson and Casey T. Weaver

Dextran sulfate sodium (DSS)-induced colitis Addition of 30-5 kDa DSS to the drinking water at 3-10% will induce colitis in hamsters, rats and mice [127], which is manifested by bloody diarrhea, weight loss, shortening of the colon, mucosal ulceration and neutrophilic infiltration. In Swiss-Webster mice a chronic colitis can be induced by feeding multiple cycles of DSS [128]. Such chronic lesions demonstrate increased lymphoid cells, prominent lymphoid aggregates in the lamina propria and serosa, a patchy distribution of inflammation and fissuring ulceration. Prolonged low-dose feeding of DSS has resulted in colitis, dysplasia and colon cancer in hamsters [129] and rats [130]. The earliest change of acute DSS-induced colitis is a progressive noninflammatory dropout of crypts [128], suggesting a primary effect on epithelial ceUs. Indeed, DSS inhibits proHferation of mouse epithelial cefls in vitro [131]. Early lesions occur mainly in the left colon and over lymphoid aggregates. In established lesions multiple inflammatory mediators are increased, including thromboxane, leukotriene B4, IL-2, IL-4 and IL-6. Colitis occurs in SCID mice fed DSS, indicating that the T cells and B cells are not required for acute coUtis to develop [131]. Luminal bacteria may play a role in the pathogenesis of these lesions in that colonic concentrations of Bacteroides spp., especially B. distasonis, are increased in acute and chronic phases of inflammation [127]; concomitant metronidazole therapy prevents DSS colitis [129]. The DSS model has some advantages and some limitations. It is a fairly simple method of inducing damage in the colon of most strains of mice. The lesions are fairly reproducible and the clinical and histologic severity can be quantitated. Because of these features, DSS colitis has been used for screening of potential therapeutic agents, and a large number of them have shown benefit in this model, indicating it is a sensitive screening system. There are also limitations of this model, mainly that it represents a nonspecific injury model that does not require either T cefls or B cells; thus, it is not well suited to address immunologic or therapeutic issues involving the acquired immune system. The significant and substantial variations in DSSinduced colon lesions among inbred strains of mice have been shown to represent a quantitative genetic trait [107]. The reproducibility of lesions produced by DSS has allowed the identification and mapping of genes important in susceptibility to colitis, genes that are likely to be of general importance in maintaining

89 mucosal homeostasis. The location of a number of candidate genes has been identified through quantitative trait locus mapping [132]. Interestingly, the same pattern of susceptibility and resistance among strains found with DSS is seen in a number of induced mutant mouse models (Tables 1-6), suggesting either that the same set of genes or different genes affecting similar pathways may be involved.

Trinitrobenzene sulfonic acid (TNBS)/ethanol enema-induced colitis This model is a hybrid one involving both chemical damage and T cell immune reactivity. The administration of an enema containing the contact sensitizing agent TNBS in 50% ethanol induces colitis in rodents. The ethanol breaks the mucosal barrier and is a crucial component; no colitis ensues if TNBS alone is given [133]. Because TNBS is a covalently reactive compound, its administration results in acute necrosis of the wall of the distal colon due to oxidative damage. The occurrence of such necrosis appears to be an important factor for the development of colitis, which may explain why the effective dose of TNBS in mice is close to the lethal dose. In a few strains of rats and mice a single enema can result in prolonged chronic colitis [133, 134]. However, in most strains of mice multiple enemas are required to generate chronicity and the duration of colitis following administration is a matter of days. Discrete foci of acute necrosis and inffammation shortly after the enema are followed by submucosal chronic inflammation that lasts a variable amount of time. In mice, TNBS colitis appears to be a classic delayed-type hypersensitivity response mediated by T cells responding to 'hapten-modified self antigen'. The latter is formed by the covalent attachment of the hapten, trinitrophenyl (TNP), to self peptides. Such reactions in mice are under complex regulation by T cefls and B ceUs. The colitic mucosa is infiltrated with CD4'^ T cefls [134] and increased IgG- and IgAproducing B cells [135], the latter being reminiscent of the changes in plasma cells that occur in the mucosa in human IBD [136]. There is a marked increase in B cells producing IgG anti-TNP in the inflamed colon, including the IgGi, IgG2a, and IgG2b subclasses. Mucosal CD4"^ T cells produce increased amounts of IFN-y and IL-2, but not IL-4 [134], consistent with a Thl eff*ector response. Administration of monoclonal anti-IL-12 to mice with TNBS/ethanol

90 colitis significantly prevents or treats the colitis [134], indicating that the interaction between antigen-presenting cells producing IL-12 and T cells producing IFN-y contributes to both the induction and progression of colitis. Antibody blockade of interactions of CD40L (CD 154) prevents induction of TNBS colitis by reducing IL-12 production [137]. Furthermore, other agents that can inhibit IL-12 production, such as IL-10 [138] or an anti-sense oligonucleotide to the transcription factor N F K B p65 [139], have similar beneficial effects. A significant finding originating from the TNBS model is that regulatory cells able to inhibit colitis can be manipulated. Contact allergens are classic oral tolerogens and thus TNBS itself [135] or TNPconjugated tissue homogenates [11] have been fed to mice to induce oral tolerance prior to the induction of colitis. The resulting colonic inflammation was less severe, the mucosal IgG anti-TNP cells were markedly reduced, and the cytokines produced by mucosal CD4"*' T cells were altered with production of less IFN-y and more TGF-p,, IL-10, and IL-4. Mucosal production of IL-12 was also reduced. Administration of antibodies to TGF-P reversed the protective effects of TNP-colon homogenate feeding in one study [11], suggesting that the regulatory cells induced by feeding were producing TGF-p.

Oxazolone/ethanol-induced colitis Similar to the induction of colitis initiated by TNBS administration, administration of the contact sensitizing agent oxazolone in 50% ethanol as an enema also induces distal colitis in mice. In contrast to the colitis initiated by TNBS administration, however, colitis induced by high-dose oxazolone given in an ethanol enema has quite distinct clinical, histopathologic and immunologic features. Colonic instillation of 6 mg oxazolone in 50% ETOH as an enema results in rapid onset of distal colitis, diarrhea, and weight loss in SJL mice which peaks on day 2. Approximately 50% of the animals dosed died by day 4. Surviving animals showed progressive clinical improvement with apparent complete resolution by 10-12 days. Histologically, oxazolone-induced lesions were confluent and characterized by superficial inflammation of the mucosa associated with severe epithelial loss, focal ulceration, and severe edema of the submucosal layers. The inflammatory infiltrate was mixed, composed predominantly of lymphocytes and neutrophils with a limited number of eosino-

Experimental mouse models of inflammatory bowel disease phils. Severe goblet cell depletion and reduction of the density of tubular glands was noted. No crypt abscesses, crypt elongation, g r a n u l o m a t o u s inflammation, or fistulas were found. Examination of the cytokine profiles of T cells isolated from the lamina propria of distal colon at day 2 demonstrated elevations of both IL-4 and IL-5 in CD3/CD28-stimulated T cell cultures, with no change in IFN-y production. TGF-P was also markedly elevated in both spontaneous and stimulated cultures. Interestingly, levels of TGF-P were even higher in T cells isolated from uninvolved proximal segments of colon of diseased animals, as were IL-4 levels. A pathologic role for IL-4 in oxazolone/ ethanol colitis was demonstrated by neutralizing antibody studies. Animals that received a single neutralizing dose of IL-4 at the time of oxazolone/ ethanol administration displayed an abbreviated and attenuated transient weight loss and only minimal inflammation. In contrast, treatment of mice with neutralizing antibodies toTGF-P or IL-12 resulted in more severe disease and weight loss. Tissues from anti-TGF-P- and anti-IL-12-treated mice demonstrated a pancolitis instead of the distal colitis that was characteristic of control animals. Both antiT G F - P and anti-IL-12 treatment resulted in sustained elevation of IL-4 production in diseased colons. The oxazolone/ethanol colitis model represents an intriguing contrast to the TNBS model, and is unusual in its expression of a Th2 pattern of response. The mechanistic basis for the very different patterns of disease induced by oxazolone and TNBS is unknown, but promises to provide insights into antigenic features that may predispose to dysregulated Thl- or Th2-polarized intestinal responses. An important caveat regarding this model, however, is the role of antigen-induced Th2 cell responses in their development. The time course of disease development is very rapid compared to the inductive phase typically required for Thl or Th2 development. It remains possible that this model represents a preferential recruitment of pre-committed Th2 cells, rather than induction of polarized effectors from naive intestinal T cell precursors. Also, the basis for the resolution of lesions in the oxazolone/ethanol colitis model is unknown. Given the exacerbation of disease that accompanies neutralization of TGF-p, it is speculated that this pleiotropic cytokine may play a central role in rapid attenuation of the inflammatory response. Further studies will be needed to define the role of TGF-p in this model, and to

Charles O. Elson and Casey T. Weaver

91

Table 6. Models as yet unclassified Bacterial flora driven

Strain variation/ genetic modifiers

IFNy, IL-1,IL-6, TNF-oc

Yes

Undefined

173

? Tcell

Undefined

Unknown

Undefined

144

Cecum, colon

Unknown

Undefined

Unknown

BALB/c

146

Cecum, colon

CD4"'

Undefined

Probable

B10,B6

145

Model

Area involved

Effector cell

Key cytokines

HLA-B27 transgenic rat

Entire intestine, other organs

Tcell CD4"'/CD8+

Wiskott-Aldrich syndrome protein deficient

Colon

Fucosyltransferase transgenic Pigeon cytochrome c T cell receptor transgenic

determine whether TGF-p may play a role in suppressing the Th2 response.

Models as yet unclassified There remain a number of models that cannot be classified easily into the other categories as yet, mostly because there are insufficient data available (Table 6).

HLA-B27/p2M transgenic rat Because the human HLA-B27 class I M H C molecule is strongly associated with ankylosing spondylitis and the spondyloarthropathies, transgenic rats expressing the human HLA-B27 and (32-microglobulin genes were generated. Certain of these transgenic rat lines develop a multiorgan disease manifested by colitis, arthritis, orchitis, and psoriasiform changes of the skin and nails in the absence of any further manipulation [140]. These features resemble the human spondyloarthropathies. Rats transgenic for other class I molecules such as HLA-A2 or HLA-B7, have not developed inflammatory disease. In the most susceptible 21-4H fine, disease occurs in all animals surviving past 10 weeks of age and is manifest primarily by watery diarrhea (Table 6). A diffuse enteritis with mononuclear cell infiltrate is present that is variable in the stomach and small intestine, but prominent in the colon where it is associated with hyperplasia of crypts and mucin

Reference

depletion. Crypt abscesses or transmural inflammation are uncommon. The mechanisms by which the human B27 molecule induces this disease are unclear [140]. The HLA-B27 transgene is not expressed on gut epithelial cells but is present on antigen-presenting cells. CD4'^ T cells are abundant in the intestinal lesions. Disease can be transferred with transgenic bone marrow and a critical role for T cells in disease is demonstrated by the observation that athymic (nude) rats bearing the transgene fail to develop disease, although bone marrow cells from these nude rats can transfer disease [141]. These results suggest disease requires B27 expression on antigen-presenting cells which then interact with host T cells. Interestingly, colitis and arthritis do not occur when B27 transgenic rats are raised under germfree conditions [142] but both develop when the flora, particularly Bacteroides spp., are restored [143].

Wiskott-Aldrich syndrome protein (WASP)deficient mice The Wiskott-Aldrich syndrome is an immunodeficiency disease due to mutations in a gene on the X chromosome encoding a cytoplasmic protein, WASP, that is expressed in lymphocytes and megakaryocytes. WASP appears to function in cell signaling and cytoskeletal interactions. WASP-deficient mice exhibit normal lymphocyte development in lymphoid tissues but have decreased numbers of lymphocytes and platelets in the blood. T cells from

92

Experimental mouse models of inflammatory bowel disease

WASP-deficient mice proliferate poorly after stimulation via the TCR which may be related to deficient receptor clustering or 'capping' which was also observed. Although colitis is not a feature in humans with WAS, most of the WASP-deficient mice developed acute and chronic colitis by 4 months of age. Large numbers of CD4"^ T cells infiltrated the mucosa of colitic mice. The effector cells mediating the colitis and cytokines involved remain unknown [144].

Lymphopenic T cell receptor transgenic mice One report described an unusual form of spontaneous colitis that arose in a subset of T cell receptor transgenic mouse lines [145]. In this study two lines of mice with rearranged, transgenic a and P TCR chains specific for the antigen cytochrome C (5C.C7 and -D TCR transgenic lines). When these TCR transgenes were crossed onto a SCID or RAG background, coHtis developed early in life and was characterized by marked mucosal hyperplasia, crypt elongation, and mixed inflammatory cell infiltrates with a predominance of mononuclear cells. Focal mucosal ulceration and crypt abscesses were also identified. Lesions were more severe in the descending colon and rectum with progressively diminished disease in the cecum and ascending colon. The stomach, small intestine, and liver were uninvolved. Intestinal inflammation in each of these transgenic lines seemed to correlate with low numbers of circulating CD4'^ T cells, suggesting that lymphopenia played a role. Curiously, the dominant T cell population found in colon lesions expresses the transgenic (3 chain, but not the transgenic a chain (Tgoc ). Presumably the transgenic (3 chain is paired with nontransgenic, endogenous oc chains. These cells have an activated phenotype (CD45RB^ CD44^\ CD69^^) and are expanded both in the intestine as well as in extraintestinal lymphoid tissues. The polyclonal nature of the pathogenic Tga" cell population correlates with 'leakiness' of endogenous oc chain rearrangement during thymic development in these mice, such that a small circulating population with potential cross-reactivity to intestinal antigens may be present. The ability of endogenous a chains to undergo rearrangement in the scid and ragr'~ background is unusual. The role of a relative lymphopenia in these mice and its association with development of colitis is unknown. It is speculated that crossreactivity with enteric bacterial antigens in mice with a 'brittle'T cell

repertoire may predispose to colitis, perhaps due to the absence of sufficient regulatory T cells.

Fucosyltransferase transgenic mice A primary abnormality in the glycosylation of colonic mucins resulting in a defective mucosal barrier function is hypothesized as a mechanism by which genetic factors may be involved in the etiology of ulcerative colitis. In a recent study, mice transgenic for al,2-fucosyltransferase (H transferase, HTF) were observed to spontaneously develop colitis with a mixed acute and chronic inflammatory infiltrate and crypt abscesses, similar to human ulcerative colitis [146]. The inflammation was limited to the colon and cecum. Changes in mucin glycosylation were demonstrated; there is no evidence of impaired immunologic function in these transgenic mice. On lectin histochemistry there was increased peanut agglutinin staining in the colonic mucus layer of pre-colitic HTF mice compared with wild-type littermates. Further, monosaccharide analysis of purified colonic mucins demonstrated marked decreases in galactose and sialic acid in non-diseased HTF mice compared with wild-type littermates. The incidence of colitis was increased with successive backcrossing onto a BALB/c background. These studies suggest a novel model of ulcerative colitis based on genetic modification of colonic mucin glycosylation.

Discussion The inflammatory bowel diseases represent a complex interplay among environmental, genetic, and immunologic components. New insight into each of these factors has been generated from experimental models. Some facets of what has been learned in each of these areas will be considered in turn.

Environmental factors The major environmental factor revealed in the different experimental models is the enteric bacterial flora. On the luminal side of the epithelial layer there are a large number of bacterial species, estimated at lO^'* organisms per gram of stool. On the other side of the epithelium are a large number of immune cells and molecules. The concept of a microbial-epithelial-lymphoid circuit with all components in communication with one another has already been presented, as have examples in which perturbations of

Charles O. Elson and Casey T. Weaver either lymphocytes or epithehum can lead to chronic intestinal inflammation (Tables 1-3). It is unknown whether abnormalities of the flora alone can lead to disease in a similar manner, but this is possible. An altered or abnormal flora has been raised as a possible explanation for the rapid increase in immune-mediated diseases, including IBD, in the Western world. Sometimes called 'the hygiene hypothesis' this theory postulates that early childhood exposure to microbes primes the immune system to respond in certain ways later in life. In the Western world such priming either does not occur or is abnormal and results in a variety of immune diseases occurring later in life, including autoimmunity, atopy, asthma, and IBD. One factor that has changed substantially in the Western world is the bacterial flora, and a leading hypothesis is that this altered flora is responsible for the abnormal priming of the immune system early in life and thus later immune-mediated diseases. This theory is being tested by interventions to alter the bacterial flora early in life, and initial results indicate a 50% reduction in atopy in young children from atopic families who were given a probiotic flora as newborns [147]. The microbial flora is very complex and comprises hundreds of different organisms [148]. Many more have never been cultured but are being identified by new molecular methods that can identify novel organisms without having to culture them. The bacterial flora is the major stimulus to the development of the intestinal immune system. For example, germfree animals have little or no intestinal lymphoid tissue or IgA despite exposure to a variety of food antigens. When germfree animals are reconstituted with an intestinal flora they develop abundant mucosal lymphoid cells along the length of the intestine, and secretory IgA is produced in large amounts [149]. However, not all organisms among the flora are equally able to stimulate the mucosal immune system [150]. One example is segmented filamentous bacteria (SFB) which colonizes the intestine of a number of species including both mice and humans [151]. Colonization of germfree mice with SFB alone stimulates a large mucosal immune response, one which is greater in extent than colonization with a more complex bacterial flora not containing SFB [152, 153]. Helicobacter comprises another species that appears to be immunostimulatory in the intestine as a member of the flora, and some Helicobacter strains can contribute to colitis in selected immunodeficient strains. In most mice with an intact immune system, Helicobacter species are

93 commensals and do not cause disease. At the other end of the spectrum there may be organisms which are delivering inhibitory signals to the mucosal immune system [154, 155]. An interesting recent observation is that a non-pathogenic Salmonella species was able to inactivate N F - K B signaling in epithelial cells, by delivery of flagellin into them, thereby down-regulating inflammatory cytokine production [156]. If this result extends to commensal bacteria it suggests that some bacteria may be actively inhibiting mucosal immune responses. Indeed, certain bacterial strains have been classified as 'probiotics', i.e. bacteria that promote health. Probiotic bacteria are being tested in both experimental models and in humans with IBD with some promising early results [157, 158]. In the future, the probiotic approach may be taken one step further by genetically altering commensal bacteria to express cytokines. This has been done in a recent report in which Lactococcus were genetically altered to express IL-10 [159]. These organisms were able to treat colitis in two different experimental models. As we learn more about the bacterial flora, its manipulation may represent an effective treatment modality, particularly for maintaining remission of disease. The variation in immune stimulating capability among organisms extends to their antigens. For example, in the colitic C3H/HeJBir mouse only a small fraction of the total enteric bacterial proteins present in the cecum are recognized by serum antibody, and this set of antigens is highly reproducible among individual members of this strain [44]. Similarly, a selective IgG reactivity has been found in other strains with colitis, although the bacterial antigens detected are different from those detected by serum from C3H/HeJBir mouse. The implication is that there are not only immunodominant organisms, but also a limited set of immunodominant antigens recognized by the host immune system which may vary from one strain to another. Such immunodominant antigens need to be defined in both mice and humans. This will be crucial to understanding the complex immune response to the flora, and particularly how this response is regulated. Availability of these dominant microbial antigens will allow us to test whether immunologic tolerance to enteric flora does indeed exist, as has been proposed [138, 160]. Identification of these microbial antigens may allow manipulation of the immune response to them, which would be a novel approach to the treatment of IBD.

94

Genetic factors Disease expression can vary greatly when the same mutant gene is bred onto different inbred mouse strains. Each inbred strain carries a set of genes that are identical within that inbred strain but differs from other inbred strains. These 'background genes' can greatly influence or modify the expression of a disease resulting from a defined mutation. With regard to experimental IBD, many of the same strains appear to be susceptible or resistant regardless of how the disease was induced (Tables 1-6). This variation has been examined in detail in DSSinduced colitis where reproducible quantitative differences were found among the inbred strains. These differences were shown to be heritable and due to the effects of multiple genes [107]. Using quantitative trait locus mapping the location of a number of genes responsible for this variation has been identified [132]. A similar effort using IL-10 deficiency as the stimulus for colitis has also revealed the influence of multiple genes determining susceptibility [46, 47]. These results in mice mirror the results of similar efforts in patients with IBD in whom multiple genetic loci contributing to susceptibility have been identified. Similar approaches are under way in humans. The first gene conferring susceptibility to Crohn's disease has been identified on chromosome 16 [113, 114]. The NOD2 gene encodes an intracellular protein that recognizes microbial products such as endotoxin. The mechanism of NOD2 function, and how its deficiency can lead to Crohn's disease, will require the generation of a NOD2-deficient line of mice, an effort that is under way. This illustrates how the genetic studies in experimental models and in human patients complement one another.

Immunologic factors The immunologic pathogenesis has been discussed in the introduction and sections on each experimental model. However, there are some overall themes that recur through multiple models that are worth considering. These include triggering factors, progression or perpetuation of disease, and immune regulation, i.e. the immunologic stages of IBD. A number of models demonstrate that an animal that is genetically susceptible to IBD may not become ill unless the process is triggered by some event that perturbs the system. Initiating or 'triggering' factors seem to operate also in humans with IBD whose disease may initiate from a viral or bacterial

Experimental mouse models of inflammatory bowel disease infection, emotional shock, etc. Very little is known about these initiating factors but several examples of their requirement for disease to occur are demonstrated in the experimental models. One example is the specific pathogen-free C57B1/6 IL-2-deficient mouse, which under strict SPF housing conditions can remain healthy. However, immunization with DNP-KLH in complete Freund's adjuvant triggers the typical disease associated with IL-2 deficiency, including a severe colitis [53]. Another is the STAT-4 transgenic mouse which remains healthy unless immunized with the same material [161]. The ITFdeficient mouse, which has a genetic defect in the epithelial healing, remains well as long as epithelium is not perturbed. Feeding of DSS to ITF-deficient mice injures the epithelium and results in a severe colitis [124]. Non-steroidal anti-inflammatory drugs have long been thought to activate human IBD, but data supporting this idea are lacking. Thus it is quite interesting that an NSAID, piroxicam, can accelerate the onset of colitis in C57Bl/IL-10-deficient mice [162]. Even from this short list it is clear that the triggering factors can be quite diverse and can involve either systemic or mucosal perturbations. Experimental models should allow further dissection and understanding of this aspect of IBD. Initiating factors and those responsible for disease perpetuation or progression are likely to be different. With regard to Thl-mediated disease a key factor in disease perpetuation is sustained IL-12 production. The latter in turn requires interactions of CD40L on T cells and CD40 on antigen-presenting cells. Antibodies to IL-12 and to CD40L have been shown to be effective in multiple models and each of these agents is currently being tested in patients. Another factor required for perpetuation of the disease is the recruitment of new effector cells into the mucosa. Effector cells, including dendritic cells, macrophages, neutrophils, and activated lymphocytes, have fairly short life spans. Thus all these different cell types need to be renewed for chronic inflammation to persist, and this involves recruitment of such cells from the circulation. Supporting this idea are the findings that agents which block trafficking of cells into the intestine are able to prevent and treat colitis in a number of experimental models. In addition, a lymphotoxin PFc fusion protein, which perturbs secondary lymphoid structure, was able to treat active colitis in two different experimental models, probably by altering cell trafficking. A third immunologic aspect important to pathogenesis of IBD is immune regulation. Deficiency of

Charles O. Elson and Casey T. Weaver regulatory cells has been implicated in a significant number of models (Table 1). One of the most exciting developments in modern immunology is the identification of these regulatory cells. Their effects have long been seen, but identification of the cells responsible for those effects have been elusive. Many different types of regulatory cells are being defined and a great deal of information will be generated in coming years. Much of the basic immunology will come from experimental systems in the mouse and will need to be translated to humans. This area has great promise for novel therapeutic approaches to IBD therapies that could potentially alter the natural history of these diseases. Based on what has been learned from the models to date, the most rational therapy for IBD is to enhance regulatory cell function, to inhibit effector cell function, or to accomplish both. Such an approach is within sight at present, but will need to be validated in experimental models before it is translated to the treatment of human IBD.

95

11.

12.

13.

14.

15. 16.

17. 18.

References 1.

Elson CO, Sartor RB, Tennyson GS, Riddell RH. Experimental models of inflammatory bowel disease. Gastroenterology 1995; 109: 1344-67. 2. Mosmann TR, Sad S. The expanding universe of T-cell subsets: Thl, Th2 and more. Immunol Today 1996; 17: 138-46. 3. Fort MM, Lesley R, Davidson NJ et al. IL-4 exacerbates disease in a Thl cell transfer model of colitis. J Immunol 2001; 166:2793-800. 4. Neurath MF, Fuss I, Pasparakis M et al. Predominant pathogenic role of tumor necrosis factor in experimental colitis in mice. Eur J Immunol 1997; 27: 1743-50. 5. Fuss IJ, Neurath M, Boirivant M et al. Disparate CD4"^ lamina propria (LP) lymphokine secretion profiles in inflammatory bowel disease. Crohn's disease LP cells manifest increased secretion of IFN-gamma, whereas ulcerative colitis LP cells manifest increased secretion of IL-5. J Immunol 1996; 157: 1261-70. 6. Bacchetta R, Bigler M, Touraine JL et al. High levels of interleukin 10 production in vivo are associated with tolerance in SCID patients transplanted with HLA mismatched hematopoietic stem ceUs. J Exp Med 1994; 179: 493-502. 7. Groux H, O'Garra A, Bigler M et al. A CD4"' T cefl subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 1997; 389: 737-42. 8. Santos LM, al-Sabbagh A, Londono A, Weiner HL. Oral tolerance to myelin basic protein induces regulatory TGFbeta-secreting T cells in Peyer's patches of SJL mice. CeU Immunol 1994; 157: 439-47. 9. Chen Y, Kuchroo VK, Inobe J, Hafler DA, Weiner HL. Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis. Science 1994; 265: 1237-40. 10. Fukaura H, Kent SC, Pietrusewicz MJ, Khoury SJ, Weiner HL, Hafler DA. Induction of circulating myelin basic protein and proteolipid protein-specific transforming

19.

20.

21.

22.

23.

24.

25.

26.

growth factor-beta 1-secreting Th3 T cells by oral administration of myelin in multiple sclerosis patients. J Clin Invest 1996;98:70-7. Neurath MF, Fuss I, Kelsall BL, Presky DH, Waegell W, Strober W. Experimental granulomatous colitis in mice is abrogated by induction of TGF-beta-mediated oral tolerance. J Exp Med 1996; 183: 2605-16. Ludviksson BR, Ehrhardt RO, Strober W. TGF-beta production regulates the development of the 2,4,6-trinitrophenol-conjugated keyhole limpet hemocyanin-induced colonic inflammation in IL-2-deficient mice. J Immunol 1997; 159: 3622-8. Simpson SJ, Mizoguchi E, Allen D, Bhan AK, Terhorst C. Evidence that CD4^, but not CD8"^ T cells are responsible for murine interleukin-2-deficient colitis. Eur J Immunol 1995;25:2618-25. Cong Y, Weaver CT, Nguyen H, Lazenby A, Sundberg JP, Elson CO. CD 8+ T cells, but not B cells inhibit enteric bacterial antigen-specific CD4"^ T ceU-induced cohtis. Gastroenterology 1999; 116: A690. Fort MM, Leach MW, Rennick DM. A role for N K ceUs as regulators of CD4'^ T cells in a transfer model of colitis. J Immunol 1998; 161:3256-61. Saubermann LJ, Beck P, De Jong YP et al. Activation of natural killer T cells by alpha-yalactosylceramide in the presence of CD Id provides protection against colitis in mice. Gastroenterology 2000; 119: 119-28. Mizoguchi E, Mizoguchi A, Preffer FI, Bhan AK. Regulatory role of mature B cells in a murine model of inflammatory bowel disease. Int Immunol 2000; 12: 597-605. Blunt T, Gell D, Fox M et al. Identification of a nonsense mutation in the carboxyl-terminal region of DNA-dependent protein kinase catalytic subunit in the scid mouse. Proc Natl Acad Sci USA 1996; 93: 10285-90. Morrissey PJ, Charrier K, Braddy S, Liggitt D, Watson JD. CD4"^ T cells that express high levels of CD45RB induce wasting disease when transferred into congenic severe combined immunodeficient mice. Disease development is prevented by cotransfer of purified CD4"^ T cells. J Exp Med 1993; 178: 2 3 7 ^ 4 . Powrie F, Leach MW, Mauze S, Caddie LB, Coff"man RL. Phenotypically distinct subsets of CD4'*" T cells induce or protect from chronic intestinal inflammation in C. B-17 scid mice. Int Immunol 1993; 5: 1461-71. Powrie F, Leach MW, Mauze S, Menon S, Caddie LB, Coffman RL. Inhibition of Thl responses prevents inflammatory bowel disease in scid mice reconstituted with CD45RBhi CD4^ T cells. Immunity 1994; 1: 553-62. Rudolphi A, Bonhagen K, Reimann J. Polyclonal expansion of adoptively transferred CD4"^ alpha beta T cells in the colonic lamina propria of scid mice with colitis. Eur J Immunol 1996; 26: 1156-63. Bregenholt S, Claesson MH. Splenic T helper cell type 1 cytokine profile and extramedullary haematopoiesis in severe combined immunodeficient (scid) mice with inflammatory bowel disease (IBD). Clin Exp Immunol 1998; 111: 166-72. Powrie F, Carlino J, Leach MW, Mauze S, Coff'man RL. A critical role for transforming growth factor-beta but not interleukin 4 in the suppression of T helper type 1-mediated colitis by CD45RB(low) CD4^ T cells. J Exp Med 1996; 183:2669-74. Asseman C, Mauze S, Leach MW, Coffman RL, Powrie F. An essential role for interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation. J Exp Med 1999; 190: 995-1004. Annacker O, Burlen-Defranoux O, Pimenta-Araujo R, Cumano A, Bandeira A. Regulatory CD4 T cells control

96

27.

28.

29. 30.

31.

32.

33.

34.

35. 36.

37. 38.

39.

40. 41.

42. 43.

44.

Experimental mouse models of inflammatory bowel disease the size of the peripheral activated/memory CD4 T cell compartment. J Immunol 2000; 164: 3573-80. Takahashi T, Tagami T, Yamazaki S et al Immunologic selftolerance maintained by CD25(+)CD4(+) regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4. J Exp Med 2000; 192: 303-10. Read S, Malmstrom V, Powrie F. Cytotoxic T lymphocyteassociated antigen 4 plays an essential role in the function of CD25(+)CD4(+) regulatory cells that control intestinal inflammation. J Exp Med 2000; 192: 295-302. Hagenbaugh A, Sharma S, Dubinett SM et al. Altered immune responses in interleukin 10 transgenic mice. J Exp Med 1997; 185:2101-10. Aranda R, Sydora BC, McAllister PL et al. Analysis of intestinal lymphocytes in mouse colitis mediated by transfer of CD4"', CD45RBhigh T cells to SCID recipients. J Immunol 1997; 158: 3464-73. Picarella D, Huribut P, Rottman J, Shi X, Butcher E, Ringler DJ. Monoclonal antibodies specific for beta 7 integrin and mucosal addressin cell adhesion molecule-1 (MAdCAM-1) reduce inflammation in the colon of scid mice reconstituted with CD45RBhigh CD4^ T cells. J Immunol 1997; 158: 2099 106. Mackay F, Browning JL, Lawton Y* et al. Both the lymphotoxin and tumor necrosis factor pathways are involved in experimental murine models of colitis. Gastroenterology 1998;115:1464 75. De Jong YP, Comiskey M, Kalled SL et al. Chronic murine colitis is dependent on the CD154/CD40 pathway and can be attenuated by anti-CD 154 administration. Gastroenterology 2000; 119:715 23. Morrisscy PJ, Charricr K. Induction of wasting disease in SCID mice by the transfer of normal CD4^/CD45RBhi T cells and the regulation of this autoreactivity by CD4"^/ C D 4 5 R B ' ' ' T cells. Res Immunol 1994; 145: 357 62. Matsuda JL, Gapin L, Sydora BC, Systemic activation and antigen-driven oligoclonal expansion of T cells in a mouse model of colitis. J Immunol 2000; 164: 2797 806. Brimncs J, Reimann J, Mogens MH, Claessen MH. Enteric bacterial antigens activate CD4"^ T cells from scid mice with inflammatory bowel disease. Eur J Immunol 2001; 31: 2331. Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. Interleukin-10-deficient mice develop chronic enterocolitis. Cell 1993;75:263-74. Sellon RK, Tonkonogy S, Schultz M et al. Resident enteric bacteria are necessary for development of spontaneous colitis and immune system activation in interleukin-10deficient mice. Infect Immun 1998; 66: 5224-31. Berg DJ, Kuhn R, Rajewsky K et al. Interleukin-10 is a central regulator of the response to LPS in murine models of endotoxic shock and the Shwartzman reaction but not endotoxin tolerance. J Clin Invest 1995; 96: 2339^7. Davidson NJ, Leach MW, Fort MM et al. T helper cell 1type CD4"^ T cells, but not B cells, mediate colitis in interleukin 10-deficient mice. J Exp Med 1996; 184: 241-51. Berg D, Davidson N, Kuhn R et al. Enterocolitis and colon cancer interleukin-10-deficient mice are associated with aberrant cytokine production and CD4"^ Thl-Uke responses. J Clin Invest 1996; 98: 1010-20. Rennick DM, Fort MM, Davidson NJ. Studies with IL1 0 - / - mice: an overview. J Leuk Biol 1997; 61: 389-96. Davidson NJ, Hudak SA, Lesley RE, Menon S, Leach MW, Rennick DM. IL-12, but not IFN-gamma, plays a major role in sustaining the chronic phase of colitis in IL-10deficient mice. J Immunol 1998; 161: 3143-9. Brandwein SL, McCabe RP, Cong Y et al. Spontaneously colitic C3H/HeJBir mice demonstrate selective antibody

45. 46. 47.

48. 49. 50. 51.

52.

53. 54.

55.

56.

57.

58.

59.

60.

61.

62.

reactivity to antigens of the enteric bacterial flora. J Immunol 1997; 159:44-52. Seibold F, Brandwein S, Simpson S, Terhorst C, Elson CO. pANCA represents a cross-reactivity to enteric bacterial antigens. J Clin Immunol 1998; 18: 153-60. Bristol IJ, Farmer MA, Cong Y et al. Heritable susceptibility for colitis in mice induced by IL-10 deficiency. Inflam Bowel Dis 2000; 6: 290-302. Farmer MA, Leiter EH, Churchill GA, Sundberg JP, Elson CO. Complex interactions among modifier genes controlling colitis severity in IL-10 deficient mice. Gastroenterology 2001; 120: A36. Spencer SD, Di Marco F, Hooley J et al. The orphan receptor CRF2-4 is an essential subunit of the interleukin 10 receptor. J Exp Med 1998; 187: 571-8. Kundig TM, Schorle H, Bachmann MF, Hengartner H, Zinkernagel RM, Horak I. Immune responses in interleukin-2-deficient mice. Science 1993; 262: 1059 61. Ma A, Datta M, Margosian E, Chen J, Horak I. T cells, but not B cells, are required for bowel inflammation in interleukin 2-deficient mice. J Exp Med 1995; 182: 1567-72. Kramer S, Schimpl A, Hunig T. Immunopathology of interleukin (IL) 2-deficient mice: thymus dependence and suppression by thymus-dependent cells with an intact IL-2 gene. J Exp Med 1995; 182: 1769 76. Sadlack B, Lohler J, Schorle H et al. Generalized autoimmune disease in interleukin-2-deficient mice is triggered by an uncontrolled activation and proliferation of CD4^ T cells. Eur J Immunol 1995; 25: 3053 9. Ehrhardl RO, Ludviksson BR, Gray B, Neurath M, Strober W. Induction and prevention of colonic inflammation in IL2-deficicnt mice. J Immunol 1997; 158: 566-73. Ludviksson BR, Strober W, Nishikomori R, Hasan SK, Ehrhardt RO. Administration of mAb against alpha E beta 7 prevents and ameliorates immunization-induced colitis in IL-2 / mice. J Immunol 1999; 162: 4975 82. Kneitz B, Herrmann T, Yonehara S, Schimpl A. Normal clonal expansion but impaired Fas-mediated cell death and anergy induction in interleukin-2-deficient mice. Eur J Immunol 1995;25:2572-7. Poussier P, NingT, Chen J, Banerjee D, Julius M. Intestinal inflammation observed in IL-2R/IL-2 mutant mice is associated with impaired intestinal T lymphopoiesis. Gastroenterology 2000; 118: 880-91. Papiernik M, de Moraes ML, Pontoux C, Vasseur F, Penit C. Regulatory CD4 T cells: expression of IL-2R alpha chain, resistance to clonal deletion and IL-2 dependency. Int Immunol 1998; 10: 371-8. Sakaguchi S, Toda M, Asano M, Itoh M, Morse SS, Sakaguchi N. T cell-mediated maintenance of natural selftolerance: its breakdown as a possible cause of various autoimmune diseases. J Autoimmun 1996; 9: 211-20. Contractor NV, Bassiri H, Reya T et al. Lymphoid hyperplasia, autoimmunity, and compromised intestinal intraepithelial lymphocyte development in colitis-free gnotobiotic IL-2-deficient mice. J Immunol 1998; 160: 385-94. Mahler M, Serreze D, Evans R, Linder CD, Leiter EH, Sundberg JP. IL-2^"''"'", an interleukin-2 gene targeted mutation: the Jackson Laboratory, Bar Harbor, ME; Fall 1996. Report No. 467. Wang B, Biron C, She J et al. A block in both early T lymphocyte and natural killer cell development in transgenic mice with high-copy numbers of the human CD3E gene. Proc Natl Acad Sci USA 1994; 91: 9402-6. Hollander GA, Wang B, Nichogiannopoulou A et al. Developmental control point in induction of thymic cortex regulated by a subpopulation of prothymocytes. Nature 1995; 373: 350-3.

Charles O. Elson and Casey T. Weaver 63. 64.

65.

66.

67.

68. 69.

70. 71.

72.

73.

74.

75.

76. 77.

78.

79. 80.

Hollander GA, Simpson SJ, Mizoguchi E et al Severe colitis in mice with aberrant thymic selection. Immunity 1995; 3: 27-38. Velkamp C, Tonkonogy SL, De Jong YP et al. Continuous stimulation by normal luminal bacteria is essential for the development and perpetuation of colitis in Tg epsilon 26 mice. Gastroenterology 2001; 120: 900-13. Simpson SJ, De Jong YP, Shah SA et al. Consequences of Fas-ligand and perforin expression by colon T cells in a mouse model of inflammatory bowel disease. Gastroenterology 1998; 115:849-55. Simpson SJ, Hollander GA, Mizoguchi E et al. Expression of pro-inflammatory cytokines by TCR alpha beta+ and TCR gamma delta+ T cells in an experimental model of colitis. Eur J Immunol 1997; 27: 17-25. Mombaerts P, Mizoguchi E, Grusby MJ, Glimcher LH, Bhan AK, Tonegawa S. Spontaneous development of inflammatory bowel disease in T cell receptor mutant mice. CeU 1993; 75: 1-20. Mizoguchi E, Mizoguchi A, Bhan AK. Role of cytokines in the early stages of chronic colitis in TCR alpha-mutant mice. Lab Invest 1997; 76: 385-97. Mizoguchi A, Mizoguchi E, Chiba C et al. Cytokine imbalance and autoantibody production in T cell receptoralpha mutant mice with inflammatory bowel disease. J Exp Med 1996; 183: 847-56. Mombaerts P, Mizoguchi E, Ljunggren HG et al. Peripheral lymphoid development and function in TCR mutant mice. Int Immunol 1994; 6: 1061-70. Das KM, Dasgupta A, Mandal A, Geng X. Autoimmunity to cytoskeletal protein tropomyosin. A clue to the pathogenetic mechanism for ulcerative cohtis. J Immunol 1993; 150: 2487-93. Bhan AK, Mizoguchi E, Smith RN, Mizoguchi A. Colitis in transgenic and knockout animals as models of human inflammatory bowel disease. Immunol Rev 1999; 169: 195207. Takahashi I, Kiyono H, Hamada S. A CD4"^ T-cell population mediates development of inflammatory bowel disease in T-cell receptor alpha-deficient mice. Gastroenterology 1997;112: 1876-82. Mizoguchi A, Mizoguchi E, Saubermann LJ, Higaki K, Blumberg RS, Bhan AK. Limited CD4 T-cell diversity associated with colitis in T-cell receptor alpha mutant mice requires a T helper 2 environment. Gastroenterology 2000; 119:983-95. Dianda L, Hanby AM, Wright NA, Sebesteny A, Hayday AC, Owen MJ. T cell receptor-alpha beta-deficient mice fail to develop colitis in the absence of a microbial environment. Am J Pathol 1997; 150:91-7. Mizoguchi A, Mizoguchi E, Chiba C, Bhan AK. Role of appendix in the development of inflammatory bowel disease in TCR-alpha mutant mice. J Exp Med 1996; 184: 707-15. Mizoguchi A, Mizoguchi E, Tonegawa S, Bhan AK. Alteration of a polyclonal to an oligoclonal immune response to cecal aerobic bacterial antigens in TCRoc mutant mice with inflammatory bowel disease. Int Immunol 1996; 8: 1387-94. Mizoguchi A, Mizoguchi E, Bhan AK. The critical role of interleukin 4 but not interferon gamma in the pathogenesis of colitis in T-cell receptor alpha mutant mice. Gastroenterology 1999; 116:320-6. Mizoguchi A, Mizoguchi E, Smith RN, Preffer FI, Bhan AK. Suppressive role of B cells in chronic colitis of T cell receptor alpha mutant mice. J Exp Med 1997; 186: 1749-56. Mizoguchi A, Mizoguchi E, Smith RN, Preffer FI, Bhan AK. Suppressive role of B cells in chronic colitis of T cell receptor a mutant mice. J Exp Med 1997; 186: 1749-56.

97 81. 82. 83.

84.

85.

86. 87.

88.

89.

90.

91.

92.

93.

94.

95.

96. 97.

Kulkarni AB, Ward JM, Yaswen L, Transforming growth factor-beta 1 null mice. An animal model for inflammatory disorders. Am J Pathol 1995; 146: 264-75. Boivin GP, O'Toole BA, Orsmby IE et al. Onset and progression of pathological lesions in transforming growth factor-beta 1-deficient mice. Am J Pathol 1995; 146: 276-88. Boivin GP, Ormsby I, Jones-Carson J, O'Toole BA, Doetschman T. Germ-free and barrier-raised TGF beta 1deficient mice have similar inflammatory lesions. Transgenic Res 1997; 6: 197-202. Letterio JJ, Geiser AG, Kulkarni AB et al. Autoimmunity associated with TGF-betal-deficiency in mice is dependent on MHC class II antigen expression. J Clin Invest 1996; 98: 2109-19. Diebold RJ, Eis MJ, Yin M et al. Early-onset multifocal inflammation in the transforming growth factor beta 1-null mouse is lymphocyte mediated. Proc Natl Acad Sci USA 1995; 92: 12215-19. Gorelik L, Flavell RA. Abrogation of TGFbeta signaling in T cells leads to spontaneous T cell diff'erentiation and autoimmune disease. Immunity 2000; 12: 171-81. Lucas PJ, Kim SJ, Melby SJ, Gress RE. Disruption of T cell homeostasis in mice expressing a T cell-specific dominant negative transforming growth factor beta II receptor. J Exp Med 2000; 191: 1187-96. Yang X, Letterio JJ, Lechleider RJ et al. Targeted disruption of SMAD3 results in impaired mucosal immunity and diminished T cell responsiveness to TGF-beta. EMBO J 1999;18:1280-91. Watanabe M, Ueno Y, Yajima T et al. Interleukin 7 is produced by human intestinal epithelial cells and regulates the proliferation of intestinal mucosal lymphocytes. J Clin Invest 1995; 95: 2945-53. Fujihashi K, Kawabata S, Hiroi T, Interleukin 2 (IL-2) and interleukin 7 (IL-7) reciprocally induce IL-7 and IL-2 receptors on gamma delta T-cell receptor-positive intraepithelial lymphocytes. Proc Natl Acad Sci USA 1996; 93: 3613-8. Fujihashi K, McGhee JR, Yamamoto M, Peschon JJ, Kiyono H. An interleukin-7 internet for intestinal intraepithelial T cell development: knockout of ligand or receptor reveal differences in the immunodeficient state. Eur J Immunol 1997; 27: 2133-8. Watanabe M, Ueno Y, Yajima T et al. Interleukin 7 transgenic mice develop chronic cohtis with decreased interleukin 7 protein accumulation in the colonic mucosa. J Exp Med 1998; 187: 389-402. Targan SR, Hanauer SB, van Deventer SJ et al. A short-term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn's disease. Crohn's Disease cA2 Study Group. N Engl J Med 1997; 337: 1029-35. Kontoyiannis D, Pasparakis M, Pizarro TT, Cominelli F, KoUias G Impaired on/off regulation of TNF biosynthesis in mice lacking TNF AU-rich elements: implications for joint and gut-associated immunopathologies. Immunity 1999; 10: 387-98. KoUias G, Douni E, Kassiotis G, Kontoyiannis D. On the role of tumor necrosis factor and receptors in models of multiorgan failure, rheumatoid arthritis, multiple sclerosis and inflammatory bowel disease. Immunol Rev 1999; 169: 175-94. Lee EG, Boone DL, Chai S et al. Failure to regulate TNFinduced NF-kappaB and cell death responses in A20-deficient mice. Science 2000; 289: 2350-4. Byrne FR, Whoriskey JS, Sarmiento U et al. Transgenic mice over-expressing the B7 related protein-1 (B7RP-1) develop an intestinal pathology similar to human Crohn's disease. A new mouse model of inflammatory bowel disease. Gastroenterology 2001: A47.

Experimental mouse models of inflammatory bowel disease

98 98. 99. 100. 101.

102. 103.

104. 105.

106.

107. 108. 109.

110. 111. 112.

113. 114. 115. 116.

117.

Foy TM, Aruffo A, Bajorath J, Buhlmann JE, Noelle RJ. Immune regulation by CD40 and its ligand GP39. Annu Rev Immunol 1996; 14: 591-617. Grewal IS, Flavell RA. CD40 and CD154 in cell-mediated immunity. Annu Rev Immunol 1998; 16: 111-35. Clegg CH, Rulffes JT, Haugen HS etal. Thymus dysfunction and chronic inflammatory disease in gp39 transgenic mice. Int Immunol 1997; 9: 1111-22. Cong Y, Weaver CT, Lazenby A, Elson CO. Colitis induced by enteric bacterial antigen-specific CD4"^ T cells requires CD40-CD40 ligand interactions for a sustained increase in mucosal IL-12. J Immunol 2000; 165: 2173-82. Sundberg JP, Elson CO, Bedigian H, Birkenmeier EH. Spontaneous, heritable colitis in a new substrain of C3H/ HeJ mice. Gastroenterology 1994; 107: 1726-35. McCabe RP, Sharmanov A, Birkenmeier E, Sundberg J, Elson CO. Mucosal immune abnormalities in C3H/HeJBir mice with susceptibility to colitis. Gastroenterology 1994; 106: A731. Ni J, Chen SF, Hollander D. Immunological abnormality in C3H/HcJ mice with heritable inflammatory bowel disease. Cell Immunol 1996; 169: 7 15. Cong Y, Brandwcin SL, McCabe RP ct al. CD4^ T cells reactive to enteric bacterial antigens in spontaneously colitic C3H/HeJBir mice: increased T helper cell type 1 response and ability to transfer disease. J Exp Med 1998; 187: 855 64. Cong Y, Weaver CT, Lazenby A, Elson CO. T-rcgulatory-1 (Trl) cells that prevent CD4^ T cell colitis inhibit the antigen-presenting function and IL-12 production of dendritic cells. Gastroenterology 2001; 120: A38. Mahler M, Bristol IJ, Leitcr EH ct al. Difl'erential susceptibility of inbred mouse strains to dextran sulfate sodiuminduced colitis. Am J Physiol 1998; 274: G544 51. Matsumoto S, Okabc Y, Sctoyama H ct al. Inflammatory bowel disease-like enteritis and caecitis in a senescence accelerated mouse PI /Yit strain. Gut 1998; 43: 71 8. Takeda K, Clausen BE, Kaisho T ct al. Enhanced Thl activity and development of chronic enterocolitis in mice devoid of Stat3 in macrophages and neutrophils. Immunity 1999; 10: 39 49. Kosiewicz MM, Nast CC, Krishnan A ct al. Thl-type responses mediate spontaneous ileitis in a novel murine model of Crohn's disease. J Clin Invest 2001; 107: 695-702. Eckmann L, Kagnofl' MF, Fierer J. Epithelial cells secrete the chemokine interleukin-8 in response to bacterial entry. Infect Immun 1993; 61: 4569-74. McCormick BA, Colgan SP, Delp-Archer C, Miller SI, Madara JL. Salmonella typhimurium attachment to human intestinal epithelial monolayers: transcellular signalling to subepithelial neutrophils. J Cell Biol 1993; 123: 895-907. Hugot J-P, Chamaillard M, Zouali H ct al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 2001; 411: 599-603. Ogura Y, Bonen DK, Inohara N ct al. A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease. Nature 2001; 411: 603-606. Katz KD, Hollander D, Vadheim CM et al. Intestinal permeability in patients with Crohn's disease and their healthy relatives. Gastroenterology 1989; 97: 927-31. Toy LS, Yio XY, Lin A, Honig S, Mayer L. Defective expression of gpl80, a novel CD8 ligand on intestinal epithelial cells, in inflammatory bowel disease. J Clin Invest 1997; 100:2062-71. Madara J, Staff'ord J. Interferon-yamma directly aff'ects barrier function of cultured intestinal epithelial monolayers. J Clin Invest 1989; 83: 724.

118. Dieleman LA, Elson CO, Tennyson GS, Beagley KW. Kinetics of cytokine expression during healing of acute colitis in mice. Am J Physiol 1996; 34: G 130-6. 119. Hermiston ML, Gordon JI. Inflammatory bowel disease and adenomas in mice expressing a dominant negative Ncadherin. Science 1995; 270: 1203-7. 120. Panwala CM, Jones JC, Viney JL. A novel model of inflammatory bowel disease: mice deficient for the multiple drug resistance gene, mdrla, spontaneously develop colitis. J Immunol 1998; 161: 5733-44. 121. Rudolph U, Finegold MJ, Rich SS ct al. Ulcerative colitis and adenocarcinoma of the colon in G alpha i2-deficient mice. Nat Genet 1995; 10: 143-50. 122. Hornquist CE, Lu X, Rogers-Fani PM et al. G(alpha)i2deficient mice with colitis exhibit a local increase in memory CD4'^ T cells and proinflammatory Thl-type cytokines. J Immunol 1997; 158: 1068-77. 123. Ohman L, Franzen L, Rudolph U, Harriman GR, HultgrenHornquist E. Immune activation in the intestinal mucosa before the onset of colitis in Galphai2-deficient mice. Scand J Immunol 2000; 52: 80 90. 124. Mashimo H, Wu DC, Podolsky DK, Fishman MC. Impaired defense of intestinal mucosa in mice lacking intestinal trefoil factor. Science 1996; 274: 262 5. 125. Baribault H, Penner J, lozzo RV, Wilson-Heiner M. Colorectal hyperplasia and inflammation in keratin 8-deficient FVB/N mice. Genes Devel 1994; 8: 2964 73. 126. Magin TM, Schroder R, Leitgeb S ct al. Lessons from keratin 18 knockout mice: formation of novel keratin filaments, secondary loss of keratin 7 and accumulation of liver-specific keratin 8-positive aggregates. J Cell Biol 1998; 140: 1441 51. 127. Okayasu I, Hatakeyama S, Yamada M, Ohkusa T, Inagaki Y, Nakaya R. A novel method in the induction of reliable experimental acute and chronic ulcerative colitis in mice. Gastroenterology 1990; 98: 694 702. 128. Cooper HS, Murthy SN, Shah RS, Sedergran DJ. Clinicopathologic study of dextran sulfate sodium experimental murine colitis. Lab Invest 1993; 69: 238-49. 129. Yamada M, Ohkusa T, Okayasu I. Occurrence of dysplasia and adenocarcinoma after experimental chronic ulcerative colitis in hamsters induced by dextran sulphate sodium. Gut 1992; 33: 1521-7. 130. Hirono I, Kuhara K, Hosaka S, Tomizawa S, Goldberg L. Induction of intestinal tumors in rats by dextran sulfate sodium. J Natl Cancer Inst 1981; 66: 579-83. 131. Dieleman LA, Ridwan BU, Tennyson GS, Beagley KW, Bucy RP, Elson CO. Dextran sulfate sodium-induced colitis occurs in severe combined immunodeficient mice. Gastroenterology 1994; 107: 1643-52. 132. Mahler M, Bristol IJ, Sundberg iV ct al. Genetic analysis of susceptibility to dextran sulfate sodium-induced colitis in mice. Genomics 1999; 55: 147-56. 133. Morris GP, Beck PL, Herridge MS, Depew WT, Szewczuk MR, Wallace JL. Hapten-induced model of chronic inflammation and ulceration in the rat colon. Gastroenterology 1989;96:795-803. 134. Neurath MF, Fuss I, Kelsall BL, Stuber E, Strober W. Antibodies to interleukin 12 abrogate established experimental colitis in mice. J Exp Med 1995; 182: 1281-90. 135. Elson CO, Beagley KW, Sharmanov AT, Hapten-induced model of murine inflammatory bowel disease - mucosal immune responses and protection by tolerance. J Immunol 1996; 157:2174-85. 136. Brandtzaeg P, Valnes K, Scott H, Rognum TO, Bjerke K, Baklien K. The human gastrointestinal secretory system in health and disease. Scand J Gastroenterol 1985; 20: 17-38. 137. Kelsafl BL, Stuber E, Neurath M, Strober W. Interleukin-12 production by dendritic cells. The role of CD40-CD40L

Charles O. Ebon and Casey T. Weaver

138.

139.

140.

141.

142.

143.

144. 145.

146.

147.

148. 149.

150.

151.

152.

153.

interactions in Thl T-cell responses. Ann NY Acad Sci 1996; 795: 116-26. Duchmann R, Schmitt E, Knolle P, Meyer zum Buschenfelde KH, Neurath M. Tolerance towards resident intestinal flora in mice is abrogated in experimental colitis and restored by treatment with interleukin-10 or antibodies to interleukin-12. Eur J Immunol 1996; 26: 934-8. Neurath MF, Pettersson S, Meyer Zum Buuschenfeld K-H, Strober W. Local administration of antisense phosphorothionate oligonucleotides to the p65 subunit of NFkB abrogates established experimental colitis in mice. Nature Med 1996; 2: 998-1004. Hammer RE, Maika SD, Richardson JA, Tang JP, Taurog JD. Spontaneous inflammatory disease in transgenic rats expressing HLA-B27 and human beta 2m: an animal model of HLA-B27-associated human disorders. Cell 1990; 63: 1099-112. Breban M, Hammer RE, Richardson JA, Taurog JD. Transfer of the inflammatory disease of HLA-B27 transgenic rats by bone marrow engraftment. J Exp Med 1993; 178:1607-16. Taurog JD, Richardson JA, Croft JT et al. The germfree state prevents development of gut and joint inflammatory disease in HLA-B27 transgenic rats. J Exp Med 1994; 180: 2359-64. Rath HC, Herfarth HH, Ikeda JS et al Normal luminal bacteria, especially Bacteroides species, mediate chronic colitis, gastritis, and arthritis in HLA-B27/human beta2 microglobulin transgenic rats. J Chn Invest 1996; 98: 94553. Snapper SB, Rosen FS, Mizoguchi E et al. Wiskott-Aldrich syndrome protein-deficient mice reveal a role for WASP in T but not B cell activation. Immunity 1998; 9: 81-91. Koh WP, Chan E, Scott K, McCaughan G, France M, Fazekas de St Groth B. TCR-mediated involvement of CD4"^ transgenic T cells in spontaneous inflammatory bowel disease in lymphopenic mice. J Immunol 1999; 162: 720816. Miller AM, Elliot PR, Council W, Desmond PV, dApice AJ. A novel model of ulcerative colitis based on genetic manipulation of the glycosylation of colonic mucins. Gastroenterology 2000; 118: A687. Kalliomaki M, Salminen S, Arvilommi H, Kero P, Koskinen P, Isolauri E. Probiotics in primary prevention of atopic disease: a randomized placebo-controlled trial. Lancet 2001; 357: 1076-9. Moore WEC, Holdeman LV. Human fecal flora: the normal flora of 20 Japanese-Hawaians. Appl Microbiol 1974; 27: 961-79. Crabbe PA, Bazin H, Eyssen H, Heremans JF. The normal microbial flora as a major stimulus for proliferation of plasma cells synthesizing IgA in the gut. The germ-free intestinal tract. Int Arch Allergy 1968; 34: 362-75. Okada Y, Setoyama H, Matsumoto S et al. Effects of fecal microorganisms and their chloroform-resistant variants derived from mice, rats, and humans on immunological and physiological characteristics of the intestines of exyermfree mice. Infect Immun 1994; 62: 5442-6. Klaasen HL, Koopman JP, Van den Brink ME, Bakker MH, Poelma FG, Beynen AC. Intestinal, segmented, filamentous bacteria in a wide range of vertebrate species. Lab Anim 1993; 27: 141-50. Klaasen HL, Van der Heijden PJ, Stok W et al. Apathogenic, intestinal, segmented, filamentous bacteria stimulate the mucosal immune system of mice. Infect Immun 1993; 61: 303-6. Umesaki Y, Okada Y, Matsumoto S, Imaoka A, Setoyama H. Segmented filamentous bacteria are indigenous intestinal bacteria that activate intraepithelial lymphocytes and in-

99

154.

155.

156. 157. 158. 159. 160.

161.

162. 163. 164.

165. 166. 167.

168. 169. 170. 171. 172. 173.

duce MHC class II molecules and fucosyl asialo G M l glycolipids on the small intestinal epithelial cells in the exyerm-free mouse. Microb Immunol 1995; 39: 555-62. Klapproth JM, Donnenberg MS, Abraham JM, James SP. Products of enteropathogenic E. coli inhibit lymphokine production by gastrointestinal lymphocytes. Am J Physiol 1996; 271: G841-8. Klapproth JM, Donnenberg MS, Abraham JM, Mobley HL, James SP. Products of enteropathogenic Escherichia coli inhibit lymphocyte activation and lymphokine production. Infect Immun 1995; 63: 2248-54. Neish AS, Gewirtz AT, Zeng H et al. Prokaryotic regulation of epithelial responses by inhibition of IkappaB-alpha ubiquitination. Science 2000; 289: 1560-3. Madsen KL, Doyle JS, Jewell LD, Tavernini MM, Fedorak RN. Lactobacillus species prevents colitis in interleukin 10 gene-deficient mice. Gastroenterology 1999; 116: 1107-14. Schultz M, Sartor RB. Probiotics and inflammatory bowel diseases. Am J Gastroenterol 2000; 95(Suppl. 1): S19-21. Steidler L, Hans W, Schotte L et al. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science 2000; 289: 1352-5. Duchmann R, Kaiser I, Hermann E, Mayet W, Ewe K, Meyer zum Buschenfelde KH. Tolerance exists towards resident intestinal flora but is broken in active inflammatory bowel disease (IBD). Clin Exp Immunol 1995; 102: 448-55. Wirtz S, Finotto S, Kanzler S et al Cutting edge: chronic intestinal inflammation in STAT-4 transgenic mice: characterization of disease and adoptive transfer by TNF- plus IFN-gamma-producing CD4(+) T cells that respond to bacterial antigens. J Immunol 1999; 162: 1884-8. Berg DJ, Weinstock J, Lynch R. Rapid induction of inflammatory bowel disease in NSAID-treated IL-10 - / - mice. Gastroenterology 2001; 120: A685. Sadlack B, Merz H, Schorle H, Schimpl A, Feller AC, Horak I. Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 1993; 75: 253-61. Willerford D, Chen J, Ferry J, Davidson L, Ma A, Alt F. Interleukin-2 receptor alpha chain regulates the size and content of the peripheral lymphoid compartment. Immunity 1995; 3: 521-30. Shull MM, Ormsby I, Kier AB et al Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature 1992; 359: 693-9. Rudolph U, Finegold MJ, Rich SS et al G(i2)alpha protein deficiency: A model for inflammatory bowel disease. J Clin Immunol 1995; 15(Suppl 6.): SlOl-5. Madara JL, Podolsky DK, King NW, Sehgal PK, Moore R, Winter HS. Characterization of spontaneous colitis in cotton-top tamarins (Saguinus oedipus) and its response to sulfasalazine. Gastroenterology 1985; 88: 13-9. Neurath M, Fuss I, Strober W. TNBS-colitis. Int Rev Immunol 2000; 19:51-62. Boirivant M, Fuss IJ, Chu A, Strober W Oxazolone coHtis: a murine model of T helper cell type 2 colitis treatable with antibodies to interleukin 4. J Exp Med 1998; 188: 1929-39. Sartor RB, Cromartie WJ, Powell DW, Schwab JH. Granulomatous enterocolitis induced in rats by purified bacterial cell wall fragments. Gastroenterology 1985; 89: 587-95. Sartor RB, Bender DE, Allen JB et al Chronic experimental enterocolitis and extraintestinal inflammation are T lymphocyte dependent. Gastroenterology 1993; 104: A775. Sartor RB, Bender DE, Holt LC. Susceptibility of inbred rat strains to intestinal inflammation induced by indomethacin. Gastroenterology 1992; 102: A690. Taurog JD, Maika SD, Simmons WA, Breban M, Hammer RE. Susceptibility to inflammatory disease in HLA-B27 transgenic rat lines correlates with the level of B27 expression. J Immunol 1993; 150: 4168-78.

5

The normal intestinal mucosa: a state of 'controlled inflammation' CLAUDIO FIOCCHI

Introduction The focus of investigation in inflammatory bowel disease (IBD) is almost invariably centered on the various cellular and soluble components of the mucosal immune system whose abnormal number or functions presumably underlies the pathogenesis of chronic intestinal inflammation. Although this approach is obviously justified, it is easy to lose sight of the peculiar conditions that allow the normal intestinal immune system to exert a protective rather than an aggressive role. As a matter of fact the intestinal immune system is not only the largest of the body, but primarily a biological shield functioning in unique ways that distinguish it from all other defense mechanisms. Because it is constantly exposed to the external environment, and carries an extremely rich and varied endogenous flora, the gut mucosa is adapted to work under intense, yet physiological, conditions of permanent antigenic pressure. This pressure requires and brings into action an enormous amount of organized (Peyer's patches and lymphoid follicles) and diffuse (intraepithelial lymphocytes and lamina propria mononuclear cells) lymphoid cells that respectively form the gut-associated lymphoid tissue (GALT) and the mucosaassociated lymphoid tissue (MALT) [1]. The existence of this population of mature immunocytes all along the lining of the gastrointestinal tract is quantitatively and qualitatively unparalleled in other organs, including those with a mucosal surface such as the oral cavity, the lungs and airways, the genitourinary tract, and the mammary glands. For this reason the term 'controlled' or 'physiological intestinal inflammation' has been coined to reflect the fact that activated immunocytes are present in large numbers and, rather than causing injury, afford instead an essential protection to the gut and indirectly the rest of the body. Despite its importance this terminology is seldom used outside of the realm

of mucosal immunology and it is difficult to find it even in comprehensive textbooks of mucosal immunology [2]. Interestingly, the majority of components and functions involved in physiological intestinal inflammation are the same responsible for pathological inflammation as found in IBD and other conditions. Thus, the question arises of what distinguishes one from the other, and the answer is the appropriateness or not of the local defense mechanisms. There is not enough information to clearly define what constitutes an inappropriate defense response in IBD and why it endures over time, but there is a reasonably good knowledge of the components responsible for physiological intestinal inflammation (Fig. 1). Since avoiding all dietary, microbial or self stimuli that are potentially harmful is practically impossible, the intestine has devised control mechanisms that aUow it to limit the quantity and quality of the antigens presented to the mucosa and GALT. According to Fig. 1 an appropriate response resulting in physiological inflammation relies on two types of control mechanisms: physical and biological. The first depends on the intrinsic properties of the gut and epithelium, while the latter involves circulatory, h u m o r a l and immune mechanisms. I m m u n e mechanisms rely on both innate and adaptive responses, and when specific immunity is required it can either incite a switch-on response by the induction of effector cells that eliminate the antigen (active immunity), or a switch-off" response resulting in a lack or suppression of response to the antigen (tolerance). An elaborate discussion of each component and function listed in the diagram is beyond the scope of this review, and some of these topics will be covered more comprehensively in other chapters of this book. Instead, the goal of this review is to provide the reader with an integrated overview of the components that together are presumably

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.J, Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 101-120. © 2003 Kluwer Academic Publishers. Printed in Great Britain

The normal intestinal mucosa

102

Stimuli Microbes, food antigens, self antigens, toxins '^ Appropriate response

^'

^^

Intestinal inflammation

Inappropriate response

Physiological / Controlled inflammation

Stimulus avoidance

Pathological / Uncontrolled inflammation

Stimulus control

Physical control Peristalsis, mucus, epithelial cells, microvilli, tight junctions Circulatory Trafficking, homing cell adhesion

Infectious, immune-mediated, autoimmune

Biological control

Humoral Defensins, lactoferrin, lysozyme, peroxidases

Innate / Non-specific immunity Neutrophils, macrophages, eosinophils, mast cells, NK cells, complement, defensins, cytokines, ROM, NO

Immune

\ Adaptive / Specific immunity T-cells, B-cells

Unresponsiveness (Tolerance)

Enteric flora

Self

Responsiveness (Active immunity)

Food antigens

Figure 1. Diagram of the various components responsible for the induction and maintenance of physiological intestinal inflammation. The relative thickness of the solid arrows indicates the proportional contribution of single components to each defensive system.

Claudio Fiocchi responsible for the induction and maintenance of physiological inflammation in the human intestine.

Physical control mechanisms Physical control mechanisms that contribute to physiological intestinal inflammation include primarily those that exclude, limit or select the amount and type of antigenic stimuli that can activate local lymphoid cells. This classification is arbitrary and somewhat artificial because even elements that exert an essentially pure mechanical function, such as peristalsis and the mucus coat, depend on and are integrated with other biological functions. These control mechanisms form what is commonly referred to as intestinal mucosal barrier.

Peristalsis Peristalsis, the coordinated contraction of the bowel propelling its luminal content aborally, prevents the unnecessary accumulation of food-derived proteins, bacteria, parasites or toxins, and therefore decreases the possibility of absorbing an excessive antigenic load that may induce inflammation [3].

IVIucus The clinical observation that mucus production and release is enhanced during gut inflammation has long been associated with a protective function. Intestinal mucus is a viscoelastic gel composed of large mucin molecules made up by a small protein core and a complex array of oligosaccharide chains. The various mucin gene products are selectively expressed by different cells along the small and large bowel, suggesting that each mucin plays a distinctive role in mucosal protection [4]. Mucus prevents mechanical damage of the mucosa, enables shedding and renewal of mucosal surfaces, functions as a selective barrier for macromolecules and microorganisms and a trap impeding the penetration of larger microbes, and retains defensive molecules on the luminal surface, such as secretory IgA (sIgA) [5]. Together, these integrated activities generate an important barrier that lowers the damaging and proinflammatory potential of the luminal contents.

103

Epithelial cells Just underneath the mucus coat lies a single cell layer of intestinal epithelial cells that offer a variety of physical and functional protective mechanisms. Intestinal epithelial cells are heterogeneous and include columnar, goblet, Paneth, enteroendocrine and undifferentiated stem cells [6]. Each type has unique morphological features translating distinct physiological and defensive roles. In addition, there are specialized epithelial cells overlying lymphoid follicles called follicle-associated epithelium. Such cells are referred to as M cells; they have less developed microvilli and play a significant role in limiting and selecting antigen sampling [7]. Epithelial cells provide diverse protective systems that include the microvilli with their large surface and negative charge, the lipid bilayer of the plasma membrane, intracellular organelles containing degradative enzymes, and the junctions between adjacent cells. Prominent among the effector mechanisms mediated by these various systems are enhanced salt and water secretion, expression of antimicrobial proteins and peptides, and production of mucin [8]. Most of these systems are regulated by locally produced cytokines [9], and epithelial cells themselves are a source of proinflammatory cytokines, especially in response to bacterial invasion [10]. This paradoxical phenomenon has been proposed to actually represent an early protective defense mechanism of the mucosa by limiting bacterial invasion through a repertoire of regulatory molecules including cytokines, chemokines, adhesion molecules, prostanoids, nitric oxide and the induction of epithelial cell apoptosis [11, 12]. Antigens can penetrate the epithelium directly through individual epithelial cells (transcellular route) or the space between two cells (paracellular route). When antigens are absorbed transcellularly they are internally processed by individual cells, destroyed or degraded for exposure on the cell surface in the context of major histocompatibility complex (MHC) class II antigens for presentation to local T cells. Although this series of events could be construed as stimulatory and potentially proinflammatory, under normal circumstances the ultimate effect is a limitation in antigen presentation and the selective activation of T cell subsets [13]. First, intestinal epithelial cells are less efficient than professional antigen-presenting cells (APC) such as macrophages and dendritic cells. Second, they activate preferentially CDS"^ cytotoxic/suppressor T cells or CD4"^ T cells putatively involved in induction of tolerance [13, 14]. Thus, the overall response is

The normal intestinal mucosa

104 actually one promoting containment of excessive immune reactivity at the mucosal level. When antigens penetrate the epithelium paracellularly the pathway utilized consists of the tight junctions (zonulae occludentes) and the subjunctional space. The tight junction is a complex and dynamic structure that constitutes a considerable barrier to large molecules and regulates the frequency, quality and quantity of antigen presentation to the adjacent and underlying immunocytes [15]. Various probes used for clinical measurement of small bowel permeability, including L-rhamnose, PEG400, and [^'Cr]EDTA, are believed to permeate the intestinal epithelial barrier through the paracellular tight junctions [16]. Their increased absorption in conditions such as Crohn's disease, celiac disease, infections, allergy and food intolerance indicates a loss of the protective function of the tight junctions under inflammatory conditions. The same apparently occurs in uninflamed ileal mucosa of Crohn's disease, and this could increase the antigen load in the mucosa and predispose to intestinal inflammation [17].

Biological control mechanisms Circulatory control GALT and MALT The impressive size and diffusiveness of the normal MALT, which forms the anatomical basis for physiological intestinal inflammation, imply the existence of highly efficient mechanisms that direct the circulation of lymphoid cells to the intestine and allow their retention in the mucosa where they can mediate a protective eff'ector function. Under physiological conditions lymphocytes circulate constantly throughout the body. This movement does not take place in a random fashion, but rather occurs under the control of a tightly regulated process coordinating the traffic of specific cell subsets to inductive (e.g. lymph nodes) and effector (e.g. mucosal surfaces) sites. In this regard there is a significant dichotomy in lymphocyte trafficking and distribution between naive (CD45A"^) and memory (antigenprimed CD45RO'^) cells: naive lymphocytes are programmed to circulate mainly among secondary lymphoid tissues (lymph nodes, tonsils, spleen and Peyer's patches) while memory cells preferentially access and recirculate to immune effector sites such as the intestinal lamina propria [18]. This distinction is of major importance to the intestinal immunity

because the vast majority of lymphocytes populating the normal intestinal mucosa is composed of mature memory cells [19]. These derive primarily from naive cells which have been primed by antigens sampled by M cells in the Peyer's patches and other organized GALT and recirculate to finally home in the lamina propria [1]. The implementation of this complex distribution system requires a series of signals and receptors on circulating leukocytes as well as the microvasculature to which immunocytes must adhere in order to penetrate the interstitial tissue. This task is accomplished through a proposed multistep paradigm in which leukocyte attachment to the vascular endothelium and the subsequent rolling, activation, arrest, spreading and transendothelial migration are mediated by specific cell adhesion and chemoattractant molecules [20]. Once in the interstitium the combined influence of local mesenchymal cells and the extracellular matrix provides a protective anti-apoptotic environment that prolongs the survival of immigrated lymphocytes [21] (Fig. 2).

Cell adhesion molecules Cell adhesion molecules are a large number of structurally and functionally related and unrelated molecules forming four major families: the selectin family which is primarily responsible for leukocyteendothelial cell interactions; the integrin family which mediates cell-cell and cell-extracellular matrix interactions, the immunoglobulin superfamily which mediates homophilic adhesion between an identical cell adhesion molecule and another cell, and the cadherin family which establishes molecular links between adjacent cells [22] (Table 1). The process of lymphocyte migration to and retention in the intestinal mucosa has two basic requirements. The first is the expression of a combination of adhesion molecules that impart tissue specificity to memory/effector cells, which for mucosal homing lymphocytes is represented by high levels of the integrin (x4p7 and ocL(32 (LFA-1, leukocyte functionassociated molecule) and low levels of L-selectin to avoid trapping in secondary lymphoid tissues. The second requirement is the coordinated participation of several cell adhesion molecules from different families, particularly those regulating the adhesion of leukocytes to the vascular endothelial cells (homing) and subsequent translocation into the interstitial space. These include L-, E-, and P-selectin of the selectin family, CD11/CD18, very late activation antigen (VLA) -4, and oc4p7 of the integrin family, intercellular cell adhesion molecule (ICAM) 1,

Claudio Fiocchi

105

Intestinal lumen

lEC TECK a. CCR9 /™^CXCR3* CCR5* / l ^ \ CCR7"

V^P ) ^^^^

> f ^ j

^^—^CXCR2" CXCR1"

FasL*

^ — ^ CD45RO*

MMEC IL-8, MCP-1

RANTES, GROa

TARC (?), SLC (?)

Intravascular space

CD45RA

PBT

Rolling CD18 Adhesion VLA-4 a4g71 P-selectin E-selectin

ICAM-1 VCAM-1 MAd-CAM-1

Transmigration PECAM

SMF

Interstitial space

Figure 2. Major chemokines, chemokine receptors, and cell adhesion molecules involved in the multiple steps necessary to attract and translocate (rolling, adhesion and transmigration) T cells from the intravascular to the Interstitial space, and to distribute and retain T cells in the intraepithelial and lamina propria compartments. In the upper part of the figure chemokines (TECK, IL-8, MCP-1, RANTES, GROa, TARC, and SLC) are shown below their respective cellular source: intestinal epithelial cells (lEC) and mucosal microvascular endothelial cells (MMEC). Receptors for CC and CXC chemokines (CCR2, CCR5, CCR7, CCR9, CXCR1, CXCR2, and CXCR3) are shown around the translocated T cells C indicates the expression and"" indicates the absence of expression of each chemokine receptor). In the lower part of the figure cell adhesion molecules expressed by T cells are shown to their left and cell adhesion molecules expressed by MMEC are shown below them. ECM, extracellular matrix; GRO, growth-related oncogene; ICAM, intercellular adhesion molecule; lEL, intraepithelial lymphocyte; IL-8, interleukin-8; LPT, lamina propria T cell; PBT, peripheral blood T cell; PECAM, platelet-endothelial cell adhesion molecule; RANTES, regulated on activation, normal T cell expressed and secreted; SLC, secondary lymphoid organ chemokine; SMF, subepithelial myofibroblast; TARC, thymus and activation-regulated chemokine; TECK, thymus expressed chemokine; VCAM, vascular cell adhesion molecule; VLA, very late activation antigen.

The normal intestinal mucosa

106

VJ

(U

o

» CD o

O

C O --o C

Q.

>, "

E

c o o ,

>,o

ex. o

'>O^ o>%"oo>» (DZ3

g

III

CO CD




E E E

C t

^

>; ^

^ ^

^

O

CO

y

CO

^

"o Q--55

•o

'ci

»- § S^ ^

^" CD

-

1,^


> > » 0 0 0 0 0 0 £Z C= C 0 0 0

CO " O > >,» !>^ >, !>> > :>N 0 0 0

-^-

E c: 0

cc "55

- ^.

E

.^

"53 0

0

-^ o

E

c o

'c/>

^

CD "T^ CD

CD -CZ •a 03

-^ .E .E .E •S "o t3 "o "o ^ ^ ^

O

"53 «f5 CO / I Th1

ThO

\ Th2

Anergy (Tolerance) High antigen dose

Low antigen dose

i

Th3 \ Active V suppression

Deletion or anergy

Tr1 Figure 3. Proposed mechanisms of adaptive immunity involved in the induction and maintenance of physiological inflammation in the normal intestine. Antigens derived from the diet and the commensal enteric flora can access the GALTand MALT through specialized cells forming the follicle-associated epithelium (FAE), intestinal epithelial cells (lEC), or the paracellular route between lEC. Antigens processed by non-professional antigen-presenting lEC bearing major histocompatibility complex (MHC) class I and CD1 molecules will preferentially activate CD8"^ T cells and result in suppression (far left). When antigens are processed by professional antigenpresenting cells (APC) bearing MHC class II antigens and B7-1/2 costimulatory molecules and are presented to 0028"^ 004"^ T cells this will result in active immunity with a non-polarized (ThO) or polarized (Th1 orTh2) profile depending on the nature of the antigen and the surrounding cytokine milieu (center left). When antigens are processed by professional APC bearing MHC class II antigens but presented through deficient B7-1 /2 costimulation or preferentially to CTLA-4"^ CD4'*' T cells this will result in tolerance mediated by active suppression (executed by ThS or Tr1 cells) or deletion/anergy depending on the dose of the antigen (center right). Antigens processed by non-professional antigen-presenting mucosal microvascular endothelial cells (MMEC) or mucosal myofibroblast (MMF)-bearing major histocompatibility (MHC) class I molecules but no costimulatory molecules will result in the generation of anergic T cells and the Induction of tolerance (far right).

Claudio Fiocchi immune responses to unrelated, but anatomically colocalized, antigens [109]. Active suppression is probably the most important mechanism of tolerance in the gut, not only because of the high likelihood of local T cells to re-encounter antigens to which they have been previously sensitized, but also because of the abundance of mature effector T cells capable of producing immunosuppressive molecules [110]. Oral tolerance The tolerance mechanisms discussed so far are based primarily on systemic immune responses to self rather than foreign protein antigens. The nature, dose and way of administration of any foreign protein are key determining factors in deciding whether active immunity or tolerance will ensue. When a foreign protein enters the body via the gastrointestinal tract this usually leads to a marked suppression of systemic humoral and cell-mediated immune responses upon rechallenge by the same protein. This phenomenon is called oral tolerance, and is believed to be essential in preventing deleterious immune responses against proteins that need to be absorbed for nutritional purposes, and developing tolerance against the antigens of the normal enteric flora that is indispensable to energy metabohsm, nutrient absorption and general health [109, 111, 112]. Various mechanisms mediate oral tolerance but generally, based on experimental animal systems, a high antigen dose induces anergy or deletion [113, 114], whereas a low antigen dose induces active T cell-mediated suppression (Fig. 3). This latter form of regulation is mediated the immunosuppressive cytokines IL-4, IL-10 and TGF-P produced by various types of CDA^ T cells with functional phenotypes of Th2, Th3, Treg (T regulatory cell) or Trl (T regulatory cell 1), all of which are involved in systemic or organ-specific autoimmune phenomena [115-117]. There is good evidence that these regulatory T cells exist in the gut of normal humans and animals with experimental intestinal inflammation [118-120], as well as evidence for the anti-inflammatory activity of locally secreted cytokines such as IL-10 and TGF-P [120-122]. Broken tolerance: impact on intestinal immune homeostasis To be fully protective oral tolerance must be developed for the two main components of the luminal content: food antigens and the endogenous flora. Ingestion of specific proteins and peptides is widely utilized in studies of systemic and organ-specific

115 immunity and autoimmunity, as well as a form of t r e a t m e n t for a u t o i m m u n e d i s e a s e s [123]. Unfortunately, except for classical food allergies, a form of immediate hypersensitivity reaction mediated primarily by mast cells and eosinophils [124], little is known about the role of normal dietary antigen in physiological or pathological intestinal inflammation. In contrast, knowledge on the role of the commensal flora on intestinal immunity and inflammation has greatly expanded in recent years, justifying some discussion on how normal enteric bacteria induce and maintain a controlled state of inflammation in the mucosa, and how loss of tolerance to these bacteria may contribute to chronic intestinal inflammation as observed in IBD. Gut microbial-immune interactions The development of a mature intestinal immune system is strictly dependent on the introduction of environmental antigens into the lumen, e.g. food and microorganisms [125]. Immediately after birth the newborn's gastrointestinal tract, previously sterile, is colonized with aerobic and anaerobic bacteria [126]. The ability to induce tolerance requires colonization by Gram-negative bacteria early in life, and germfree conditions or exposure to certain aerobic Grampositive bacteria may impair development of oral tolerance mechanisms [127]. With advancing age, and feeding of complex diets, the gut acquires a full spectrum of microflora that increases in number and variety from the almost sterile stomach to the luxuriant milieu of the colon, where Bacteroides, Bifidobacteria, Peptostreptococcus and Eubacterium spp. predominate. The type of feeding has a significant impact on colonization: for example, distinct from breast-fed infants, Bacteroides count increases in formula-fed infants, and anaerobic counts remain high after introduction of solid food [128]. The normal human intestinal biota consists of hundreds of bacterial species that vary considerably from person to person, though the predominant phylogenetic groups do not change [129]. The interdependence of the bacterial and immune systems in the gut is mutual and lifelong, and enteric bacteria exert a wide range of modulatory effects on specific and non-specific immune responses. In animals they range from mitogenic effects on lymphocytes to induction of N K and cytotoxic cell activity, enhanced antibody production, stimulation of macrophage phagocytosis, cytokine production, and induction of oxygen free radicals [128]. In humans the enteric bacteria affect both systemic and

116 mucosal immunity, including modulation of IL-1, IL-2, IL-6, TNF-a, and IFN-y production, antigenand mitogen-induced lymphocyte proliferation, and macrophage phagocytic and killing activity [130]. Normal enteric flora and pathological intestinal inflammation There is compelling evidence for a decisive role of the commensal flora in pathological gut inflammation. Four lines of evidence support this claim. The strongest evidence for a role of the normal enteric flora in IBD comes from experimental models of bowel inflammation in transgenic and gene-deleted rodents [131]. For instance, HLA-B27/human P2microglobulin transgenic rats and IL-10-deficient mice fail to develop colitis under germfree conditions, but do so when exposed to normal luminal bacteria, and the load and composition of commensal bacteria influences the degree of inflammation [132-134]. Additional support comes from human studies in which the fecal content has been removed from or put in contact with intestinal mucosa. Crohn's disease patients with fecal stream diversion after ileal resection have no inflammation in the neoterminal ileum, but inflammation quickly reappears after reanastomosis [135]. Furthermore, infusion of intestinal luminal contents into histologically normal excluded ileal loops after ileocolonic resection induces inflammatory changes within a few days [136]. Another example of how luminal content affects the health of the intestinal mucosa derives from ulcerative colitis patients with a reconstructed ileo-anal pouch who develop pouchitis, a recurrent inflammation in the remodeled small bowel loops. This condition is believed to be caused by a bacterial dysbiosis induced by the presence of colonic-type bacteria in the small bowel loops forming the pouch [137]. Finally, antibiotics can be beneficial in the management of IBD and particularly CD [138], and probiotics appear to be a promising new form of treatment for IBD and pouchitis [139, 140]. Loss of tolerance to normal enteric flora in intestinal inflammation If gut-induced tolerance is crucial to prevent systemic immune responses, the same must be true for local responses, and it seems logical that gut inflammation may result from loss of local tolerance. This concept finds support in a series of human and animal studies. LPMC isolated from the inflamed, but not uninflamed, intestine of adult IBD patients proliferate strongly in response to autologous, but

The normal intestinal mucosa not heterologous, intestinal bacteria [141]. LPMC from control and IBD patients in remission fail to proliferate to autologous bacteria. This suggests that tolerance to commensal flora normally exists in the healthy intestine, and that tolerance is broken during inflammation. Loss of tolerance to commensal flora is also demonstrable in murine hapten-induced colitis, and tolerance can be restored by treatment with IL-10 or neutralization of IL-12 [142]. Gut bacteria-reactive T cells are found in patients with IBD and murine models of IBD [119, 143-145], and the adoptive transfer of bacterial antigen-specific CD4"*" T cells from diseased into naive animals induces colitis [119, 145]. Taken together, these results indicate that loss of tolerance to autologous enteric flora is instrumental in the pathogenesis of some forms of pathological intestinal inflammation.

Implications of physiological Intestinal inflammation for health and disease Having defined some of the physical and biological control mechanisms responsible for the development and maintenance of a state of controlled inflammation in the intestine, few final considerations are in order to conclude what physiological inflammation really is and what its role might be in health and disease. This may be best done by posing a series of intriguing questions. How much physiological inflammation is enough? Why does the degree of mucosal leukocyte infiltration vary among normal individuals? Why is physiological inflammation symptom-free? Why does a certain degree of mucosal lymphoid cell infiltration represent physiological inflammation in one individual when the same degree may be accompanied by symptoms of pathological inflammation in another? In other words, when does gut inflammation cease to be physiological and become pathological? Clinical experience shows that IBD patients in remission often display significant inflammatory changes at endoscopic or histological examination, and that a routine screening colonoscopy reveals the presence of florid inflammation in an individual totally free of gastrointestinal symptoms. These practical observations indicate that physiological intestinal inflammation cannot be simply equated to a certain number of cells per volume of mucosa just as much as pathological inflammation cannot be defined only by clinical

Claudio Fiocchi symptoms. Therefore, it is reasonable to assume that, depending on genetic make-up, a healthy subject can have disparate amounts of immunocytes in the gastrointestinal mucosa. This amount is almost certainly modified by the quality of his or her diet, as well as the concentration and type of commensal flora, as shown by differences between individuals living in northern vs tropical areas. While these concepts may be intuitively acceptable, on the other hand they complicate the definition of physiological intestinal inflammation. Since it appears that what is physiological in one person may be pathological in another, perhaps it would be best to assume that what constitutes physiological inflammation in response to normal stimuli is as tailor-made as the response each person puts forward in the face of pathological stimuli.

References 1.

2. 3. 4.

5.

6.

7. 8. 9. 10.

11. 12. 13. 14.

Brandtzaeg P, Farstad IN, Haraldsen G. Regional specialization in the mucosal immune system: primed cells do not always home along the same track. Immunol Today 1999; 20: 267-77. Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999. Mayer EA, Raybould H, Koelbel C. Neuropeptides, inflammation and motility. Dig Dis Sci 1988; 33: 71S-7S. Chang S-K, Dohrman AF, Basbaum CB et al Localization of mucin (MUC2 and MUC3) messenger RNA and peptide expression in human normal intestine and colon cancer. Gastroenterology 1994; 107: 28-36. Cone RA, Mucus. In: Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, editors. Mucosal Immunity, 2nd edn. San Diego: Academic Press, 1999: 4 3 64. Kato T, Owen RL. Structure and function of intestinal mucosal epithelium. In: Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, editors. Mucosal Immunity, 2nd edn. San Diego: Academic Press, 1999: 11532. Gebert A, Rothkotter HJ, Pabst R. M cells in Peyer's patches of the intestine. Int Rev Cytol 1996; 167: 91-159. Hecht G. Innate mechanisms of epithelial host defense: spothght on intestine. Am J Physiol 1999; 277: C351-68. McKay DM, Baird AW. Cytokine regulation of epithelial permeability and ion transport. Gut 1999; 44: 283-9. Jung HC, Eckmann L, Yang S-K, et al. A distinct array of proinflammatory cytokines is expressed in human colon epithelial cells in response to bacterial invasion. J Clin Invest 1995; 95: 55-65. Kagnoff MF, Eckmann L. Epithehal cells as sensors for microbial infection. J Clin Invest 1997; 100: 6-10. Kim JM, Eckmann L, Savidge TC, Lowe DC, Witthoft T, Kagnoff" MF. Apoptosis of human intestinal epithelial cells after bacterial invasion. J Clin Invest 1998; 102: 1815-23. Hershberg RM, Mayer LF. Antigen processing and presentation by intestinal epithelial cells - polarity and complexity. Immunol Today 2000; 21: 123-8. Mayer L, Shlien R. Evidence for function of la molecules on gut epithehal cells in man. J Exp Med 1987; 166: 1471-83.

117 15. Madara JL. Maintenance of the macromolecular barrier at cell extrusion sites in intestinal epithelium: physiological rearrangements of tight junctions. J Membr Biol 1990; 116: 177-84. 16. Bjarnason I, MacPherson A, Hollander D. Intestinal permeability: an overview. Gastroenterology 1995; 108: 156681. 17. Soderholm JA, Peterson KM, Olaison G et al. Epithelial permeability to proteins in the noninflamed ileum of Crohn's disease? Gastroenterology 1999; 117: 65-72. 18. Butcher EC, Picker LJ. Lymphocyte homing and homeostasis. Science 1996; 272: 60-6. 19. Schieferdecker HL, Ullrich R, Hirseland H, Zeitz M. T cell differentiation antigens on lymphocytes in the human intestinal lamina propria. J Immunol 1992; 149: 2816-22. 20. Springer TA. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 1994; 76: 301-14. 21. Akbar AN, Salmon M. Cellular environments and apoptosis: tissue microenvironments control activated T cell death. Immunol Today 1997; 18: 72-6. 22. Frenette PS, Wagner DD. Adhesion molecules - Part I. N Engl J Med 1996; 334: 1526-9. 23. Panes J, Granger DN. Leukocyte-endothelial interactions: molecular mechanisms and implications in gastrointestinal disease. Gastroenterology 1998; 114: 1066-90. 24. Briskin M, Winsor-Hines D, Shyjan A et al. Human mucosal addressin cell adhesion molecule-1 is preferentially expressed in intestinal tract and associated lymphoid tissue. Am J Pathol 1997; 151: 97-110. 25. Salmi M, Jalkanen S. Molecules controlling lymphocyte migration to the gut. Gut 1999; 45: 148-53. 26. Kunkel SL. Through the looking glass: the diverse in vivo activities of chemokines. J Chn Invest 1999; 104: 1333-4. 27. Cyster JG. Chemokines and cell migration in secondary lymphoid organs. Science 1999; 286: 2098-102. 28. Luster AD. Chemokines - chemotactic cytokines that mediate inflammation. N Engl J Med 1998; 338: 436-45. 29. Utgaard JO, Jahnsen FL, Bakka A, Brandtzaeg P. Rapid secretion of prestored interleukin 8 from Weibel-Palade bodies of microvascular endothelial cells. J Exp Med 1998; 188:1751-6. 30. Premack BA, Schall TJ. Chemokine receptors: gateways to inflammation and infection. Nature Med 1996; 2: 1174-8. 31. Ward SG, Bacon K, Westwick J. Chemokines and Tlymphocytes: more than an attraction. Immunity 1998; 9: 1-11. 32. Campbell JJ, Haraldson G, Pan J et al. The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memeory T cells. Nature 1999; 400: 776-80. 33. Mackay CR. Dual personality of memory T cells. Nature 1999;401:659-60. 34. Zabel BA, Agace WA, Campbell JJ et al. Human G proteincoupled receptor GPR-9-6/CC chemokine receptor 9 is selectively expressed on intestinal homing T lymphocytes, mucosal lymphocytes, and thymocytes and is required for thymus-expressed chemokine-mediated chemotaxis. J Exp Med 1999; 190: 1241-55. 35. Papadakis KA, Prehn J, Nelson V et al. The role of thymusexpressed chemokine and its receptor CCR9 in the regional specialization of the mucosal immune system. J Immunol 2000; 165: 5069-76. 36. Bonecchi R, Bianchi G, Bordignon PP et al. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (This) and Th2s. J Exp Med 1998;187: 129-34. 37. Agace WW, Roberts Al, Wu L, Greineder C, Ebert EC, Parker CM. Human intestinal lamina propria and intraepithelial lymphocytes express receptors specific for chemo-

The normal intestinal mucosa

118

38.

39.

40. 41. 42.

43. 44. 45. 46. 47. 48. 49.

50. 51.

52. 53. 54.

55. 56.

57.

kines induced by inflammation. Eur J Immunol 2000; 30: 819-26. Salmi M, Granfors K, MacDermott RP, Jalkanen S. Aberrant binding of lamina propria lymphocytes to vascular endothelium in inflammatory bowel disease. Gastroenterology 1994; 106: 596-605. MacDermott RP, Sanderson IR, Reinecker H-C. The central role of chemokines (chemotactic cytokines) in the immunopathogenesis of ulcerative colitis and Crohn's disease. Inflamm Bowel Dis 1998; 4: 54-67. Selsted ME, Oullette AJ. Defensins in granules of phagocytic and phagocytic cells. Trends Cell Biol 1995; 5: 114-19. Ouellette A J. Paneth cells and innate immunity in the crypt microenvironment. Gastroenterology 1997; 113: 1779-84. Lehrer RI, Bevins CL, Ganz T. Defensins and other antimicrobial peptides. In: Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, editors. Mucosal Immunity, 2nd edn. San Diego: Academic Press, 1999: 8999. Yang D, Chertov O, Bykovskaia SN et al. P-Defensins: linking innate and adaptive immunity through dendritic and T cell CCR6. Science 1999; 286: 525-8. Tomita M, Takase M, Bellamy W, Shimamura S. A review: the active peptide of lactoferrin. Acta Paediatr Jpn 1994; 36: 585 91. Fearon DT, Locksley RM. The instructive role of innate immunity in the acquired immune response. Science 1996; 272: 50 3. Medzhitov R, Janeway C. Innate immunity. N Engl J Med 2000;343: 338 44. Medzhitov R, Janeway CA. Innate immunity: impact on the adaptive immune response. Curr Opin Immunol 1997; 9: 4 9. Rock FL, Hardiman G, Timans JC, Kastelein RA, Bazan JF. A family of human receptors structurally related to Dwsophila JoW. Proc Natl Acad Sci USA 1998; 95: 588 93. Selby WS, Poulter LW, Hobbs S, Jewell D, Janossy G. Heterogeneity of HLA-DR positive histiocytes of human intestinal lamina propria: a combined histochemical and immunological analysis. J Clin Pathol 1983; 36: 379 84. Allison MC, Cornwall S, Poulter LW, Dhillon AP, Pounder RE. Macrophage heterogeneity in normal colonic mucosa and in inflammatory bowel disease. Gut 1988; 29: 1531-8. Hara J, Ohtani H, Matsumoto T et al. Expression of costimulatory molecules B7-1 anf B7-1 in macrophages and granulomas of Crohn's disease: demonstration of cellto-cell contact with T lymphocytes. Lab Invest 1997; 77: 175-84. Golder JP, Doe WF. Isolation and preliminary characterization of human intestinal macrophages. Gastroenterology 1983;84:795-802. Mahida YR, Wu KC, Jewell DP. Respiratory burst activity of intestinal macrophages in normal and inflammatory bowel disease. Gut 1989; 30: 1362-70. Rugtveit J, Nilsen EM, Bakka A, Carlsen H, Brandtzaeg P, Scott H. Cytokine profiles differ in newly recruited and resident subsets of mucosal macrophages from inflammatory bowel disease. Gastroenterology 1997; 112: 1493-505. Abbas AK, Lichtman AH, Pober JS, editors. Cellular and Molecular Immunology, 4th edn. Philadelphia: W.B. Saunders, 2000. Simmonds NJ, Allen RE, Stevens TRJ et al. Chemiluminescence assay of mucosal reactive oxygen metabolites in inflammatory bowel disease. Gastroenterology 1992; 103: 186-96. Oshitani N, Kitano A, Okabe H, Nakamura S, Matsumoto T, Kobayashi K. Location of superoxide anion generation in human colonic mucosa obtained by biopsy. Gut 1993; 34: 936-8.

58. 59.

60.

61. 62. 63.

64.

65.

66.

67. 68. 69.

70. 71.

72. 73. 74. 75. 76. 77.

Buffington GD, Doe WF. Depleted mucosal antioxidant defences in inflammatory bowel disease. Free Radical Biol Med 1995; 19:911-18. Dijkstra G, Moshage H, VanDulleman HM et al. Expression of nitric oxide synthases and formation of nitrotyrosine and reactive oxygen species in inflammatory bowel disease. J Pathol 1998; 186:416-21. Singer II, Kawka DW, Scott S et al. Expression of inducible nitric oxide synthase and nitrotyrosine in colonic epithelium in inflammatory bowel disease. Gastroenterology 1996; 111: 871-85. Wallace JL, Miller MJS. Nitric oxide in mucosal defense: a little goes a long way. Gastroenterology 2000; 119: 512-20. Carroll MC. The role of complement and complement receptors in induction and regulation of immunity. Annu Rev Immunol 1998; 16: 545-68. Halstensen TS, Mollnes TE, Brandtzaeg P. Persistent complement activation in submucosal vessels of active inflammatory bowel disease: immunohistochemical evidence. Gastroenterology 1989; 97: 10-19. Ahrenstedt O, Knutson L, Nilsson B, Nilsson-Ekdahl K, Odlind B, Hallgren R. Enhanced local production of complement components in the small intestines of patients with Crohn's disease. N Engl J Med 1990; 322: 1345 9. Russell MW, Kilian M, Lamm ME. Biological activities of IgA. In: Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, editors. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999. Carbonare SB, Silva MLM, Trabulsi LR, Carneiro-Sampaio MMS. Inhibition o^ HEp-2 cell invasion by entroinvasive Escherichia coli by human colostrum IgA. Int Arch Allergy Immunol 1995; 198: 113 18. Childers NK, Bruce MG, McGhee JR. Molecular mechanisms of immunoglobulin A defense. Annu Rev Microbiol 1989; 43: 503 36. Walker WA, Issclbacher KJ, Bloch KJ. Intestinal uptake of macromolecules: effect of oral immunization. Science 1972; 177:608 10. Nikolova EB, Tomana M, Russell MW. All forms of human IgA antibodies bound to antigen interfere with complement (C3) fixation induced by IgG or by antigen alone. Scand J Immunol 1994; 39: 275-80. VanKerckhove C, Russell GJ, Deusch K et al. Oligoclonality of human intestinal intraepithelial T cells. J Exp Med 1992; 175: 57-63. Gross GC, Schwartz VL, Stevens C, Ebert EC, Blumberg RS, Balk SP Distribution of dominant T cell receptor P chain in human intestinal mucosa. J Exp Med 1994; 180: 1337-44. Ladel CH, Blum C, Dreher A, Reifenberg K, Kaufmann SH. Protective role of Y / 5 T cells and a/p T cells in tuberculosis. Eur J Immunol 1995; 25: 2877-81. Ebert EC. Proliferative responses of human intraepithelial lymphocytes to various T cell stimuli. Gastroenterology 1989; 97: 1372-81. Cerf-Bensussan N, Guy-Grand D, Griscelli C. Intraepithelial lymphocytes of human gut: isolation, characterization and study of natural killer activity. Gut 1985; 26: 81-8. Taunk J, Roberts Al, Ebert EC. Spontaneous cytotoxicity of human intrapithelial lymphocytes against epithelial cell tumors. Gastroenterology 1992; 102: 69-75. Ebert EC. Human intestinal intraepithelial lymphocytes have potent chemotactic activity. Gastroenterology 1996; 109: 1154-9. James SP, Fiocchi C, Graeff" AS, Strober W. Phenotypic analysis of lamina propria lymphocytes. Predominance of helper-inducer and cytolytic T cell phenotype and deficiency of suppressor-inducer phenotype in Crohn's disease and control patients. Gastroenterology 1986; 91: 1483-9.

Claudio Fiocchi 78.

79.

80. 81. 82.

83.

84.

85.

86.

87. 88. 89.

90. 91. 92.

93. 94.

95.

Saubermann LJ, Probert CSJ, Christ AD et al. Evidence of T cell receptor p-chain patterns in inflammatory and noninflammatory bowel disease states. Am J Physiol 1999; 276: G613-21. Pallone F, Fais S, Squarcia O, Biancone L, Pozzilli P, Boirivant M. Activation of peripheral blood and intestinal lamina propria lymphocytes in Crohn's disease. In vivo state of activation and in vitro response to stimulation as defined by the expression of early activation antigens. Gut 1987; 28: 745-53. Matsuura T, West GA, Youngman KR, Klein JS, Fiocchi C. Immune activation genes in inflammatory bowel disease. Gastroenterology 1993; 104: 448-58. DeMaria R, Boirivant M, Cifone MG et al. Functional expression of Fas and Fas ligand on human gut lamina propria T lymphocytes. J Clin Invest 1996; 97: 316-22. Boirivant M, Pica R, DeMaria R, Testi R, Paflone F, Strober W. Stimulated human lamina propria T cells manifest enhanced Fas-mediated apoptosis. J Chn Invest 1996; 98: 2616-22. Ina K, Itoh J, Fukushima K et al. Resistance of Crohn's disease T cells to multiple apoptotic stimuli is associated with a Bcl-2/Bax mucosal imbalance. J Immunol 1999; 163: 1081-90. Boirivant M, Marini M, DiFelice G et al. Lamina propria T cells in Crohn's disease and other gastrointestinal inflammation show defective CD2 pathway-induced apoptosis. Gastroenterology 1999; 116: 557-65. Fiocchi C, Battisto JR, Farmer RG. Studies on isolated gut mucosal lymphocytes in inflammatory bowel disease. Detection of activated T cells and enhanced proliferation to Staphylococcus areus and lipopolysaccharides. Dig Dis Sci 1981;26:728-36. MacDermott RP, Franklin GO, Jenkins KM, Kodner IJ, Nash GS, Weinrieb IJ. Human intestinal mononuclear cells. I. Investigation of antibody-dependent, lectin-induced, and spontaneous cell-mediated cytotoxic capabilities. Gastroenterology 1980; 78: 47-56. Fiocchi C, Tubbs RR, Youngman KR. Human intestinal mucosal mononuclear cells exhibit lymphokine-activated killer ceU activity. Gastroenterology 1985; 88: 625-37. Shanahan S, Brogan M, Targan S. Human mucosal cytotoxic effector cells. Gastroenterology 1987; 92: 1951-7. MacDermott RP, Bragdon MJ, Jenkins KM, Franklin GO, Shedlofsky S, Kodner IJ. Human intestinal mononuclear cells. II. Demonstration of a naturally occurring subclass of T cells which respond in the allogeneic mixed lymphocyte reaction but do not effect cell-mediated lympholysis. Gastroenterology 1981; 80: 748-57. Fiocchi C. Cytokines. In: Targan SR, Shanahan F, editors. Inflammatory Bowel Disease. From Bench to Bedside. Baltimore: Williams & Wilkins, 1994: 106-22. Fuss IJ, Neurath M, Boirivant M et al. Disparate CD4+ lamina propria lymphokine secretion profiles in inflammatory bowel disease. J Immunol 1996; 157: 1261-70. Pirzer UC, Schurmann G, Post S, Betzler M, Meuer SC. Differential responsiveness to CD3-Ti vs CD2-dependent activation of human intestinal lymphocytes. Eur J Immunol 1990; 20: 2339-42. Qiao L, Schurmann G, Betzler M, Meuer SC. Activation and signaling status of human lamina propria T lymphocytes. Gastroenterology 1991; 101: 1529-36. Qiao L, Schurmann G, Betzler M, Meuer SC. Downregulation of protein kinase C activation in human lamina propria T lymphocytes: influence of intestinal mucosa on T cell reactivity. Eur J Immunol 1991; 21: 2385-9. Qiao L, Schurmann G, Autschbach F, Wallich R, Meuer SC. Human intestinal mucosa alters T cell reactivities. Gastroenterology 1993; 105: 814-19.

119 96.

97. 98. 99.

100.

101. 102. 103. 104. 105. 106. 107. 108. 109. 110.

111. 112. 113.

114. 115.

116.

117.

Meuer SC, Autschbach F, Schurmann G, Golling M, Braunstein J, Qiao L. Molecular mechanisms securing 'unresponsiveness' in lamina propria T lymphocytes. NY Acad Sci 1996; 778: 174^84. Bell GM, Imboden JB. CD2 and the regulation of T cell anergy. J Immunol 1995; 154: 2805-6. Quill H. Anergy as a mechanims of peripheral T cell tolerance. J Immunol 1996; 156: 1325-7. Zeitz M, Quinn TC, Graeff" AS, James SP. Mucosal T cells provide helper function but do not proliferate when stimulated by specific antigen in lymphogranuloma venereum proctitis in nonhuman primates. Gastroenterology 1988; 94: 353-66. McGhee JR, Lamm ME, Strober W. Mucosal immune responses. In: Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, editors. Mucosal Immunity,2nd edn. San Diego: Academic Press, 1999: 485-506. VanParijs L, Abbas AK. Homeostasis and self-tolerance in the immune system: turning lymphocytes off". Science 1998; 280: 243-8. Bluestone JA. Is CTLA-4 a master switch for peripheral T cell tolerance? J Immunol 1997; 158: 1989-93. Salojin KV, Zhang J, Madrenas J, Delovitch XL. T cell anergy and altered T cell receptor signaling: eff'ects on autoimmune disease. Immunol Today 1998; 19: 468-73. Reiser H, Stadecker MJ. Costimulatory B7 molecules in the pathogenesis of infections and autoimmune diseases. N Engl J Med 1996; 335: 1369-77. Marelli-Berg FM, Lechler RI. Antigen presentation by parenchymal cells: a route to peripheral tolerance? Immunol Rev 1999; 172:297-314. Kabelitz D, Pohl T, Pechhold K. Activation-induced cell death (apoptosis) of mature peripheral blood T lymphocytes. Immunol Today 1993; 14: 338-9. Lynch DH, Ramsdell F, Alderson MR. Fas and FasL in the homeostatic regulation of immune responses. Immunol Today 1995;16:569-74. Shevach EA. Regulatory T cells in autoimmunity. Annu Rev Immunol 2000; 18:423-49. Strobel S, Mowat AM. Immune responses to dietary antigens: oral tolerance. Immunol Today 1998; 19: 173-81. Podolsky DK, Fiocchi C. Cytokines, chemokines, growth factors, eicosanoids and other bioactive molecules in IBD. In: Kirsner JB, editor. Inflammatory Bowel Disease. Philadelphia: WB. Saunders, 1999: 191-207. Husby S. Normal immune responses to ingested foods. J Pediatr Gastroenterol Nutr 2000; 30: SI3-19. Simecka JW. Mucosal immunity of the gastrointestinal tract and oral tolerance. Adv Drug Deliv Rev 1998; 34: 235-59. Whitaker CC, Gienapp IE, Orosz CG, Bitar D. Oral tolerance in experimental autoimmune encephalomyelitis. III. Evidence for clonal anergy. J Immunol 1991; 147: 215563. Chen Y, Inobe J-I, Marks R, Gonnella P, Kuchroo VK, Weiner HL. Peripheral deletion of antigen-reactive T cells in oral tolerance. Science 1995; 376: 177-80. Saoudi A, Seddon B, Heath V, Fowell D, Mason D. The physiological role of regulatory T cells in the prevention of autoimmunity: the function of the thymus in the generation of the regulatory T cell subset. Immunol Rev 1996; 149: 195-216. Chen Y, Kuchroo VK, Inobe J-I, Hafler DA, Weiner HL. Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis. Science 1994; 265: 1237-40. Groux H, O'Garra A, Bigler M et al. A CD4'^ T ceU subset inhibits antigen-specific T cell responses and prevents colitis. Nature 1997; 389: 737-42.

120 118. Khoo UY, Proctor IE, Macpherson AJS. CD4^ T cell down-regulation in human intestinal mucosa. Evidence for intestinal tolerance to luminal bacterial antigens. J Immunol 1997; 158:3626-34. 119. Cong BY, Brandwein SL, McCabe RP et al CD4+ T cells reactive to enteric bacterial antigens in spontaneously colitic C3H/HeJBir mice: increased T helper cell type 1 response and ability to transfer disease. J Exp Med 1998; 187: 85564. 120. Read S, Malmstrom V, Powrie F. Cytotoxic T lymphocyteassociated antigen 4 plays an essential role in the function of CD25+CD4+ regulatory cells that control intestinal inflammation. J Exp Med 2000; 192: 295-302. 121. Asseman C, Mauze S, Leach MW, Coff'man RL, Powrie F. An essential role for interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation. J Exp Med 1999; 190:995-1003. 122. Gorelik L, Flavell RA. Abrogation of TGF(3 signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 2000; 12: 171-81. 123. Weiner HL. Oral tolerance: immune mechanisms and treatment of autoimmune diseases. Immunol Today 1997; 18: 335-43. 124. Crowe SE. Food allergies. In: Ogra PL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, editors. Mucosal Immunity, 2nd edn. San Diego: Academic Press, 1999: 1129 39. 125. Fiocchi C. The immunological resources of the large bowel. In: Kirsner JB, Shorter RG, editors. Diseases of the Colon, Rectum, and Anal Canal. Baltimore: Williams & Wilkins, 1988:95 117. 126. Fuller R. A review: Probiotics in man and animals. J Appl Bacteriol 1989; 66: 365-78. 127. Kolb H, Pozzilli P. Cow's milk and type I diabetes: the gut immune system deserves attention. Immunol Today 1999; 20: 108 110. 128. Naidu AS, Bidlack WR, Clemens RA. Probiotic spectra of lactic acid bacteria (LAB). Critical Rev Food Sci Nutr 1999; 38: 13 126. 129. Wilson KH, Ikeda JS, Blitchington RB. Phylogenetic placement of community members in human colonic biota. Clin Infect Dis 1997; 25: SI 14-16. 130. Famularo G, Moretti S, Marcellini S, DeSimone C. Stimulation of immunity by probiotics. In: Fuller R, editor. Probiotics 2. Applications and Practical Aspects. London: Chapman & Hall, 1997: 133-61. 131. Elson CO, Sartor RB, Tennyson GS, Riddell RH. Experimental models of inflammatory bowel disease. Gastroenterology 1995;109: 1344-67. 132. Rath HC, Herfarth HH, Ikeda JS et al. Normal luminal bacteria, especially Bacteroides species, mediate chronic colitis, gastritis, and arthritis in HLA-B27/human p2 microglobulin transgenic rats. J Clin Invest 1996; 98: 945-53.

The normal intestinal mucosa 133. Rath HC, Ikeda JS, Linde H-J, Scholmerich J, Wilson KH, Sartor RB. Varying cecal bacterial loads influences colitis and gastritis in HLA-B27 transgenic rats. Gastroenterology 1999; 116:310-19. 134. Sellon RK, Tonkonogy S, Schultz M et al. Resident enteric bacteria are necessary for development of spontaneous colitis and immune system activation in interleukin-10deficient mice. Infect Immun 1998; 66: 5224-31. 135. Rutgeerts P, Geboes K, Peeters M et al. Effect of faecal stream diversion on recurrence of Crohn's disease in the neoterminal ileum. Lancet 1991; 2: 771-4. 136. D'Haens G, Geboes K, Peeters M, Baert F, Penninckx F, Rutgeerts P. Early lesions caused by infusion of intestinal contents in excluded ileum of Crohn's disease. Gastroenterology 1998; 114:262-7. 137. Ruseler-van-Embden JGH, Schouten WR, van Lieshout LMC. Pouchitis: result of microbial imbalance? Gut 1994; 35: 658-64. 138. Prantera C, Scribano ML, Berto E, Zannoni F. Antibiotic use in Crohn's disease - why and how? Biodrugs 1997; 8: 293-306. 139. Turunen UM, Farkkila MA, Hakala K et al. Long-term treatment of ulcerative colitis with ciprofloxacin: a prospective, double-blind, placebo-controlled study. Gastroenterology 1998; 115: 1072 8. 140. Gionchelti P, Rizzello F, Venturi A et al. Oral bacteriotherapy as maintenance treatment in patients with chronic pouchitis: a double-blind, placebo-controlled trial. Gastroenterology 2000; 119: 305-9. 141. Duchmann R, Kaiser I, Hermann E, Mayet W, Ewe K, Meyer zum Buschenfelde KH. Tolerance exists towards resident intestinal flora but it is broken in active inflammatory bowel disease (IBD). Clin Exp Immunol 1995; 102: 448-55. 142. Duchmann R, Schmitt E, Knolle P, Meyer zum Buschenfelde KH, Neurath M. Tolerance towards resident intestinal flora in mice is abrogated in experimental colitis and restored by treatment with interlukin-10 or antibodies to interleukin-12. Eur J Immunol 1996; 26: 934-8. 143. Pirzer U, Schonhaar A, Fleischer B, Hermann E, Meyer zum Buschenfelde KH. Reactivity of infiltrating T lymphocytes with microbial antigens in Crohn's disease. Lancet 1991:338: 1238-9. 144. Takahashi I, lijima H, Katashima R, Itakura M, Kiyono H. Clonal expansion of CD4^ T cells in TCR a-chain-deficient mice by gut-derived antigens. J Immunol 1999; 162: 184350. 145. Wirtz S, Finotto S, Kanzler S et al. Chronic intestinal inflammation in STAT-4 transgenic mice: characterization of disease and adoptive transfer by TNF- plus IFN-yproducing CD4+ T cells that respond to bacterial antigens. J Immunol 1999; 162: 1884-8.

6

The lymphocyte-epithelial-bacterial interface ROBERT HERSHBERG AND RICHARD S. BLUMBERG

Introduction The immunologic and antigenic challenge faced along the gastrointestinal tract is truly extraordinary. Across a surface area that approximates a tennis court, it is here that we separate the highest concentration of foreign antigen (mostly food and bacteria) from the largest complement of lymphocytes in the body by a single cell layer of polarized epithelium. The gut-associated lymphoid tissue (or GALT) has evolved highly complex mechanisms to permit life in this 'open ecosystem'. Although the dominant 'response' of the GALT to orally administered antigen or non-invasive microbes is that of immunologic tolerance (or oral tolerance), the GALT can also initiate both regional and systemic responses to certain antigens and pathogens. The precise molecular mechanisms by which the mucosal immune system regulates this intricate balance between tolerance and responsiveness remain incompletely understood [1, 2]. While most individuals can maintain 'homeostasis' and health in the face of this extreme immunologic challenge, others cannot. Indeed, the clinical entities collectively known as inflammatory bowel disease, or IBD (i.e. Crohn's disease and ulcerative colitis) are conditions in which inflammation persists at various anatomic sites along the gastrointestinal tract [3]. IBD is the second most common disease of potential autoimmune etiology in the United States affecting approximately 1 in 2 to 1 in 10 individuals [4]. Although many details regarding the etiology of IBD remain unclear, a central hypothesis has emerged. Specifically, IBD can be said to represent an aberrant immune response to bacteria from the lumen of the intestinal tract in a subset of individuals with a genetic predispostion or propensity to develop chronic, mucosal inflammation. This is a broad hypothesis that needs to be viewed within the context of the current genetic data describing the marked heterogeneity of human IBD. Still, the microbial focus inherent to the hypothesis

highlights a series of important cell types at the mucosal interface (T lymphocytes and antigen-presenting cells including intestinal epithelial cells) and their associated questions. For example, what is the nature of the bacterial species that are responsible? Do the predisposing genes relate to immunologic function within the GALT or to intrinsic properties of the gastrointestinal tract (e.g. intestinal barrier function)? What are the immune effector cells that mediate IBD and how are they modulated by bacteria? In this chapter we address these and other questions and early events relevant to the central hypothesis outlined; in particular, those involving interaction between bacteria and the lymphocytes of the GALT at the epithelial interface.

The microbial ecosystem We live in a complex microbial environment, with the diversity and number of prokaryotic cells far outnumbering the eukaryotic cells of their hosts. In no instance of human health or disease is this fact more important than in a discussion of immunity at mucosal surfaces and the pathogenesis of IBD. Indeed, we exist in a completely open ecosystem with the anatomy of the entire gastrointestinal tract from the mouth to the rectum in continuity with the external environment. In this context one might consider the gastrointestinal tract as 'outside-in' with an extensive microbiota derived from the outside world resident inside the body, generally with highly beneficial consequences [5]. Studies from both humans and mice have underscored the critical nature of the intestinal microbiota in the estabhshment and maintenance of IBD. The data from human studies include the potential efficacy of antibiotics in treating patients with IBD [6], the resolution of inflammation distal to fecal diversion [7], and the recent data showing a beneficial effect of 'probiotic' therapy [8]. The mouse data are even more compelling with data from many labora-

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.J, Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 121-146. © 2003 Kluwer Academic Publishers. Printed in Great Britain

122 tories demonstrating a lack of mucosal inflammation in established mouse models of IBD raised in either a 'germfree' (GF) or a clean 'specific pathogen free' (SPF) environment. These data have extended across the spectrum of mouse models described [3], including models dominated by T h l ' responses (e.g. IL10^~, IL-2-^ , CD45RB^* transfer model, SAMPl/ Yit) models dominated by T h 2 ' responses (e.g. T C R a ^ , WASP^ ) and models characterized by intrinsic 'barrier' defects (e.g. m d r l a ^ ) (reviewed in ref. 9 and Chapter 4). Recent data from a number of laboratories indicate that repopulation of these mice with limited numbers of organisms (e.g. murine Helicobacter) in these 'clean' mice can trigger the inflammation previously seen when the mice were housed in conventional ('dirty') colonies [10, 11]. Interestingly, the same bacterial challenge in normal mice has no untoward consequences. Although the molecular details as to how these bacteria induce mucosal inflammation in these genetically susceptible hosts remain to be determined, the mounting evidence favors the 'bacterial hypothesis'. Given the estimates of there being approximately 10*^ organisms per milliliter of fecal material (closely approximating the maximum packing density), it is perhaps surprising that more individuals do not develop chronic mucosal inflammation. Clearly, the mucosal immune system has evolved to balance the need to respond to pathogens while maintaining active 'tolerance' to commensal bacteria (and food antigens) [2]. Consistent with the data from mice and family studies in humans highlighting the genetic basis of susceptibility to IBD in certain individuals (addressed in more detail in Chapter 3), most individuals exist in a homeostatic balance with the complex microbiota in their intestinal tract [12]. Understanding why and how certain people show a 'dysregulated' (or pathogenic) response to bacteria that are not universally pathogenic remains central to our detailed understanding of IBD pathogenesis. Still, the microbiota itself can be considered a dynamic, living, organism - and is likely to shape the host mucosal immune response and balance [13]. Hence, a more detailed understanding of the bacterial populations of the gastrointestinal tract is likely also to off'er important and critical clues to the cause(s) of IBD. The microbiota of the gastrointestinal tract have been the topic of several recent thorough and insightful reviews from authors with extensive experience in this area [5, 13-16]. In this portion of the current chapter we focus on microbial aspects most relevant

The lymphocyte-epithelial-bacterial

interface

to IBD, and not on providing the comprehensive microbiological overview seen in these recent summaries.

General considerations Although precise estimates are difficult, there appear to be > 350 different species of bacteria present in the colon of the normal human. In actuality the number is likely to be dramatically higher when viewed in the context that conventional culture techniques result in growth of less than 50% of the organisms present [17]. There are clear regional differences in bacterial populations identified along the cephalocaudal axis of the gastrointestinal tract (see below), but the overall representation of bacterial genera is extremely broad. In the colon, where bacterial counts are four to five orders of magnitude higher than proximal to the ileocecal valve, all major groups of bacterial species are present (Table 1). Specifically, one finds a diverse mixture of Gram-positive, Gram-negative, and atypical bacteria (e.g. mycobacteria) that grow under varying conditions (e.g. aerobic, facultatively anaerobic and obligate anaerobic conditions). Hence, in order to interpret the literature in this complex area, and its potential relationship to IBD, one needs to consider a few important technical considerations. First, what are the conditions used to culture the organisms? For example, culture conditions for aerobes or facultative anaerobes are widely available, while anaerobic culture is still difficult and problematic. Second, what is the nature of the patient sampling? Certainly, the clinical and medication history (e.g. bowel preparation for surgery or endoscopy) will bias the results obtained from microbiological analysis. In addition, there are likely to be marked differences when biopsies and/or surgical specimens are used as a source for culture compared to stool samples. Finally, what is the cost of such detailed analysis? Given the biodiversity present, the answer is likely to be high. This final consideration has limited the detailed analysis of large numbers of individual patients and detailed 'species' analysis. The increasing widespread use of 16S rRNA sequence analysis for microbial detection and analysis [18] is a promising approach now being extended to studies of the gastrointestinal tract [19]. In this context it should not be surprising that there are few data on the 'complete' analysis of the gastrointestinal mucosal microbiota - either in normal individuals or individuals with IBD. The 'best' data set now dates back to the seminal work of

Robert Hershberg and Richard S. Blumberg

123

Table 1. Common bacteria in human large intestine Organism

Description

Range in log scale

Bacterioides

Anaerobic Gram-negative rods, most common species include B. thetaiotaomicron,

9-12

B. vulgatus, B. ovatus and B. fragilis Fusobacterium

Anaerobic Gram-negative rods, most common species include F. mortiferum and F necrophorum

Peptostreptococcus

Anaerobic Gram-positive cocci, most common species is Ps. productus

5-12

Bifidobacterium

Anaerobic Gram-positive, non-spore-forming rods, most common species include B. adolescentis, B. infantis and B. longum Anaerobic Gram-positive, non-spore-forming rods, most common species include E aerofaciens, £ lentum and E contortum

6-12

Eubacterium

Lactobacillus

Facultative and obligately anaerobic Gram-positive rods, most common species include L. acidophilus, L fermentum and L. plantarum

Clostridium

Anaerobic, spore-forming. Gram-positive rods, most common species include C. perfringens,

7-12

9-12

4-12

7-12

C. bifermentans and C. ramosum Enterococcus

Facultative Gram-positive cocci; E faecalis and E faecium are most common species

5-10

Streptococcus

Facultative Gram-positive cocci; S. lactis and S. intermedius are most common species

5-9

Staphylococcus

Facultative Gram-positive cocci

Eschericha coli

Facultative Gram-negative cocci

3-6 3-9

Enterobacter

Facultative Gram-negative cocci

3-9

Klebsiella

Facultative Gram-negative cocci

3-9

Reproduced with permission from ref. 14

Gorbach and colleagues [20], i.e. over 20 years. The current consensus will be briefly summarized.

Stomach/small intestine Previously thought to be 'sterile', the stomach and small intestine harbor considerable microbiologic diversity. The proximal gastrointestinal tract contains abundant flora, derived predominantly from the mouth. These include, but are not limited to, alpha-hemolytic Streptococcus, Staphylococcus epidermidis, and various yeast species. The hostile acidic environment of the stomach results in a rapid decline of culturable organisms. Still, the obvious presence and functional consequences of Helicobacter pylori infection in the stomach and proximal small intestine highlight the i m p o r t a n c e of bacterially-driven mucosal inflammation even in the proximal gastrointestinal tract [21].

The small intestine is marked by a 'transition' from oral to colonic flora. It is here that one needs to consider ingested food as a microbiological reservoir and source for colonization. For example, dairy products, meats and fruits/vegetables are rich sources of bacteria. The substantial antimicrobial effect of low gastric pH is likely responsible for the fact that bacterial counts are not higher in the proximal gastrointestinal tract. Still, as one moves more distally, one sees increasing concentrations of bacterial counts and increasing numbers of Gramnegative species (e.g. Enterobacteriaceae, Bacteroides, Enterococci species) as one approaches the terminal ileum. The total counts in the jejunum are 10^-10"^ times higher than in the jejunum, and one study has estimated the counts at the terminal ileum approximating those seen in the proximal colon [22].

124

The lymphocyte-epithelial-bacterial

interface

Colon

Relevance of microbial populations to IBD

An enormous biodiversity is present within the colonic microbiota, both in terms of mass (estimated to be 10^ ^ organisms per gram of dry weight of stool) and number (estimated to be >350 species) of bacterial organisms. Consistent 'themes' regarding the nature of colonic bacteria exist. First, obligate anaerobes are the predominant bacterial species in the colon comprising >99% of the total population. Considerable genus and species variation in this microbiological population exists between individuals, reflecting a delicate balance between diet, environmental influences, and host genetic factors. Still, species of the Gram-negative, obligate anaerobe, Bacteriodes spp. (e.g. B. vulgatus, B.fragilis, B. theatomicron) comprise the major subset of bacteria in humans and several other carnivores, and represent a large portion of the the so-called 'normal' flora which is always present in large numbers in the population. Facultative aerobes (e.g. Escherichia coli, Enterococcus sp., Staphylococcus sp.) are also common in the colon, but are present at concentrations typically two logs less than the obligate anaerobes. As mentioned, a large variety of other bacteria, including the Gram-positive rods of the genera Lactobacillus and Bifidobacteria (common bacteria used as 'probiotics') are also seen in the colon and are summarized in Table 1. Two important points should be mentioned in the context of this microbial diversity. First, the mucosal micobiota is a dynamic 'structure' - with changes capable of resulting in significant clinical impact. A notable example is the expansion and 'overgrowth' of Clostridium difficile in a subset of patients treated with broad-spectrum antibiotics [23]. The resulting diarrhea and pseudomembranous colitis that results from the elaboration of C difficile toxin in these patients is a reminder of how a 'dysregulated' microbiota may result in untoward mucosal inflammation. On the other hand, the use and limited efficacy of 'probiotic' therapy [8] highlights the possibility of manipulating the colonic flora to suppress mucosal inflammation. Second, one must realize that our current view of the colonic microbiota is truly an 'overview'. Subpopulations of microbes within specific microenvironments or 'niches' are likely to exist, and are overlooked by even sophisticated methods of analysis of the isolated samples currently obtained in clinical and/or experimental studies. Both the changing flora and local microenvironments are likely to have significant implications in the microbial pathogenesis of IBD.

Given the obvious relevance of the microbiota to the development and maintenance of IBD, one needs to consider potential diff'erences in gastrointestinal microbial p o p u l a t i o n s in patients with IBD compared to 'normal' controls. Although critically important in understanding the pathogenesis of IBD, this analysis is replete with potentially misleading information and needs to be viewed with caution. For example, to date the analysis of mucosal microbiota in IBD has been performed in patients with chronic disease that have been treated medically and/or surgically [24]. While an important first step, it remains to be seen if these data reflect the early events associated with IBD pathogenesis. One notable exception is a study from Scandinavia looking at microbial differences in children with IBD [25]. In addition, one needs to consider phenotypic differences between the same genus/species of bacteria isolated from a patient with IBD compared to control patients. The report of increased adhesion of bacteria from patients with IBD is notable in this regard [26].

Mechanisms of bacterial-driven IBD Although a relative 'consensus' is emerging regarding the fact that luminal bacteria are involved in the establishment and maintenance of IBD, still, various hypotheses have been put forth to explain the mechanism(s) underlying this bacterial 'trigger'. These disparate (and often overlapping) hypotheses have been thoroughly detailed in recent reviews [15, 27] and are only briefly summarized here. In general, these can be broadly defined into those that are directly related to pathogenic factors of microbes, those that are related to alterations of luminal energy metabolism, those that are related to activation of innate immune factors through interaction between microbial products and 'pattern-recognition' receptors of the host, and those that are related to specific nominal antigens (peptide, glycolipid) which stimulate adaptive (or acquired) immune responses associated with the interactions between antigen-presenting cells and lymphocyte populations (T and B cells). Although repeated attempts to prove that IBD is due to mucosal infection have consistently proven negative [15], our current level of understanding of the pathogenesis of IBD does not completely rule out an infectious etiology or component of a pathogenic organism. A pathogen can be defined as a micro-

Robert Hershberg and Richard S. Blumberg organism that employs a particular virulence determinant that allows the microorganism to bypass host protective mechanisms and directly elicit cellular and tissue injury. This direct tissue injury is further promulgated by the effects of host immune (innate and adaptive) responses. Given that a genetically susceptible host such as one predisposed to the development of IBD may either lack or maintain a dysfunctional component of host defense, this definition of a pathogen becomes largely semantic and relative to the genetic context within which the microorganism impinges. Specifically, it could be argued, for example, that in the genetically susceptible host the normal (putatively non-pathogenic) microbiota of the intestine is perceived as if it were a pathogen [3]. Regardless, it is clear that 'so-called' pathogenic microorganisms utilize a wide variety of mechanisms to either parasitize and/or inflict injury upon the host [27]. These include the ability to adhere to intestinal epithelial cells (lEC) through a variety of mechanisms, to directly invade lEC, to spread between lEC and to secrete a variety of injurious toxins (Table 2). One interesting aspect of the luminal microbiota that is important relative to IBD is the concept that bacteria may affect the energy metabolism of lEC [28]. Specifically, it has been proposed that subsets of anaerobic bacteria, through their production of hydrogen sulfide, inhibit the metabolism of shortchain fatty acids (SCFA), an important energy source for lEC. As such, it has been hypothesized that IBD is a disease caused by lEC starvation. Interestingly, there are some data supporting this notion in that a beneficial role for topical SCFA therapy has been observed in human IBD [29]. Pattern-recognition receptors are components of the host immune system (either soluble or membrane-associated) that recognize characteristic structural features of the microbial world [30]. As such they are part of the innate, or so-called natural immune, system; a 'hard-wired' system that allows for rapid responses to microbes in contrast to the delayed responses associated with the adaptive immune system (see below). These receptors (summarized in Table 2) recognize specific structures, usually carbohydrates, associated with microbes that either indirectly through stimulation of other innate and adaptive components or directly through initiation of phagocytosis, promote the removal of microbes and their products. Arguably, these host structures may also conceivably aid in the maintenance of immune tone such as that associated with

125 the GALT Indeed, in genetically susceptible hosts, bacterial products such as peptidoglycan-polysaccharides from Eubacterial spp. can elicit IBD-like pathology, perhaps in part through innate immune mechanisms [31]. Finally, specific immune responses, which are initiated by presentation of nominal antigens from microbiota to T and B cells, provide long-term protection to microbial pathogens through generation of memory responses and maintenance of immune homeostasis (tolerance) to non-pathogens (normal microbiota) [1, 32]. These include the classical antigen-presentation pathways associated with the major histocompatibility complex (MHC) class I and II molecules for presentation of peptides and the non-classical MHC class I-like molecule, C D l , for presentation of glycolipid antigens (see below) [33, 34]. Dysregulation of this aspect of the immune system, as clearly seen in a large variety of animal models and implicated in h u m a n s , leads to inappropriate tissue damage and/or autoimmunity through molecular mimicry between microbial and host antigens [35-39].

Establishment of the intestinal epithelial barrier In the context of the growing body of evidence linking luminal bacteria to the pathogenesis of IBD, one needs to consider the nature of the anatomic barrier that separates the overwhelming number of bacterial organisms from the underlying mucosal immune system. In addition to the marked differences between the microbiota of the small intestine and that of the colon, there are clear distinctions in the microarchitecture and organization of lymphoid tissue that are likely to affect the interaction between bacteria and immune cells at these sites. For example, in the small intestine one sees classical lymphoid follicles (Peyer's patches or PP) in addition to villous structures composed of columnar epithelial cells. It appears that certain pathogenic bacteria selectively gain access to the mucosal immune system via the modified epithelial cells (M cells, or follicle-associated epithelium, FAE [40]) overlying these follicles [41]. It remains to be determined whether 'nonpathogenic' bacteria, which are likely to be responsible for stimulating IBD in genetically susceptible hosts, contact the underlying immune system via FAE or by crossing the conventional epithelial surface. Given the fact that the epithelial surface area

The lymphocyte-epithelial-bacterial

126

interface

Table 2. Comparison of human MHC class i, MMC class II, and CD1 MHC class I

MHC class II

CD1

Structure Gene locus

Chromosome 6

Chromosome 6

Chromosome 1

N-glycosylation

Yes

Yes

Yes;No(37kDaCD1d)

Cytoplasmic tail

Long:3 exons

Variable lengths: 1 exon

Short: 2 exons with motif for endosomal targeting in CD1b,c,d*

Genomic structure

Exons 2 - 4 encode the a 1 , a2, aS domains

Exons 2 - 3 encode the a 1 , a2(or p i , P2) domains

Exons 2 - 4 encode the a 1 , a2, a3 domains

Homology in a 1 , a 2 , a3 domains

—

—

25-30% with a2 domain of MHC class 1; 30-35% with p2 domain of MHC Class 11

Allelic polymorphism in a 1 - a 2 domains

Polymorphic

Polymorphic

Nonpolymorphic

Biosynthesis/assembly P2-microglobulin dependence

Yes

No

Yes (CD1a-c); yes/no (CD1d)

TAP dependence

Yes

No

No

Associated chaperones

Calnexin, calreticulin, tapasin

Invariant chain, HLA-DM

Calnexin (CD1 b) prolyl-4hydroxylase(CDId)

Function Cellular expression

Ubiquitous

B-cells, dendritic cells, Langerhans cells, activated monocytes, lEC

CD1a-c: thymocytes, B cells, Langerhans cells, activated monocytes, lELs (?); CD1d: lEC, hepatocytes, B cells, monocytes, dendritic cells

Peptide presentation

9 amino acids

14-22 amino acids

22 amino acids (murine CD1d)

Lipid presentation

No

No

Lipoglycan: CD1 b,c; glycosylceramide and glycophosphatidylinositol:CD1d

T cell recognition

CDS'", TCR aP""

CD4'', TCR aP""

CD8-CD4-TCRaP"'(CD1a,b,c,d); CD8-CD4-TCRy6"'(CD1c); CD8"'TCRap"'ilELs(CD1a,c,d); CD8-CD4''TCRotp"'(CD1d)

*YXXZ -* Tyrosine-amino acid-amino acid-hydrophobic amino acid.

is orders of magnitude greater than that of the FAE, and that many bacteria adhere to and invade columnar epithehal cells, it is likely that bacterial entry is not restricted to lymphoid follicles. Indeed, in the colon, 'classical' lymphoid follicles (PP) are absent, although small lymphoid aggregates, presumably with modified FAE, can be seen (see below), yet

bacteria are able to cross the mucosal barrier and trigger IBD. The barriers to bacterial entry across the intestinal mucosal surface are numerous and varied. From an anatomical standpoint there is an extensive mucinous glycocalyx extending from the apical (i.e. luminal) surface of epithelial cells that can trap bacteria and prevent adherance [42]. In addition, epithelial

Robert Hershberg and Richard S. Blumberg cells are connected to one another by so-called tight junctions which effectively limit the paracellular transport of macromolecules across the epithelial barrier [43]. Furthermore, various populations of cells along the gastrointestinal tract (e.g. Paneth cells) are capable of secreting small peptides (e.g. defensins) that have potent, local, antimicrobial properties (see below). How do bacteria breach this formidable barrier and stimulate mucosal inflammation? The answer is complex, but derives in part from the growing realization that alterations in intestinal barrier function are seen in IBD - and that this 'dysregulation' of barrier function may contribute to the establishment of IBD. Indeed, the term 'dysregulation' (and not simply disruption) underscores the fact that the barrier is a dynamic structure that is actively regulated and maintained. Based on seminal studies using polarized monolayers of intestinal epithelial cells in vitro, it is clear that various cytokines present in the intestinal mucosa regulate intestinal barrier function. Notably, interferon gamma (IFN-y), IL-4, and IL-13 all attenuate barrier function [44, 45], while IL-10 appears to 'enhance' the barrier [46]. While the precise mechanisms underlying this regulation are not known, recent data suggest that alterations in the apical actin cytoskeleton might disrupt the structural integrity of the tight junction [47]. As the molecular architecture of the tight junction becomes more clear, establishing the direct link(s) between signaling events resulting from proinflammatory cytokines (for example, IFN-y) and specific tight junction proteins will be possible. These proteins are likely to include the growing number of MAGUK-like molecules (ZO-1, -2, -3, claudins, occludins) that are concentrated at these junctions [48]. Finally, it is important to consider how alterations in intestinal barrier function (with the untoward effect of increased transit of bacteria into mucosal tissues) may contribute to an underlying genetic predisposition to the development of IBD. As noted, decreased barrier function is a hallmark of established IBD [49]. Interestingly, some investigators have suggested that alterations in barrier function actually precede the appearance of IBD, and are seen to a greater degree in healthy, first-degree relatives of patients with IBD [50]. These human data are supported by mouse data showing alterations in barrier function preceding histological inflammation in the IL-10 knockout mice [51] and the enhanced mucosal inflammation that could be induced in mice deficient

ni for intestinal trefoil factor, a molecule important in intestinal epithelial repair and restitution [52]. Taken together, the information suggests that subtle alterations in barrier function (e.g. abnormal regulation of tight junctions, impaired epithelial restitution) may contribute to the underlying genetic susceptibility to IBD seen in humans and some experimental animals. One might predict that the function of one of more of the IBD susceptibility genes being mapped by various groups will relate to intestinal barrier maintenance.

Immune cell populations Phenotype of T cells Exemplary of the fact that the mucosal surfaces of the intestine must confront a large microbial antigenic burden, it is not surprising that it has been estimated that approximately 10% of the T lymphocytes in the normal host are associated with the GALT. These GALT-associated T cells are organized into three compartments connected by distinctive and highly regulated trafficking pathways. These include Peyer's patches (PP) and related isolated lymphoid follicles that are considered important sites of immune response induction and the loosely affiliated lamina propria and intraepithelial lymphocyte compartments, which are considered effector compartments [53]. Consistent with this, the vast majority of T cells within the PP are naive, thymically derived T cells that are characterized as CD3"^ TCRap"" CD45RA"' a^Pv^^ L-selectin (CD62L)"' CD44^^ a^Py- with a CD4:CD8 ratio of 3.5:1. These naive T cells, as well as their naive (IgM"^ IgD^) B cell counterparts, are directed to the PP via interactions between a'^Pv and L-selectin on the T cell and the protein and carbohydrate constituents of MadCAM1 on the high endothelial venule (HEV), respectively. In addition, contributions are provided from interactions between LFAl (a^P2, C D l l a / C D 1 8 ) and either ICAMl (CD54) or ICAM2 (CD 120) on the lymphocyte and HEV, respectively. Consistent with this naivety, most of the T cells possess a diverse array of TCR-ap chains [54]. Few TCR-yS"" T cells can be observed within the PP, consistent with the notion that they have received their priming earlier in ontogeny within the thymus [55]. At any one time a significant subset of these naive T cells within the PP are likely the recipients of their cognate antigen-derived signal from the luminal milieu as transported to the lymphoreticular struc-

128 tures of the PP by specialized microfold villous (M) cells. This is based upon the fact that the ratio of CD45RA:CD45RO within the organized GALT is 1:1 and the majority of these CD40RO"^ (memory) cells are congregated within the M cell pocket immediately adjacent to the lumen in the context of professional APC [56]. The range of luminal antigens, including antigens from commensal bacteria, that are specifically sampled by the M cell and/or breach the M cell barrier through pathologic mechanisms (e.g. pathogenic microbes) thus dictates the generation of antigen-specific B and T cell blasts which emigrate from the PP via efferent lymphatics. The antigen-specific €04"^ and CDS'*" T cell blasts including memory populations, which are largely a^p7^" L-selectin^'' CD45RO'', presumably migrate from the PP and disseminate widely to the loosely affiliated tissues associated with the lamina propria and epithelium. This occurs through interactions between protein components of MadCAMl on the postcapillary venule (PCV) and a'^Pv on the T cell together with interactions between LFA-1 on the T cell and ICAM1 /ICAM2 on the PCV. Whereas both CD4"*" and CD8"^ memory cells and blasts populate the lamina propria, the epithelium is preferentially populated by CD8^ T cells. The basis for this is unclear, but presumably reflects differences in the presumed effector functions of the T cells within the lamina propria and epithelium [57].

Lamina propria lymphocytes (LPL) LPL represent a tightly regulated effector compartment. The vast majority of these T cells express the memory marker, CD45RO (>80%), with a ratio of CD4:CD8 cells which approximates peripheral blood (2:1) [32, 53]. Based upon TCR repertoire analysis the CD4"^ LPL are directed at a broad range of antigens with evidence of a limited number of clonally expanded cell populations [36]. Although the nature of the antigens to which these small numbers of clonal expansions is directed is unknown, it is presumed that they relate to secondary responses to previously remote antigenic encounters. This is consistent with the fact that LPL, which are responsive to remote enteral viral infections and presumably microbial infections, for example, can be identified [58]. The CD8'' LPL, on the other hand, exhibit a significantly greater proportion of clonal expansions as defined by complementarity determining region 3 (CDR3) analyses suggesting, perhaps, either a larger number of secondary exposures for

The lymphocyte-epithelial-bacterial

interface

this subset and/or differences in their susceptibility to activation-induced cell death in comparison to the CD4^ T cell subset [36]. The entire LPL compartment, both CD4 and CD8, are highly activated. This activation is manifest by a phenotype which is uniformly L-selectin^° CD44^^ CD69"'. Yet, at the same time, the activation LPL is restrained since only a limited proportion of these cells express the high-affinity IL-2 receptor oc chain, CD25 ( < 20-25%) [53]. Moreover, LPL exhibit a limited evidence of spontaneous proliferation based upon expression of Ki67, a nuclear marker of proliferation, or uptake of bromodeoxyuridine. Additionally, LPL are hypoproliferative to TCR/ CD3 complex-mediated signals in vitro [59]. However, at the same time, LPL respond to antigenic signals with significant cytokine production (as measured by IL-2). These characteristics suggest that, under physiologic circumstances, LPL are chronically stimulated by antigens ubiquitously present in the intestinal environment but which elicit predominantly cytokine production and little proliferation. This maintains a tight control on LPL T cell numbers. The cytokines produced are consistent with a predominantly T helper (Th) 2-tone under normal physiologic conditions [60]. In the physiologic context it is likely that the antigens responsible for this unique state of T cell activation are derived from the intestinal microenvironment and, presumably, components of the normal microbiota. Specifically, mice expressing a transgenic T cell receptor specific for ovalbumin in the context of MHC class II I-A^ do not exhibit this activated LPL phenotype when the transgenic animal is back-crossed onto a recombinase activating gene (RAG)-deficient background [61]. In the absence of the RAG gene the expression of non-allelically excluded T cell receptor (TCR)-oc chains is prevented. Thus, it can be argued that the LPL are maintained at a certain level of restrained activation by components of the normal microbiota; poised for a pathologic assault. The loosely affiliated lamina propria also contains varieties of organized structures: lymphocyte filled villi (LFV), isolated lymphoid follicles (ILF) and cryptopatches. LFY are congregations of predominantly CD4'' CD45RO'' T cells with variable numbers of B cells and MHC class 11"^ dendritic cells with an overlying epithelium similar to FAE [62]. ILF are submucosal lymphoid aggregates that contain a B cell follicle and memory T cells, suggesting a region undergoing a cognate immune response. Finally, cryptopatches, which are present in rodent but not

Robert Hershberg and Richard S. Blumberg adult human intestinal tissues, consist of lamina propria structures that appear to be the source of extrathymic T cell lymphopoiesis within the intestine. These structures are independent of luminal bacteria (occur in germfree animals) and a thymus (occur in nu/nu mice) but are dependent on IL-7 and consist of a limited number (2000-5000) of ckit"'lL7R^CD44^Thyr/-CD4^/-CD25^^-oc^P^Lin" (hneage markers such as CD3, B220, Mac-1 and Gr-1 associated with T cell, B cell, macrophage and granulocyte development, respectively) cells which give rise to TCR-ap and TCR-y5 cells [63].

ilEL Intestinal intraepithelial lymphocytes (ilEL) are predominantly T cells with few, if any, B cells [57]. Although lEL are present in the epithelium during antenatal life as early as 11 weeks of gestation, suggesting an initial antigen-independent origin, their numbers increase rapidly postnatally. This massive increase (more than 10-fold) appears to be driven by the luminal microbiota as it is not evident in germfree mice and rats [64]. Although it is possible that these developmental changes may be due to the induction of phenotypic changes in the epithelium and associated structures, there is some reason to believe that these changes are antigen-driven as the effects of the microbiota on lEL are also reflected in changes inTCR repertoire [65]. Most lEL of mouse and human origin are CD45RO'' (memory) CDS'" T cells which express a limited array of aP TCR, suggesting restriction to a narrow range of MHC class I-related molecules and their antigenic ligands. Whether these putative antigens are derived from the normal luminal microbiota, and thus reflect important immunoregulatory functions of these cells such as the MHC class I chain-related genes [MICA; 66] or are directed to either non-cognate 'stress signals' on altered intestinal epithelial cells or remote pathogenic exposures remains to be defined. It is likely that all these possibilities are operative, although their relative contributions at any given time are unknown. lEL have been shown to have immunoregulatory functions associated with oral tolerance, to exhibit recognition of non-classical MHC-like molecules (see below) and to exhibit antigen-specific, MHC recognition of remote viral antigens, for example [57]. Like LPL, lEL are in a unique state of activation. They are almost uniformly CD69"', CD44^^ and Lsel^^ consistent with previous exposure to an activa-

129 tion signal delivered presumably through the TCR [53]. Given that this phenotype is evident in the 'physiologically' inflamed intestine, it seems plausible to suggest that these antigenic signals are derived from the luminal microbiota. Whether the oligoclonality normally associated with human and rodent lEL is related to these same antigenic signals is unknown [67-69]. In addition to these activation markers, lEL also express a unique constellation of homing markers consistent with their mucosal localization within the epithelium (oc^Pv"^, CCR9"^) and natural killer markers [70-72]. The latter are quite interesting as this group of molecules (killer inhibitory and activatory receptors) affect the regulation and activation state of cells. This suggests one possible mechanism by which lEL may be tightly controlled despite a constant exposure to their cognate, perhaps luminal bacterial, antigens as reflected by their anatomic location and activation state. Control of lEL and likely LPL is also likely mediated by a large array of soluble mediators including TGF-p and unknown factors many of which are derived from the lEC itself [73]. Whether these factors are directly regulated by the luminal microbiota is unknown.

Regulation of lymphocyte populations by luminal microbiota Given the apparent central role of the normal luminal microbiota in regulating the development of IBD, and that colonization of the mammalian intestine occurs prior to the onset of disease in genetically susceptible animals (rodent and human), it is important to consider how the phenotype and function of mucosal lymphocyte populations are regulated by such factors. There are two potential roles of the luminal microbiota: promoting the development of the normal GALT and maintenance of the state of physiologic inflammation. Given that both of these are likely operative in the normal mammalian host, it is possible that, in the genetically susceptible host, abnormalities of one and/or the other process may be existent. It is therefore important to consider the supportive evidence for these processes in the normal host and what is known in the genetically susceptible host. Most of the work in this area has been performed with respect to the development and responses of B lymphocytes [74]. Germfree rodents exhibit functional hypotrophy of the ceUs and tissues associated

130 with the GALT. The PP are small, poorly developed structures that primarily contain B cells which exhibit diminished responses to B cell mitogens and antigens. Most of the B cells which are present have a naive (IgM"^IgD"^) phenotype with few or no IgG"^ or IgA"^ cells, indicating that these cells have switched their immunoglobulin locus under stimulation of cognate antigen. Consistent with this, IgA"^ antibody secreting plasma cells are markedly diminished in the spleen and lamina propria. In comparison, rodents raised in conventional conditions exhibit PP with secondary lymphoid follicles with reactive germinal centers as would be observed in an antigen-activated peripheral lymph node, high levels of switched B cells (IgG"" and IgA"") in the PP and significant number of IgA-secreting plasma cells in the spleen and lamina propria. Interestingly, when adult rodents which had previously been raised under germfree conditions are later colonized with non-pathogenic microorganisms, their immune responses are often augmented [75]. This suggests that, if mucosal priming does not occur either at the correct time of life or with the appropriate microbial antigen, hyperresponsiveness to the microbial antigen may ensue; a phenomenon of obvious relevance toIBD. It is also important to recognize that not all organisms have the ability to drive the development and activation state of the GALT. Some organisms which are components of the normal microbiota elicit no detectable immune response as assessed by the presence of IgA antibodies. Such organisms are considered part of the authochthonous microbiota which have presumably coevolved with the host to such a degree that they are truly tolerated nonpathogens. Other organisms, on the other hand, which are also components of the normal microbiota, elicit a non-pathogenic secretory IgA immune response. This immune response is, however, selflimited in that secondary exposures with the same microorganism are muted. At the same time the immunizing microorganism can be found intraluminally coated with IgA. Thus the normal microbiota of the mammalian intestine stimulates a state of tolerance that is manifest by the production of specific secretory IgA antibodies that likely prevents the further uptake of the relevant microbe. Similar regulation of the development and activation of the cellular immune system has also been elucidated [74]. As noted above, the numbers of TCR-oc(3'' lEL, but not TCR-y5'' lEL, increase dramatically with bacterial colonization [64]. Moreover,

The lymphocyte-epithelial-bacterial

interface

the spontaneous cytotoxic activity of these lEL, a characteristic feature of these cells, is also markedly up-regulated. Thus, the cytotoxicity of lEL obtained from conventionally reared animals is markedly elevated over those observed with animals reared under germfree conditions. Not only the CDS"^ compartment, but also the CD4"^ compartment, is affected. CD4"^ T cells from germfree animals exhibit a diminution in the autologous mixed lymphocyte reaction. Consistent with these functional observations, most PP T cells in germfree animals are naive (CD45RB^*), similar to the levels observed in a sterile peripheral lymph node. Upon colonization, increased levels of activated or memory (CD45RB''') cells are observed. Given that all of these characteristics are observed in healthy animals, it can only be assumed that the normal microbiota establishes a unique state of inflammation which is associated with the health of the host. Moreover, this healthy state not only affects the physiology of the intestine but extends to the whole host [76].

The intestinal epitlielial cell as an antigen-presenting cell The gastrointestinal tract represents a unique immunologic compartment which subserves an important role in host defense against a wide variety of microorganisms and the regulation of responses to foreign antigen [2, 32]. The management of antigen responsiveness is accomplished through complex pathways of antigen uptake and processing by antigen-presenting cells (APC) and their subsequent presentation to T lymphocytes. In the intestine, APC include not only cells of the monocyte-macrophage lineage and B cells, but also potentially non-conventional APC such as the lEC [77]. The gastrointestinal tract epithelial cells may have a unique role in immunoregulation, since luminal events are likely important in normal gut homeostasis and may be prerequisite for the initiation and/or perpetuation of inflammation as occurs in IBD. The lEC are the first host cells to come in contact with dietary and microbial antigens and are therefore in a unique position to function as an early host-signaling system to the immune cells located adjacent to and in the underlying intestinal mucosa. Therefore, antigen-presenting molecules that are expressed on the lEC cell surface may function in the regulation of gut immune responses by presenting specific antigen-containing ligands to T cells which function in the direct activa-

Robert Hershberg and Richard S. Blumberg tion and/or down-regulation of local T lymphocytes. In the former case this activation may be associated with engagement of regulatory T cells which mediate down-regulatory signals. These MHC-restricted signals which are delivered to local T cells by lEC are geared toward the self-perpetuation of the intestinal barrier, the removal of altered epithelial cells injured by stress (including hypoxia), infection and/or neoplasia and the tight regulation of responses to antigen at the mucosal surfaces [77]. As such the lEC and local gut T cells within the epithelium and lamina propria utilize the fine specificity provided by MHCrelated molecules on intestinal epithelia and the T cell receptor (TCR) and costimulatory molecules on T cells to regulate barrier function, antigen absorption and processing, immunosurveillance and local immunoregulation.

Antigen presentation It is now well established that there are two major pathways of antigen presentation: the MHC class I pathway and the MHC class II pathway (reviewed in ref. 33). T cells recognize processed nominal antigens, usually peptides, derived from proteolytic degradation of a larger polypeptide chain. These processed peptides are recognized in the context of an MHC class I or II molecule on the cell surface of an APC (summarized in Table 2). MHC class I consists of a 43-45 kDa glycosylated heavy chain non-covalently associated with P2-microglobulin which presents a nonapeptide, acquired from the transporter associated with antigen presentation (TAP), to CDS"" T cells. MHC class II consists of a 32-34 kDa ocP heterodimer that presents much longer peptides (14-22 amino acids) to CD4'^ T cells. Since MHC class I generally presents an array of peptides derived from the degradation of intracellular proteins and MHC class II from extracellular proteins, these antigen-presenting pathways have likely evolved to a large extent for the protection of the host from unforeseen intracellular and extracellular deleterious events, respectively. As described above, T lymphocytes are abundant in the gastrointestinal tract and are implicated in a wide variety of 'physiological' immune responses (both oral tolerance and responses to invasive pathogens) and 'pathophysiological' responses (for example, the induction/maintenance of chronic inflammation, or IBD). T cells are stimulated by foreign protein antigens via specific molecules on their surface (the so-called T cell receptor, or TCR) which

131 recognize processed nominal fragments of the antigen. These fragments are 'presented' to the T cell by molecules of the MHC complex which are expressed on the surface of so-called antigen-presenting cells, or APC. These APC also express additional surface molecules (e.g. co-stimulatory molecules such as CD80 and/or CD86) that are required for maximal stimulation of T cells via counter-receptor expressed on the surface of the T cell other than the specific TCR. The majority of APC are bone marrow-derived cells (e.g. dendritic cells, or DC) with specialized mechanisms for regulated and efficient antigen uptake, processing and presentation. In addition, within specific anatomical contexts, other 'non-professional' APC exist. These are cells which express MHC antigens and have a limited, but significant, capacity to process and present antigen to T cells. One such 'non-professional' APC relevant to the gastrointestinal tract is the polarized epithelial cell which separates the extremely high concentration of foreign antigen from the underlying lymphoid tissue. Consistent with a potential role in antigen presentation, these lEC are in intimate contact with the various T cell populations previously described. Specifically, lEC have extensive contact along their lateral and basal surface with lEL, and contact underlying LPL via basolateral projections through the semiporous basement membrane. As outlined, considerable reactivity against bacterial antigens exists within the T cell compartments of the GALT. The mechanisms underlying the processing of intact bacteria and/or bacterial fragments and the generation of specific bacteria-derived T cell epitopes remain poorly defined, however. The GALT and regional nodes (e.g. mesenteric lymph nodes) are replete with potential APC, including apparently distinct populations of DC, which may be involved in processing of bacterial antigens [78]. In addition, the anatomic considerations noted above, and the fact that lEC of the colon are exposed to the highest concentration (by several orders of magnitude) of luminal bacteria and bacterial antigens of any cell in the gastrointestinal tract highlight the possibility that these cells may participate in the processing and/or presentation of bacterial (or other) antigens to T cells in the GALT. In this section we review some of the recent data consistent with this supposition, focusing on several of the unique features of lEC antigen processing and presentation that distinguish these cells from other more conventional APC.

132

The lymphocyte-epithelial-bacterial

interface

General features of lEC as antigen-presenting cells In order for lEC to act as APC, they must be able to internalize and process antigen. The apical mucous layer and glycocalyx in the intestine can restrict the size of particles capable of being internalized by lEC favoring antigen uptake by the modified 'dome' epithelium (M cells) which overlie Peyer's patches [79]. The highly specialized M cells facilitate transport of macromolecules without significant 'cellular processing' to the underlying lymphoid tissue which contains several different types of professional APC including dendritic cells and B cells [80]. However, the surface area of the villous epithelium is extraordinary (the surface area of the human small intestine being approximately equal to that of a tennis court) and lEC have a well-documented role in nutrient and solute uptake. lEC are exposed to a wide variety of antigens of diverse sizes and varying biochemical properties, including some antigens such as gliadin with known pathologic significance. In addition, a wide variety of organisms of various sizes (e.g. rotavirus. Salmonella, and spp.) can enter lEC from their luminal (i.e. apical) surface. Collectively, these observations highlight a physiologic role for the I EC in the uptake and processing of luminal antigens and pathogens. Several unique characteristics related to antigen uptake by the epithelium are worthy of note. First, lEC express a variety of molecules on their apical surface that might serve as 'antigen receptors' (e.g. FcRn [81, 82], DEC-205 [83], ganglioside Ml (GMl), which may enhance antigen processing by directing internalization via receptor-mediated (instead of fluid-phase) endocytosis. The polarized expression of these molecules on lEC is likely to dramatically modulate processing of antigens by enhancing antigen uptake and/or 'targeting' antigens to certain intracellular compartments. In addition, the rate and efficiency of antigen uptake and endocytic processing by lEC can be modulated by a variety of inflammatory mediators, such as interferon (IFN)-Y.

A limited, but significant, number of lEL express CD4 (most notably in the colon [88, 89]) and approximately two-thirds of LPL express CD4 [53]. Because the expression of MHC class II antigens is required on an APC in order to stimulate CD4"^ T cells, it is noteworthy that MHC class II expression has been observed on lEC from human, mouse and rat with elevated levels consistently seen in the context of mucosal inflammation [90-92]. Indeed, using a variety oUn-vitro models, several groups have demonstrated the capacity of I EC to process and present antigen via HLA class II [91, 93]. There are marked regional differences in the composition (and presumably the function) of the GALT along the cephalo-caudal axis of the gastrointestinal tract. Given the limitation of the in-vitro models it is difficult to address potential distinctions in MHC class II processing between lEC of the small and large intestine. In fact, little information is available regarding specific regional differences (both crypt to villous, and small intestine to colon) in the co-expression of MHC class II antigens and other molecules (for example, invariant chain and HLA-DM) essential for efficient class II processing in conventional APC (reviewed in ref. 94), which are typically induced along with MHC class II at the transcriptional level following activation of the APC with IFN-y [95]. Data from I EC cell lines transfected with HLA-DR alone or with li and HLA-DM are consistent with the observations that the generation of specific class II peptide epitopes shows a variable dependence on the expression of li and HLA-DM in lEC - with some epitopes processed via other class II pathways [96, 97]. Several groups have described one such 'alternative' class II processing pathway that uses MHC class II molecules recycled from the cell surface, which presumably bind peptide antigens in an early endosomal compartment [98, 99]. These alternative pathways may be particularly relevant to MHC class II processing by lEC, especially when MHC class II expression may be limited (i.e. in the absence of inflammation) and the antigen may be denatured or fragmented following luminal 'preprocessing'.

Interaction of lEC with CD4^ T cells - MHC class II processing by lEC CD4"^ T cell responses are critical for both the establishment of oral tolerance [84-86] and in several experimental models of IBD [87]. Two distinct populations of CD4'^ T cells are present within the intestinal mucosa and in intimate contact with lEC.

Influence of lEC polarity on antigen presentation by lEC class II nnolecules lEC are highly polarized cells, with distinct apical and basolateral domains with very different physicochemical properties. The highly polarized morphology of lEC highlights several important distinctions in class II processing between lEC and other more

Antigen presentation by lEC

Robert Hershberg and Richard S. Blumberg conventional APC. First, the expression of HLA class II antigens in vivo and in cell culture models of polarized lEC is mostly restricted to the basolateral membrane (where the cell contacts both ilEL and LPL) [96, 100, 101]. Hence, antigen presentation occurs in a highly polarized manner. In addition, invitro data suggest that the polarized surface from which antigen is internalized dramatically affects the functional outcome with regard to the generation of T cell epitopes [96]. This notion is consistent with the biochemical differences observed between endocytosis from the apical and basolateral surface of polarized epithelial cells [102, 103]. Consequences of MHC class II expression and processing by lEC There is limited experimental data regarding the invivo function of MHC class II on lEC. MHC class IIrestricted antigen presentation can be identified in uninflamed small intestinal [104] and colonic epithelium [105], and is increased in colonic epithelium in the setting of inflammation as observed in IL-2deficient animals [105]. It is thus not unreasonable to suggest that, under pathological circumstances, when the balance shifts toward uncontrolled inflammation such as in IBD (for reasons that remain obscure), the lEC functions as an important accessory APC, stimulating mucosal CD4"^ T cell responses. Within the setting of the inflamed mucosa the unique features of polarized MHC class II processing may come into play. In particular, the processing of luminal antigens (such as bacterial or food antigens) normally exposed only to the apical surface might have especially untoward effects when these antigens gain access to the basolateral surface of lEC via 'leaky' tight junctions [106]. Conceivably, 'pathogenic epitopes' within an antigen (that normally elicits no significant response or a tolerogenic response when processed apically) may be unmasked via internalization and trafficking from the basolateral surface and presented to underlying CD4'^ T cells. Whether lEC can stimulate 'naive' CD4'^ T cells or CD28-dependent T cells in the intestinal mucosa (if either of these populations exist in this anatomical location) remains a matter for speculation. CD4"^ T cells have been widely implicated in the development of IBD [87]. Hence the possibility exists that the processing of intact bacteria and/or bacterial fragments by HLA class II positive professional (dendritic cells, macrophages) and non-professional (intestinal epithelial cells) APCs results in the presentation of specific bacterial T cell epitopes to

133 pathogenic CD4"^ T cells. This hypothesis, which does not preclude the potential role of bacteria or bacteria-derived products such as LPS in modulating mucosal immune responses, is an attractive explanation for the dependence of mucosal inflammation on a subset of bacteria present in the lumen of the gastrointestinal tract. Data supporting this hypothesis exist in mice and humans. It has recently become appreciated that IBD-like pathology can be initiated by CD4"^ T cells with a T h l cytokine profile which are restricted to MHC class II and specific for peptide antigens associated with the normal luminal microbiota [107]. In a corollary manner, CD4'^ regulatory T cells with a cytokine profile consistent with T regulatory 1 cells (IL-10) can be similarly identified and established which can antagonize these effector Thl cells in adoptive transfer models [108]. Although their site of origin is unclear, based upon studies of oral tolerance induction and work in theTCR-a-deficient animal model of colitis, it can be suggested that these agonist and antagonist cell populations originate within the inductive sites of the distal small intestine (appendix, Peyer's patches) and carry out their effector functions at the epithelial barrier of the colon after subsequent reactivation by either homologous or heterologous microbial antigen at this location [2, 38]. Whether the subsequent encounters are through interaction with the lEC or professional APC subjacent to the epithelium remains unclear. Nonetheless, these studies suggest that MHC class II antigen presentation pathways in the context of the immunologic milieu associated with the inductive sites of the GALT lead to the generation of agonist and antagonist (regulatory) T cell populations. Furthermore, it can be hypothesized that, under normal circumstances, these regulatory populations prevail over the agonist cells leading to homeostasis. In IBD, however, this tolerance to the antigens of the normal microbiota is not estabhshed [12].

Interaction of lEC with CD8^ cells In the earliest studies assessing the capacity of lEC to act as non-professional APC, the T cells activated were predominantly CD 8"^, which functionally suppressed immune responses in an antigennon-specific fashion [109]. Consistent with the somewhat unusual features of these CD8'^ T cells stimulated by lEC (i.e. the lack of cytotoxicity and the ability of these cells to suppress a mixed lymphocyte reaction), recent data have emerged revealing atypical characteristics of

134 MHC class I function in lEC. For example, although lEC Hnes have been shown to be good targets for class I-restricted virus-specific cytotoxic T lymphocytes (CTL) and appear to express a normal proteasome repertoire, they fail to prime an antiviral CTL response (K. Becker and L. Mayer, unpublished). These data suggest that MHC class I-mediated processing may be somehow 'altered' in lEC compared to professional APC like dendritic cells, or imply that the lack of conventional costimulation may preclude priming of a naive CDS"^ T cell responses [110]. Several lines of evidence point to novel interactions between CDS"^ T cells and lEC. One stems from the early observation that antibodies specific for HLA class 1 did not inhibit the lEC-induced proliferation of CD8"^ T cells in vitro [109]. One intriguing hypothesis proposes that 'non-classical' or class lb molecules function in antigen presentation by I EC to CD8"^ T cells in the intestinal mucosa. This hypothesis is supported by data demonstrating the expression of these molecules by lEC in mice and humans, often in a pattern highly restricted to the intestinal epithelium [111, 112] (reviewed in ref. 113). In humans, the list includes CDld [114], MICA and MICB [115], HLA-E [116], and HFE (which is involved in iron metabolism) all of which are expressed by I EC [117].

CD1D, a novel ligand on lEC CD1 family of proteins Recent studies have revealed that, in addition to MHC class I and class II, the host also utilizes a third pathway of antigen presentation which is represented by the CDl gene family. This pathway (or pathways) functions in specific tasks of immunologic recognition distinct from MHC class I and II, yet which draws structural and functional features of both. First identified on cortical thymocytes, the CDl gene family encodes a group of proteins which are structurally most similar to MHC class I proteins but have several features in common with MHC class II [34]. This suggests that CDl emerged from an ancient ancestor that diverged into MHC class I, MHC class II, and CDl (Table 2). The human CDl locus on chromosome 1 contains five genes, C D I A - E [by convention, CDl genes are in capital letters (A-E) and CDl proteins in lower-case letters (a-e)]. The human CDl gene family falls into two groups based upon sequence homology: C D I A - C (group 1) and

The lymphocyte-epithelial-bacterial

interface

CD ID (group 2). A gene product for CD IE has not been defined. C D l a - c subserves a specific role in presenting lipoglycan antigens from mycobacteria (and bacteria) to discrete subsets of CDS"^ and double-negative ( C D r C D S ) T cells [118, 119]. These T cells utilize a wide variety of conventional ocp TCR. C D l a - c are expressed by the majority of immature, double-positive (CD4"^CD8"^) thymocytes and professional APC such as B cells, dendritic cells and activated monocytes. Rodents do not express group 1 CDl proteins and appear to have deleted these genes [34].

CDld and natural l(illerT cells CD 1 d is expressed by the majority of thymocytes and at low levels by resting B cells and monocytes [34, 120-122]. In addition, we have also found that CD 1 d is expressed by a variety of epithelial cell types including I EC and hepatocytes [120, 121]. CDld functions in the presentation of lipids. These include glycosylphosphatidyl inositol, which has been eluted from murine CDld in transfectants [123], and ocgalactosylceramide (oc-GalCer), which has been extracted from marine sponges [124]. CDld has also been shown to present hydrophobic peptides [125]. These observations are consistent with the hydrophobic nature of the CDld groove [126]. Recognition of CDld appears to be primarily accomplished by a discrete subset of T cells, which share similarities with natural killer (NK) cells [127]. These NK-T cells, which are also commonly called natural T cells, were initially described in mouse systems [128]. Mice express the human homologue of CDld but not C D l a - c [129, 130]. These murine CD Id-restricted T cells are characterized by their expression of the N K l . l marker (NKR-PIC or CD161), a C-type lectin, and an invariant TCRa-chain (Val4-Joc281), which preferentially pairs with Vp8.2, 7 or 2 in hierarchal order [127, 131, 132]. These cells are also either CD4"^CD8" or double-negative since the forced expression of CDS in mouse results in the deletion of invariant NK-T cells [133]. NK-T cells represent a major fraction of mature thymocytes, liver-associated T cells and up to 5% of splenic cells [134]. In addition, they have been identified in the epithelium amongst lEL [135] and among LPL at low levels (R.S.B., personal observation). Although their function is not entirely clear, they have been recognized to be the major source of IL-4 and IFN-y in vivo in response to anti-CD3 stimulation, suggesting an important immuno-

Robert Hershberg and Richard S. Blumberg regulatory role during early phases of an immune response [136]. CD Id-deficient mice lack NK-T cells and the early cytokine response associated with antiCD 3 activation, but maintain their ability to mount most T helper (Th2)-type responses [137-139]. Murine NK-T cells are also involved in IL-12mediated pathways as they express IL-12 receptors and, when stimulated by IL-12, exhibit significant cytolytic activity and regulation of IFN-y responses; responses which are lacking in CD Id-deficient mice [140, 141]. These observations with murine NK-T cells suggest that they play an important role in immunoregulation of type 1 and type 2 cytokine responses through their abiHty to generate high levels of cytokines and immunosurveillance through high cytolytic activity. Similar populations of NK-T cells are also expressed in humans. Analogous to mice, these human NK-T cells express an invariant TCR composed of Voc24-JocQ which preferentially pairs with V p l l , the human homologue of mouse VpS [142, 143]. Most human NK-T cells are double-negative while a minority express CD4 [144]. They uniformly express NKR-PIA (CD 161), the only human homologue of mouse N K l . l , which may provide crucial costimulatory signals [145]. NK-T cells also express two other C-type lectins, CD94 (the ligand for HLAE) and CD69, but lack expression of most other NK markers including CD 16, CD56, CD57 or killer inhibitory receptors [142]. Similar to the mouse NKT cells, human NK-T cells function as potent cytolytic effectors and secrete cytokines with a ThO phenotype in response to CDld-transfected APC [142, 145]. These functional activities are significantly enhanced by the lipid antigen, oc-GalCer isolated from marine sponge [146-148]. Although much is known about the cellular immunology of CDla-d, little is known about the role of CD Id in human diseases. The information to date points toward a crucial role of CD Id in immunoregulation to microbial antigens given the vigorous cytokine production of CD Id-responsive, NK-T cells and observations that NK-T cells are significantly diminished in diabetes mellitus [149], systemic lupus erythematosus [150] and systemic sclerosis [151] and immunosurveillance to microbial antigens functions given the potent anti-tumor responses of NK-T cells in mouse models [152]. The CD Id pathway is particularly relevant to IBD pathogenesis given its biologic relevance to the central hypothesis raised above. CD Id is expressed by lEC [153, 154] and professional APC [122], and on

135 lEC is able to present model glycolipid antigens to NK-T cells [155]. This presentation is polarized in that it is most efficient basally relative to apically, suggesting a role for CD Id in presenting glycolipid antigens from luminal microbial antigens to subepithelial CD Id-restricted T cells such as the NK-T cells [155]. Indeed, activation of NK-T cells in a CD Id-restricted manner leads to the amelioration of colitis in the dextran sodium sulfate colitis model [156] and promotion of Th2-mediated colitis [W. Strober and R. Blumberg, unpublished observation] suggesting complex interactions of NK-T cells in mucosal tissues relative to IBD induction.

Mechanisms of innate immunity tlirougliout tlie intestinal mucosa The intestinal lumen, particularly that of the colon, contains a large variety of bacteria and bacterially derived products [157, 158]. As a consequence a number of 'broadly specific' defense mechanisms have evolved to guard against the risk of attack from invasive microorganisms. These factors are important because they can be mobilized rapidly in comparison to the time taken for mobilization of the adaptive responses described in the sections above. These mechanisms can be classified as intrinsic or extrinsic in nature. Intrinsic mechanisms of immunity are derived from the physical presence of an epithelial barrier and are dependent upon the unique structural properties exhibited by the lEC. Extrinsic defenses are defined as those processes which act outside of the monolayer to resist microbial interaction with the mucosa. In addition to these two mechanisms, cells constituting the epithelial monolayer have developed an ability to interact with immune cell populations of the underlying intestinal mucosa, thereby alerting the immune system to the presence of luminal pathogens. These three levels of innate immunity are described below.

Intrinsic barriers The formation of a selectively permeable epithelial barrier is essential in preventing the uncontrolled passage of pathogenic antigens from the external environment to the internal tissue. Estabhshment of the epithelial monolayer by contributing lEC is dependent upon a considerably high degree of intracellular and intercellular organization [159]. Besides

136 allowing the formation of tight intercellular junctions, lEC exhibit a number of structural and functional features that help physically restrict the passage of potentially harmful microorganisms into the underlying mucosa. Interaction of macromolecules with the apical surface of lEC is reduced by the presence of large heavily glycosylated proteins, or mucins, anchored to the epithelial plasma membrane [160]. In addition, the organization of the apical membrane into microvilli is thought to minimize the contact area available to luminal macromolecular antigens [161]. Cell trafficking processes within the I EC may also reduce the amount of intact microbial antigen crossing the monolayer. The contents of epithelial cell endocytic vesicles are, for example, primarily directed to proteolytic lysosomal compartments. Thus, internalization of many microorganisms is liable to lead to their degradation.

Extrinsic barriers Mucus The surface of the intestinal epithelium is lined with a layer of mucus produced as a result of mucin secretion by goblet cells [162, 163]. Mucins, a group of glycoproteins of which the mucus lining is composed, protect the epithelium in several ways. By creating a hydrated viscous layer they act as a physical barrier, separating I EC from the turbulent luminal environment ofthe intestinal tract [164, 165]. In addition, the carbohydrate moieties present in mucins display the ability to adhere to the surfaces of many microoganisms, preventing microbial association with the monolayer [166-168]. With time, mucus is propelled down the intestinal tract thereby facilitating the removal of bound microbial components away from the epithelia. The mucus lining is therefore considered to contribute to innate immune responses within the intestine. In support of this role it has been observed that mucus secretion is induced in the presence of agents potentially harmful to the epithelium including noxious chemical and bacterial toxins [169-171]. Such a response may act to counter the presence of these agents by reducing their capacity to interact with the monolayer. Defensins A second extrinsic mechanism of mucosal defense occurs by the secretion of agents that directly exhibit microbicidal or antiviral activities. Investigations performed on sections of the mouse epithelium have

The lymphocyte-epithelial-bacterial

interface

revealed the presence of mRNA encoding cyptdins or a-defensins, molecules sharing a conserved cysteine-rich motif with the family of antimicrobial peptides known as defensins [172]. Expression of cryptdin mRNA was confined to Paneth cells, located within intestinal crypts. Further studies revealed the presence of two peptides in these cells, termed cryptdin 1 and cryptdin 2 [173]. Both these peptides display a high degree of sequence homology to the defensins secreted by polymorphonuclear leukocytes and exhibit potent microbicidal activity on a variety of bacteria including strains of Escherichia coli. Salmonella typhimurium and Listeria monocytogenes. The potential importance of cryptdins in conferring resistance to bacterial infection is highlighted by the observation that, while avirulent strains of S. typhimurium are susceptible to the antimicrobial properties of cryptdins, the growth of pathogenic S. typhimurium is not affected by these peptides [173]. Although unclear, it remains possible that the loss of cryptdin sensitivity in the pathogenic strain contributes to its enhanced virulence. Intestinal epithelial cells also express defensins. Recent studies have revealed that human lEC express two functionally related peptides termed (3defensin 1 and P-defensin 2 (hBDl and hBD2 respectively) [174]. hBDl was shown to be constitutively expressed while hBD2 was induced following stimulation with the cytokine IL-1. Like the a-defensins, both hBDl and hBD2 exhibit antimicrobial activity against a range of bacteria including E. coli and strains o^ Salmonella spp. and Pseudomonas spp. Lysozyme and lactoferrin Lysozyme and lactoferrin represent two proteins which, like defensins, are thought to contribute to the innate defense of mucosal surfaces [175]. Both lysozyme and lactoferrin are found in exocrine secretions and can be detected within the lumen of the human gastrointestinal tract. Lactoferrin (Lf), a molecule that is thought to facilitate iron transport across the epithelium of suckling infants, is cleaved by gastric pepsin to form a peptide exhibiting potent antimicrobial properties [176]. In vitro, Lf inhibits replication of a number of viruses including cytomegalovirus (CMV), herpes simplex virus (HSV) and members of the rotavirus family [177, 178]. Lf interacts with both the lipoglycan-coated cell wall of Gram-positive bacteria and the outer membrane of Gram-negative bacteria, processes that are thought to mediate its bactericidal function. In addition, Lf inhibits the adhesion of enterotoxigenic E. coli to

Robert Hershberg and Richard S. Blumberg human epithelial cells, thereby preventing its colonization of the surface monolayer [179]. Lactoferrin is also capable of inhibiting fungal growth, and has been shown to be cytotoxic to several protozoan parasites such as trophozoites of Giardia lamblia and Toxoplasma gondii [\^0, 181]. Human lysozyme consists of a single polypeptide chain of 129 amino acids, which is crossHnked by four stabilizing disulfide interactions. The polypeptide is a constituent of both saliva and pancreatic juice. Luminal concentrations within the intestinal tract range from 43 to 106 micrograms per milliliter [175]. Lysozyme exerts its microbicidal activity at a number of levels. These include the hydrolysis of bacterial cell wall peptidoglycans, the activation of bacterial autolysins, the induction of bacterial cell aggregation and the abrogation of bacterial adhesion to host epithelial cells. However, while lysozyme may possess multiple enzymatic activities, it is thought that its cationic properties are primarily responsible for its ability to interact strongly with bacterial membranes and may contribute to the disruption of membrane processes [182]. In many cases lysozyme alone does not alter bacterial viability but does render the bacteria more susceptible to lysis by other environmental factors. In this regard a combination of lysozyme with lactoferrin is microbicidal for Streptococcus mutans while lysozyme alone has no visible effect on the growth of this bacterial strain [183]. Similarly, lysozyme synergizes with peroxides in preventing glucose uptake by S. mutans [175]. In vivo therefore, lysozyme is likely to function in association with additional antimicrobial agents. Other microbicidal factors, which function as innate mechanisms of defense at mucosal surface, include peroxidases and components of the complement pathway. Studies using the human epithelial cell line Caco-2 have demonstrated that these cells express and secrete complement components including C3, C4 and factor B [184]. Peroxidases, secreted from exocrine glands onto mucosal surfaces, catalyze the peroxidation of halides such as chloride, bromide and thiocyanide into oxidative products, which exhibit potent antimicrobial activity.

Epithelial cell-Immune cell cross-talk The exposed location of the epithelia implies that lEC are likely to play an important role in alerting components of the mucosal immune system to the presence of foreign antigen. Intestinal epithelial cells coexist with immune cell populations present along

137 the epithelium itself and those localized to the lamina propria. ilEL reside at the basolateral surface of the epithelium. Immune cell populations within the lamina propria include lymphocytes, macrophages, granulocytes and mast cells. Within this environment lEC have developed an ability to communicate with regional immune cell populations in a manner proposed to directly influence their growth, migration and state of responsiveness to antigenic stimuli. Some of the methods by which lEC affect the mucosal immune system are categorized and discussed below.

Chemokines Under conditions that promote infection or insult to the epithelial monolayer, an influx of immune and inflammatory cells into the mucosa is observed as these cells are recruited to the site of injury. This recruitment can occur within hours of infection and is mediated, partly, by a group of chemotactic cytokines known as chemokines [185, 186]. Several studies have revealed the ability for lEC to express and secrete chemokines including IL-8, ENA 78, gro-oc, MIP-loc, and MCP-1 [187]. Infection of the epithelial cell lines T84, HT29 and Caco-2 with Salmonella dublin, Yersinia enterocolitica or Shigella dysentariae, for example, selectively induces expression of IL-8 and MCP-1 [188-191]. Following cellular invasion by S. typhimurium, IL-8 is secreted from the basolateral surface establishing a chemical gradient and facilitating the migration of neutrophils into the paracellular spaces of the monolayer [189]. IL-8 expression is also induced in gastric epithelial cells in response to Helicobacter pylori infection [192, 193]. Taken together, these data demonstrate the ability of lEC to actively participate in recruiting immune cells, to sites of infection, through the release of chemotactic factors upon exposure to luminal pathogens.

Cytokines Cytokines are a group of factors that exert profound influence on the functional state of immune cell populations through inducing an altered pattern of gene expression upon binding to specific cellular receptors [194]. Intestinal epithelial cells are capable of expressing a number of cytokines, thereby possessing the ability to affect local immune responses. The cell lines T84 and Caco-2 constitutively express mRNA for the proinflammatory cytokines, IL-la, IL-ip, IL-8, IL-15 and TNF-a [187, 191, 195, 196]. These cells also express TGF-P, a cytokine thought

138 to play a role both in promoting lEC growth and in regulating lEC barrier function and responses of lEL to antigenic stimuh [197-199]. IL-6, a cytokine thought to be important in generating IgA-secreting plasma cells, has been shown to be expressed in freshly isolated human lEC, and can be induced in a variety of epithelial cell lines by cholera toxin secreted from Vibrio cholerae [2]. IL-6 release by I EC may therefore contribute to the production of secretory IgA. IL-10 represents another cytokine expressed by I EC [46]. Like IL-6 it is thought to promote IgA secretion by B cells [201]. In addition to this, IL-10 acts as a suppressor of Thl responses through inhibiting macrophage function and cytokine secretion from Thl cells [202]. Recently, studies have shown that IL-10 secretion from T84 cells can protect the epithelial monolayer from IFN-yinduced permeabilization [46, 203, 204]. Its ability both to suppress Thl responses and to maintain epithelial barrier function implicates IL-10 as a regulator of intestinal inflammation. The production of IL-10 from I EC may therefore act to protect the barrier against injury resulting from an influx of inflammatory effector cells. A number of additional factors secreted from lEC are thought to be important in influencing the growth and development of surrounding ilEL. Mice lacking the functional gene encoding IL-7 exhibit reduced levels of yS"^ T cells in the epithelia, while IL-7 receptor knockout mice do not possess any intestinal yS"*" T cells [205, 206]. IL-7 has also been shown to play a role in activating T cell cytolytic function against various intracellular parasites including HIV [207, 208]. Such findings, together with the observation that freshly isolated lEC produce IL-7 [209], suggest that epithelial-derived IL-7 may contribute to maintaining ilEL growth and responsiveness to foreign antigens. In addition to IL-7, two other molecules secreted from epithelial cells are important in influencing ilEL activity. Stem cell factor, or SCF, is required for sustaining levels of yS"^ T cells in the intestine [210]. Furthermore, invitro studies suggest that SCF can protect lEC from bacterial infection [211]. Murine epithelial cells have also been shown to express thyroid-stimulating hormone (TSH) [212]. Secreted following the action of thyrotropin-releasing hormone on its receptor, TSH is critical in maintaining normal levels of CD8aP'^ T cells within lEL populations [212, 213]. Intestinal epithelial cells are also able to synthesize a variety of prostanoids. Rabbit colonic lEC consti-

The lymphocyte-epithelial-bacterial

interface

tutively express prostaglandin E2 (PGE2), 6-ketoprostaglandin Floe (6-keto PDFia), prostaglandin D2 (PGD2) and prostaglandin F2a (PGF2a) [214]. PGE2 has been shown to reduce IL-3 secretion from lamina propria mononuclear cells, inhibit the effects of local T cells on the epithelial barrier and reduce IL-2 secretion by activated T cell lines [214, 215]. Based on such data it seems likely that the release of PGE2 from lEC functions to regulate leukocyte responses within the intestinal mucosa. Cytokine receptors Besides exhibiting an ability to influence mucosal immune responses, lEC are also able to alter their own phenotype in response to cytokines released from local immune cell populations. Freshly isolated human I EC, as well as a number of lEC lines, express mRNA for the common gamma chain and specific alpha chains of the receptors for IL-2, IL-4, IL-7, IL9, and IL-15 [196, 197, 216]. Northern blot analysis has further revealed the presence of the IL-2 receptor P chain while studies of rat I EC demonstrate the expression of the IL-1 receptor (IL-IR), a molecule homologous to the human type I IL-IR [216, 217]. Through the expression of cytokine receptors the epithelium is sensitive to immunologic changes within its environment and can subsequently respond in a way that either facilitates or regulates those responses. In the presence of IL-4, IFN-y or TNF-oc, for example, lEC respond by enhancing levels of the polymeric immunoglobulin receptor, thus facilitating the release of secretory IgA [218220]. IFN-y also induces the expression of MHC class II molecules and the epithelial cell adhesion molecule, ICAMl [221-224]. IFN-y is capable of stimulating cell surface expression of the co-stimulatory factor CD86 while IL-6 and IL-1 trigger the release of complement components C3, C4 and factor B from lEC [184, 225, 226]. Receptiveness of the epithelium to the local cytokine environment, therefore, allows the mucosal immune system to enhance both the epithelial cells' innate capacity for defense and their ability to interact with and influence local immune cell populations. Toll-like receptors The mucosal immune system represents a dominant interface between the two main arms of the immune system - so-called 'innate' and 'adaptive' immunity. We have outlined the various cellular components of the adaptive immune system, in particular, the diverse T cell populations with somatically rear-

Robert Hershberg and Richard S. Blumberg

139

Table 3. Mechanisms of bacteria-driven IBD Organism, molecule or cellular subset

Mechanism of action

lEC adherence

ETEC EPEC Yersinia enterocolitica

Specific colonization factors Attachment and effacement via Bfp pill Intimin binds to oc4(31 integrin

lEC invasion

Shigella spp. Salmonella spp.

Type III secretion system Type III secretion system

Cytotoxins

Clostridium difficile EHEC

C. difficile toxins A and B Shiga-Wke toxins

Luminal dysbiosis

Abnormal metabolism

Eubacteria spp.

Hydrogen-sulfide inhibition of SOFA utilization

Pattern recognition molecules of innate immune system

Microbial poly-saccharide recognition

C-reactive protein

Activates complement and phagocytosis Facilitates phagocytosis Binds 019 receptor and promotes phagocytosis

Pathogenic category

Functional category

Microbial virulence factors

Serum amyloid protein Mannose binding protein Microbial glyco-lipid recognition

DEC-205 Lipipolysaccharide binding protein SolubleCD14andCD14 Surfactant protein A

Adaptive (specific immune responses

Cell surface protein that targets MHO class II pathway Inactivates LPS Binds and regulates proinflammatory response to LPS Binds and aggregates lipids

Cellular

004"" T cell

Recognition of bacterial peptide antigens on APO in context of MHO class II leading to excess proinflammmatory (TNF-oc, IFN-y) relative to anti-inflammatory (TGF-P) cytokines

Humoral

Bcell

Production of autoantibodies that cross-react with microbial antigens (e.g. ANOA cross-reactivity with 12 sequence of Pseudomonas spp.

Adapted from refs 15, 27, 30. ETEO, enterotoxygenic E coll; EHEO, enterohemorrhagic £ coli; EPEO, enteropathogenic E coli; ANOA, antineutrophil cytoplasmic antibody.

ranged T cell receptors that recognize specific processed protein fragments displayed by histocompatibility molecules on the surface of APC. While it is likely that 'specific' T cell (and B cell) responses to bacteria and other luminal microorganisms will be increasingly implicated in the pathogenesis of IBD, a growing body of literature suggests that 'innate' responses to the mucosal microbiota are likely to have a dramatic influence on the immunological tone at mucosal surfaces. As detailed, the mucosal microbiota is extremely diverse, with most major classes of bacteria, yeast

and, often, helminthes represented. In this complex microbiological ecosystem the challenge of the innate system is daunting. It has become increasingly clear that, in order for the innate immune system to be able to recognize and respond to such a large number of diverse microbial stimuli, it has evolved an intricate system of 'pattern recognition' [30]. Specifically, this involves the ability to respond to structural motifs shared by diverse groups of organisms (e.g. lipid A from Gram-negative bacteria, peptidoglycans from Gram-positive bacteria (Table 3).

The lymphocyte-epithelial-bacterial

140 Recent data on the Toll-like receptors (TLR) in humans and mice have provided keen insight regarding the mechanisms by which this diverse pattern recognition occurs [227, 228]. The term Toll-like derives from structural homology to the gene product of the Toll locus in the fruitfly Drosophila melanogaster [229]. Mutations in Toll, and genes relevant to the signaling events downstream from this surface receptor (with homology to the receptor for IL-1), result in marked defects in innate immunity (particularly to yeast and Gram-positive bacteria). Homologues to Drosophila Toll have been found in mice and humans, with a growing family of proteins with homology to Toll, currently with 10 distinct members comprising the list of TLR. Interestingly, the different TLR appear to have distinct ligand specificity - in the cases so far determined, representing structural disparate 'bioactive' components of microorganisms. This rapidly changing area of scientific inquiry has been the topic of several recent insightful reviews [230, 231]. One particular example of TLR biology illustrates both the importance of these molecules in innate immunity and the potential relevance to inflammatory responses of the gastrointestinal tract. First, in an elegant series of genetic studies in mouse strains hyporesponsive to bacterial LPS (e.g. C3H/HeJ), mutations were identified in a conserved region of TLR4 [232]. Subsequent studies from a number of laboratories have demonstrated that (in concert with LPS binding protein, LBP and surface or soluble CD 14) TLR4 is the 'dominant' surface molecule that confers a signal by LPS [233]. Recent data are emerging to indicate that TLR2 may be involved in signaling by LPS of certain bacteria, for example, Porphomonas gingivalis. The pattern of expression of TLR4 is variable, and includes macrophages and dendritic cells. Interestingly, functional TLR4 has also been shown to be expressed on non-hematopoietic cells, including lEC [234]. Recently it has been demonstrated that TLR4 is selectively upregulated on lEC from patients with IBD. Based on these studies one might predict an 'exaggerated' response to LPS (a potent immunomodulatory molecule driving 'proinflammatory' responses) in IBD. Additionally, these studies highlight TLR4 as a potential target for antagonism therapeutically in IBD. The complexity is likely to increase further as one considers the role of other bacterially derived immunomodulators (e.g. CpG islands in bacterial DNA that signal via TLR9 [235]) and potential interactions between the TLR within the anatomi-

interface

cally varied sites along the gastrointestinal tract. This is likely to be an extremely important area with considerable therapeutic relevance in the coming years.

References 1.

2. 3.

4. 5. 6. 7. 8. 9. 10. 11. 12.

13.

14.

15.

16. 17. 18. 19.

McGhee JR, Lamm ME, Strober W. Mucosal immune responses: an overview. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 485-506. Strobel S, Mowat AM. Immune responses to dietary antigens: oral tolerance. Immunol Today 1998; 19: 173 81. Blumberg RS, Saubermann LJ, Strober W. Animal models of mucosal inflammation and their relation to human inflammatory bowel disease. Curr Opin Gastroenterol 1999; 11:648 56. Strober W, Blumberg RS. Inflammatory bowel diseases. J Am Med Assoc 2001; 285: 643 7. Drasar BS, Barrow PA. Intestinal microbiology. In: Aspects of microbiology 10 [monograph]. Am Soc Microbiol. Berkshire, UK, 1985. Sutherland L, Singleton J, Sessions J ct al. Double blind, placebo-controlled trial of metronidazole in Crohn's disease. Gut 1991; 32: 1071. Rutgcerts P, Gcboes K, Peeters M cf al. Effect of faecal stream diversion on recurrence of Crohn's disease in the neolcrminal ileum. Lancet 1991; 338: 771. Shanahan F. Probiotics and inflammatory bowel disease: is there a scientific rationale? Inflam Bowel Dis 20; 6: 107 15. Sartor RB. Enteric microflora in IBD: pathogens or commensals? Inflam Bowel Dis 1997; 3: 230. Fox J. Enterohepatic helicobacters: natural and experimental models. Ital J Gastroenterol Hepatol 1998; 30: S264 9. Dielman LA, Arends A, Tonkonogy SL ct al. Helicobacter hepaticus does not induce or potentiate colitis in interleukin10-deficient mice. Infect Immun 20; 68: 5107 13. Duchmann R, Kaiser I, Hermann E, Mayet W, Ewe K, Meyer zum Buschenfelde K-H. Tolerance exists towards resident intestinal flora but is broken in active inflammatory bowel disease (IBD). Clin Exp Immunol 1995; 102: 448-55. Savage DC. Mucosal microbiota. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 19-30. Onderdonk AB. Intestinal microflora and inflammatory bowel disease. In: Kirsner, J and Shorter RG, eds. Inflammatory Bowel Disease, 5th edn. Baltimore: Williams & Wilkins, 2000: 144-52. Sartor RB. Microbial factors in the pathogenesis of Crohn's Disease, ulcerative colitis, and experimental intestinal inflammation. In: Kirsner, J and Shorter RG, eds. Inflammatory Bowel Disease, 5th edn. Baltimore: Williams & Wilkins, 2000: 153-78. Theron J, Cloete TE. Molecular techniques for determining microbial diversity and community structure in natural environments. Crit Rev Microbiol 20; 26: 37-57, Schmidt TM, Relman DA. Phylogenetic identification of uncultured pathogens using ribosomal RNA sequences. Methods Enzymol 1994; 235: 205-22. Kolbert CP, Persing DH. Ribosomal DNA sequencing as a tool for identification of bacterial pathogens. Curr Opin Microbiol 1999; 2: 299-305. Tannock GW. Analysis of the intestinal microflora: a renaissance. Antonie Van Leeuwenhoek 1999; 76: 265-78.

Robert Hershberg and Richard S. Blumberg 20. Gorbach SL, Plaut AG, Nahas L et al Studies of intestinal microflora. II. Microorganisms of the small intestine and theirrelations to oral and fecal flora. Gastroenterology 1967; 73: 856. 21. Warren JR, Marshall BJ. Unidentified curved bacilli on gastric epithelium in active gastritis. Lancet. 1983; 1: 12735. 22. Drasar BS, Shiner M, McLeod GM. Studies on the intestinal flora. I. The bacterial flora of the gastrointestinal tract in healthy and achlorhydric persons. Gastroenterology 1969; 56: 71-9. 23. Banerjee S, LaMont JT. Treatment of gastrointestinal infections. Gastroenterology 20; 118: 548-67. 24. Onderdonk AB, Hermos JA, Bartlett JG. The role of the intestinal microflora in experimental colitis. Am J Clin Nutr 1977; 30: 1819. 25. Van de Merwe JP, Schroder AM, Wesninck F et al. The obligate anaerobic faecal flora of patients with Crohn's disease and their first-degree relatives. Scand J Gastroenterol 1988; 23: 1125. 26. Burke DA, Axon ATR. Adhesive Escherichia coli in inflammatory bowel disease and infective diarrhea. Br Med J 1988; 297: 102. 27. Lee C, Mekalanos J. Bacterial interactions with intestinal epithelial cells. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 65769. 28. Roediger WEW, Duncanb A, Kapaniris OK et al. Reducing sulfur compounds of the colon impair colonocyte nutrition: implications for ulcerative colitis. Gastroenterology 1993; 104: 803. 29. Breuer RI, Soergel KH, Lashner B et al. Short chain fatty acid rectal irrigation for left-sided ulcerative colitis: a randomized, placebo controlled trial. Gut 1997; 40: 485. 30. Fearon DT, Locksley RM. The instructive role of innate immunity in the acquired immune response. Science. 1996; 272: 50-4. 31. Sartor RB, Rath HC, Sellon RK. Microbial factors in chronic intestinal inflammation. Curr Opin Gastroenterol 1996; 12: 327. 32. Abreu-Martin MT, Targan SR. Regulation of immune responses of the intestinal mucosa. Crit Rev Immunol 1996;16: 277-309. 33. Germain R. MHC-dependent antigen processing and peptide presentation: providing ligands for T lymphocyte activation. Cell. 1994; 76: 287-99. 34. Blumberg R, Gerdes D, Chott A, Porcelli S, Balk S. Structure and function of the CD 1 family of MHC-like cell surface proteins. Immunol Rev. 1995; 147: 5-29. 35. Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis. Gastroenterology 1998; 115: 182-205. 36. Chott A, Gross GG, Probert C, Schwartz VL, Blumberg RS, Balk SP. Analysis of T cell antigen receptor expression by intestinal mucosa lymphocytes demonstrates a common junctional motif among CD8'^ T cells in ulcerative colitis. J Immunol 1996a; 156: 3024-35. 37. Probert CS, Chott A, Turner JR, Bodinaku K, Elson CO, Balk SP, Blumberg RS. Persistent clonal expansions of peripheral blood CD4^ lymphocytes in chronic inflammatory bowel disease. J Immunol 1996b; 157: 3182-91. 38. Mizoguchi A, Mizoguchi E, Saubermann SJ, Higaki K, Blumberg RS, Bhan AK. Limited CD4'' T cell diversity associated with colitis in T cell receptor a mutant mice requires a T helper 2 environment. Gastroenterology 20; 119:983-95. 39. Cohavy O, Harth G, Horwitz M et al. Identification of a novel mycobacterial histone HI homologue (HupB) as an antigenic target of pANCA monoclonal antibody and serum

141

40. 41. 42.

43.

44. 45. 46.

47. 48. 49.

50. 51.

52. 53. 54.

55. 56. 57.

58. 59.

immunoglobulin A from patients with Crohn's disease. Infect Immun 1999; 67: 7510-17. Kraehenbuhl JP, Neutra MR. Epithelial M cells: diff'erentiation and function. Annu Rev Cefl Dev Biol 20; 16: 301-32. Davis IC, Owen RL. The immunopathology of M cells. Springer Sem Immunopathol 1997; 18: 421-48. Cone RA. Mucus. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 4 3 64. Mitic LL, Van Itallie CM, Anderson JM. Molecular physiology and pathophysiology of tight junctions I. Tight junction structure and function: lessons from mutant animals and proteins. Am J Physiol Gastrointest Liver Physiol 120; 279: G250-4. Colgan SP, Parkos CA, Matthews JB et al. Interferon-y induces a cell surface phenotype switch on T84 intestinal epithehal cefls. Am J Physiol 1994; 267: C402-10. Zund G, Madara JL, Dzus AL, Awtrey CS, Colgan SP. Interleukin-4 and interleukin-13 diff'erentially regulate epithehal chloride secretion. J Biol Chem 1996; 271: 7 4 6 0 ^ . Colgan SP, Hershberg RM, Furuta GT, Blumberg RS. Ligation of epithelial CD Id by antibody crossHnking induces bioactive IL-10; critical role of the cytoplasmic tail in autocrine signaling. Proc Natl Acad Sci USA 1999; 96: 13938-43. Liu Y, Nusrat A, Schnell FJ et al. Human junction adhesion molecule regulates tight junction reseahng in epithelia. J CeUSci20; 113:2363-74. Kinugasa T, Sakaguchi T, Gu X, Reinecker HC. Claudius regulate the intestinal barrier in responseto immune mediators. Gastroenterology 20; 118: 1001-11. Miki K, Moore DJ, Butler RN, Southcott E, Couper RTG, Davidson GP. The sugar permeability test reflects disease activity in children and adolescents with inflammatory bowel disease. J Pediatr 1998; 133: 750-4. Munkholm P, Langholz E, Hollander D et al. Intestinal permeability in patients with Crohn's disease and ulcerative cohtis and their first degree relatives. Gut 1994; 35: 68-72. Madsen KL, Malfair D, Gray D, Doyle JS, Jewell LD, Fedorak RN. Interleukin-10 gene-deficient mice develop a primary intestinal permeability defect in response to enteric microflora. Inflam Bowel Dis 1999; 5: 262-70. Mashimo H, Wu DC, Podolsky DK, Fishman MC. Impaired defense of intestinal mucosa in mice lacking intestinal trefoil factor. Science 1996; 274: 262-5. Brandtzaeg P, Farstad IN, Helgeland L. Phenotypes of T cefls in the gut. Chem Immunol. 1998; 71: 1-26. Dogan A, Dunn-Walters DK, MacDonald TT. Demonstration of local clonality of mucosal T cells in human colon using DNA obtained by microdissection of immunohistochemically stained tissue sections. Eur J Immunol 1996; 26: 1240-5. Itohara S, Farr AG, Lafaille JJ et al. Homing of y5 thymocyte subset with homogeneous T-cell receptors to mucosal epithelia. Nature 1990: 343: 754-7. Farstad IN, Norstein J, Brandtzeg P. Phenotypes of B and T cells in human intestinal and mesenteric lymph. Gastroenterology 1997; 112: 163-73. Lefrancois L, Fuller B, Huleatt JW, Olson S, Puddington L. On the front lines: intraepithelial lymphocytes as primary effectors of intestinal immunity. Springer Sem Immunopathol. 1997; 18: 463-76. Molberg O, Nilsen EM, Soflid LM et al. CD4"' T cells with specific reactivity against astrovirus isolated from normal human smafl intestine. Gastroenterology 1998; 114: 115-22. James SP, Graeff" AS, Zeitz M. Predominance of helperinducer T cells in mesenteric lymph nodes and intestinal

The lymphocyte-epithelial-bacterial

142

60. 61. 62.

63. 64.

65.

66. 67. 68.

69.

70. 71.

72.

73.

74.

75.

76.

lamina propria of normal nonhuman primates. Cell Immunol 1987; 107: 372-83. Brandtzaeg P, Haraldsen G, Rugtveit J. Immunopathology of human inflammatory bowel disease. Springer Sem Immunopathol. 1997; 18: 555-89. Hurst SD, Sitterding SM, Ji S, Barrett TA. Functional differentiation of T cells in the intestine of T cell receptor transgenic mice. Proc Natl Acad Sci USA 1997; 94: 3920-5. Moghaddami M, Cummins A, Mayrhofer G. Lymphocytefilled villi: comparison with other lymphoid aggregations in the mucosa of the human small intestine. Gastroenterology 1998;115: 1558. Suzuki K, Oida T, Hamada H et al. Gut cryptopatches: direct evidence of extrathymic anatomical sites for intestinal T lymphopoiesis. Immunity 20; 13: 691-702. Umesaki Y, Setoyama H, Matsumoto S, Okada Y. Expansion of ocp T-cell receptor-bearing intestinal intraepithelial lymphocytes after microbial colonization in germfree mice and its independence from thymus. Immunology 1993; 79: 32 7. Helgeland L, Vaage JT, Rolstad B, Midtvedt T, Brandtzaeg P. Microbial colonization influences composition and T-cell receptor Vp repertoire of intraepithelial lymphocytes in rat intestine. Immunology 1996; 89: 494 501. Bahram S, Bresnahan M, Geraghty DE, Spies T. A second lineage of mammalian major histocompatibility complex class I genes. Proc Natl Acad Sci USA 1994; 91: 6259 63. Balk S, Ebert E, Blumenthal R et al. Oligoclonal expansion and CDl recognition by human intestinal intraepithelial lymphocytes. Science 1991; 253: 1411 15. Blumberg RS, Yockey CE, Gross GG, Ebert EC, Balk SP Human intestinal intraepithelial lymphocytes are derived from a limited number of T cell clones that utilize multiple v p T cell receptor genes. J Immunol. 1993; 150: 5144 53. Gross GG, Schwartz VL, Stevens C, Ebert EC, Blumberg RS, Balk SP. Distribution of dominant T cell receptor p chains in human intestinal mucosa. J Exp Med. 1994; 180: 1337 44. Cepek KL, Parker CM, Madara JL, Brenner MB. Integrin cx^Pv mediates adhesion of T lymphocytes to epithelial cells. J Immunol 1993; 150: 3459 70. Papadakis KA, Prehn J, Nelson V et al. The role of thymusexpressed chemokine and its receptor CCR9 on lymphocytes in the regional specialization of the mucosal immune system. J Immunol 20; 165: 5069-76. Jabri B, de Serre NP, Cellier C et al. Selective expansion of intraepithelial lymphocytes expressing the HLA-E-specific natural killer receptor CD94 in celiac disease. Gastroenterology 20; 118:867-79. Christ AD, Colgan SP, Balk SP, Blumberg RS. Human intestinal epithelial cell lines produce factor(s) that inhibit CD3-mediated T-lymphocyte proliferation. Immunol Lett 1997; 58: 159-65. Cebra JJ, Jiang H-Q, Sterzl J, Tlaskalova-Hogenova H. The role of mucosal microbiota in the development and maintenance of the mucosal immune system. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 267-80. Berg RD, Savage DC . Immune responses of specific pathogen-free and gnotobiotic mice to antigens of indigenous and nonindigenous microorganisms. Infect Immun 1975; 11: 320-9. Nieuwenhuis ESS, Visser MR, Kavelaars A et al. Oral antibiotics as a novel therapy forarthritis; evidence of a beneficial eff'ect of intestinal Escherichia colL Arthritis Rheum 20; 43: 2583-9.

77. 78. 79.

80.

81. 82.

83.

84.

85.

86. 87. 88.

89. 90.

91. 92.

93. 94.

95. 96.

interface

Christ AD, Blumberg RS. The intestinal epithelial cell: immunological aspects. Springer Sem Immunopathol. 1997; 18:449-62. Kelsall BL, Strober W. Dendritic cells of the gastrointestinal tract. Springer Sem Immunopathol. 1997; 18: 409-20. Prey A, Giannasca KT, Weltzin R et al. Role of the glycocalyx in regulating access of microparticles to apical plasma membranes of intestinal epithelial cells: implications for microbial attachment and oral vaccine targeting. J Exp Med 1996; 184,1045-59. 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. Dickinson BL, Badizadegan K, Wu Z et al. Bidirectional FcRn-dependent IgG transport in a polarized human intestinal epithelial cell line. J Clin Invest 1999; 104: 903-11. Zhu X, Meng G, Dickinson BL et al. MHC class I-related neonatal Fc receptor for IgG is functionally expressed in monocytes, macrophages and dendritic cells. J Immunol 2001; 166: 3266 76. Witmer-Pack MD et al. Tissue distribution of the DEC-205 protein that is detected by the monoclonal antibody NLDC145. II. Expression in situ in lymphoid and nonlymphoid tissues. Cell Immunol 1995; 163: 157 62. Barone KS, Jain SL, Michael JG et al. Eff'ect of in vivo depletion of CD4^ and CD8'^ cells on the induction and maintenance of oral tolerance. Cell Immunol . 1995; 163: 19 29. Chen Y, Inobe J, Weiner HL. Induction of oral tolerance to myelin basic protein in CD8-depleted mice: both CD4^ and CDS"*^ cells mediate active suppression. J Immunol 1995; 155:910 16. Garside P, Steel M, Liew FY, Mowat AM. CD4^ but not CD8"^ T cells are required for the induction of oral tolerance. Int Immunol 1995; 7: 501 4. Powrie F. T cells in inflammatory bowel disease: protective and pathogenic roles. Immunity 1995; 3: 1714. Beagley KW, Fujihashi K, Lagoo AS et al. Differences in intraepithelial lymphocyte T cell subsets isolated from murine small versus large intestine. J Immunol 1995; 154: 5611 19. Camerini V, Panwala C, Kronenberg M. Regional specialization of the mucosal immune system. J Immunol 1993; 151:1765-76. Bland PW, Whiting CV. Induction of MHC class II gene products in rat intestinal epithelium during graft-versushost disease and effects on the immune function of the epithelium. Immunology 1992; 75: 366-71. Kaiserlian D, Vidal K, Revillard JP. Murine enterocytes can present soluble antigen to specific class Il-restricted CD4^ T cells. Eur J Immunol. 1989; 19: 1513-16. Mayer L, Eisenhardt D, Salomon P et al. Expression of class II molecules on intestinal epithelial cells in humans. Differences between normal and inflammatory bowel disease. Gastroenterology 1991; 1: 3-12. Hershberg RM, Framson PE, Cho DH et al. Intestinal epithelial cells utilize two distinct pathways for HLA class II antigen processing. J Clin Invest. 1997; 1: 204-15. Wolf PR, Ploegh HL. How MHC class II molecules acquire peptide cargo: biosynthesis and trafficking through the endocytic pathway. Annu Rev Cell Dev Biol 1995; 11: 267306. Chang C-H, Flavell RA. Class II transactivator regulates the expression of multiple genes involved in antigen presentation. J Exp Med 1995; 181: 765-7. Hershberg RM, Cho DH, Youakim A et al. Highly polarized HLA class II antigen processing and presentation by human intestinal epithelial cells. J Clin Invest 1998; 102: 792-803.

Robert Hershberg and Richard S. Blumberg 97.

98. 99.

100.

101.

102.

103.

104.

105.

106. 107.

108. 109. 110. 111. 112. 113.

114. 115.

Katz JF, Stebbins C, Appella E, Sant AJ. Invariant chain and DM edit self-peptide presentation by major histocompatibility complex (MHC) class II molecules. J. Exp. Med. 1996;184: 1747-53. Pinet V, Vergelli M, Martin R et al. Antigen presentation mediated by recycling of surface HLA-DR molecules. Nature 1995;375: 603-6. Zhong G, Romagnoli P, Germain RN. Related leucinebased cytoplasmic targeting signals in invariant chain and major histocompatibility complex class II molecules control endocytic presentation of distinct determinants in a single protein. J Exp Med 1997; 185: 429-38. Hirata I, Austin LL, Blackwell WH et al Immunoelectron microscopic localization of HLA-DR antigen in control small intestine and colon and in inflammatory bowel disease. Dig Dis Sci 1986; 31: 1317-30. Mayrhofer G, Spargo LD. Distribution of class II major histocompatibility antigens in enterocytes of the rat jejunum and their association with organelles of the endocytic pathway. Immunology 1990; 70: 11-19. Gottlieb TA, Ivanov IE, Adesnik M, Sabatini DD. Actin microfilaments play a critical role in endocytosis at the apical but not the basolateral surface of polarized epithelial cells. J Cell Biol 1993; 120: 695-710. Jackman MR, Shurety W, Ellis JA, Luzio JP. Inhibition of apical but not basolateral endocytosis of ricin and folate in Caco-2 cells by cytochalasin D. J Cell Sci 1994; 107: 254756. Brandeis JM, Sayegh, Gallon L, Blumberg RS, Carpenter CB. Rat intestinal epithelial cells present major histocompatibihty complex allopeptides to primed T cells. Gastroenterology 1994; 107: 1537-42. Telega GW, Baumgart DC, Carding SR. Uptake and presentation of antigen to T cells by primary colonic epithelial cells in normal and diseased states. Gastroenterology 20; 119:1548-59. Madara J L, Stafford J. Interferon-gamma directly affects barrier function of cultured intestinal epithelial monolayers. JClinInvestl989;83:724-7. Cong J, Brandwein SL, McCabe RP et al CD4^ T cells reactive to enteric bacterial antigens in sponataneous colitis C3H/HeJBir mice: increased T helper cell type 1 response and ability to transfer disease. J Exp Med 1998; 187: 85564. 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. Mayer L, Shlien R. Evidence for function of la molecules on gut epitheHal cells in man. J. Exp. Med. 1987; 166: 1471-83. Hershberg R, Blumberg RS. What's so (co)stimulating about the intestinal epithelium? Gastroenterology. 1999; 117:726-36. Bleicher PA, Balk SP, Hagen SJ et al Expression of murine CDl on gastrointestinal epithelium. Science 1990; 250: 67982. Hershberg R, Eghtesady P, Sydora B et al Expression of the thymus leukemia antigen in mouse intestinal epithelium. Proc Natl Acad Sci USA 1990; 87: 9727-31. Blumberg RS. Current concepts in mucosal immunity. II. One size fits all: nonclassical MHC molecules fulfill multiple roles in epithelial cell function. Am J Physiol 1998; 274: G227-31. Balk SP, Burke S, Polischuk JE et al Beta 2-microglobulinindependent MHC class lb molecule expressed by human intestinal epithelium. Science 1994; 265: 259-62. Groh V, Bahram S, Bauer S et al Cell stress-regulated human major histocompatibility complex class I gene expressed in gastrointestinal epithelium. Proc Natl Acad Sci USA 1996; 93: 12445-50.

143 116. Brand VM, Allan DS, McMichael AJ. Functions of nonclassical MHC and non-MHC-encoded class I molecules. Curr Opin Immunol 1999; 11: 100-8. 117. Parkkila S, Waheed A, Britton RS et al Immunohistochemistry of HLA-H, the protein defective in patients with hereditary hemochromatosis, reveals unique pattern of expression in gastrointestinal tract. Proc Natl Acad Sci USA 1997; 94: 2534-9. 118. Beckman E, Porcelli S, Morita C, Behar S, Furlong S, Brenner M. Recognition of a lipid antigen by CDl-restricted aP"" T cells. Nature 1994; 372: 691-4. 119. Sieling PA, Chatterjee D, Porcelli SA et al CDl-restricted T cell recognition of microbial ligoplycan antigens. Science 1995; 269:227-30. 120. Blumberg R, Terhorst C, Bleicher P et al Expression of a nonpolymorphic MHC class I-like molecule, CD Id by human intestinal epithehal cells. J Immunol 1991; 147: 251824. 121. Canchis P, Bhan A, Landau S, Yang L, Balk S, Blumberg R. Tissue distribution of the non-polymorphic major histocompatibility complex class I-like molecule, CD Id. Immunology 1993; 80: 561-5. 122. Exley M, Garcia J, Wilson SB et al Developmental and activation regulated expression of CD Id in human lymphoid and myeloid lineages. Immunology 20; 1: 37-47. 123. Joyce S, Woods AS, Yewdell JW et al Natural ligand of mouse C D l d l : cellular glycosylphosphatidylinositol. Science 1998; 279: 1541-4. 124. Kawano T, Cui J, Koezuka Y et al CD Id-restricted and TCR-mediated activation of Val4 NKT cells by glycosyleeramides. Science 1997; 278: 1626-9. 125. Castano A, Tangri S, Miller S et al Peptide binding and presentation by mouse C D l . Science 1995; 269: 223-6. 126. Zeng A-H, Castano AR, Segelke BW, Stura EA, Peterson PA, Wilson lA. Crystal structure of mouse C D l : an MHClike fold with a large hydrophobic binding groove. Science 1997; 277: 339-45. 127. Bendelac A, Lantz O, Quimby M, Yewdell J, Bennink J, Brutkiewicz R. CD Id recognition by mouse NKl"^ T lymphocytes. Science. 1995; 268: 863-5. 128. Bendelec A, Rivera MN, Park S-H, Roark JH. Mouse C D l specific N K l T cells: development, specificity, and function. Annu Rev Immunol 1997; 15: 535-62. 129. Balk S, Bleicher P, Terhorst C. Isolation and expression of cDNA encoding the murine homologues of CD 1. J Immunol. 1991;146:768-74. 130. Calabi R, Bradbury A. A review. The CDl system. Tissue Antigens. 1991; 37: 1-9. 131. MacDonald HR. NKl.l"" T cell receptor-a/p^ cells: new clues to their origin, specificity, and function. J Exp Med 1995; 182: ; 633-8. 132. Chen H, Paul WE. Cultured NK1.1^CD4^ T cells produce large amounts of IL-4 and IFN-y upon activation by antiCD3 or C D l . J Immunol 1997; 159: 2240-9. 133. Lantz O, Bendelac A. An invariant T cell receptor a chain is used by a unique subset of MHC class I-specific CD4"^ and CD4"CD8~ T cells in mice and humans. J Exp Med 1994; 180:1097-106. 134. Toyabe S, Seki S, Li3.iT etal Requirement of IL-4 and liver N K l ^ T cells for concanavalin A-induced hepatic injury in mice. J Immunol 1997; 159: 1537-42. 135. Matsuda JL, Naidenko OY, Gapin L et al Tracking the response of natural killer T cells to a glycolipid antigen using CDld tetramers. J Exp Med 2000; 192: 741-54. 136. Yoshimoto T, Bendelac A, Watson C, Hu-Li J, Paul WE. Role of N K l . 1"^ T cells in a T H 2 response and in immunoglobulin E production. Science 1995; 270: 1845-7.

144 137. Smiley ST, Kaplan MH, Grusby MJ. Immunoglobulin E production in the absence of interleukin-4-secreting CD Iddependent cells. Science 1997; 275: 977-9. 138. Mendiratta SK, Martin WD, Hong S, Boesteanu A, Joyce S, Van Kaer L. C D l d l mutant mice are deficient in natural T cells that promptly produce IL-4. Immunity 1997; 6: 46977. 139. Chen YH, Chiu MM, Mandel M, Wang CR. Impaired NKT cells development and early IL-4 production in C D l deficient mice. Immunity 1997; 459-67. 140. Kawamura T, Takeda K, Mendiratta SK et al. Cutting edge: critical role of N K l ^ T cells in IL-12-induced immune responses m vivo. J Immunol 1998; 160: 16. 141. Seki S, Hashimoto W, Ogasawara K et al. Antimetastatic effect of NKl T cells on experimental haematogenous tumour metastases in the liver and lungs of mice. Immunology 1997; 92: 561. 142. Exley M, Garcia J, Balk SP, Porcelli S. Requirements for CD Id recognition by human invariant Va24"^ CD4 CD8 T cells. J Exp Med 1997; 186: 1. 143. Prussin C, Foster B. TCR Va24 and Vpi 1 coexpression defines a human NKl T cell analog containing a unique ThO subpopulation. J Immunol 1997; 159: 5862 70. 144. Porcelli S, Gerdcs D, Fcrtig A, Balk SP Human T cells expressing an invariant Va24-JaQ TCRoc are CD4 and heterogeneous with respect to TCRp expression. Human Immunol 1996; 48: 63. 145. Exley M, Porcelli S, Furman M, Garcia J, Balk S. CD161 (NKR-PIA) costimulation of CDld-dcpcndcnt activation of human T cells expressing invariant Voc24JaO T cell receptor a chains. J Exp Med 1998; 188: 867. 146. Brossay L, Chioda M, Burdin N ef al. CDld-mediatcd recognition of an a-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution. J Exp Med 1998; 188: 1521 8. 147. Nieda M, Nicol A, Koczuka Y c( al. Activation o^ human VaNKT cells by cx-glycosylceramide in a CD Id-restricted and Va24 TCR-mediated manner. Human Immunol 1999; 60: 10 19. 148. Spada FM, Koczuka Y, Porcelli SA. CD Id-restricted recognition of synthetic glycolipid antigens by human natural killer T cells. J Exp Med 1998; 188: 1529 34. 149. Wilson SB, Kent SC, Patton KT et al. Extreme Thl bias of invariant Va2JaQ T cells in type 1 diabetes. Nature 1998; 391: 177. 150. Mieza MA, Itoh T, Cui JQ et al. Selective reduction of Val4"^ NK T cells associated with disease development in aujtoimmune-prone mice. J Immunol 1996; 156: 4035. 151. SumidaT, Sakamoto A, Murata H et al. Selective reduction of T cells bearing invariant Va24JaQ antigen receptor in patients with systemic sclerosis. J Exp Med 1995: 182: 1163. 152. Cui J, Shin T, Kawano T et al. Requirement for Val4 NKT cells in IL-12-mediated rejection of tumors. Science 1997; 278:1623 9. 153. Somnay-Wadgaonkar K, Nusrat A, Kim HS et al. Immunolocalization of CD Id in human intestinal epithelial cells and identification of a (32-microglobulin associated form. Int Immunol 1999; 383-92. 154. Blumberg RS, Terhorst C, Bleicher P et al. Expression of human CD Id on gastrointestinal epithelium. J Immunol 1991; 147:2518-24. 155. van de Wal Y, Pitman RS, Hershberg RM et al. Human and mouse intestinal epithelial cells (lEC) present glycolipid antigens to natural killer (NK)-T cells in a CD Id-restricted manner. Gastroenterology 2001 (in press). 156. Saubermann LJ, Beck P, de Jong YP et al. Natural killer-T cells activated by a-galactosylceramide in the presence of CD Id provide protection against colitis in mice. Gastroenterology 20; 119: 119-28.

The lymphocyte-epithelial-bacterial

interface

157. Bleday R, Braidt J, Ruoff K, Shellito PC, Ackroyd FW. Quantitative cuhures of the mucosal-associated bacteria in the mechanically prepared colon and rectum. Dis Colon Rectum 1993; 36: 844-9. 158. Roediger WE. Anaerobic bacteria, the colon and colitis. AustNZJSurgl980;50:73-5. 159. Madara JL. Epithelia: biologic principles of organization. In: Yamada T, ed. Textbook of Gastroenterology. Philadelphia: Lippincott, 1995: 141. 160. Perez-Vilar J, Hill RL. The structure and assembly of secreted mucins. J Biol Chem 1999; 274: 31751-4. 161. Sanderson IR, Walker A. Mucosal barrier: an overview. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 5. 162. Jabbal I, Kells DI, ForstnerG, Forstner J. Human intestinal goblet cell mucin. Can J Biochem 1976; 54: 707-16. 163. Gum JR Jr. Mucin genes and the proteins they encode: structure, diversity, and regulation. Am J Respir Cell Mol Biol 1992; 7: 557 64. 164. Loomes KM, Senior HE, West PM, Roberton AM. Functional protective role for mucin glycosylated repetitive domains. Eur J Biochem 1999; 266: 105 11. 165. Kindon H, Pothoulakis C, Thim L, Lynch-Devaney K, Podolsky DK. Trefoil peptide protection of intestinal epithelial barrier function: cooperative interaction with mucin glycoprotein. Gastroenterology 1995; 109: 516 23. 166. Lindahl M, Carlstedt I. Binding of pig small intestinal mucin glycopeplidcs to fimbriated enterotoxigenic Escherichia coli. Symp Soc Exp Biol 1989; 43: 423 8. 167. Piotrowski J, Slomiany A, Murty VL, Fekete Z, Slomiany BL. Inhibition of Helicobacter pylori colonization by sulfated gastric mucin. Biochem Int 1991; 24: 749 56. 168. Smith C J, Kapcr JB, Mack DR. Intestinal mucin inhibits adhesion of human enteropathogenic Escherichia coli to HEp-2 cells. J Pediatr Gastroenterol Nutr 1995; 21: 269 76. 169. Epple HJ, Krcuscl KM, Hanski C, Schulzke JD, Riecken EO, Fromm M. Ditferential stimulation of intestinal mucin secretion by cholera toxin and carbachol. Pflugers Arch 1997:433:638 47. 170. Choi J, Klinkspoor JH, YoshidaT, Lee SP. Lipopolysaccharide from Escherichia coli stimulates mucin secretion by cultured dog gallbladder epithelial cells. Hepatology 1999; 29: 1352 7. 171. McCool DJ, Marcon MA, Forstner JF, Forstner GG. The T84 human colonic adenocarcinoma cell line produces mucin in culture and releases it in response to various secretagogues. Biochem J 1990; 267: 491-500. 172. Ouellette, AJ, Greco RM, James M, Frederick D, Naftilan J, Fallon JT. Developmental regulation of cryptdin, a corticostatin/defensin precursor mRNA in mouse small intestinal crypt epithelium. J Cell Biol 1989; 108: 1687-95. 173. Eisenhauer PB, Harwig SS, Lehrer RI. Cryptdins: antimicrobial defensins of the murine small intestine. Infect Immun 1992; 60: 3556-65. 174. O'Neil DA, Porter EM, Elewaut D et al. Expression and regulation of the human beta-defensins hBD-1 and hBD-2 in intestinal epithelium. J Immunol 1999; 163: 6718-24. 175. Pruitt KM, Rahemtulla B, Rahemtula F, Russel MW. Innate humoral factors. In: Ogra RL, Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 65. 176. Bellamy W, Takase M, Wakabayashi H, Kawase K, Tomita M. Antibacterial spectrum of lactoferricin B, a potent bactericidal peptide derived from the N-terminal region of bovine lactoferrin. J Appl Bacteriol 1992; 73: 472-9. 177. Harmsen MC, Swart PJ, de Bethune MP et al. Antiviral effects of plasma and milk proteins: lactoferrin shows potent

Robert Hershberg and Richard S. Blumberg

178. 179.

180. 181.

182.

183.

184.

185.

186. 187.

188. 189.

190.

191.

192.

193.

194.

activity against both human immunodeficiency virus and human cytomegalovirus repHcation in vitro. J Infect Dis 1995;172:380-8. Fujihara T, Hayashi K. Lactoferrin inhibits herpes simplex virus type-1 (HSV-1) infection to mouse cornea. Arch Virol 1995; 140: 1469-72. Kawasaki Y, Tazume S, Shimizu K et al. Inhibitory effects of bovine lactoferrin on the adherence of enterotoxigenic Escherichia coli to host cells (In process citation). Biosci Biotechnol Biochem 2000; 64: 348-54. Turchany JM, Aley SB, Gillin FD. Giardicidal activity of lactoferrin and N-terminal peptides. Infect Immun 1995; 63: 4550-2. Tanaka T, Omata Y, Saito A, Shimazaki K, Igarashi I, Suzuki N. Growth inhibitory effects of bovine lactoferrin to Toxoplasma gondii parasites in murine somatic cells. J Vet Med Sci 1996; 58: 61-5. Wang YB, Germaine GR. Effect of lysozyme on glucose fermentation, cytoplasmic pH, and intracellular potassium concentrations in Streptococcus mutans 10449. Infect Immun 1991; 59: 6 3 8 ^ 4 . Soukka T, Lumikari M, Tenovuo J. Combined inhibitory effect of lactoferrin and lactoperoxidase system on the viability of Streptococcus mutans, serotype c. Scand J Dent Res 1991; 99: 390-6. Andoh A, Fujiyama Y, Bamba T, Hosoda S. Differential cytokine regulation of complement C3, C4, and factor B synthesis in human intestinal epithelial cell line, Caco-2. J Immunol 1993; 151: 4239-47. MacDermott RP, Sanderson IR, Reinecker HC. The central role of chemokines (chemotactic cytokines) in the immunopathogenesis of ulcerative colitis and Crohn's disease. Infiam Bowel Dis 1998; 4: 54-67. MacDermott RR Chemokines in the inflammatory bowel diseases. J CUn Immunol 1999; 19: 266-72. Eckmann L, Jung HC, Schurer-Maly C, Panja A, Morzycka-Wroblewska E, Kagnoff MF. Differential cytokine expression by human intestinal epithelial cell lines: regulated expression of interleukin 8. Gastroenterology 1993; 105:1689-97. Eckmann L, Kagnoff MF, Fierer J. Epithelial cells secrete the chemokine interleukin-8 in response to bacterial entry. Infect Immun 1993; 61: 4569-74. McCormick BA, Hofman PM, Kim J, Carnes DK, Miller SI, Madara JL. Surface attachment of Salmonella typhimurium to intestinal epithelia imprints the subepithelial matrix with gradients chemotactic for neutrophils. J Cell Biol 1995; 131: 1599-608. Schulte R, Autenrieth IB. Yersinia enterocolitica-inducQd interleukin-8 secretion by human intestinal epithelial cells depends on cell differentiation. Infect Immun 1998; 66: 1216-24. Jung HC, Eckmann L, Yang SK et al. A distinct array of proinflammatory cytokines is expressed in human colon epithelial cells in response to bacterial invasion. J Clin Invest 1995; 95: 55-65. Crowe SE, Alvarez L, Dytoc M et al. Expression of interleukin 8 and CD 54 by human gastric epithelium after Helicobacter pylori infection in vitro. Gastroenterology 1995;108: 65-74. Jung HC, Kim JM, Song IS, Kim CY. Helicobacter pylori induces an array of pro-inflammatory cytokines in human gastric epithelial cells: quantification of mRNA for interleukin-8, -1 alpha/beta, granulocyte-macrophage colonystimulating factor, monocyte chemoattractant protein-1 and tumour necrosis factor-alpha. J Gastroenterol Hepatol 1997; 12: 473-80. Husband AJ, Beagley KW, McGhee JR. Mucosal Cytokines. In: Ogra RL, Mestecky J, Lamm ME, Strober W,

145

195.

196.

197. 198. 199. 200. 201.

202.

203. 204.

205.

206. 207. 208.

209. 210.

211. 212.

Bienenstock J, McGhee JR, eds. Mucosal Immunology, 2nd edn. San Diego: Academic Press, 1999: 541. Schuerer-Maly CC, Eckmann L, Kagnoff MF, Falco MT, Maly FE. Colonic epithelial cell lines as a source of interleukin-8: stimulation by inflammatory cytokines and bacterial lipopolysaccharide. Immunology 1994; 81: 85-91. Reinecker HC, MacDermott RP, Mirau S, Dignass A, Podolsky DK. Intestinal epithelial cells both express and respond to interleukin 15. Gastroenterology 1996; 111: 1706-13. Ciacci C, Mahida YR, Dignass A, Koizumi M, Podolsky DK. Functional interleukin-2 receptors on intestinal epithelial cells. J Clin Invest 1993; 92: 527-32. Ciacci C, Lind SE, Podolsky DK. Transforming growth factor beta regulation of migration in wounded rat intestinal epithehal monolayers. Gastroenterology 1993; 105: 93-101. Kim PH, Kagnoff" MF. Transforming growth factor beta 1 increases IgA isotype switching at the clonal level. J Immunol 1990; 145: 3773-8. Bromander AK, Kjerrulf M, Holmgren J, Lycke N. Cholera toxin enhances alloantigen presentation by cultured intestinal epithelial cells. Scand J Immunol 1993; 37: 452-8. Goodrich ME, McGee DW. Effect of intestinal epithelial cell cytokines on mucosal B-cell IgA secretion: enhancing effect of epithelial-derived IL-6 but not TGF-p pm IgA"^ B cells. Immunol Lett 1999; 67: 11-14. Defrance T, Vanbervliet B, Briere F, Durand I, Rous set F, Banchereau J. Interleukin 10 and transforming growth factor beta cooperate to induce anti-CD40-activated naive human B cells to secrete immunoglobulin A. J Exp Med 1992; 175:671-82. Moore KW, O'Garra A, de Waal Malefyt R, Vieira P, Mosmann TR. Interleukin-10. Annu Rev Immunol 1993; 11: 165-90. Madsen KL, Lewis SA, Tavernini MM, Hibbard J, Fedorak RN. Interleukin 10 prevents cytokine-induced disruption of T84 monolayer barrier integrity and limits chloride secretion. Gastroenterology 1997; 113: 151-9. Fujihashi K, McGhee JR, Yamamoto M, Peschon JJ, Kiyono H. An interleukin-7 internet for intestinal intraepithelial T cell development: knockout of ligand or receptor reveal differences in the immunodeficient state. Eur J Immunol 1997;27:2133-8. He YW, Malek TR. Interleukin-7 receptor alpha is essential for the development of gamma delta + T cells, but not natural killer cefls. J Exp Med 1996; 184: 289-93. Maki K, Sunaga S, Komagata Y et al. Interleukin 7 receptor-deficient mice lack gammadelta T cells. Proc Natl Acad Sci USA 1996; 93: 7172-7. Carini C, Essex M. Interleukin 2-independent interleukin 7 activity enhances cytotoxic immune response of HIV-1infected individuals. AIDS Res Hum Retroviruses 1994; 10: 121-30. Kasper LH, Matsuura T, Khan I A. IL-7 stimulates protective immunity in mice against the intracellular pathogen. Toxoplasma gondii. J Immunol 1995; 155: 4798-804. Watanabe M, Ueno Y, Yajima T et al. Interleukin 7 is produced by human intestinal epithelial cells and regulates the proliferation of intestinal mucosal lymphocytes. J Clin Invest 1995; 95: 2945-53. Puddington L, Olson S, Lefrancois L. Interactions between stem cell factor and c-Kit are required for intestinal immune system homeostasis. Immunity 1994; 1: 733-9. Klimpe, GR, Langley KE, Wypych J, Abrams JS, Chopra AK, Niesel DW. A role for stem cell factor (SCF): c-kit interaction(s) in the intestinal tract response to Salmonella typhimurium infection. J Exp Med 1996; 184: 271-6.

146 213. Wang J, Whetsell M, Klein JR. Local hormone networks and intestinal T cell homeostasis. Science 1997; 275: 19379. 214. Hata Y, Ota S, Nagata T, Uehara Y, Terano A, Sugimoto T. Primary colonic epithelial cell culture of the rabbit producing prostaglandins. Prostaglandins 1993; 45: 1 2 9 ^ 1 . 215. Barrera S, Lai J, Fiocchi C, Roche JK. Regulation by prostaglandin E2 of interleukin release by T lymphocytes in mucosa. J Cell Physiol 1996; 166: 130-7. 216. Reinecker HC, Podolsky DK. Human intestinal epithelial cells express functional cytokine receptors sharing the common gamma c chain of the interleukin 2 receptor. Proc Natl Acad Sci USA 1995; 92: 8353-7. 217. Sutherland DB, Varilek GW, Neil GA. Identification and characterization of the rat intestinal epithelial cell (IEC-18) intcrleukin-1 receptor. Am J Physiol 1994; 266: CI 198-203. 218. McGee DW, Vitkus SJ, Lee P The effect of cytokine stimulation on IL-1 receptor mRNA expression by intestinal epithelial cells. Cell Immunol 1996; 168: 276 80. 219. Sollid LM, Kvale D, Brandtzaeg P, Markussen G, Thorsby E. Interferon-gamma enhances expression of secretory component, the epithelial receptor for polymeric immunoglobulins. J Immunol 1987; 138: 4303-6. 220. Kvale D, Lovhaug D, Sollid LM, Brandtzaeg P Tumor necrosis factor-alpha up-regulates expression of secretory component, the epithelial receptor for polymeric Ig. J Immunol 1998; 140: 3086 9. 221. Phillips JO, Evcrson MP, Moldoveanu Z, Lue C, Mestecky J. Synergistic effect of IL-4 and IFN-gamma on the expression of polymeric Ig receptor (secretory component) and IgA binding by human epithelial cells. J Immunol 1990; 145: 1740 4. 222. Ccrf-Bensussan N, Quaroni A, Kurnick JT, Bhan AK. Intraepithelial lymphocytes modulate la expression by intestinal epithelial cells. J Immunol 1984; 132: 2244 52. 223. Hoang P, Crotty B, Dalton HR, Jewell DP Epithelial cells bearing class II molecules stimulate allogeneic human

The lymphocyte-epithelial-bacterial

interface

colonic intraepithelial lymphocytes. Gut 1992; 33: 1089-93. 224. Lowes JR, Radwan P, Priddle JD, Jewell DP. Characterisation and quantification of mucosal cytokine that induces epithelial histocompatibility locus antigen-DR expression in inflammatory bowel disease. Gut 1993; 33: 315-9. 225. Kvale D, Krajci P, Brandtzaeg P. Expression and regulation of adhesion molecules ICAM-1 (CD54) and LFA-3 (CD58) in human intestinal epithelial cell lines. Scand J Immunol 1992;35:669-76. 226. Ye G, Barrera C, Fan X, Gourley WK, Crowe SE, Ernst PB, Reyes VE. Expression of B7-1 and B7-2 costimulatory molecules by human gastric epithelial cells: potential role in CD4^ T cell activation during Helicobacter pylori infection. J CHn Invest 1997; 99: 1628-36. 227. Medzhitov R, Janeway C Jr. The toll receptor family and microbial recognition. Trends Microbiol 20; 8: 452-6. 228. Medzhitov R, Janeway C Jr. Innate immunity. N Engl J Med 20; 343: 338-44. 229. Aderem A, Ulevitch RJ. Toll-like receptors in the induction of the innate immune response. Nature 20; 406: 782-7. 230. Anderson KV. Toll signaling pathways in the innate immune response. Curr Opin Immunol 20; 12: 13 19. 231. Poltorak A, He X, Smirnova I ct ai Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice; mutations in Tlr4 gene. Science 1998; 282: 2085 8. 232. Carlo E, Rosenberg IM, Brandwein SL, Beck PL, Reinecker HC, Podolsky DK. Lipopolysaccharide activates distinct signaling pathways in intestinal epithelial cell lines expressing Toll-like receptors. J Immunol 20; 164: 966 72. 233. Beutler B. Tlr4: central component of the sole mammalian LPS sensor. Curr Opin Immunol 20; 12: 20 6. 234. Cario E, Podolsky DK. Difierential alteration in intestinal epithelial cell expression of toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel disease. Infect Immun 20; 68: 7010 17. 235. Hemmi H, Takeuchi O, Kawai J ct al. A Toll-like receptor recognizes bacterial DNA. Nature 20; 408: 740 5.

The mucosal inflammatory response. Cytokines and chemokines FABIO COMINELLI, KRISTEN O. ARSENEAU AND THERESA T. PIZARRO

Introduction Cytokines and chemokines play an important role in inflammatory bowel disease (IBD) pathogenesis, as demonstrated by the efficacy of cytokine-targeted immunomodulatory therapy in treating IBD. The precise etiology of IBD is unknown, but disease pathogenesis clearly involves genetic, immunologic, and environmental factors. The current hypothesis for IBD etiology suggests that the disease is initiated and perpetuated by a dysregulated immune response to an unknown environmental antigen in a genetically susceptible host. A working hypothesis for the pathogenesis of Crohn's disease, the most debilitating form of IBD, is depicted in Fig. 1. In order to maintain gut homeostasis, the normal mucosal immune system must maintain a delicate balance between a network of proinflammatory and/ or inductive cytokines and anti-inflammatory and/or regulatory cytokines. During bacterial invasion the proinflammatory and inductive type I T-helper cell (Thl)-polarizing cytokines exert pleiotropic effects upon release from antigen-presenting cells (macrophages, dendritic ceUs, etc.) in the intestinal mucosa. Proinflammatory cytokines (i.e. interleukin [IL]-1, IL-6, and tumor necrosis factor [TNF]) activate immune cells in the lamina propria and stimulate intestinal epithelial cells (lEC) and macrophages to produce chemokines which, when secreted, estabUsh a chemotactic gradient across the intestinal mucosa, promoting leukocyte recruitment and extravasation across the intestinal endothelium. Meanwhile, Thlpolarizing cytokines (i.e. IL-12, IL-18, T N F ) released from antigen-presenting cells (primarily macrophages) induce activation of naive intraepithelial and lamina propria T lymphocytes and Thl differentiation. These activated Thl cells release Thl cytokines (i.e. IL-2, interferon gamma [IFN-y], and lymphotoxin [LT]), which further perpetuate

inflammatory and delayed hypersensitivity immune responses in the gut. In the normal gut mucosa this proinflammatory/ inductive immune response is kept in balance by a network of anti-inflammatory and regulatory cytokines. Regulatory cytokines (i.e. IL-10, IL-4, IL-5, IL-13, and transforming growth factor beta [TGFp]) are derived from many cellular sources within the intestine, including macrophages as well as type IIThelper (Th2) cells and type I T regulatory (Trl) cells. Along with anti-inflammatory cytokines (i.e. IL-1 receptor antagonist [IL-lra] and IL-11) derived primarily from lEC and macrophages, these Th2 cytokines serve to counterbalance the proinflammatory effects of Thl-mediated immune responses. For unknown reasons, patients with IBD cannot maintain normal gut homeostasis. An understanding of the cytokine network and its role in promoting IBD pathogenesis is a crucial step toward finding a cure for this devastating disease. This chapter provides an overview of the cytokines and chemokines that have been implicated in IBD, with specific focus on what is known about each cytokine in relation to IBD pathogenesis and treatment.

Proinflammatory/inductive cytokines The first broad class of cytokines includes proinflammatory/inductive cytokines. For organizational purposes we have divided this class into Thl-polarizing cytokines, Thl-derived cytokines, and other proinflammatory/inductive cytokines. These cytokines are innately involved in initiating and promoting inflammatory immune responses within the intestinal mucosa after antigen presentation. Over-production of these cytokines leads to a Thl-type inflammatory response characteristic of human Crohn's disease (CD).

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 147-176. © 2003 Kluwer Academic Publishers. Printed in Great Britain

The mucosal inflammatory response. Cytokines and chemokines

148 P-SELEC

Genetically-Susceptible Host Regeneration Cell Death -• Apoptosis Necrosis Differentiation INFLAMMATION]

/ I V

iL-18 ulcerative colitis

Serum Fe,TIBC ratio >15% (serum erythropoietin levels are not useful)

Supplement with recombinant erythropoietin (150 U/kgt.i.w. subqfor up to 12 weeks or 10 000 U subq q week)

Vitamin B12 deficiency

Megaloblastic

Crohn's disease (ileal disease or ileal resection)

Vitamin B12 levels

Monthly IM B12 injections 1000 jig dose

Anti-RBC antibodies

Hemolytic anemia

Ulcerative colitis > Crohn's disease

Coombs test positive

Corticosteroids, usually self-limited

*Parenteral iron dextran replacement is based on the following formula: 0.0476 x (normal hemoglobin - observed hemoglobin in g/dl) ( ± ) 1 ml/5 kg body weight (up to maximum of 14 ml) which accounts for storage iron = total dosage of iron-dextran in ml.

Konstantinos A. Papadakis and Maria T. Abreu Active IBD is associated with changes in hemostatic factors which may lead to hypercoagulability. Increases in coagulation factors are known risk factors for the development of thromboembolic disease [106]. The most common hemostatic abnormality identified in patients with IBD is thrombocytosis, which correlates with underlying bowel disease activity [107] (Table 3). In a study of 92 thromboembolic events occurring in 7199 (1.3%) patients with IBD, 60% of patients had thrombocytosis and 73%) had elevated ESR values, suggesting active IBD [89]. Patients with IBD have increased levels of thrombin generation (prothrombin fragment 1 + 2 and thrombin-antithrombin III complex) compared with control populations, and these levels correlate with disease activity [107-110]. Fibrinogen, factor V and factor VIII are also commonly elevated in patients with IBD whereas the anticoagulant factors antithrombin III, and proteins S and C, are decreased, which may contribute to the propensity for thrombosis [89, 107, 110-119]. Other plateletrelated abnormalities identified in patients with IBD are increased spontaneous and induced platelet aggregation which correlated with a history of thromboembolism in seven of eight patients [107, 120]. Surgical resections in patients with CD result in significant decreases in platelet counts, fibrinogen levels and spontaneous platelet aggregation supporting the concept that the underlying bowel disease contributes to the hypercoagulable state [121]. Increased platelet mass in IBD is due to systemic increases in thrombopoietin and IL-6 [122]. Specifically, patients with active IBD have significantly increased levels of thrombopoietin compared with patients in remission and this increase in thrombopoietin is associated with increased platelet counts. Increased megakaryocyte maturation occurs in response to endogenous production of IL-11 and increases with IL-11 therapy for IBD [123]. Acute inflammation can also result in increased hepatic production of fibrinogen and increased platelet aggregation, which is another risk factor in thrombosis. A study of thrombotic risk factors in IBD found that patients with IBD had increased plasma factor VII coagulant activity (a marker of thrombin generation), lipoprotein (a) and fibrinogen concentrations compared with a normal population [124]. With respect to acquired disorders of coagulation, anticardiolipin antibodies or the lupus anticoagulant are associated with an increased risk of thromboembolic events, and have been found with increased frequency in patients with IBD [107, 118, 125].

241 Studies examining the prevalence of anticardiolipin antibodies in patients with IBD found that both CD and UC patients had increased antibody titers compared to a control group, but the presence of anticardiolipin antibodies was not associated with a higher risk of thromboembolic events in these patients [87, 126]. Others, however, have described the presence of a lupus anticoagulant in the setting of severe thrombotic events such as dural sinus thrombosis or hepatic vein thrombosis in patients with IBD [94, 104, 127]. Based on these studies, a screen for a lupus anticoagulant should be part of the evaluation of a patient with IBD and a recognized thrombotic event, but has little predictive value in a patient without demonstrated hypercoagulability. Patients with hypercoagulability and IBD may have inherited disorders of coagulation. In combination with inflammation-associated changes in hemostasis these inherited disorders may become manifest [128]. Activated protein C resistance from a mutation in the prothrombin gene (factor V Leiden) is the most common inherited disorder leading to thrombosis, and accounts for 30-40% of episodes of idiopathic venous thrombosis [128-130]. Other inherited causes of hypercoagulability include hyperhomocysteinemia and mutations in the prothrombin gene [131]. Since the description of the relatively common factor V Leiden mutation as an inherited cause of thrombophilia, several groups have investigated the frequency of this mutation in IBD patients with and without a history of thromboembolism. A study evaluating the prevalence of factor V Leiden, methylene tetrahydrofolate reductase (resulting in hyperhomocysteinemia) and prothrombin gene mutations in IBD patients without thrombosis compared with an age-matched control group found no increase in these inherited disorders [132]. Several studies examining the prevalence of activated protein C resistance or factor V Leiden mutations in adults or children with IBD and without a history of thrombosis have not found an increased prevalence of these mutations compared with healthy controls [119, 133, 134]. A large Greek series identified factor V Leiden mutations in 8.3%) of IBD patients, which was not significantly different when compared with a healthy control group with a 4.9% mutation rate [111]. Two studies found a slightly increased allelic frequency of factor V Leiden mutations in UC [135] or CD patients [136], but the sample sizes were limited. Although the prevalence of factor V Leiden mutations or resistance to activated protein C does not appear to be increased in an unselected population of IBD patients, the frequency of this

242 mutation is increased in those IBD patients with a history of thromboemboHc events. An Austrian study evaluated the presence of resistance to activated protein C in patients with IBD compared with healthy controls. In IBD patients without a history of thromboembolism, the frequency of activated protein C resistance was similar to that of healthy controls (7% versus 5.9%) [137]. By contrast, 31.3% of IBD patients with a history of thromboembolism had activated protein C resistance, suggesting that patients with thromboembolism and IBD are just as likely to have activated protein C resistance as patients with thromboembolism in the general population [129]. Similar results were observed in an American study that found four of 11 IBD patients (36%) with thrombosis and two of 51 IBD controls (4%) were heterozygotes for the factor V Leiden mutation [138]. In a group of 20 patients with IBD complicated by thrombosis, a screen for the most common inherited and acquired disorders of coagulation including protein S, protein C levels, and antithrombin III levels, antiphospholipid antibodies, and activated protein C resistance were negative, with only one patient found to be heterozygous for factor V Leiden mutation [90]. In patients with IBD there is an increased prevalence of hyperhomocysteinemia, a risk factor for thrombophilia, which can be corrected by the administration of folate, cobalamin and pyridoxine [139]. Recent data also suggest that patients with IBD have increased prevalence of the C677T variant of the methylenetetrahydrofolate reductase gene (17% versus 7.3% in healthy controls) which is associated with thromboembolic disease [140]. In a retrospective study of 231 patients with IBD, hyperhomocysteinemia was more prevalent in patients than in healthy controls, but was not higher in IBD patients with a history of thromboembolic disease [141]. These studies suggest that the majority of patients with thromboembolic events and IBD have the same risk as the rest of the population for inherited disorders of coagulation, but the majority will have no identifiable risk factor except active inflammation. Although the risk of inherited disorders predisposing to thrombophilia is not increased in IBD patients, there is an epidemiologic study suggesting that hemophilia and VonWillebrand's disease are under-represented in the IBD population [142]. As the molecular mechanisms regulating disorders of coagulation are elucidated, additional patients with IBD and hypercoagulability will be recognized to have an inherited disorder of coagulation.

Systemic consequences of intestinal inflammation Management of thromboses in patients with IBD requires a multi-faceted approach (see Fig. 1). Even in patients with active IBD, a search for an underlying disorder of coagulation is required. If active IBD is the only identifiable risk factor for the thrombosis, and it is a single thrombotic event, short-term anticoagulation combined with treatment of the underlying IBD is appropriate [143]. Heparin therapy can generally be given safely, and has a small therapeutic benefit in patients with UC [144]. Thrombolytic therapy may be given cautiously in IBD patients with extensive thromboses or if the condition is life-threatening [145, 146]. In patients with recurrent thrombotic episodes or life-threatening thromboses, long-term anticoagulation is required. This may be difficult in the setting of active IBD because of associated intestinal bleeding. In addition to hypercoagulability due to inherited genetic mutations in coagulation factors or inflammation, there are multiple case reports of disseminated intravascular coagulation in patients with UC generally associated with a flare of the UC [148-150]. Thrombocytopenia may also complicate IBD, and may be immune-mediated [151, 152] or associated with drug-induced bone marrow suppression as described above. Cases of sulfasalazine-induced and mesalamine-induced immune-mediated thrombocytopenia have been reported [153].

Leukocytosis As with inflammation of any type, leukocytosis is often present in patients with IBD [122]. Elevated leukocyte counts prior to surgical resection for CD are associated with an increased risk of recurrence [154]. White blood cell counts above 18 000 or the presence of an elevated band count, should prompt an investigation for a septic process. Neutrophil and monocyte maturation and release from the bone marrow is regulated by granulocyte colony-stimulating factor (GCSF) and monocyte colony-stimulating factor (MCSF), respectively. These factors are derived from bone marrow stromal cells and monocytes which are activated by IL-1, TNF-a and LPS to release these trophic factors. In addition, GM-CSF is derived from activated T lymphocytes and IL-1/ TNF-a-activated stromal cells and monocytes. In patients with active IBD, leukocytosis correlates with serum concentrations of IL-6 and thrombocytosis [122]. The presence of leukocytosis in patients with active IBD is thus related to increased circulating levels of proinflammatory cytokines or a septic

Konstantinos A. Papadakis and Maria T. Abreu

243

Thrombotic event

1 Screen for disorder of coagulation (see Table 3)

No disorder of coagulation identified

Disorder of coagulation identified Long-term Anticoagulation

Multiple thrombotic events or lifethreatening; inflammatory bowel disease well-controlled

Single thrombotic event; associated with active inflammatory bowel disease

Control intestinal inflammation

Short-term Anticoagulation (6 months)

Figure 1. Evaluation of the IBD patient with a thrombotic event.

Table 3. Causes of hypercoagulability in patients with inflammatory bowel disease [128,147] Inherited causes of hypercoagulability

Acquired causes of hypercoagulability

Activated protein C resistance (Arg 506 changed to Gin in factor V) (factor V Leiden mutation)

Thrombocytosis (active inflammation)

Prothrombin G2021OA polymorphism (results in increased levels of prothrombin)

Increased platelet aggregation

Methylene tetrahydrofolate reductase mutations (hyperhomocysteinemia)

Elevated factor VIII plasma levels

Deficiency or mutations in protein C

Elevated fibrinogen levels

Deficiency or mutations in protein S

Anticardiolipin antibodies

Antithrombin III deficiency

Deficiency in protein S Secondary hyperhomocysteinemia (due to folate, pyridoxine or cobalamin deficiency)

244 complication such as a microperforation, frank perforation or abscess. Corticosteroids also lead to leukocytosis secondary to down-regulation of intercellular adhesion molecule expression and granulocyte demargination.

Hematologic malignancies In spite of multiple case reports of leukemias or lymphomas associated with IBD, it remains controversial whether patients with IBD are at increased risk for the development of hematologic malignancies either as a result of chronic inflammatory disease or its treatment [155-157]. Other chronic inflammatory disorders such as rheumatoid arthritis do pose an increased risk of hematologic malignancies [158]. Based on several case series and population-based studies, it seems that patients with CD are at a slightly increased risk for the development of nonHodgkin's lymphomas than the average population search. A recent study based on Olmstead County, Minnesota, records found a standardized incidence ratio of lymphomas of 2.4 (95% confidence intervals 0.1-13) in patients with CD and no cases of lymphomas occurred in patients with UC [155]. Only three patients out of 61 who developed lymphomas were receiving purine analogs. Few patients with IBD in this study were on immunomodulator therapy thus this may represent the true incidence of lymphomas in IBD patients in the absence of immunosuppressive therapy. There was no increased risk of leukemia in these patients. In another large series of IBD patients, five lymphomas occurred among 1156 patients (0.43%) with UC and four lymphomas among 1480 patients (0.27%) with CD and the risk of lymphomas correlated with disease duration in patients with CD [159]. A recent population-based study performed in Florence, Italy, reported a ninefold increased incidence of Hodgkin's lymphoma in patients with UC compared with the general population [160]. The same study found a reduced risk (standardized incidence ratio 0.6) of respiratory tract cancers in patients with UC and increased risk in patients with CD. Severity of the IBD has not been established as an independent risk factor in the development of lymphomas. Immunosuppressive therapy has been implicated as a risk factor in the development of hematologic malignancies, but a causal role of immunosuppressive therapy in the development of lymphomas associated with IBD has not been estabhshed [161, 162]. To date there have been three published

Systemic consequences of intestinal inflammation cases of brain lymphomas (one each from three large series) in patients with IBD on long-term AZP or 6MP in a total of 1701 patients [156, 161, 163]. Although the overall risk of brain lymphomas is low in patients with IBD treated with purine antimetabolites, immunosuppression is clearly a risk factor for this otherwise rare malignancy [164]. In a series of 550 patients with IBD on long-term 6-MP therapy, two patients developed non-Hodgkin's lymphoma, one leukemia and one brain lymphoma [161]. The overall rate of these malignancies was not higher than in the general population. Two other large series of 755 and 396 patients found no excess in the risk of any hematologic malignancies in IBD patients treated with AZP or 6-MP [156, 163]. A decision analysis comparing the eflficacy of AZP at maintaining remission in IBD compared with the risk of non-Hodgkin's lymphoma found that AZP was favorable and resulted in improved quality of life [165]. In spite of the encouraging data in patients with IBD, children treated with 6-MP for acute lymphoblastic leukemia have an increased risk of secondary myelodysplasia or acute myeloid leukemia (5/493), and this risk is associated with low TPMT enzymatic activity and high erythrocyte 6-thioguanine nucleotides [166]. 6MP and AZP have also been associated with rare cases of acute myeloid leukemia in patients with autoimmune diseases including IBD treated for prolonged periods of time [167, 168]. In all the cases a prolonged pancytopenic phase preceded the onset of the leukemia suggesting an antecedent myelodysplastic phase. Immunocompromised hosts, especially patients who have undergone solid organ transplantation, are at significantly increased risk for the development of B cell lymphomas related to Epstein-Barr virus (EBV) infection [169, 170]. The risk in patients post-transplantation is thought to be high because of the use of multiple immunosuppressive drugs, as well as the long-term nature of the immunosuppression. In a series of four patients with gastrointestinal Hodgkin's lymphoma and IBD, all four lymphomas were found to be EBV-mediated with only two patients on immunomodulator therapy [171]. An additional case of an EBV-driven lymphoproliferative disorder in a CD patient on long-term AZP regressed after discontinuation of the AZP [172, 173]. Finally, one B cell lymphoma and one Hodgkin's lymphoma have been reported in patients with CD after receiving infliximab anti-TNF-oc therapy [174]. These lymphomas occurred 3 weeks and 10 months after receiving infliximab. It is too soon to

Konstantinos A. Papadakis and Maria T. Abreu determine the relationship of anti-TNF-a strategies to the development of hematologic malignancies.

Amyloidosis in patients witli IBD An uncommon but morbid systemic consequence of IBD is the development of amyloidosis [30,175-177]. Amyloidosis is a group of diseases characterized by the extracellular deposition of pathologic insoluble fibrillar proteins in organs and tissues [178, 179]. There are three principal forms of amyloidosis. Systemic AL amyloidosis is associated with blood cell dyscrasias and monoclonal gammopathies. There are multiple familial forms of amyloidosis or type ATTR amyloidosis. As with other chronic inflammatory states such as rheumatoid arthritis, IBD may rarely result in secondary systemic amyloidosis type A A [180, 181]. In general, rheumatologic diseases are more commonly associated with amyloidosis than IBD, and CD is more often complicated by amyloidosis than UC [175, 178, 179, 182]. In a large series of IBD followed at the Mount Sinai Hospital, amyloidosis occurred in 15 of 1709 patients with CD (0.9%) and one of 1341 patients with UC (0.07%) [30]. Amyloidosis was more often associated with CD of the colon than with pure small bowel disease. The diagnosis of amyloidosis is made by appropriate pathology demonstrating Congo red-positive amyloid deposits on fat-pad biopsy, rectal biopsy or renal biopsy [178, 179]. Scintigraphy has recently been developed as a non-invasive and quantitative alternative to histology. A radioactive tracer ^^^Iserum amyloid P component is injected and specifically targets amyloid deposits in vivo. The technique has almost 100% sensitivity for systemic AA amyloidosis. In four patients with CD and amyloidosis, ^^^Iserum amyloid P nuclear medicine scanning demonstrated the increased amyloid content and correlated with disease improvement following renal transplantation [180]. In patients with CD and AA amyloidosis, proteinuria is the most common presentation [30, 175,177, 179, 180, 183]. The overall 5-year mortality for AA amyloidosis is 50%) [178, 179]. In one series, one of four patients died prior to renal transplantation demonstrating the high morbidity and mortality from this complication [180]. Nephropathy is the most common lethal manifestation of IBD-associated amyloidosis. Nephrotic syndrome was responsible for 10 deaths out of 25 patients with IBDassociated amyloidosis in a large series [30].

245 Therapy for type AA amyloidosis is directed at limiting the acute-phase response generating the amyloid protein [178, 179]. Unfortunately, this type of amyloidosis is generally advanced by the time of diagnosis with extensive amyloid deposits and renal failure [175]. Only limited case report information is available on the management of IBD complicated by amyloidosis. Several cases of UC complicated by AA amyloidosis have been effectively treated with colchicine, which led to reduction in proteinuria [182,184187]. AZP and colchicine combination therapy has also been eff*ective in improving amyloidosis-induced renal failure and controlling CD in one patient [188]. An elemental diet has been reported as effective in preventing ongoing renal damage in a CD patient with amyloidosis [189]. Regression of amyloid has also been reported in patients with CD after bowel resection or after treatment with dimethylsulfoxide (DMSO) [190, 191]. Based on the limited data, periodic tests of renal function are warranted in patients with IBD to assess nephrotoxicity from medications, as well as the rare occurrence of amyloidosis.

References 1. 2. 3.

4. 5.

6. 7. 8. 9. 10.

Chrousos GP. The hypothalamic-pituitary-adrenal axis and immune-mediated inflammation. N Engl J Med 1995; 332:1351-62. Papanicolaou DA, Wilder RL, Manolagas SC, Chrousos GP. The pathophysiologic roles of interleukin-6 in human disease. Ann Intern Med 1998; 128: 127-37. van Deventer SJ, Buller HR, ten Gate JW, Aarden LA, Hack CE, Sturk A. Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood 1990; 76: 2520-6. Kotler D. Cachexia. Annals Intern Med 2000; 133: 622-34. Tracey KJ, Morgello S, KopHn B et al. Metabolic effects of cachectin/tumor necrosis factor are modified by site of production. Cachectin/tumor necrosis factor-secreting tumor in skeletal muscle induces chronic cachexia, while implantation in brain induces predominantly acute anorexia. J Clin Invest 1990; 86: 2014-24. Tracey KJ, Cerami A. Tumor necrosis factor: a pleiotropic cytokine and therapeutic target. Annu Review of Med 1994; 45: 491-503. Papadakis KA, Targan SR. Role of cytokines in the pathogenesis of inflammatory bowel disease. Annu Rev Med 2000; 51: 289-98. Papadakis KA, Targan SR. Tumor necrosis factor: biology and therapeutic inhibitors. Gastroenterology 2000; 119: 1148-57. O'Neill LA, Dinarello CA. The IL-1 receptor/toll-like receptor superfamily: crucial receptors for inflammation and host defense. Immunol Today 2000; 21: 206-9. Howard AD, Kostura MJ, Thornberry N et al. IL-1-converting enzyme requires aspartic acid residues for processing of the IL-1 beta precursor at two distinct sites and does not cleave 31-kDa IL-1 alpha. J Immunol 1991; 147: 29649.

Systemic consequences of intestinal inflammation

246 11. 12. 13. 14. 15. 16.

17.

18. 19. 20.

21.

22. 23. 24.

25. 26. 27. 28.

29.

30. 31.

Rosenwasser LJ. Biologic activities of IL-1 and its role in human disease. J Allergy Clin Immunol 1998; 102: 344^50. Dinarello CA. Proinflammatory cytokines. Chest 2000; 118: 503-8. Ershler WB, Keller ET. Age-associated increased interleukin-6 gene expression, late-life diseases, and frailty. Annu Rev Med 2000; 51: 245-70. Cressman DE, Greenbaum LE, DeAngelis RA et al. Liver failure and defective hepatocyte regeneration in interleukin6-deficient mice. Science 1996; 274: 1379-83. Streetz KL, Luedde T, Manns MP, Trautwein C. Interleukin 6 and liver regeneration. Gut 2000; 47: 309-12. Nielsen OH, Vainer B, Madsen SM, Seidelin JB, Heegaard NH. Established and emerging biological activity markers of inflammatory bowel disease. Am J Gastroenterol 2000; 95: 359-67. Mahida YR, Ceska M, Efl^enberger F, Kurlak L, Lindley I, Hawkey CJ. Enhanced synthesis of neutrophil-aclivating peptide-l/interleukin-8 in active ulcerative colitis. Clin Sci 1992; 82: 273-5. Farmer RG, Hawk WA, TurnbuU RB Jr. Clinincal patterns in Crohn's disease: a statistical study of 615 cases. Gastroenterology 1975:68:627 35. Sparberg M, Fennessy J, Kirsner JB. Ulcerative proctitis and mild ulcerative colitis: a study of 220 patients. Medicine 1966:45: 391 412. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation [published erratum appears in N Engl J Med 1999; 340: 1376]. N Engl J Med 1999:340:448 54. Moldawer LL, Copcland EM Illrd. Proinflammatory cytokines, nutritional support, and the cachexia syndrome: interactions and therapeutic options. Cancer 1997; 79: 1828 39. Plala-Saiaman CR. Food intake suppression by growth factors and platelet peptides by direct action in the central nervous system. Ncurosci Lett 1988; 94: 161 6. Sarraf P, Frederich RC, Turner EM et al. Multiple cytokines and acute inflammation raise mouse leptin levels: potential role in inflammatory anorexia. J Exp Med 1997; 185: 1715. Both H,Torp-Pedersen K, Kreiner S, Hendriksen C, Binder V. Clinical appearance at diagnosis of ulcerative colitis and Crohn's disease in a regional patient group. Scand J Gastroenterol 1983; 18:987-91. Dinarello CA. Cytokines as endogenous pyrogens. J Infect Dis 1999; 179 (Suppl 2): S294-304. Castell JV, Gomez-Lechon MJ, David M et al. Interleukin-6 is the major regulator of acute phase protein synthesis in adult human hepatocytes. FEBS Lett 1989; 242: 237-9. Fattori E, Cappelletti M, Costa P et al. Defective inflammatory response in interleukin 6-deficient mice. J Exp Med 1994;180: 1243-50. Betz UAK, Bloch W, van den Broek M et al. Postnatally induced inactivation of gpl30 in mice results in neurological, cardiac, hematopoietic, immunological, hepatic, and pulmonary defects. J Exp Med 1998; 188: 1955-65. Gershov D, Kim S, Brot N, Elkon K. C-reactive protein binds to apoptotic cells, protects the cells from assembly of the terminal complement components, and sustains an antiinflammatory innate immune response: implications for systemic autoimmunity. J Exp Med 2000; 192: 1353-64. Greenstein AJ, Sachar DB, Panday AK et al.\ Amyloidosis and inflammatory bowel disease. A 50-year experience with 25 patients. Medicine 1992; 71: 261-70. Kurzrock R, Redman J, Cabanillas F, Jones D, Rothberg J, Talpaz M. Serum interleukin 6 levels are elevated in lymphoma patients and correlate with survival in advanced Hodgkin's disease and with B symptoms. Cancer Res 1993; 53:2118-22.

32. 33.

34.

35. 36. 37. 38.

39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

49. 50. 51. 52. 53. 54.

Gasche C. Anemia in IBD: the overlooked villain. Inflammat Bowel Dis 2000; 6: 142-50; discussion 151. Gasche C, Dejaco C, Waldhoer T et al. Intravenous iron and erythropoietin for anemia associated with Crohn disease. A randomized, controUed trial. Ann Intern Med 1997; 126: 782-7. Horina JH, Petritsch W, Schmid CR et al. Treatment of anemia in inflammatory bowel disease with recombinant human erythropoietin: results in three patients. Gastroenterology 1993;104:1828-31. Gasche C, Reinisch W, Lochs H et al. Anemia in Crohn's disease. Importance of inadequate erythropoietin production and iron deficiency. Dig Dis Sci 1994; 39: 1930-4. Beeken WL. Remediable defects in Crohn disease: a prospective study of 63 patients. Arch Intern Med 1975; 135: 686-90. Gold Y, Reif S. [Aphthous stomatitis as a first manifestation of Crohn's disease in a 5 year-old boy]. Harefuah 1998; 135: 364 6, 407. Froom P, Benbassat J, Kiwelowicz A, Erel J, Davidson B, Ribak J. Significance of low hematocrit levels in asymptomatic young adults: results of 15 years follow-up. Aviat Space Environ Med 1999; 70: 983 6. Menachem Y, Weizman Z, Locker C, Odes S. Clinical characteristics of Crohn's disease in children and adults. Harefuah 1998; 134: 173 5,247. Smith AD, Cochran KM. Serum ferritin: it may guide the diagnosis of the anaemic patient. Scot Med J 1997; 42: 182 3. Gasche C, Dejaco C, Reinisch W et ^//.Sequential treatment of anemia in ulcerative colitis with intravenous iron and erythropoietin. Digestion 1999; 60: 262 7. Macdougall IC. Strategies for iron supplementation: oral versus intravenous. Kidney Int (Suppl.) 1999; 69: S61 6. Hamstra RD, Block MH, Schockct AL. Intravenous iron dextran in clinical medicine. J Am Med Assoc 1980; 243: 1726 31. Dyer N, Dawson A: Malnutrition and malabsorption in Crohn's disease with references to the eflect of surgery. Br J Surg 1973; 60: 134 40. Dyer NH, Child JA, Mollin DL, Dawson AM. Anaemia in Crohn's disease. Q J Med 1972; 41: 419 36. Thompson W, Wrathell E. The relation between ileal resection and vitamin B12 absorption. Can J Surg 1977; 20: 4614. Fone D et al. ["''^]Co B-12 absorption after gastrectomy, ileal resection and in coeliac disorders. Gut 1961; 2: 218. Lee G. Pernicious anemia and other causes of vitamin B12 (cobalamin) deficiency. In: Lee G (ed.), Wintrobe's Clinical Hematology, 10th edn. Baltimore: Lippincott, Williams & Wilkins, 1999; 956. Spiegel JE, Willenbucher RF. Rapid development of severe copper deficiency in a patient with Crohn's disease receiving parenteral nutrition. J Parent Ent Nutr 1999; 23: 169-72. Means RT Jr. Erythropoietin in the treatment of anemia in chronic infectious, inflammatory, and malignant diseases. Curr Opin Hematol 1995; 2: 210-3. Means RT Jr. Advances in the anemia of chronic disease. Int J Hematol 1999;70:7-12. Schreiber S, Howaldt S, Schnoor M et al. Recombinant erythropoietin for the treatment of anemia in inflammatory bowel disease. N Engl J Med 1996; 334: 619-23. Koury M, Bondurant M. The molecular mechanism of erythropoietin action. Eur J Biochem 1992; 210: 649-63. Jelkmann WE, Fandrey J, Frede S, Pagel H: Inhibition of erythropoietin production by cytokines. Implications for the anemia involved in inflammatory states. Ann NY Acad Sci 1994; 718: 300-9; discussion 309-11.

247

Konstantinos A. Papadakis and Maria T. Abreu 55. 56. 57.

58.

59. 60. 61.

62.

63.

64.

65.

66.

67. 68. 69.

70. 71. 72. 73.

Jelkmann W. Proinflammatory cytokines lowering erythropoietin production. J Interferon Cytokine Res 1998; 18: 555-9. Faquin WC, Schneider TJ, Goldberg MA. Effect of inflammatory cytokines on hypoxia-induced erythropoietin production. Blood 1992; 79: 1987-94. Gunnell J, Yeun JY, Depner TA, Kaysen GA. Acute-phase response predicts erythropoietin resistance in hemodialysis and peritoneal dialysis patients. Am J Kidney Dis 1999; 33: 63-72. Nordstrom D, Lindroth Y, Marsal L et al Availability of iron and degree of inflammation modifies the response to recombinant human erythropoietin when treating anemia of chronic disease in patients with rheumatoid arthritis. Rheumatol Int 1997; 17:67-73. Tarng DC, Huang TP, Chen TW, Yang WC. Erythropoietin hyporesponsiveness: from iron deficiency to iron overload. Kidney Int (Suppl.) 1999; 69: S107-18. Lennard L, Murphy M, Maddocks J: Severe megaloblastic anaemia associated with abnormal azathioprine metabolism. Br J Clin Pharmacol 1984; 17: 171. Weinshilboum RM, Sladek SL. Mercaptopurine pharmacogenetics: monogenic inheritance of erythrocyte thiopurine methyl transferase activity. Am J Hum Genet 1980; 32: 65162. Schutz E, Gummert J, Armstrong VW, Mohr FW, Oellerich M. Azathioprine pharmacogenetics: the relationship between 6-thioguanine nucleotides and thiopurine methylt r a n s f e r a s e in p a t i e n t s after h e a r t a n d k i d n e y transplantation. Eur J Clin Chem Clin Biochem 1996; 34: 199-205. Lennard L, Van Loon JA, Lilleyman JS, Weinshilboum RM. Thiopurine pharmacogenetics in leukemia: correlation of erythrocyte thiopurine methyltransferase activity and 6thioguanine nucleotide concentrations. Clin Pharmacol Ther 1987;41: 18-25. Lowry PW, Szumlanski CL, Weinshilboum RM, Sandborn WJ. Balsalazide and azathiprine or 6-mercaptopurine: evidence for a potentially serious drug interaction [Letter; comment]. Gastroenterology 1999; 116: 1505-6. Lewis LD, Benin A, Szumlanski CL et al. Olsalazine and 6mercaptopurine-related bone marrow suppression: a possible drug-drug interaction [Published erratum appears in Clin Pharmacol Ther 2000; 67:431]. Clin Pharmacol Ther 1997; 62: 464-75. Szumlanski CL, Weinshilboum RM. Sulphasalazine inhibition of thiopurine methyltransferase: possible mechanism for interaction with 6-mercaptopurine and azathioprine. Br J Chn Pharmacol 1995; 39: 456-9. Swinson CM, Perry J, Lumb M, Levi AJ. Role of sulphasalazine in the aetiology of folate deficiency in ulcerative colitis. Gut 1981; 22: 456-61. Longstreth G, Green R. Folate status in patients receiving maintenance doses of sulfasalazine. Arch Intern Med 1983; 143: 902. Das KM, Eastwood MA, McManus JP, Sircus W. Adverse reactions during salicylazosulfapyridine therapy and the relation with drug metabolism and acetylator phenotype. N Engl J Med 1973; 289: 491-5. Abboudi ZH, Marsh JC, Smith-Laing G, Gordon-Smith EC. Fatal aplastic anaemia after mesalazine [Letter]. Lancet 1994; 343: 542. Dunn AM, Kerr GD. Pure red ceU aplasia associated with sulphasalazine [Letter]. Lancet 1981; 2: 1288. Daneshmend T. Mesalazine-associated thrombocytopenia. Lancet 1991; 337: 1297-8. Wyatt S, Joyner M, Daneshmend T Filgrastim for mesalazine-associated neutropenia. Lancet 1993; 341: 1476.

74. 75. 76.

77. 78. 79.

80. 81.

82. 83.

84. 85. 86. 87. 88. 89. 90.

91. 92. 93.

94. 95.

Gumaste V, Greenstein AJ, Meyers R, Sachar DB. Coombspositive autoimmune hemolytic anemia in ulcerative colitis. DigDisSci 1989; 34: 1457-61. Bell DW, Urban E, Sears DA, Walder Al, Ostrower VS. Ulcerative colitis complicated by autoimmune hemolytic anemia. S Med J 1981; 74: 359-61. Altman AR, Maltz C, Janowitz H D Autoimmune hemolytic anemia in ulcerative colitis: report of three cases, review of the literature, and evaluation of modes of therapy. Dig Dis Sci 1979; 24: 282-5. Ma A, Datta M, Margosian E, Chen J, Horak I: T cells, but not B cells, are required for bowel inflammation in interleukin 2-deficient mice. J Exp Med 1995; 182: 1567-72. Ramakrishna R, Manoharan A. Auto-immune haemolytic anaemia in ulcerative coHtis. Acta Haematol 1994; 91: 99102. Giannadaki E, Potamianos S, Roussomoustakaki M, Kyriakou D, Fragkiadakis N, Manousos ON. Autoimmune hemolytic anemia and positive Coombs test associated with ulcerative colitis. Am J Gastroenterol 1997; 92: 1872-4. Teplitsky V, Virag I, Halabe A. Immune complex haemolytic anaemia associated with sulfasalazine. Br Med J 2000; 320: 1113. Mechanick JI. Coombs' positive hemolytic anemia following sulfasalazine therapy in ulcerative colitis: case reports, review, and discussion of pathogenesis. Mount Sinai J Med 1985; 52: 667-70. van Hees PA, van Elferen LW, van Rossum JM, van Tongeren JH. Hemolysis during salicylazosulfapyridine therapy. Am J Gastroenterol 1978; 70: 501-5. Yates P, Macht LM, Williams NA, Elson CJ. Red cell autoantibody production by colonic mononuclear cells from a patient with ulcerative colitis and autoimmune haemolytic anaemia. Br J Haematol 1992; 82: 753-6. Murphy PT, Cunney R, Nolan A, O'Donnell JR. Autoimmune haemolytic anaemia associated with ulcerative colitis. Irish Med J 1996; 89: 172-3. Cohen SM, Rosenthal DS, Karp PJ. Ulcerative colitis and erythrocyte G6PD deficiency. Salicylazosulfapyridine-provoked hemolysis. J Am Med Assoc 1968; 205: 528-30. Dahlback B, Blood coagulation. Lancet 2000; 355: 1627-32. Aichbichler BW, Petritsch W, Reicht GA et al. Anti-cardiolipin antibodies in patients with inflammatory bowel disease. Dig Dis Sci 1999; 44: 852-6. Schapira M, Henrion J, Ravoet C et ^/.Thromboembolism in inflammatory bowel disease. Acta Gastroenterol Belg 1999;62:182-6. Talbot RW, Heppell J, Dozois RR, Beart RW, Jr. Vascular complications of inflammatory bowel disease. Mayo Clin Procl986;61: 140-5. Jackson LM, O'Gorman PJ, O'Connell J, Cronin CC, Cotter KP, Shanahan F. Thrombosis in inflammatory bowel disease: clinical setting, procoagulant profile and factor V Leiden. Q J Med 1997; 90: 183-8. Koenigs KP, McPhedran P, Spiro HM. Thrombosis in inflammatory bowel disease. J Clin Gastroenterol 1987; 9: 627-31. Johns DR. Cerebrovascular complications of inflammatory bowel disease. Am J Gastroenterol 1991; 86: 367-70. Carmona MA, Jaume Anselmi F, Ramirez Rivera J. Cerebral thrombosis and vasculitis: an uncommon complication of ulcerative colitis. Bol Asoc Med Puerto Rico 2000; 92: 911. Papi C, Ciaco A, Acierno G et al. Severe ulcerative colitis, dural sinus thrombosis, and the lupus anticoagulant. Am J Gastroenterol 1995; 90: 1514-7. Musio F, Older SA, Jenkins T, Gregorie EM. Case report: cerebral venous thrombosis as a manifestation of acute ulcerative colitis. Am J Med Sci 1993; 305: 28-35.

Systemic consequences of intestinal inflammation

248 96.

97. 98. 99. 100.

101. 102. 103.

104.

105. 106. 107. 108.

109.

110. 111.

112.

113. 114. 115. 116.

Markowitz RL, Ment LR, Gryboski JD. Cerebral thromboembolic disease in pediatric and adult inflammatory bowel disease: case report and review of the literature. J Pediatr Gastroenterol Nutr 1989; 8: 413-20. Bansal R, Goel A. Ulcerative colitis with sagittal sinus thrombosis with normal coagulation profile. Ind J Gastroenterol 2000; 19:88-9. Derdeyn CP, Powers WJ. Isolated cortical venous thrombosis and ulcerative cohtis. Am J Neuroradiol 1998; 19: 48890. Crowe A, Taffinder N, Layer GT, Irvine A, Nicholls RJ. Portal vein thrombosis in a complicated case of Crohn's disease. Postgrad Med J 1992; 68: 291-3. Miyazaki Y, Shinomura Y, Kitamura S et al. Portal vein thrombosis associated with active ulcerative colitis: percutaneous transhepatic recanalization. Am J Gastroenterol 1995; 90: 1533 4. Chesner IM, Muller S, Newman J. Ulcerative colitis complicated by Budd-Chiari syndrome. Gut 1986; 27: 1096 100. Maccini DM, Berg JC, Bell GA. Budd-Chiari syndrome and Crohn's disease. An unreported association. Dig Dis Sci 1989; 34: 1933 6. Brinson RR, Curtis WD, Schuman BM, Mills LR. Recovery from hepatic vein thrombosis (Budd Chiari syndrome) complicating ulcerative colitis. Dig Dis Sci 1988; 33: 1615 20. Praderio L, Dagna L, Longhi P, Rubin G, Sabbadini M G Budd Chiari syndrome in a patient with ulcerative colitis: association with anticardiolipin antibodies. J Clin Gastroenterol 2000; 30: 203 4. Halliday CE, Farthing MJ. Arterial thrombosis in Crohn's disease. Med J Austr 1988; 149: 559 60. Kyrle PA, Minar E, Hirschl M ct al. High plasma levels of factor VI11 and the risk of recurrent venous thromboembolism. N Engl J Med 2000; 343: 457 62. Chiarantini E, Valanzano R, Liotta AA. Hemostatic abnormalities in inflammatory bowel disease. Thromb Res 1996; 82: 137 46. Smith CJ, Haire WD, Kaufman SS, Mack DR. Determination of prothrombin activation fragments in young patients with inflammatory bowel disease. Am J Gastroenterol 1996; 91: 1221-5. Chamouard P, Grunebaum L,Wiesel MLctai Prothrombin fragment 1 + 2 and thrombin-antithrombin III complex as markers of activation of blood coagulation in inflammatory bowel diseases. Eur J Gastroenterol Hepatol 1995; 7: 11838. Souto JC, Martinez E, Roca M et ^^/.Prothrombotic state and signs of endothelial lesion in plasma of patients with inflammatory bowel disease. Dig Dis Sci 1995; 40: 1883-9. Koutroubakis IE, Sfiridaki A, Mouzas lA et al. Resistance to activated protein C and low levels of free protein S in Greek patients with inflammatory bowel disease. Am J Gastroenterol 2000; 95: 190-4. Lee LC, Spittell JA, Jr, Sauer WG, Owen CA, Jr, Thompson JH, Jr. Hypercoagulability associated with chronic ulcerative colitis: changes in blood coagulation factors. Gastroenterology 1968; 54: 76-85. Braverman D, Bogoch A. Arterial thrombosis in ulcerative colitis. Am J Dig Dis 1978; 23: 1148-50. Aadland E, Odegaard OR, Roseth A, Try K. Free protein S deficiency in patients with Crohn's disease. Scand J Gastroenterol 1994; 29: 333-5. Talstad I, Rootwelt K, Gjone E. Thrombocytosis in ulcerative colitis and Crohn's disease. Scand J Gastroenterol 1973; 8: 135-8. Lam A, Borda I, Inwood M, Thomson S. Coagulation studies in ulcerative colitis and Crohn's disease. Gastroenterology 1975; 68: 245-51.

117. Morowitz D, Allen L, Kirsner J. Thrombocytosis in chronic inflammatory bowel disease. Ann Intern Med 1968; 68: 1013-21. 118. Vecchi M, Cattaneo M, de Franchis R, Mannucci PM. Risk of thromboembolic complications in patients with inflammatory bowel disease. Study of hemostasis measurements. Int J Clin Lab Res 1991; 21: 165-70. 119. Heneghan MA, Cleary B, Murray M, O'Gorman TA, McCarthy CF. Activated protein C resistance, thrombophilia, and inflammatory bowel disease. Dig Dis Sci 1998; 43: 1356-61. 120. Webberley MJ, Hart MT, Melikian V. Thromboembolism in inflammatory bowel disease: role of platelets. Gut 1993; 34: 247-51. 121. Chiarantini E, Valanzano R, Liotta A A et al. Persistence of hemostatic alterations in patients affected by Crohn's disease after bowel surgery. Thromb Res 1997; 87: 539-46. 122. Heits F, Stahl M, Ludwig D, Stange EF, Jelkmann W. Elevated serum thrombopoietin and interleukin-6 concentrations in thrombocytosis associated with inflammatory bowel disease. J Interferon Cytokine Res 1999; 19: 757 60. 123. Sands BE, Bank S, Sninsky CA et al. Preliminary evaluation of safety and activity of recombinant human interleukin 11 in patients with active Crohn's disease. Gastroenterology 1999; 117:58 64. 124. Hudson M, Chitolie A, Hutton RA, Smith MS, Pounder RE, Wakefield AJ. Thrombotic vascular risk factors in inflammatory bowel disease. Gut 1996; 38: 733 7. 125. Chamouard P, Grunebaum L, Wiesel ML et al. Prevalence and significance of anticardiolipin antibodies in Crohn's disease. Dig Dis Sci 1994; 39: 1501 4. 126. Koutroubakis IE, Petinaki E, Anagnostopoulou E et al. Anti-cardiolipin and anti-beta2-glycoprolein I antibodies in patients with inflammatory bowel disease. Dig Dis Sci 1998; 43:2507 12. 127. Vianna JL, D'Cruz DP, Khamashta MA, Asherson RA, Hughes GR. Anticardiolipin antibodies in a patient with Crohn's disease and thrombosis. Clin Exp Rheumatol 1992; 10: 165 8. 128. Olds RJ, Fitches AC, Geary CP. The multigenic basis for venous thrombosis. Br J Haematol 2000; 109: 508-11. 129. Dahlback B. New molecular insights into the genetics of thrombophilia. Resistance to activated protein C caused by Arg (506) to Gin mutation in Factor V as a pathogenic risk factor for venous thrombosis. Thromb Haemostas 1995; 74: 139-48. 130. Sheppard DR. Activated protein C resistance: the most common risk factor for venous thromboembolism. J Am Board Fam Pract 2000; 13: 111-5. 131. De Stefano V, Martinelli I, Mannucci PM, Paciaroni K, Chiusolo P, Casorelli I, Rossi E, Leone G. The risk of recurrent deep venous thrombosis among heterozygous carriers of both factor V Leiden and the G20210A prothrombin mutation. N Engl J Med 1999; 341: 801-6. 132. Vecchi M, Sacchi E, Saibeni S et al. Inflammatory bowel diseases are not associated with major hereditary conditions predisposing to thrombosis. Dig Dis Sci 2000; 45: 1465-9. 133. Zauber NP, Sabbath-Solitare M, Rajoria G, Mogan G. Factor V Leiden mutation is not increased in patients with inflammatory bowel disease. J Clin Gastroenterol 1998; 27: 215-6. 134. Levine A, Lahav J, Zahavi I, Raz A, Dinari G. Activated protein C resistance in pediatric inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1998; 26: 172-4. 135. Haslam N, Standen GR, Probert CS. An investigation of the association of the factor V Leiden mutation and inflammatory bowel disease. Eur J Gastroenterol Hepatol 1999; 11: 1289-91.

Konstantinos A. Papadakis and Maria T. Abreu 136. Over HH, Ulgen S, Tuglular T et al. Thrombophilia and inflammatory bowel disease: does factor V mutation have a role? Eur J Gastroenterol Hepatol 1998; 10: 827-9. 137. Novacek G, Miehsler W, Kapiotis S, Katzenschlager R, Speiser W, Vogelsang H. Thromboembolism and resistance to activated protein C in patients with inflammatory bowel disease. Am J Gastroenterol 1999; 94: 685-90. 138. Liebman HA, Kashani N, Sutherland D, McGehee W, Kam AL. The factor V Leiden mutation increases the risk of venous thrombosis in patients with inflammatory bowel disease. Gastroenterology 1998; 115: 830-4. 139. Cattaneo M, Vecchi M, Zighetti ML, et al. High prevalence of hyperchomocysteinemia in patients with inflammatory bowel disease: a pathogenic link with thromboembolic complications? Thromb Haemostas 1998; 80: 542-5. 140. Mahmud N, Molloy A, McPartlin J et al. Increased prevalence of methylenetetrahydrofolate reductase C677T variant in patients with inflammatory bowel disease, and its clinical implications. Gut 1999; 45: 389-94. 141. Oldenburg B, Fijnheer R, van der Griend R, vanBergeHenegouwen GP, Koningsberger JC. Homocysteine in inflammatory bowel disease: a risk factor for thromboembohc complications? Am J Gastroenterol 2000; 95: 2825-30. 142. Thompson N, Wakefield A, Pounder R. Inherited disorders of coagulation appear to protect against inflammatory bowel disease. Gastroenterology 1995; 108: 1011-15. 143. Kearon C, Gent M, Hirsh J et al. A comparison of three months of anticoagulation with extended anticoagulation for a first episode of idiopathic venous thromboemboHsm [published erratum appears in N Engl J Med 1999 22; 341: 298]. N Engl J Med 1999; 340: 901-7. 144. Gaff'ney PR, Doyle CT, Gaff'ney A, Hogan J, Hayes DP, Annis P. Paradoxical response to heparin in 10 patients with ulcerative colitis. Am J Gastroenterol 1995; 90: 220-3. 145. Van Woert JH, Thompson RC, Cangemi JR, Metzger PP, Blackshear JL, Fleming CR. Streptokinase therapy for extensive venous thromboses in a patient with severe ulcerative colitis. Mayo Clin Proc 1990; 65: 1144-9. 146. Kermode AG, Ives FJ, Taylor B, Davis SJ, Carroll WM. Progressive dural venous sinus thrombosis treated with local streptokinase infusion [Letter]. J Neurol Neurosurg Psychiatry 1995; 58: 107-8. 147. Nguyen A. Prothrombin G20210A polymorphism and thrombophilia. Mayo Clin Proc 2000; 75: 595-604. 148. Mufler S, Chesner IM, Sheridan J, Newman J. Ulcerative colitis complicated by disseminated intravascular coagulation. Postgrad Med J 1987; 63: 689-91. 149. Ryan FP, Timperley WR, Preston FE, Holdsworth CD. Cerebral involvement with disseminated intravascular coagulation in intestinal disease. J Clin Pathol 1977; 30: 551-5. 150. Wong TZ, Welch JP, Holt JB. Intraoperative disseminated intravascular coagulation in a patient with ulcerative colitis. Connect Med 1989; 53: 577-8. 151. Zlatanic J, Korelitz BI, Wisch N et al. Inflammatory bowel disease and immune thrombocytopenic purpura: is there a correlation? Am J Gastroenterol 1997; 92: 2285-8. 152. Mones RL. Thrombocytopenia and hypofibrinogenemia in association with inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1983; 2: 175-7. 153. Gremse DA, Bancroft J, Moyer MS. Sulfasalazine hypersensitivity with hepatotoxicity, thrombocytopenia, and erythroid hypoplasia. J Pediatr Gastroenterol Nutr 1989; 9: 261-3. 154. Caprilli R, Corrao G, Taddei G, Tonelh F, Torchio P, Viscido A. Prognostic factors for postoperative recurrence of Crohn's disease. Gruppo Italiano per lo Studio del Colon e del Retto (GISC). Dis Colon Rectum 1996; 39: 335-41. 155. Loftus EV, Jr., Tremaine WJ, Habermann TM, Harmsen WS, Zinsmeister AR, Sandborn WJ. Risk of lymphoma in

249

156.

157. 158.

159. 160. 161.

162.

163.

164. 165.

166. 167.

168. 169. 170. 171.

172.

173. 174.

175.

inflammatory bowel disease. Am J Gastroenterol 2000; 95: 2308-12. Connell WR, Kamm MA, Dickson M, Balkwill AM, Ritchie JK, Lennard-Jones JE. Long-term neoplasia risk after azathioprine treatment in inflammatory bowel disease. Lancet 1994; 343: 1249-52. Caspi O, Polliack A, Klar R, Ben-Yehuda D. The association of inflammatory bowel disease and leukemia - coincidence or not? Leukemia Lymphoma 1995; 17: 255-62. Georgescu L, Quinn GC, Schwartzman S, Paget SA. Lymphoma in patients with rheumatoid arthritis: association with the disease state or methotrexate treatment. Semin Arthritis Rheum 1997; 26: 794-804. Greenstein AJ, Mullin GE, Strauchen JA et al. Lymphoma in inflammatory bowel disease. Cancer 1992; 69: 1119-23. PaUi D, Trallori G, Bagnoli S et al. Hodgkin's disease risk is increased in patients with ulcerative colitis. Gastroenterology 2000; 119:647-53. Korelitz BI, Mirsky FJ, Fleisher MR, Warman JI, Wisch N, Gleim GW Malignant neoplasms subsequent to treatment of inflammatory bowel disease with 6-mercaptopurine. Am J Gastroenterol 1999; 94: 3248-53. Present DH, Korelitz BI, Wisch N, Glass JL, Sachar DB, Pasternack BS. Treatment of Crohn's disease with 6-mercaptopurine. A long-term, randomized, double-blind study. N Engl J Med 1980; 302: 981-7. Present DH, Meltzer SJ, Krumholz MP, Wolke A, Korelitz BI. 6-Mercaptopurine in the management of inflammatory bowel disease: short- and long-term toxicity. Ann Intern Med 1989; 111:641-9. Schabet M. Epidemiology of primary CNS lymphoma. J Neuro-Oncol 1999; 43: 199-201. Lewis JD, Schwartz JS, Lichtenstein GR. Azathioprine for maintenance of remission in Crohn's disease: benefits outweigh the risk of lymphoma. Gastroenterology 2000; 118: 1018-24. Bo J, Schroder H, Kristinsson J et al. Possible carcinogenic effect of 6-mercaptopurine on bone marrow stem cells: relation to thiopurine metaboUsm. Cancer 1999; 86: 1080-6. Kwong YL, Au WY, Liang RH. Acute myeloid leukemia after azathioprine treatment for autoimmune diseases: association with -7/7q. Cancer Genet Cytogenet 1998; 104: 947. Heizer WD, Peterson JL. Acute myeloblastic leukemia foUowing prolonged treatment of Crohn's disease with 6mercaptopurine. Dig Dis Sci 1998; 43: 1791-3. DeMario MD, Liebowitz DN. Lymphomas in the immunocompromised patient. Semin Oncol 1998; 25: 492-502. Nalesnik MA. Clinicopathologic features of posttransplant lymphoprohferative disorders. Ann Transplant 1997; 2: 3 3 40. Kumar S, Fend F, QuintaniUa-Martinez L et a/.Epstein-Barr virus-positive primary gastrointestinal Hodgkin's disease: association with inflammatory bowel disease and immunosuppression. Am J Surg Pathol 2000; 24: 66-73. Calaminici MR, Sheaff MT, Norton AJ, Feakins RM. Ileocaecal Epstein-Barr virus-positive lymphoprohferative disorder complicating Crohn's disease [Letter]. Histopathology 1999; 35: 388-90. Larvol L, Soule JC, Le Tourneau A. Reversible lymphoma in the setting of azathioprine therapy for Crohn's disease [Letter]. N Engl J Med 1994; 331: 883-4. Bickston SJ, Lichtenstein GR, Arseneau KO, Cohen RB, ComineUi F. The relationship between infliximab treatment and lymphoma in Crohn's disease. Gastroenterology 1999; 117:1433-7. Pardi DS, Tremaine WJ, Sandborn WJ, McCarthy JT. Renal and urologic complications of inflammatory bowel disease. Am J Gastroenterol 1998; 93: 504-14.

250 176. Kahn E, Markowitz J, Simpser E, Aiges H, Daum F. Amyloidosis in children with inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1989; 8: 447-53. 177. Lowdell CP, Shousha S, Parkins RA. The incidence of amyloidosis complicating inflammatory bowel disease. A prospective survey of 177 patients. Dis Colon Rectum 1986;29:351-4. 178. Falk RH, Comenzo RL, Skinner M. The systemic amyloidoses. N Engl J Med 1997; 337: 898-909. 179. Gillmore JD, Hawkins PN, Pepys MB. Amyloidosis. A review of recent diagnostic and therapeutic developments. Br J Haematol 1997; 99: 245-56. 180. Lovat LB, Madhoo S, Pepys MB, Hawkins PN. Long-term survival in systemic amyloid A amyloidosis complicating Crohn's disease. Gastroenterology 1997; 112: 1362-5. 181. Edwards P, Cooper DA, Turner J, O'Connor TJ, Byrnes DJ. Resolution of amyloidosis (AA type) complicating chronic ulcerative colitis. Gastroenterology 1988; 95: 810-15. 182. Gcrtz MA, Kyle RA. Secondary systemic amyloidosis: response and survival in 64 patients. Medicine 1991; 70: 246 56. 183. Fausa O, Nygaard K, Elgjo K. Amyloidosis and Crohn's disease. Scand J Gastroenterol 1977; 12: 657 62. 184. Meyers S, Janowitz HD, Gumaste VV ct al. Colchicine therapy of the renal amyloidosis of ulcerative colitis. Gastroenterology 1988; 94: 1503 7.

Systemic consequences of intestinal inflammation 185. Menges M, Steffen HM. Secondary amyloidosis in ulcerative colitis - successful treatment with colchicine. Z Gastroenterol 1996;34:753-6. 186. Gertz MA, Kyle RA. Amyloidosis: prognosis and treatment. Semin Arthritis Rheum 1994; 24: 124-38. 187. Ravid M, Shapira J, Kedar I, Feigl D. Regression of amyloidosis secondary to granulomatous ileitis following surgical resection and colchicine administration. Acta Hepato-Gastroenterol 1979; 26: 513-5. 188. Larvol L, Cervoni J, Besnier M, Dupouet L, Beaufils H, Clauvel J, Levecq H. Reversible nephrotic syndrome in Crohn's disease complicated with renal amyloidosis. Gastroenterol Chn Biol 1998; 22: 6 3 9 ^ 1 . 189. Horie Y, Chiba M, Miura K ct al. Crohn's disease associated with renal amyloidosis successfully treated with an elemental diet. J Gastroenterol 1997; 32: 663 7. 190. Mandelstam P, Simmons DE, Mitchell B. Regression of amyloid in Crohn's disease after bowel resection. A 19-year follow-up. J Clin Gastroenterol 1989; 11: 324-6. 191. Iwakiri R, Sakemi T, Fujimoto K. Dimethylsulfoxide for renal dysfunction caused by systemic amyloidosis complicating Crohn's disease [Letter; comment]. Gastroenterology 1999; 117: 1031 2.

Section II THE BEDSIDE

13 Understanding symptoms and signs in inflammatory bowel disease CORNELIUS C. CRONIN AND FERGUS SHANAHAN

Introduction The natural history of Crohn's disease (CD) and ulcerative colitis (UC) is highly variable, but most typically follows a course of relapses and remissions. Some patients have chronically active disease with no or few apparent remissions. In some, the condition appears to 'burn out', and they enter long term remission. Series from specialized centers may tend to over-estimate clinical severity [1]. Generally, symptoms of UC tend to be uniform; most patients complain of abrupt onset passage of blood, diarrhea and weight loss. Each acute relapse typically has similar clinical features. Because of its greater anatomical distribution potentially involving any part of the gastrointestinal tract, its transmural distribution and its propensity to give raise to complications such as strictures and fistulae, CD shows greater variability between patients in clinical features. Also, as the disease evolves involving different parts of the gastrointestinal tract, the clinical features in any one patient may also change through time. Symptoms and signs of inflammatory bowel disease (IBD) ultimately depend on the extent, distribution and severity of the gastrointestinal inflammation. Many of the clinical features of CD and UC are related to the anatomical location of disease. Abdominal colic is caused by intestinal strictures and diarrhea by intestinal inflammation. Other features such as anorexia, weight loss and malnutrition, anemia and constitutional features are largely due to the systemic consequences of intestinal inflammation. In recent years, the role of soluble mediators of intestinal inflammation in the pathogenesis of non-intestinal features of IBD has been appreciated. Cytokines are produced by many different tissues in response to immune stimulation and mediate a multiplicity of immunologic and nonimmunologic functions [2, 3]. As well as local (autocrine-paracrine) actions, cytokines have systemic

(endocrine) effects, many mediated by the central nervous system [4, 5]. While the short-term, local effects of cytokines may be beneficial to the organism in the acute phase of an immune reaction, prolonged systemic cytokine activity as in IBD is often deleterious (Table 1) [6, 7].

Disease vs illness There is an imperfect association between disease activity and patient disability. The objective assessment of disease activity often provides a poor guide to the subjective impact of the condition on the Table 1. Adaptions associated with pro-inflammatory cytokines Behavioral Anorexia Fatigue IVIalaise Altered sleep pattern Altered level of consciousness Physiologic Elevated body temperture Increased resting energy expenditure Stress hormone response Skeletal muscle wasting Hepatic acute phase response Trace mineral sequestration Decreased gastric emptying, intestinal transit time Bone marrow suppression Diuresis Nutritional Weight loss Negative nitrogen balance Hypoalbuminemia Hyperinsulinemia Hypertriglyceridemia Hypocholesterolemia Adapted from ret. 7

Stephan R. Targan, Fergus Siianahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 253-267. © 2003 Kluwer Academic Publishers. Printed in Great Britain

254

Understanding symptoms and signs in inflammatory bowel disease

patient: disease does not equate with illness. The biomedical model - the notion that the disease activity determines the clinical outcome - had long been dominant in western medicine. The biopsychosocial model, as described by Engel [8] and further expanded in relation to chronic gastrointestinal disease by Drossman [9], proposes that illness and disease result from simultaneously interacting systems at the cellular, tissue, organism, interpersonal and environmental levels. This approach integrates biological science with the uniqueness of the individual, to determine the degree to which biological and psychosocial factors interact to explain the disease, illness and outcome. Patients with chronic gastrointestinal conditions may develop a variety of psychosociologic responses such as depression, decreased activities, increased illness behaviour, dependent relationships, and decreased responsibilities. Patients often have difficulties in employment, recreation and marital relations. The problems consequent on constitutional symptoms such as malaise and loss of energy may contribute more to patient disability than gastrointestinal symptoms. Illness, particularly that arising from chronic disease, is therefore best understood in terms of a multifactorial model, integrating biologic, psychologic and sociologic variables. Such a model facilitates and optimizes diagnosis, treatment and ultimately patient care. The aphorism 'treat the patient, not the disease' is particularly relevent in the long-term management of chronic IBD. The unknown primary cause (or causes) of IBD and multiple secondary variables produce the heterogenous pathological and clinical manifestations of the illness. A representative list of such factors is shown in Table 2. The clinical features of IBD are numerous and varied. No symptom or sign exists in isolation. A single mechanism or etiopathogenesis rarely explains any one symptom or sign; nearly all are multifactorial. This chapter reviews aspects of the mechanisms, etiopathogenesis and clinical features of the major clinical features of IBD: abdominal pain, malnutrition and diarrhea.

Pain Pain is a subjective experience that cannot be accounted for solely by consideration of disturbances in structure and function. The perception of pain involves interaction between pathophysiologic and psychosocial events. The primary pathophysio-

Table 2. A representative list of variables underlying clinical features in inflammatory bowel disease Demographic factors: age, gender and age of onset Distribution, extent and severity of inflammation Disease 'activity' as evidenced by pain, diarrhea, anorexia etc Anatomical complications such as abscesses, fistulas and obstruction, either mechanical or inflammatory, perianal disease Nutritional status and related factors Extraintestinal manifestations affecting the liver, joints, skin, eye etc Presence of associated diseases Iatrogenic factors, medical and surgical Psychological, family and social factors

logic factor is the nature, intensity and extent of the stimulus that causes pain, but the perception of the painful stimulus is also dependent on the type and number of the receptor involved, on the organisation of the anatomical pathways transmitting the stimuli, and also on the complex role of modifying influences on the transmission, interpretation and reaction to the painful stimuli. Psychosocial factors are of particular importance in chronic conditions such as IBD and include the setting in which the pain occurs, personality factors, and family, ethnic and cultural background. All of these factors by influencing the stoicism of the patient affect the reaction to, and description of, pain and discomfort. The goal of management in patients complaining of pain is to identify and treat the underlying cause. This however is often not possible in patients with chronic IBD. Although a multi-disciplinary approach is necessary to help the patient cope with the physical and psychologic manifestations of chronic pain, the central role in long-term management should be taken by a supportive, caring clinician with an unhurried approach. The propensity to addictive behaviour in IBD patients with chronic pain should not be underestimated. In IBD, pain may arise from non-obstructed inflamed bowel, from bowel distension due to obstruction or from extension of the inflammatory process beyond the bowel wall, from perforation, fistula, or abscess. Temporal aspects, such as its frequency, tempo of onset, duration and its progress over time, often provide a guide to the pathological process causing the pain. Because of the greater extent of inflammation, its transmural distribution, the tendency for stricture formation and obstruction,

Cornelius C. Cronin and Fergus Shanahan and involvement of extraintestinal tissues, abdominal pain is more a feature of CD than UC. However, persistent or severe abdominal pain in patients with UC should alert the clinician to the possibility of the presence of complications such as fulminant colitis. Intermittent, recurrent or chronic pain is a feature of most patients with CD. As a general rule, the degree of pain runs parallel to disease activity, although there are important exceptions. Pain may result from fibrous strictures in the absence of active inflammation. Frequently, irritable bowel syndromelike symptoms of bloating, gaseous distension and sensation of incomplete evacuation after defecation are present. CD involving the upper gastrointestinal tract may cause odynophagia from esophageal involvement, or when the antrum and duodenum are involved, peptic ulcer disease-like symptoms. Typically, patients with CD of the iliocecal region have steady, mild-to-moderate discomfort in the right iUac fossa. A mass may be palpable. Peritoneal signs are absent. Often, there is superimposed colicky pain, referred to the umbilicus, due to obstruction, or in the absence of obstruction, from hyperalgesia. With bacterial overgrowth, bacterial metabolism of carbohydrate to hydrogen and carbon dioxide may cause abdominal pain from gaseous distension. When obstruction supervenes from either a stricture or from intra-luminal material impacting on a narrowed segment, the classical symptoms of post-prandial colicky cramps, often severe, with nausea and vomiting are found. With colonic inflammation in either CD or UC, there may be crampy pain in the iliac fossae. Often, the patient has the sensation that bowel opening might provide relief; a similar sensation may be present with small bowel inflammation but is generally less urgent. Inflammation of the distal colon tends to cause tenesmus and pain in the lower back. Fulminant, diffuse abdominal pain with signs of peritoneal inflammation is suggestive of toxic megacolon. Pain and discomfort from perianal CD is mediated by somatic nerves and is sharply localized. Perianal pain usually implies the presence of sepsis. The extrinsic innervation of the gastrointestinal tract is through the sympathetic and parasympathetic nerves [10, 11]. Although sensory information is carried by both spinal afferent (sympathetic) and vagal (parasympathetic) fibers, pain is mediated primarily by spinal aff'erents. Free nerve ending mediate both physiological and nociceptive inputs. In health, afferent activity in the gastrointestinal tract seldom reaches the level of consciousness.

255 In contrast to skin receptors which are highly specialized and complex structures, visceral receptors do not show any morphologic specialization and consist mainly of relatively simple free nerve endings. The abdominal viscera are relatively insensitive to stimuli, such as cutting, tearing and crushing that when applied to, for example, the skin would evoke severe pain. While stimuli from the skin may give raise to a variety of sensations, the only consciously perceived noxious sensation from the viscera is of pain or discomfort. While a range of appropriate and protective reflexes are available to respond to somatic pain, the repetoire of response to visceral pain is limited and generally ineffectual. Among the stimuli that may give raise to pain in the abdominal viscera are irritation or ulceration of the mucosa or serosa, gross visceral distension, torsion or traction of the mesentery, and forceful c o n t r a c t i o n s , especially when the lumen is obstructed. Receptors are located within the mucosa, submucosa and walls of hollow organs, on the serosal peritoneum and capsules of the solid organs, and in the mesentery. Free nerve endings respond to both mechanical and chemical stimuli. The principal mechanical signal to which visceral nocioceptors respond is stretch. Tension receptors are in series with the smooth muscle of hollow organs. Pain results when intestinal smooth muscle undergoes a sufficient change in tension. Pain receptors may be activated by chemical mediators generated or released during intestinal injury and inflammation, tissue hypoxia or necrosis. Mucosal receptors respond primarily to chemical stimuli. After local tissue injury, the release of inflammatory and chemical mediators such as prostaglandins and bradykinin can directly activate nerve endings, and also trigger the release from afferent nerves and other cells of algesic mediators such as histamine, 5-HT, nerve growth factor (NGF) and prostanoids [12, 13]. This assembly of chemicals sensitizes endings of afferent nerve terminals, resulting in an increased response to painful stimuli, and furthermore to a secondary sensitization of nearby nociceptors, in which neuromediators in the extracellular space, such as substance P, histamine, 5-HT and cytokines play a role. Substance P released from nerve endings in intimate contact with mast cells causes degranulation and further histamine release, in turn amplifying release of substance P and NGF. In chronic inflammation, nerve remodelling occurs, increasing nerve sensitivity and lowering the pain

256

Understanding symptoms and signs in inflammatory bowel disease

Table 3. Embryologic origin of intestinal innervation Intestinal structures

Embryologic origin

Spinal segments

Pain location

Distal esophagus, gastric, proximal duodenum

Foregut

T5-6toT8-9

Epigastric

Small intestine, cecum, ascending and proximal transverse colon

Midgut

T8-11 to L1

Periumbilical

Distal transverse, descending and rectosigmoid colon

Hindgut

T11 toL1

Suprapubic

Adapted from ref. 10.

threshold. Pain may be mediated in part by smooth muscle contraction induced by these mediators. Somatic structures are densely innervated, and somatic pain is distinct and precisely localized. In contrast, visceral pain is ill-defined and poorly localized. The greater part of the extensive enteric nervous system is involved in local physiological control of secretion, absorption and motility, and does not have a role in the mediation of pain perception. There is no separate visceral sensory pathway. Furthermore, there is a relative paucity of visceral afferent nerves that project to the central nervous system. Also, splanchnic nerves from a single tissue or organ enter the spinal cord through several levels, further diluting the accuracy of localisation. Most abdominal structures are embryologically midline, and receive bilateral and symmetrical innervation. Visceral pain generally localizes to the midline. Clearly lateralized pain arising from the abdominal viscera is usually from those few structures whose innervation is predominantly onesided; or from structures with somatic rather than visceral innervation. The parietal peritoneum is innervated by somatic nerves, and when involved in the inflammatory process, pain is more precisely localized. In addition, between individuals, there is considerable variability of innervation, and therefore the same stimulus may be differently localized. The tendency to perceive visceral abdominal pain at a different location from its origin further impairs accurate localisation of visceral pain. In this phenomenon of referred pain, pain is perceived in the skin dermatomes whose afferent nerve roots enter the same levels of the spinal cord as the involved abdominal structure. There are no separate pathways within the spinal cord for visceral and somatic pain. Because somatic afferents are more numerous, the brain tends to associate activation of the second order neurones with a somatic source, irrespective of its origin.

The embryologic origin of abdominal tissues and organs provides a general guide to the localization of gastrointestinal pain [10] (Table 3). Pain arising from foregut structures is generally perceived in the epigastrium, from midgut structures in the periumbilical region, and from the hindgut, suprapubically.

Malnutrition Malnutrition is common in patients with IBD (Table 4) [14, 15]. Nutritional deficiencies are related to the extent and the activity of disease. Because of small bowel involvement, malnutrition tends to be more common in patients with CD. Many of the studies assessing malnutrition have been in hospitalized patients or in those attending specialized clinics, which may tend to overstate its frequency and severity. In addition, with improved medical and surgical treatment and with the appropriate emphasis on ensuring adequate nutrition in patients with IBD, the prevalence of serious malnutrition may have declined in recent years. Nonetheless, malnutrition remains a major problem in those with IBD, particularly in pediatric patients, and is a major contributer to poor quality of life, morbidity and mortality. Malnutrition in patients with IBD is multifactorial (Table 5). Several mechanisms may be involved in any one patient, chief among which however in most patients is curtailment of caloric intake due to any combination of anorexia, nausea, post-prandial abdominal pain and vomiting, and dietary restrictions [16]. The importance of poor caloric intake as a cause of weight loss is best illustrated by the sometimes dramatic effect supplemental enteral feeding has on reduced weight in children with IBD [17-19]. Maldigestion and malabsorption of ingested nutrients is likely to be a factor only in those with extensive and chronically active CD or in those who have had major small bowel excisional surgery.

Cornelius C. Cronin and Fergus Shanahan

257

Table 4. Incidence of malnutrition in patients with Crohn's disease

Table 5. Causes of malnutrition in patients with Crohn's disease

Feature

Percentage

Inadequate intake

Weight loss Growth failure in children Muscle wasting Fat depletion Hypoalbuminemia Iron deficiency Bi2 deficiency Folate deficiency Hypokalemia Hyponatremia Hypomagnesemia Zinc deficiency Fat soluble vitamins: D K A Clinical osteomalacia

40-80 15-88 59

Nausea Anorexia Fear of pain after food Fear of diarrhea after food Restrictive diets

Malabsorption

Loss of absorptive mucosa (disease/ surgery) Stagnant loops/bacterial overgrowth Bile salt deficiency after ileal resection Rapid Gl transit - mucosal disease/ fistulas Lymphangiectasia

Increased requirements/ decreased synthesis

Active inflammation Increased cell turnover Sepsis

Enteric loss of nutrients

Protein-losing enteropathy Interrupted enterohepatic circulation Gastrointestinal blood loss Electrolyte, mineral loss in diarrhea

Drugs

Sulfasalazine Corticosteroids Cholestyramine Immunosuppressives

15-30 25-76 25-50 20-37 13-37 33 10 14-88 40 75 30-50 21 36

Adapted from ref. 15.

Inadequate caloric intake Postprandial symptoms may occasionally be a disincentive to the maintenance of adequate nutritional intake in patients with IBD, particularly the fear of diarrhea or, in patients with stricturing disease, of colicky abdominal pain. Many of the drugs used in the treatment of IBD may themselves cause nausea or anorexia. The use of any nutritionally restrictive dietary therapy is now outmoded. However, the major cause of poor caloric intake is inflammationrelated anorexia and loss of appetite. Weight loss often predates the clinical onset of CD and in children, decreased growth velocity precedes diagnosis [20]. Not infrequently, anorexia and weight loss is the predominant initial feature in IBD. Anorexia, particularly in young patients, may occasionally be wrongly attributed to anorexia nervosa. Rapid weight loss is a feature of intra-abdominal sepsis, which may be occult or, less comonly, of intraabdominal neoplasia.

Physiology of appetite and hunger Anorexia is the diminution or absence of hunger and appetite. Hunger and appetite both refer to the desire to eat, but whereas the determinants of hunger are related to physiological mechanisms present at a given time, appetite is in addition influenced by ongoing pathophysiological, psychological and

environmental processes. Satiety is the gratification of hunger and appetite. The drive to eat is basic and essential to survival. Its control is complex, multifactorial and ill-understood. The generally accepted working model proposes that eating is regulated by a central feeding drive that is held in check during feeding by a peripheral satiety system. Various gastrointestinal, endocrine/metabolic and neurological stimuli are perceived by peripheral and central receptors and, centrally integrated, ultimately produce the sensations of hunger and satiety, and determine eating behaviour. The hypothalamus plays a key role in interpreting and integrating these signals [21]. From animal experiments, the lateral nucleus of the hypothalamus has been proposed as a center for hunger, and the ventromedial nucleus a center for satiety. Whether such concepts of anatomical centers are applicable to humans is unclear; appetite regulation may be more appropriately viewed as the hypothalmic integration of neurotransmittermediated neuropharmacologic systems with opposing effects on hunger/appetite and on satiety.

258

Understanding symptoms and signs in inflammatory bowel disease

Many peptides and neurotransmitters acting centrally and peripherally have been proposed as candidates for control of food intake (Table 6) [22-24]. Dopaminergic tracts appear to be vital in initiating feeding, and norepinephric systems and neuropeptide Y (NPY) and peptide YY also promote feeding. NPY, a 36 amino acid peptide is the most potent orexigenic peptide in the hypothalamus, and appears to be of fundamental importance in the regulation of energy homeostasis [25]. Most anorectic peptides exert their effects through interruption of NPY release or action. Among the satiety factors acting on the hypothalamus, serotonin (5-hydroxytryptamine [5-HT]) plays a central role [26]. Most of the clinically available drugs affecting appetite act on serotonin [27]. Injection of serotonin into specific areas of the hypothalamus inhibits feeding in a variety of species. Plasma free tryptophan is the precursor of serotonin, and tryptophan availability parallels synthesis and release of brain serotonin. Carbohydrate ingestion increases circulating tryptophan, mediated in part by insulin. The consequent increased post-prandial brain serotonin levels is a proposed mechanism by which the macronutrient content of a meal is sensed by the brain, promoting satiety. Other proposed appetite-inhibiting neuropeptides are cholecystokinin, corticotrophin-releasing peptide (CRP), calcitonin gene-related peptide and neurotensin. Several peripheral factors influence appetite, hunger and satiety [22], including the composition of food and not least taste and smell (palatability). Gastric distension is an obvious and important satiety signal, mediated by neural (vagal) and humoral (bombesin) signals, and explains why bulky meals give rise to greater satiety than meals of refined carbohydrates of equal caloric value. Intestinal cholecystokinin (CCK) by sensing intestinal nutrient concentration and the volume and rate of luminal emptying may be an important peripheral regulator of satiety [24]; CCK causes pyloric contraction with resultant gastric distension and binds to the gastric afferent vagus which ultimately courses to the ventromedial hypothalamus. Peripherally released CCK may also exert a central satiety effect. Meal size is reduced by exogenously administered CCK at physiological concentrations, and is increased by specific CCK antagonists, implying that meal size may be physiologically limited by these factors [28, 29]. Ingestion of food presents a large antigen load to the intestine. The substantial amounts of cytokines

that are produced in response may also play a role in the control of feeding. Although these complex neuroendocrine signals may provide a basis for an understanding of how feeding is initiated and terminated, they do not explain how most individuals maintain a stable body weight over long periods of time. The lipostat model hypthesized that signals proportional to the size of body fat stores become integrated with other regulators of food intake [30]. How the size of body fat stores might influence appetite and hunger has been at least in part recently elucidated. Leptin, the protein product of the obese {oh) gene, is secreted into the circulation by adipose cells [31]. A rising circulating leptin (from the Greek leptos, meaning thin) concentration with increasing adiposity is proposed to serve as a negative feedback signal to the hypothalamus. Mice with a mutated oh gene {oh/oh mouse) lack functional leptin and develop extreme Table 6. Neurotransmitters affecting appetite and satiety Stimulatory Norepinephrine GABA Neuropeptide Y Opiods Galanin Growth-hormone-releasing factor Melanin concentrating hormone Orexin a and b CART Inhibitory 5-HT Dopamine Histamine a-Meianocyte-stimulating hormone Glucagon-like peptide 1 Agouti-related protein Neurotensin Cholecystokinin Bombesin Calcitonin-gene related peptide Amylin Adrenomedullin Glucagon Oxytocin Anorectin Thyrotropin-releasing hormone Pituitary adenylate-cyclase activating polypeptide Acidic fibroblast growth factor Interleukin 1 Leptin

Cornelius C. Cronin and Fergus Shanahan obesity at a young age [32]. In humans, plasma leptin concentrations closely correlate with adipose mass. It is proposed that starvation results in a reduction in fat mass and a fall in circulating leptin, promoting appetite and food seeking behaviour, and conversely increasing obesity increases plasma leptin concentration, which acts as a negative signal on the hypothalamus, decreasing appetite. The action of leptin on appetite is independent of its numerous other actions [33]. Insulin as a regulator of adipose tissue mass has many similarities with leptin, including a concentration proportional to adipose tissue mass and receptors in the same key hypothalmic areas. Both when centrally administered reduce food intake and body weight, without producing symptoms of malaise. The hypothalamic effects of leptin and insulin are mediated by a variety of neuropeptides: those that promote feeding and body energy stores include neuropeptide Y and melaninconcentrating hormone, and those that suppress feeding include corticotrophin-releasing factor, the oc-melanocyte-stimulating hormone/melanocortin-4 receptor and the cocaine-amphetamine-regulated transcript [34, 35]. The control of feeding is complex, and has variously been described as multiple overlapping systems, each of which may be employed differently under differing circumstances, or as a cascade with the emphasis on driving feeding modulated by a regulatory feedback signals promoting satiety. There may be separate short-term (5HT) and long-term (leptin) satiety signalling systems [36]. Both modulate the action of NPY, which may function as a common output pathway for the expression of appetite [34].

Anorexia of chronic disease Anorexia accompanies many forms of chronic disease whether due to infection, inflammation, neoplasia or tissue injury. The presence of continuing anorexia despite weight loss implies a failure of the normal compensatory mechanisms which operate after a period of reduced feeding. Increasingly, the anorexia and weight loss of such conditions are attributed to the actions of increased secretion of proinflammatory cytokines [37]. A variety of cytokines inhibits feeding after central or peripheral administration, but IL-1 and TNF-oc have the most profound effects. Administration of TNF and IL-1 induces anorexia and weight loss in experimental animals and in humans [38]. The anor-

259 ectic effect is specific as it is blocked by receptor antagonists or by specific antibodies, and is independent of other systemic effects of these cytokines. Cytokines appear to act by causing a reduction in the body weight set-point, which is then maintained by reduced feeding. In chronic disease, cytokines are produced mainly in the periphery. To inhibit feeding, cytokines or signals derived from their peripheral actions must influence central mechanisms affecting appetite [38]. Cytokine-mediated visceral afferent nerve activity contributes but is not essential to their action as their effects are not abolished by experimental nerve section experiments [38]. Cytokines increase cholecystokinin and leptin release and activity [39, 40]. Circulating cytokines may also be taken up by the central nervous system - cytokines increase blood brain barrier permeability [4]. In addition, there may be increased de novo brain cytokine synthesis [41]. Cytokines have a direct effect on glucose sensitive neurons in the lateral hypothalamus; IL-1 increases central seratonergic activity and other appetite suppressant transmitters such as CRF, histamine and oc-MSH and reduces NPY mRNA levels, all features that promote satiety [42].

Anorexia of IBD Anorexia is the dominant cause of poor nutritional intake and of malnutrition in IBD. The elucidation of the mechanisms is in the early stages of investigation. Most studies thus far have been in experimental models. Classical early experiments showed that the induction of experimental acute colitis by intrarectal administration of trinitrobenzene sulfonic acid (TNBS) causes a 70-80% reduction in food intake, both solid and liquid, during the first three days, which, as the inflammation recedes, reverses within a week [43]. Water consumption and meal frequency are unchanged, but the amount consumed at each meal is reduced, implying that the reduction in intake depends on satiety signals in the brain, and are not due to a state of general malaise. It is likely that in experimental colitis that alterations in multiple pathways regulating food intake are altered. Increased IL-1 activity may play a pivotal role. The pattern of anorexia with TNBS colitis is similar to that of long-term IL-1 administration; intestinal IL-1 concentrations peak when anorexia is maximal; and administration of IL-1 receptor antagonists significantly attenuates anorexia and weight loss [44]. Immediately after induction of

260

Understanding symptoms and signs in inflammatory bowel disease

experimental colitis, reduced food intake is associated with hypothalamic IL-1 receptor activation and increases in prostaglandin synthesis. Serotonin release, a potent inhibitor of feeding, from the hypothalamus is stimulated by IL-1. The peak phase of anorexia induced by TNBS-colitis is associated with increased hypothalmic serotonin release, but again serotonin antagonism has only a partial effect on reversal of anorexia [45], suggesting that other factors are also important. It has been proposed as a mechanism of anorexia and weight loss in chronic IBD that proinflammatory cytokines cause the release of leptin from adipose tissue, giving inappropriately high plasma concentrations for the percentage fat mass. In both animals and human subjects, a number of cytokines, especially tumor necrosis factor (initially termed cachexin), has been shown to increase leptin concentrations [40]. Animal experiments using a variety of methods to induce intestinal inflammation suggest that cytokine-mediated increased circulating leptin during the initial phase of inflammation contributes to the suppression of feeding, and that subsequently as leptin concentration falls, there is an associated and appropriate increase in feeding [46]. There are numerous differences between spontaneously occurring IBD in humans and experimentally induced intestinal inflammation in rodents, including the extent, severity and chronicity of the induced inflammation. Experimental findings in animal models may have limited relevance to human disease. In human IBD, plasma leptin concentration is similar in both patients with active disease and in remission as in control subjects; increased plasma concentrations of soluble TNF-receptor, a marker of intestinal inflammation, are not correlated with plasma leptin [47, 48]. While not excluding a role for disturbed leptin physiology in the causation of intestinal inflammation-related anorexia [35], such mechanisms do not appear to play a similar role in human as in experimental intestinal inflammation. The study of anorexia and satiety in chronic disease is in its infancy. The unravelling of the mechanisms of anorexia in human disease holds the promise of attenuation or amelioration of the malnutrition associated with chronic IBD, with all its attendent benefits.

Malabsorption Malabsorption occurs if mucosal inflammation or resection involves a sufficent length of small bowel,

from the combined eff'ects of lack of sufficient absorptive surface, decreased transit time and diminution of the bile salt pool. Less commonly, other mechanisms may be involved, all a consequence of complications of intestinal inflammation. Each of the three major nutrients - carbohydrate, protein and fat - may be malabsorbed, but clinical symptoms are primarily due to fat malabsorption. In general, malabsorption of carbohydrate and of protein is uncommon and as a factor in decreased energy availability plays only a secondary role to reduced intake. Jejunal function as measured by dxylose absorption is preserved in most patients, except in patients with extensive small bowel CD [18]. The prevalence of lactose malabsorption is probably not increased in patients with IBD. Albumin concentration, often used as a marker of protein status, may be further reduced from intestinal protein loss, reduced hepatic albumin synthesis, malabsorption and anorexia. Fat malabsorption occurs in approximately 30% of patients with CD [15]. The most severe steatorrhea and the greatest malnutrition occurs in those with extensive jejunal disease or resection. Normally, because of its large reserve, up to 50% of the small bowel can be resected without serious impairment of nutrition [49]. In CD, however, much of the bowel may be already diseased so that resection of, or extension of inflammation to, even short lengths may be poorly tolerated. Because the small bowel is shorter in women, they may be more prone to develop such problems. Extensive jejunoileal disease or resection is always associated with malabsorption of fluid, electrolyte, calorie and protein intake, whereas an isolated ileal or jejunal disease resection in isolation may be better tolerated. Transit time through the small bowel is reduced by disease or resection and may occasionally be further compromized by the presence of fistulae, reducing the time available for nutrient absorption. Bile acids are absorbed in the ileum. If the length of functioning ileum is so compromized that inadequate amounts are reabsorbed, hepatic synthesis may be unable to compensate for bile salt loss. If the concentration of conjugated bile acids falls below the micellar concentration critical for fat absorption, steatorrhoea supervenes. Other mechanisms may be involved in malabsorption. Small bowel bacterial overgrowth is a syndrome characterized by nutrient malabsorption associated with excessive numbers of bacteria in the small intestine [50]. Small intestinal peristalsis has

Cornelius C. Cronin and Fergus Shanahan the major role in preventing bacterial overgrowth. The most common cause of bacterial overgrowth in IBD is stasis from small intestinal strictures. Seedling of colonic bacteria and colonization of the small intestine is further facilitated by resection of the iliocecal valve, and less commonly, by fistulous tracts between the large and small intestines. Other potential factors that may promote bacterial overgrowth are advanced age, hypochlorhydria from gastric atrophy or from surgical resection, drugs, and relative immunodeficiency from malnutrition or drug therapy. Bacterial deconjugation of bile acids is the primary mechanism of fat malabsorption in bacterial overgrowth. In florid cases, malabsorption of fat soluble vitamins A, D and E may occur. Generally, such patients are protected from the coagulopathy of vitamin K deficiency by bacterial vitamin synthesis. Bacterial overgrowth results in carbohydrate malabsorption from reduction of brush border disacharidases and decreased uptake of monosaccharides. Hypoproteinemia is common, arising from a number of mechanisms: bacteria compete for protein substrates; bacteria may induce mucosal injury; and decreased intestinal enzyme activity. Vitamin B12 deficiency is a chacteristic finding in bacterial overgrowth. Also, patients with bacterial overgrowth may curtail their nutrient intake to minimize postprandial symptoms from gaseous distension. Protein losing enteropathy may contribute to malnutrition in patients with IBD. Enteric loss of plasma proteins is normally of minor importance in the metabolism of body proteins. Once plasma proteins pass from the circulation into the gastrointestinal tract, they are degraded in the same manner as orally ingested protein into their constituent amino acids and absorbed. Three mechanisms underlie increased intestinal nutrient loss in IBD: increased mucosal permeability; mucosal erosions or ulceration; and lymphatic obstruction. When excessive amounts of proteins are lost into the gastrointestinal tract, synthetic capacity may not compensate. A new steady state is achieved, characterized by hypoproteinemia, in which the daily losses are balanced by synthesis. Such protein depletion primarily affects proteins with a low turnover rate, with consequent hypoalbuminemia and decreased levels of immunoglobulins, fibrinogen, caeruloplasmin and transferrin. Except for dependent edema, these subnormal protein concentrations infrequently have clinical consequences. Iron, trace elements and lipids may also be lost. If lymphatic obstruction is the major mechan-

261 ism in the pathogenesis of protein losing enteropathy, as it may be in CD, lymphocytopenia develops.

Growth retardation Growth retardation represents a specific form of malnutrition in children and adolescents with IBD, particularly in those with CD. Decreased growth velocity often precedes diagnosis [20]. Growth retardation is the presenting symptom in up to a third of young patients with IBD [51] and many others subsequently develop impaired growth. Studies using a variety of definitions of growth impairment based on height for age percentiles have reported frequencies of impaired growth during childhood and adolescence ranging from 13% to 58% in CD patients and from 3% to 21% for UC [52]. Growth retardation may be associated with permanent stunting; as many as 50%) of young adults with childhood onset IBD have height percentiles on or less than the 10th percentile [53]. An attendant problem is delayed menarche and lack of development of secondary sexual characteristics. The final common pathway for growth is the epiphyseal growth plate of long bones. The increase in length of bone is determined by a complex interplay between chondrocyte proliferation, maturation and hypertrophy and matrix synthesis and degradation. Among the factors that may interfere with the process of growth in youngsters with IBD are malnutrition, inflammation and the effects of treatment, particularly corticosteroids [52, 54]. Growth failure in pediatric IBD is primarily due to inadequate caloric intake: estimates of caloric intake have been repeatedly recorded to be between 50%) and 75% of recommendations [17]. Rapid and significant height and weight gain have been shown in these patients when they receive oral or enteral caloric supplementation [17-19]. However, optimal growth is not always attained, suggesting factors other than nutritional deprivation may be involved. Maldigestion and malabsorption play a less significant role except in those who have had extensive small bowel inflammation or resection. Zinc deficiency may be associated with growth retardation and delayed bone age in children [55], and has been proposed as a mediator of growth failure with IBD. Zinc deficiency has been documented in some children with IBD and growth retardation [56, 57], arising from malabsorption or from excessive stool losses. There are however no reports of zinc supple-

262

Understanding symptoms and signs in inflammatory bowel disease

mentation reversing growth retardation in children with IBD. As in the human form, growth retardation is a feature of experimental IBD. However, even when adequate nutritional supplementation is administered, animals with experimental colitis do not achieve optimal growth. It has been estimated that up to 40% of growth retardation in experimental colitis may be attributable to the effects, direct or indirect, of inflammation [58], many mediated by cytokines. A prerequisite for normal linear growth is a functioning growth hormone/insulin-like growth factor-1 axis. Pituitary growth hormone stimuates production of insulin-like growth factor-1 (IGF-1) from the hver, from which most of the circulating IGF-1 is derived, which in turn acts on the epiphyseal growth plate. Young patients with chronic inflammatory states associated with growth retardation (including IBD) have normal growth hormone secretion but suppressed plasma IGF-1 concentrations [59 61]. In animal models of intestinal inflammation, exogenous administration of IGF-1 significantly attenuates growth retardation [58]. The deleterious effects of proinflammatory cytokines on growth may be mediated through interruption of the growth hormone/IGF-1 axis. Transgenic mice overexpressing either IL-1 or TNF have growth retardation [62-64]. Excessive IL-6 activity either by genetic manipulation or by exogenous administration [62] is associated with normal growth hormone levels and with suppression of IGF-1 concentrations, the same pattern as in pediatric IBD. Administration of TNF and IL-1 P inhibits IGF-1 production from rat hepatocytes [65]. Immunoneutralization of IL-6 in animal models of intestinal inflammation increases both IGF-1 concentrations and linear growth [58]. In addition to the indirect endocrine-mediated effects of cytokines on bone growth, cytokines may have direct negative effects on the epiphyseal growth plate [64]. Since nutritional supplementation does not fully reverse growth retardation in children with IBD, it may that the achievement of optimal growth may be obtained only by both adequate nutritional and antiinflammatory therapy. Corticosteroids used in supraphysiological dosages retard growth in children with a variety of conditions [66]. The use of high dosage steroid therapy is closely linked to disease and inflammatory activity, and it is difficult to separate the effect of one from the other. Dose, type and timing of glucocorticoid exposure each influence the degree of growth

suppression observed. Pre-pubertal children exposed to glucocorticoids may be particularly susceptible. 66 Resumption of growth follows discontinution of steroid therapy. However, chronic use of glucocorticoids has a significant effect on final height attained. The pathogenesis of growth suppression by glucocorticoids is complex and multifactorial, involving several steps in the cascade of events leading to linear growth at the epiphyseal growth plate, including the growth hormone/IGF-1 axis [67, 68]. Precisely which effects are most responsible for growth suppression remains unclear (Table 7).

Table 7. Effects of glucocorticoids on linear growth On bone Interfere with the nitrogen and mineral retention required for the growth process Inhibition of chondrocyte mitosis Inhibition of collagen synthesis Inhibition of osteoclastic activity On calcium metabolism Decrease intestinal calcium absorption Increase urinary calcium excretion Secondary hyperparathyroidism On endocrine function Antagonism of growth hormone Reduces growth hormone receptor expression Reduces plasma growth hormone-binding protein Suppression of IGF-1 activity Suppression of sex steroid secretion

Diarrhea Normal stool frequency varies from between three times a day to three times a week, with formed consistency and a volume of about 100 ml per day. Patients tend to report any increase in their usual frequency, fluidity or volume as diarrhea. Diarrhea is formally defined as an increase in daily stool weight to above 200 g per day. It is important that when a complaint of diarrhea is made that the patient is questioned and the symptom analysed. Diarrhea must be distingished from hyperdefecation, from urgency and from fecal incontinence. Alterations from the usual bowel habit that are reported as diarrhea but do not fulfill formal diagnostic criteria are still worthy of evaluation and assessment.

Cornelius C. Cronin and Fergus Shanahan The cardinal symptom of IBD and nearly always present in those with active disease, diarrhea varies from being a nuisance to being life-threatening. Diarrhea is related to the severity and extent of intestinal inflammation, and may be used as a marker of activity. Ulcerative colitis severe enough to cause diarrhea almost always contains blood. If blood is absent, another diagnosis is likely. Patients with colitis typically complain of frequent, small volume stools. If inflammation is confined to the rectum, fresh blood, or blood on the surface of often otherwise normal appearing stool, or blood stained mucus may be passed. Urgency, tenesmus and a sensation of incomplete evacuation are common, and incontinence may occur. Occasional patients with proctitis may complain of constipation. With more extensive colitis, diarrhea is grossly bloody, with pus and fecal matter, sometimes likened to anchovy sauce. Passing clots is unusual and suggests another diagnosis. Patterns of diarrhea vary in CD, depending on the anatomical distribution. With colonic or rectal involvement, symptoms are similar to those of UC involving the same areas. In disease confined to the small intestine, stools are larger in volume and are not generally associated with tenesmus and urgency and are free of blood. If steatorrhea is present, stools are typically greasy, pale, malodorous and difficult to flush away. Fluidity of the intestinal contents is essential for digestion, for diffusion of digested nutrients to the absorptive epithelial surface, and for absorption itself. Being critical for a variety of processes, the amount of water in the intestine is closely regulated. The daily intestinal fluid load approximates to 9000 ml. Oral intake is generally about 2000 ml, with saliva, gastric juice, bile, and pancreatic and intestinal secretion accounting for the remainder (Table 8). Of the 9000 ml that enters the duodenum every day, about 7000 ml is absorbed in the small intestine, with 1500-2000 ml reaching the ihocaecal valve. Most of this volume is then reabsorbed in the colon, with the daily stool containing about 100 ml of water [69] (Table 9). The absorptive capacity of the normal small bowel has not been defined, but presumably is considerably greater than the 7000 ml absorbed on average every day. Under experimental conditions, the normal colon has an absorptive capacity of up to 5000 ml per day, but under real life conditions is probably less. The intestine has a tremendous capacity for both water secretion and absorption. Diarrhea is ulti-

263 Table 8. Daily fluid load entering duodenum Origin

ml/24 h

Oral intal<e Saliva Gastric juice Bile Pancreatic juice Intestinal secretion

2000 1500 2500 500 1500 1000

Total

9000

Table 9. Intestinal water and electrolyte distribution Water (ml/24 h)

Na (mmol/24h)

K (mmol/24 h)

CI (mmol/24 h)

Entering duodenum

80-10 000

800

100

700

1500

200

10

100

100

3

8

2

Entering colon Stool

Adapted from ref. 69.

mately an expression of deranged intestinal water regulation and may result from increased secretion or reduced absorption or both, originating in either the small or large intestine or both. If small intestinal function is deranged so that its absorptive capacity is reduced, the volume of fluid presented to the colon may exceed that organ's reabsorptive capacity. If the function of the colon is deranged so that it cannot absorb the normal load of 2000 ml presented to it every day, diarrhea will also result. The intestines also have a great potential for secretion of fluid, and abnormal secretion in either the small or large intestine, by overwhelming absorptive capacity, will also result in diarrhea. More commonly in IBD, both functions of the bowel are impaired, and diarrhea results from a combination of impaired absorption and increased secretion, each defect amplifying the effect of the other. The intestinal epithelium is composed of a single layer of columnar cells, serving as a barrier, absorbing water, electrolytes, and nutrients and secreting water and electrolytes [70]. By convolutions of folds, villi and microvilli, the small intestine has a surface area of greater than 200 m^, with a tremendous reserve for both absorption and secretion. Water moves passively to maintain isoosmolarity between

264

Understanding symptoms and signs in inflammatory bowel disease

lumen and tissue. Electrolytes are actively transported, and their movement plays a pivotal role in water regulation. Absorptive and secretory functions are regulated by both intracellular and intercellular mechanisms by an array of endogenous and exogenous factors. Small intracellular molecules act as second messengers, which by controlling the activity of protein kinases, regulate activity of various transport pathways. Intercellular regulatory mechanisms include hormones from endocrine cells, neurotransmitters from nerve cells and mediators released by immune cells [70].

Table 10. InflammatJon and diarrhea in inflammatory bowel disease Decreased absorption of water and electrolytes: Reduced surface area - mucosal injury/resection Neural/hormonal inhibition of NaCI absorption Rapid intestinal transit - increased motor activity Increased osmolar load from nutrient malabsorption Increased secretion of water and electrolytes: Inflammation-derived secretagogues Loss of mucosal barrier function Denuded mucosa leaking plasma and blood

Table 11. Representative list of Intestinal absorptive and secretory stimuli Absorptive stimuli

Secretory stimuli

Endogenous

a-Adrenergic agonists Dopamine Enkephalins and other opioids Somatostatin Glucocorticoids Angiotensin Peptide YY Neuropeptide Y Prolactin

Acetylcholine Histamine 5-Hydroxytryptamine Substance P Cholecystokinin Prostaglandins Bradykinin Vasoactive intestinal polypeptide Secretin Adenosine Gastrin Motilin

Exogenous

Nutrients: glucose, amino acids

Bacterial enterotoxins Bile salts

Adapted from ref. 70.

Table 12. Causes and mechanisms of diarrhea in inflammatory bowel disease Consideration

Mechanism

Treatment

Bacterial overgrowth Bile acid diarrhea Bile acid deficiency Lactase deficiency Short bowel Internal fistula Antibiotic related

Deconjugation Secretion Steatorrhea Osmotic Malabsorption Bypass C. difficile

Antibiotics Cholestyramine Low fat diet Avoid lactose Low fat diet Surgery Stop antibiotics

Cornelius C. Cronin and Fergus Shanahan

Diarrhea in IBD The physiology of fluid and electrolyte transport in the healthy individual is complex and the relative importance of the various regulatory processes is unclear. When diarrhea results from intestinal inflammation, complexity is increased by an order of magnitude. Diarrhea may be related to the inflammatory process or to specific comphcations, usually the consequence of inflammation. Pathogenetic processes vary from patient to patient and several mechanisms may be operative in the individual patient.

Inflammation-related diarrhea Inflammation in the intestinal tract may give raise to diarrhea by inhibiting absorption and/or increasing secretion (Table 10). Loss of absorptive surface area is the major factor in diarrhea in IBD. Diarrhea results when mucosal injury or resection is such that the absorptive capacity of the intestine is diminished or overwhelmed. A major mechanism is inhibition of electrolyte and water absorption by the products of immune and inflammatory processes, mediated by direct effects on enterocytes and by indirect effects on cellular elements in the lamina propria. Immune and inflammatory mediators initiate a cascade of biochemical events that inhibit absorption of sodium and chloride. Inhibition of absorption may also occur in areas where mucosal integrity and structure appear to be intact. In CD, if there is extensive small bowel involvement, malabsorbed and maldigested nutrients increase the osmolarity of the intestinal fluid, exacerbating diarrhea through an osmotic process. Rapid intestinal transit decreases the time available for absorption to occur. Increased anion secretion is the other major cause of diarrhea in IBD. An ever increasing Hst of immune and inflammatory mediators have a role in increasing intestinal secretion [70] (Table 11). These mediators directly affect the epithelium, but also stimulate the enteric nervous system. Diarrhea is exacerbated by exudation of plasma-like fluid and blood from denuded epithelium.

265 terminal ileum and the bile acid pool is recirculated six times each day - the enterohepatic circulation. In patients with terminal ileal disease or resection, dihydroxy bile salts are not fully absorbed. If a sufficient concentration reaches the colon, net ffuid and electrolyte secretion occurs through multiple mechanisms and watery diarrhea ensues. If there is more extensive terminal ileal inflammation or resection ( > 1 0 0 cm), bile salt loss may be such that hepatic synthesis cannot compensate; bile salt concentration falls below the level critical for adequate fat absorption and steatorrhea ensues. Bacterial overgrowth may, as outlined above, contribute to malabsorption and weight loss in CD. Diarrhea results through several mechanisms. Bacterial deconjugation of bile acids and consequent malabsorption of fat is the primary mechanism of diarrhea and steatorrhea. Also, bacterial metabolites may have a toxic effect on the small bowel mucosa, acting as secretagogues. Bacterial overgrowth reduces brush border disaccharidases and the increased delivery of osmotically active carbohydrate fragments to the small intestine further exacerbates diarrhea. In patients with extensive small bowel resection, diarrhea develops as part of the short bowel syndrome. The major mechanism is loss of absorptive surface leading to malabsorption of water, electrolytes and nutrients. Several other factors may be involved: decreased transit time; an osmotic induced diarrhea due to malabsorption of lactose and other carbohydrates; concomitant bacterial overgrowth; and bile acid induced diarrhea from ileal resection. That the diarrhea persists when fasting implies a secretory component, possibly due to absence of signals from the distal small bowel that inhibit proximal intestinal secretion.

References 1. 2. 3.

Mechanisms other than inflammation-related Indirect consequences of inflammation may give rise to diarrhea (Table 12). Conjugated bile acids are required for efficient digestion and absorption of dietary fat. Bile acids are actively reabsorbed in the

4.

Munkholm P, Langholz E, Davidsen M, Binder V. Disease activity courses in a regional cohort of Crohn's disease patients. Scand J Gastroenterol 1995; 30: 699-706. Curfs JH, Meis JF, Hoogkamf-Korstanje JA. A primer on cytokines: sources, receptors, effects and inducers. Clin Microbiol Rev 1997; 10: 742-80. Lucey DR, Clerici M, Shearer GM. Type 1 and type 2 cytokine dysregulation in human infectious, neoplastic and inflammatory diseases. Clin Microbiol Rev 1996; 9: 532-62. Rivest S, Lacroix S, Vallieres L, Nadeau S, Zhang J, Laflamme N. How the blood talks to the brain parenchyma and the paraventricular nucleus of the hypothalamus during systemic inflammatory and infectious stimuli. Proc Soc Exp Biol Med 2000; 223: 22-38.

266 5.

6. 7. 8. 9.

10.

11. 12. 13.

14. 15.

16.

17.

18.

19.

20. 21. 22. 23. 24. 25. 26. 27.

Understanding symptoms and signs in inflammatory bowel disease Faggioni R, Benigni F, Ghezzi P. Proinflammatory cytokines as pathogenetic mediators in the central nervous system: brain-periphery connections. Neuroimmunomodulation 1995;2:2-15. Slifka MK, Whitton JL. Clinical implications of dysregulated cytokine production. J Mol Med 2000; 78: 74-80. Kotler DR Cachexia. Ann Intern Med 2000; 133: 622-34. Engel GL. The need for a new medical model: a challenge for biomedicine. Science 1977; 196: 129-46. Drossman DA. Psychosocial factors in gastrointestinal disorders. In: Feldman M, Scharschmidt BF, Sleisenger MH, eds. Gastrointestinal and Liver Disease, 6th ed. Philadelphia: W.B. Saunders, 1998: 69-79. Klein KB. Approach to the patient with abdominal pain. In: Yamada T, Alpers DH, Owyang C, Powell DW, Silverstein FE, eds. Textbook of Gastroenterology. Philadelphia: J.B. Lippincott, 1995:750 71. Ccrvero F, Laird JM. Visceral pain. Lancet 1999; 353: 2145 8. Bucno L, Fioramonti J, Delvaux M, Frexinos J. Mediators and pharmacology of visceral sensitivity: from basic to clinical implications. Gastroenterology 1997; 112: 1714 43. Bueno L, Fioramonti J, Garcia-Villar R. Pathobiology of visceral pain: molecular mechanisms and therapeutic implications HI visceral afferent pathways: a source of new therapeutic targets for abdominal pain. Am J Physiol Gastrointest Liver Physiol 2000; 278: G670 6. Stokes MA. Crohn's disease and nutrition. Br J Surg 1992; 79: 391 4. O'Keefc SJD, Rosser BG. Nutrition and inflammatory bowel disease. In: Targan SR, Shanahan F, eds. Inflammatory Bowel Disease: From Bench to Bedside. Baltimore: Williams&Wilkins, 1994:461 77. Rigaud D, Angel LA, Cerf M ct al. Mechanisms of decreased food intake during weight loss in adult Crohn's disease patients without obvious malabsorption. Am J Clin Nutr 1994; 60: 775 81. Kirschncr BS, Klich JR, Kalman SS. Reversal of growth retardation in Crohn's disease with therapy emphasizing oral nutritional restitution. Gastroenterology 1981; 80: 10 15. Kelts DG, Grand RJ, Shen G, Watkins JB, Werlin SL, Boehme C. Nutritional basis of growth failure in children and adolescents with Crohn's disease. Gastroenterology 1979; 76: 720 7. Aiges H, Markowitz J, Rosa J, Daum F. Home nocturnal supplemental nasogastric feedings in growth-retarded adolescents with Crohn's disease. Gastroenterology 1989; 97: 905-10. Kanof ME, Lake AM, Bayless TM. Decreased height velocity in children and adolescents before the diagnosis of Crohn's disease. Gastroenterology 1988; 95: 1523-7. Lawrence CB, Turnbull AV, Rothwell NJ. Hypothalamic control of feeding. Curr Opin Neurobiol 1999; 9: 778-83. Stubbs RJ. Peripheral signals affecting food intake. Nutrition 1999; 15:614-25. Woods SC, Seeley RJ, Porte D, Schwartz MW. Signals that regulate food intake and energy homeostasis. Science 1998; 280:1378-83. Bray GA. Afferent signals regulating food intake. Proc Nutr Soc 2000; 59:373-84. Inui A. Neuropeptide Y: a key molecule in anorexia and cachexia in wasting disorders. Mol Med Today 1999; 5: 7985. Curzon G. Serotonin and appetite. Ann NY Acad Sci 1990; 600: 521-30. Halford JC, Blundell JE. Pharmacology of appetite suppression. Prog Drug Res 2000; 54: 25-58.

28. 29.

30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

41. 42.

43.

44. 45. 46.

47.

48.

49.

Lieverse RJ, Jansen JBMJ, Masclee AAM, Lamers CBHW. Satiety effects of cholecystokinin in humans. Gastroenterology 1994;106:1451-4. Ballinger A, McLoughlin L, Medbak S, Clark M. Cholecystokinin is a satiety hormone in humans at physiological post-prandial plasma concentrations. Clin Sci (Colch) 1995;89:375-81. Kennedy GC. The role of depot fat in the hypothalamic control of food intake in the rat. Proc R Soc Lond 1953; 140: 578-92. Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature 1998; 395: 763-70. Zhang Y, Proenca R, Maffel M, Barone M, Leopold L, Friedman J. Positional cloning of the mouse obese gene and its human homologue. Nature 1994; 372: 425-32. Fruhbeck G, Jebb SA, Prentice AM. Leptin: physiology and pathophysiology. Clin Physiol 1998; 18: 399-419. Inui A. Feeding and body-weight regulation by hypothalamic neuropeptides mediation of the actions of leptin. Trends Neurosci 1999; 22: 62 7. Ballinger A, Farthing M. Intestinal disease, feeding behaviour and leptin role of the gut-brain axis. CNS 1999; 2: 12 16. Halford JC, Blundell JE. Separate systems for serotonin and leptin in appetite control. Ann Med 2000; 32: 222 32. Plata-Salaman CR. Cytokine-induced anorexia: behavioral, cellular and molecular mechanisms. Ann NY Acad Sci 1998;856:160 70. Langhans W, Hrupka B. Interlcukins and tumor necrosis factor as inhibitors of food intake. Neuropeptides 1999; 33: 415 24. Daun JM, McCarthy DO. The role of cholecystokinin in interleukin-l induced anorexia. Physiol Behav 1993; 54: 237 41. Finck BN, Kclley KW, Dantzcr R, Johnson RW. In vivo and in vitro evidence for the involvement of tumor necrosis factor alpha in the induction of leptin by lipopolysaccharide. Endocrinology 1998; 139: 2278 83. Laye S, Gheusi G, Cremona S ct al. Endogenous brain IL-l mediates LPS-induced anorexia and hypothalamic cytokine expression. Am J Physiol 2000; 279: R93-8. Laviano A, Cangiano C, Fava A, Muscaritoli M, Mulieri G, Fanelli FR. Peripherally injected IL-l induces anorexia and increases brain tryptophan concentrations. Adv Exp Med Biol 1999; 467: 105 8. McHugh KJ, Weingarten HP, Keenan C, Wallace JL, Collins SM. On the suppression of food intake in experimental models of colitis in the rat. Am J Physiol 1993; 264: R871-6. McHugh KJ, Collins SM, Weingarten HR Central interleukin-l receptors contribute to suppression of feeding after acute colitis in the rat. Am J Physiol 1994; 266: R1659-63. Ballinger A, El-Haj T, Perrett D et al. The role of medial hypothalamic serotonin in the suppression of feeding in a rat model of colitis. Gastroenterology 2000; 118: 544-53. Barbier M, Cherbut C, Aube AC, Blottiere HM, Galmiche JP. Elevated plasma leptin concentrations in early stages of experimental intestinal inflammation in rats. Gut 1998; 43: 783-90. Ballinger A, Kelly P, Hallyburton E, Besser R, Farthing M. Plasma leptin in chronic inflammatory bowel disease and HIV: implications for the pathogenesis of anorexia and weight loss. Clin Sci (Colch) 1998; 94: 479-83. Hoppin AG, Kaplan LM, Zurakowski D, Leichtner A, Bousvaros A. Serum leptin in children and young adults with inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1998; 26: 500-5. WiUiamson RC. Intestinal adaptation. N Engl J Med 1978; 298: 1393^02.

Cornelius C. Cronin and Fergus Shanahan 50.

51. 52. 53. 54. 55. 56. 57.

58.

59.

60. 61.

Li E. Bacterial overgrowth. In: Yamada T, Alpers DH, Owyang C, Powell DW, Silverstein FE, eds. Textbook of Gastroenterology. Philadelphia: J.B. Lippincott, 1995: 1673-80. Kanof ME, Lake AM, Bayless TM. Decreased height velocity in children and adolescents before the diagnosis of Crohn's disease. Gastroenterology 1988; 95: 10-15. Markowitz J, Daum F. Growth impairment in pediatric inflammatory bowel disease. Am J Gastroenterol 1994; 89: 319-26. Kirschner BS. Growth and development in chronic inflammatory bowel disease. Acta Pediatr Scand 1990; 366(Suppl): 98-104. Ballinger AB, Camacho-Hubner C, Croft NM. Growth failure and intestinal inflammation. Q J Med 2001; 94: 121-5. Solomans NW, Rosenfield RL, Jacobs RA. Growth retardation and zinc deficiency. Pediatr Res 1976; 10: 923-7. Sturniolo GC, Molokhia MM, Shields R, et al . Zinc absorption in Crohn's disease. Gut 1980; 21: 387-3-1. Nishi Y, Lifshitz F, Bayne MA et al. Zinc status and its relation to growth retardation in children with chronic inflammatory bowel disease. Am J Clin Nutr 1980; 33: 2613-21. Ballinger AB, Azooz O, El-Haj T, Poole S, Farthing MJG. Growth failure occurs through a decrease in insuHn-like growth factor 1 which is independent of undernutrition in a rat model of coHtis. Gut 2000; 46: 694-700. Beattie RM, Camacho-Hubner C, Wacharasindhu S, Cotterill AM, Walker-Smith JA, Savage MO. Responsiveness of IGF-1 and IGFBP-3 to therapeutic intervention in adolescents with Crohn's disease. Clin Endocrinol 1998; 49: 483-9. Tenore A, Berman WF, Parks IS, Bongiovannie AM. Basal and stimulated GH concentrations in inflammatory bowel disease. J Clin Endocrinol Metab 1977; 44: 622-8. Braegger CP, Torresani T, Murch SH, Savage MO, WalkerSmith JA, MacDonald TT. Urinary growth hormone in growth-impaired children with chronic inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1993; 16: 49-52.

267 62.

63.

64.

65.

66. 67.

68. 69.

70.

De Benedetti F, Alonzi T, Moretta A et al. Interleukin-6 causes growth impairment in transgenic mice through a decrease in insulin-like growth factor-1. J Clin Invest 1997; 99: 643-50. Siagel SA, Sheahy DJ, Nakada MX. The mouse human chimeric monoclonal antibody CA2 neutralizes TNF in vitro and protects transgenic mice from cachexia and T N F lethality in vivo. Cytokine 1995; 7: 15-25. Enomoto M, Pan HO, Kinoshita A, Yutani Y, Suzuki F, Takigawa M. Effects of tumor necrosis factor alpha on proliferation and expression of differentiated phenotypes in rabbit costal chondrocytes in culture. Calcif Tissue Int 1990; 47: 145-51. Wolf M, Bohm S, Brand M, Kreymann C. Pro-inflammatory cytokines interleukin 1 beta and tumor necrosis factor alpha inhibit growth hormone stimulation of insulin-like growth factor-1 synthesis and growth hormone receptor mRNA levels in cultured rat liver cells. Eur J Endocrinol 1996; 35: 729-37. Allen DB. Growth suppression by glucocorticoid therapy. Endocrinol Metabol Clin N Am 1996; 25: 699-717. Guistina A, Wehrenberg WB. The role of glucocorticoids in the regulation of growth hormone secretion - mechanisms and clinical significance. Trends Endocrinol Metab 1992; 3: 306-11. Untermann TG, Phillips LS. Glucocorticoid eff"ects on somatomedins and somatomedin inhibitors. J Clin Endocrinol Metab 1985; 61: 618-26. Powell DW. Approach to the patient with diarrhea. In: Yamada T, Alpers DH, Owyang C, Powell DW, Silverstein FE, eds. Textbook of Gastroenterology, 2nd edn. Philadelphia: J.B. Lippincott, 1995: 813-63. Kaunitz DK, Barrett KE, McRoberts JA. Electrolyte secretion and absorption: small intestine and colon. In: Yamada T, Alpers DH, Owyang C, Powell DW, Silverstein FE, eds. Textbook of Gastroenterology, 2nd edn. Philadelphia: J.B. Lippincott, 1995: 326-61.

14 Clinical course and complications of ulcerative colitis and ulcerative proctitis REMO PANACCIONE AND LLOYD R. SUTHERLAND

Introduction Ulcerative colitis (UC) is a chronic inflammatory condition of the colon that may affect individuals of any age. It generally begins at the anus and extends a variable length from the rectum in a continuous fashion. Patients usually present with a constellation of symptoms including diarrhea (often bloody) and associated with urgency, lower abdominal cramping, and tenesmus. Extensive involvement of the colon is often accompanied by systemic symptoms of anorexia, fatigue, low-grade fever, and weight loss. Knowledge of the natural history of UC and its associated complications forms the cornerstone upon which clinicians can build techniques for the proper education and clinical management of patients with the disease. Questions of interest to most patients include those pertaining to etiology, prognosis, dietary issues, medical and surgical therapy, cancer risk, and genetics. Fortunately, several studies regarding the clinical course of UC and ulcerative proctitis are available to aid the clinician in answering these questions. This chapter will review and highlight the existing literature in UC which will form a foundation upon which clinicians may answer patient inquiries and base management decisions.

Bias An important concept to understand before reviewing the relevant literature is the problem of bias (Table 1). Decisions regarding patient care are often based on studies that examine patient groups that may have little relevance to physicians in smaller communities. Bias may exist in several forms; however, for the purposes of this discussion, referral, temporal, therapeutic, technological, and statistical biases will be taken into account and examined. A

firm understanding of bias will allow the physician to review the studies and decide which are most relevant to his or her patients.

Referral bias Although patients may have similar diseases, patient populations may differ greatly depending on their source. Differences in population base may alter the clinical course predicted in these different studies. The physician should be aware of the population that most resembles his or her own patient population. Problems in reviewing the literature concerning outcome have been discussed elsewhere [1]. Most of the available data are derived from two major sources, either referral or population-based, and the degree of bias which influences the generalizability of this information varies depending on the source. The first source of data (referral) is reported by pioneers in inflammatory bowel disease. These studies often reflect the experience of a small number of individuals working in large inflammatory bowel disease referral centers and include series by Kirsner [2], the Mt. Sinai (NY) group, [3], Truelove [4], Lennard-Jones [5, 6] and others [7-9]. The patient population is often biased toward more difficult cases with unusual or rare presentations and complications. Due to the nature of the patient population these series tend to represent the worst-case scenarT a b l e l Sources of bias Referral Temporal Therapeutic Technological

Statistical

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 269-290. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Ulcerative colitis and ulcerative proctitis

270 ios with the worst dinical outcomes. These studies will be referred to as non-population-based or referral studies. The second source of data is derived from population-based studies. These are studies which include all patients within a geographically defined area [1017]. The clinical outcomes in these studies are often more favorable for several reasons. These populations are not 'contaminated' by excessive numbers of patients with refractory disease who need to be seen by an expert in the field, or by the cases with the unusual presentations or rare complications that gravitate toward tertiary centers. Secondly, the mix of disease tends to be more uniform with a higher proportion of patients with disease limited to the rectum or rectosigmoid. Lastly, follow-up in population-based studies tends to be more complete because these studies originate in countries where the burden of health-care cost is not the responsibility of the patient, and because patients tend to be less mobile and are available for more complete follow-up. An example of a population-based study comes from Copenhagen where the vast majority (99%) of patients with UC have been followed at one center since 1962 [16). This center, the Herlev Hospital, serves approximately 10% of the Danish population and is staffed by a small group of gastroenterologists with a dedicated interest in UC. Patient management is quite similar to what would be considered standard gastroenterological practice. Patients with UC are on 5-ASA agents for maintenance of remission purposes. Disease flares are treated with oral or topical corticosteroids, and surgical management with colectomy is reserved for patients who fail medical management. In their latest report [16] the long-term prognosis of 1161 patients is presented. Of the 1161 patients, 44% had rectal involvement, 18% had pancolitis, and the remaining 36%) had disease between these extremes. In approximately 2% the localization of disease extent was incomplete. Follow-up was complete and ranged from 1 to 26 years. Patients were assessed annually and the median duration of follow-up was 12 years. This type of data most likely represents the practice of most gastroenterologists. An example of the differences in prognosis between referred as compared to community-based patients is illustrated in the report of Truelove and Pena [18]. Their study describes the course and prognosis of patients with Crohn's disease (CD) seen at the Radcliffe Infirmary between 1938 and 1970. Of

the 221 patients, 55 (25%) were already diagnosed with CD prior to assessment at the Radcliffe Infirmary; the remainder were patients who lived in the Oxford area. The cumulative mortality for this group utilizing actuarial methods was 21.4% at 15 years, which is much higher than the age-adjusted expected mortality rate of 8.4%. However, when the data are analyzed separately for referred cases and cases within the health district significant differences in the observed mortality are seen (Fig. 1). Differences in disease extent and severity between population-based and non-population-based studies are shown in Figs 2 and 3. The referral center at Oxford [4] has a higher proportion of patients with pancolitis (Fig. 2) and severe disease activity compared to the population-based studies (Fig. 3).There-

Survival compared with age-specific mortality Truelove and Pena (1976)

100

Q.

•

Observed (New cases)

A Expected (New cases) •

Observed (referred)

•

Expected

(referred) — I —

10

15

Survival (Years after diagnosis) Figure 1. Comparison of life expectancy for patients specifically referred to the Radcliffe Infirmary for treatment of their Crohn's disease compared with those drawn from the local health unit. The life expectancy for the referred patients differs markedly from that of the general population. The difference in mortality for the patients drawn from the local unit compared with the general population is not as marked [18].

I Distal i Substantial Pancolitis Oxford

Scotland

Figure 2. Comparison of the proportion of patients with pancolitis between a population-based study (Scotland) and a referral-based study (Oxford). The Oxford study contains a greater proportion of patients with extensive disease.

Remo Panaccione and Lloyd R. Sutherland

271 • MM H Moderate ki Severe

Oxford

Scotland

Figure 3. Comparison of the proportion of patients with disease activity characterized as severe in a populationbased study (Scotland) and a referral-based study (Oxford). The Oxford study contains a greater proportion of patients characterized as having severe disease.

fore, one can anticipate that outcomes in the Oxford Group would be worse than that of the Scottish group. Information drawn from non-population-based studies may assist in management decisions for the difficult patient, and better define the range of complications that patients may have. However, information on outcomes is generally more pessimistic and will not aid in counseling individual patients in terms of prognosis.

Temporal bias Many of the early non-population-based studies which outline the natural course of UC span many decades [3, 4]. During that time advances have occurred in pharmacotherapy, surgical innovation, and postoperative care. If these advances are not taken into account inappropriate conclusions regarding new interventions may be drawn. These early studies are also influenced by referral bias as they originate in centers that are highly specialized and recognized for the excellent care of patients with inflammatory bowel disease.

same is true regarding sulfasalazine. Although sulfasalazine was available in the United Kingdom as of 1955, a review of the fatalities in the Radcliff'e Infirmary study (1953-1962) demonstrates that seven of the 13 were not on any specific therapy at the time of admission [4]. Similarly, although sulfasalazine was shown to be effective for the maintenance of remission in 1965 [20], many of the early long-term prognosis studies included patients who were not on maintenance therapy with a 5-aminosalicylate (5-ASA) agent. The impression that UC is less severe than originally reported may be due to the shift in belief that all patients should be on maintenance therapy [21].

Technological bias Technological advances have allowed for better characterization of disease severity and extent. Endoscopic evaluation was not available until the late 1960s. Prior to this, extent of disease was based on the findings of rigid sigmoidoscopy combined with a single-contrast enema. Single-contrast studies do not provide the mucosal detail of the air-contrast technique. Disease at presentation and during follow-up may have been underestimated. This is demonstrated in the Ritchie study in which only one of the 269 patients seen between 1969 and 1975 had undergone a full colonoscopic examination [5].

Statistical bias The technique of life-table analysis, which adjusts for variable lengths of follow-up, has greatly improved our understanding of the natural history of disease. This technique was not used extensively until the 1970s. Data analyzed prior to this in the original natural history studies will therefore contain more limited information.

Therapeutic bias Therapeutic bias is an important factor to consider when assessing outcomes. Over time there are changes in what is considered to be the standard of care. This standard of care may differ when comparing community-based practices to highly specialized centers. For example, the utility of corticosteroids in the treatment of UC was reported in 1952 [19]. However, corticosteroids did not enter general use until 1958 [4]. This may have arisen due to lack of agreement on patient selection for therapy and controversy regarding dose and duration of therapy. The

Presentation of ulcerative colitis and proctititis Patients with UC present with a variety of symptoms depending on the disease extent and severity. However, it has also been shown that patients may have endoscopic and histological evidence of active disease but be symptom-free. The history and physical examination may assist the physician in differentiating between disease limited to the rectum and more

Ulcerative colitis and ulcerative proctitis

272 extensive disease. Laboratory evaluation at the time of presentation may be useful in defining severity as well as ruling out other causes of colitis (especially infectious colitis) that may mimic the clinical picture of active UC. Endoscopic evaluation has aided in accurately defining disease extent as well as differentiating between UC and CD in the majority of cases. Despite an increasing effort to educate physicians regarding inflammatory bowel disease the mean and median intervals between onset of symptoms and diagnosis are 1.8 years and 5 months respectively [12].

Presenting symptoms In the earlier published population-based Copenhagen series of 783 patients 40% had proctitis and 16% had pancolitis, with the remainder having disease extent between these two extremes. In this patient population 75% of the individuals presented with diarrhea and bleeding while 15% had blood-perrectum alone [22]. The remaining 10% of patients presented with either diarrhea alone or with neither blood nor diarrhea. Fifty-three percent of patients had either abdominal or rectal cramping, 43% described weight loss, and 25% complained of fever. When surveyed, 13% had extracolonic manifestations consisting of joint, eye, or skin manifestations. Joint symptoms were more prevalent in women than men, particularly in those between ages 45 and 59. Constipation is a well-recognized presentation in those patients with ulcerative proctitis. Of the 100 patients with ulcerative proctitis in the St Mark's series, 30% complained of infrequent stools or passage of hard stools during a flare of their proctitis [6]. The passage of blood alone, or the combination of blood and mucus, was described by the majority of patients (99%). Although systemic complaints are usually associated with more extensive disease, 13% of patients in this series described either fatigue or weight loss. A few patients with UC may be relatively asymptomatic but have endoscopic evidence of disease. In a study evaluating a screening program for colorectal cancer, six of the 481 patients who tested positive for fecal occult blood had UC at colonoscopy. Four had pancolitis, one had proctosigmoiditis, and the other had proctitis. Two of the patients reported occasional loose stools but otherwise these patients had no intestinal or systemic symptoms [23]. The clinical features at the time of diagnosis may aid in defining patients with so-called acute self-

limited colitis and those with true UC. Shumacher and colleagues evaluated 42 patients presenting with acute colitis and followed the patients for at least 5 years [24]. Patients with acute self-limited colitis related to known enteropathogens reported: (1) a shorter duration of symptoms, (2) more bowel motions, and (3) associated fever when compared with those subsequently found to have UC. Patients with UC reported a more insidious onset of symptoms with gradually increasing diarrhea. When symptoms were analyzed using logistic regression analysis this insidious onset of symptoms successfully identified patients with UC. Unfortunately, endoscopic appearance alone could not distinguish between acute self-limited colitis and UC. Histopathological review suggests that basilar plasmacytosis and crypt distortion were found more often in the rectal biopsies of patients with UC. This distinction, based on histopathological features, has been confirmed by others [25].

Physical findings There have been no studies specifically addressing the physical findings at the time of presentation in patients with UC. The more extensive the disease, and the more severe the symptoms, the more likely the patient will have abnormal findings on physical examination. Patients with ulcerative proctitis often have a completely normal examination except for the digital rectal examination which may reveal bogginess or irregularity of the rectal mucosa or perhaps blood or/and mucus on the examining finger. More extensive and severe disease signs of systemic toxicity may become apparent, including elevated temperature, tachycardia, signs associated with volume depletion, and documented loss of weight. Patients with long-standing symptoms may show signs of malnutrition. The physical findings associated with the extracolonic manifestations associated with UC are discussed in Chapters 16, 38, 39, and 40. The physical findings of toxic megacolon are discussed later in this chapter.

Laboratory The laboratory can be used in three distinct areas: first to define disease severity and exclude other diagnoses (infectious colitis, ischemia); second to monitor side-effects of therapeutic agents; and third

273

Remo Panaccione and Lloyd R. Sutherland Table 2. Truelove and Witts classification of ulcerative colitis Mild

Moderate

Severe

Less than 5 bowel motions per day Small amounts of blood in the stool No fever No tachycardia Mild anemia ( > 75%, approximately 100 g/L) ESR < 3 0

Intermediate between mild and severe

Six or more bowel motions per day Large amounts of blood in the stool Fever (>99.5°F,37.5°C) Pulse ( > 90 beats/min Anemia (hemoglobin > 75%) ESR > 3 0

to evaluate new serological markers which may help differentiate between the forms of inflammatory bowel diseases and aid in diagnosis. Laboratory monitoring of therapeutic agents is covered in Chapter 30 and serological markers are discussed in Chapter 20. At presentation most patients should have a complete blood count, electrolyte profile, blood urea nitrogen, sedimentation rate, and albumin. Stools should be collected for typical pathogens as well as Clostridium difficile toxin. Stool studies aid in excluding infectious colitides that may mimic UC. The rest of the blood work serves as a framework to assess the severity of the presentation. Patients with severe or extensive disease may demonstrate anemia, leukocytosis, thrombocytosis, and an elevation of the sedimentation rate. A rise in the blood urea nitrogen and perhaps the creatinine will be present if dehydration exists. In one study of patients with severe attacks, hypoalbuminemia ( < 3 0 g/L), hypokalemia (200 units/L) were associated with a poor prognosis [8]. Various laboratory abnormalities may accompany the various extracolonic manifestations and are discussed in Chapters 16, 38, 39, and 40. A complete review of the radiological and pathological features of UC is found in Chapters 19 and 17, respectively.

Assessment of severity Although there have been numerous disease activity indices to measure disease activity (severity) in CD, this is not the case in UC. There have been two indices developed for clinical trial use but these indices are not informally used [26,27]. The development of a more precise tool to measure disease severity would greatly improve the ability to compare

the various patient populations reported in the literature. Truelove and Witts developed the first classification of disease activity in 1955 [19] for the trial evaluating the efficacy of corticosteroids in the treatment of UC. The Truelove and Witts classification was developed to describe the severity of presentation for patients with extensive disease. However, it has been used frequently in studies of prognosis to classify severity in patients with variable disease extent including proctitis. The classification scheme proposed by Truelove and Witts characterizes attacks of UC as mild, moderate, or severe, depending on the number of bowel motions per day, presence of blood in the stool, fever, tachycardia, anemia, and the sedimentation rate (Table 2). The classification scheme is easy to use and understand but has never been formally validated. It may lack important variables in defining severity of disease. This is demonstrated in the Oxford series in which several patients classified as having mild or moderate disease at presentation died. A second scoring system was based on a regression analysis described by the group from Scotland [8]. Different weights are applied to temperature (X2), number of daily bowel motions (X4), and sedimentation rate {X^). Y= OMX2 + 0.12X4 + 0.35^6 - 0.59 Classification is mild, moderate, or severe according to the value of Y. All patients who died as a result of UC during their first attack were identified as severe by this scoring system according to the authors. However, this tool has not been used extensively outside of Scotland and has also never been validated.

Ulcerative colitis and ulcerative proctitis

274 Table 3. Characteristics of population-defined studies that deal with prognosis Source Sinclair ef a/. [10] Hendriksen et al. [31 ] Brostrom et al. [49] Stonnington et al. [14] Jones etal. [13] Ekbomefa/. [15]

No. of patients

Years of entry

Male/female ratio

Rectum/ sigmoid (%)

Intermediate

Pancolitis

(%)

(%)

537 783

1967-1976 1960-1978 1955-1979 1960-1979 1975-1984 1965-1983

275:262 342:441 681:593 75:63 160:153 1429:1080

74 41 25 46 56 42

15 41 38 18 25 28

11 16 37 36 19 30

1274

182 313 2509

Distribution of disease The distribution of disease at the time of presentation varies widely depending on the study. This is due to the heterogeneity of the populations being studied (i.e. population- or non-population-based). As previously mentioned, the population-based studies give a more accurate reflection of the natural distribution of the disease. The major population-based studies are summarized in Table 3. Despite all these studies being population-based they may not be comparable, for a variety of reasons. The most obvious reason for inability to accurately assess the precise frequencies of disease extent is the lack of uniformity in actually defining the extent of disease. Different investigators have used different definitions to define disease extent. For example, proctitis has been variably defined as affecting 12-14 cm [2], 15 cm [28], or the distal 20 cm [29]. Various studies have found that 14-36.7% of patients have pancolitis, 3641% have disease extending beyond the rectum, and 44-49% have proctosigmoiditis [30-32]. JVloreover, disease extent is most likely underestimated because the extent was determined by a combination of flexible sigmoidoscopy and barium enema instead of complete colonoscopy. In the recent populationbased study from southeastern Norway 525 patients were identified as having UC [33]. All patients underwent a colonoscopy at the time of diagnosis. Eightyfour percent had the colon visualized proximal to the splenic flexure and 73% had a complete colonoscopy with visualization to the cecum. Thirty-two percent had proctitis, 33% had left-sided colitis, and 35% had extensive colitis. This is the largest population-based study in which colonoscopy was used to define extent; it therefore should be considered as representing the true distribution of disease at presentation. However, one must also take into account that disease extent has always been defined by the macroscopic changes on endoscopy and alteration of the

mucosal pattern on barium enema. The microscopic extent of disease and its influence on disease course, has never been formally studied. The actual endoscopic margin of disease extent is not always well demarcated. In many instances there may be a gradual transition form abnormal to normal mucosa. In a study by Geboes and Ectors [34], 75% of patients diagnosed as having proctitis or left-sided disease were found to have segmental inflammation extending into the cecum when ileocolonoscopy was performed. This has also been demonstrated by other investigators who have described the 'patchy' nature of untreated UC [35]. The finding of a patch of cecal inflammation adjacent to the appendiceal orifice (cecal patch) in patients with left-sided disease is also well recognized [36]. These patients for all intents and purposes are thought to have UC, but whether there may be a later transition to CD in a few patients is not known. The number of patients with ulcerative proctitis may be increasing while the number of cases of UC is decreasing or remaining the same. In the Cleveland Clinic series, 124 patients were described as having ulcerative proctitis during the period 1950-1963. However, 32 newly diagnosed cases were described in 1969 alone [28]. This trend was confirmed in the population-based study from Uppsala, Sweden, in which the increase in total cases of UC was found to be due to an increase in the documented cases of ulcerative proctitis [37]. The age-adjusted incidence rate tripled between 1965-1969 and 1975-1978. The exact reason why the incidence of ulcerative proctitis is climbing in these population-based studies is uncertain. One explanation put forward was possibly an increased incidence of sexually transmitted disease. However, in one study the increase in incidence was due to a rise not only in males but also in females.

275

Remo Panaccione and Lloyd R. Sutherland

Diagnostic accuracy It is reassuring to note that interobserver variability in the assessment of the rectal mucosa is low at the time of endoscopic examination [38]. The accuracy of colonoscopic diagnosis of inflammatory bowel disease is reported to be 80-90% [39, 40]. In the Copenhagen study, of 1161 patients, 36 initially diagnosed as having UC eventually developed definite signs of CD later in the disease course, whereas seven patients initially diagnosed as having CD were later re-classified as having UC [32]. Only 3.7% of patients had a change in their diagnosis. It is uncertain whether this represents inaccuracy in diagnosis or change in disease over time. These findings are consistent with those of Pera and colleagues, in which 357 patients with inflammatory bowel disease were followed prospectively with endoscopic evaluation [41]. The authors found that the diagnosis was correct in 89% of cases, 4% had been misclassified, and 7% were not easily classified as either CD or UC. Difficulties and errors in diagnosis appeared to occur more often when severe inflammation was present. UC is characteristically a mucosal disease, usually beginning in the rectum and extending proximally in a continuous fashion for a variable length. However, rectal sparing, even in the absence of local enema therapy, is well described [42]. In a series from the Mayo Clinic there appeared to be a higher incidence of rectal sparing in patients with concurrent primary sclerosing cholangitis [43]. Although the mucosa may appear normal, endoscopic biopsies often show histological abnormalities. This is highlighted in a small study in which 12 patients with diarrhea and intermittent blood and mucus per rectum had normal endoscopic appearance in the rectum, but histological examination of rectal biopsies revealed evidence of UC [44].

Prognosis The prognosis of UC is the single most important issue from a patient perspective. Although many studies have been published which deal with prognosis, there is a lack of uniformity in these studies. Earlier non-population-based studies may be overly pessimistic with respect to prognosis. Once again the best clinically applicable data come from the larger population-based studies. However, even in these studies the clinician must be aware of differences in prognosis based on disease extent, presentation (first attack or relapse), and temporal trends with respect

to therapy. The past four decades have been accompanied by the use of corticosteroids in severe flares, maintenance therapy in the form of a m i n o salicylates, and more recently increased utilization of immunomodulators. The introduction of corticosteroids has greatly reduced the mortality of severe disease. Most patients will be on maintenance therapy with aminosalicylates. Patient acceptance of surgical therapy has also increased with the introduction of proctocolectomy with ileal pouchanal anastamosis. Therefore, more recent studies reflect the natural history of what may be deemed as treated UC [45].

Placebo response Examining the outcomes of patients randomized to the placebo arm of a randomized clinical trial may represent the natural history of patients with untreated UC. However, this concept has recently been challenged with the increasing awareness of the relationship between the neuropsychiatric system and the immune response controlling inflammation. Ilnyckyj and colleagues recently reviewed in detail the placebo response in patients with active UC [46]. In this study the placebo response in 38 double-blind randomized controlled studies was evaluated. The clinical remission rate in the placebo groups was 9.1% (CI95 6.6-11.6%) and the benefit rate was 26.7% (CI95 24.1-29.2%). This clinical improvement was accompanied by both endoscopic and histological benefits (endoscopic remission 13.5% (CI95 10.0-17.1%), endoscopic improvement 30.3% (CI95 26.6-34.0%), histological remission 8.6% (CI95 5.012%), histological improvement 25.2% (CI95 20.829.6%)). Meyers and Janowitz [47] noted similar findings when randomized placebo-controlled trials of maintenance therapy for UC were evaluated. Fifty-one percent (CI95 36-66%) of patients receiving placebo remained in remission when evaluated at 6 months. Patients with less extensive disease and those already in remission for at least 1 year were more likely to remain in remission compared to patients with more extensive disease and those who had recently entered remission.

Prognosis for the first attacic There are a large number of studies that have examined the issue of prognosis in UC following the

276 first attack. These include both population-based and non-population-based studies. Regardless of the source of information, the major determinant of prognosis appears to be extent and severity of disease. The discussion below will include both nonpopulation-based and population-based studies. Non-population-based studies from tertiary centers once again point to a more pessimistic outlook. Non-population-based studies There is some consistency in the reports of the early non-population-based studies in that the majority of first attacks were mild and associated with low mortality [4, 8, 10, 48]. Severe disease occurred in 625% of patients and affected predominantly patients with pancolitis. These severe cases were associated with a mortality rate of 20-33%. Edwards and Truelove described the course and prognosis of 624 patients with UC seen in Oxford between 1938 and 1962 [4]. All patients were available for follow-up during the observation period, which ranged from 1 to 24 years. Two hundred and fifty patients were seen during what is described as their first attack. Most of the 250 patients were seen within 6 months of onset of symptoms with only a minority (22%) having had symptoms between 6 and 12 months. The course of these 250 patients is outlined. Using the Truelove and Witts criteria, 54% had mild disease, 27% had moderate disease, and 19% had severe disease. The mortality rate during the first attack was 10%. Mortality was directly related to severity with most deaths occurring in patients with severe disease. Of the 25 reported deaths, 72% of patients were classified as having severe disease, 24% as having moderate disease, and one patient (4%) had mild disease. The mortality rate of those presenting with severe disease was 33%. Overall, mortality declined over the three decades of the study with no deaths occurring in the mild to moderate disease groups after 1952. There was also a trend toward decreased mortality in patients with severe disease, but this did not reach statistical significance. Prognostic factors for mortality included age at presentation, short duration of symptoms, extent of disease, and severity. Patients over the age of 60 had a mortality rate of 16.3% compared to 8.7% in patients under 60. A small subset of patients who presented with symptoms of less than 1 month and severe disease had a mortality rate of 18.4%. Increased mortality associated with increasing extent of disease. Twenty-five percent of patients with pancolitis died compared to 7% with disease defined as sub-

Ulcerative colitis and ulcerative proctitis stantial colitis (i.e. left-sided colitis). No deaths were reported in patients with distal disease. Overall, 84% of patients entered clinical remission, 25% died, and two patients underwent colectomy. Watts and colleagues reported their experience in treating 204 patients for their first attack of UC between 1952 and 1963 [9]. This group was diagnosed early in the course of disease with 90% of patients having had symptoms for 6 months or less. Thirty-six percent of patients were classified as having severe disease according to Truelove and Witts criteria with a mortality rate of 8.2% in this group. The overall mortality rate was 4%. Severe disease, pancolitis, and age over 60 were associated with increased risk of death or surgery. Complete remission was seen in 70% and a clinical improvement seen in an additional 16.2%. The rate of colectomy in the Leeds series was 11.3%. This is higher than the 0.8% colectomy rate reported in the Oxford series. The reason for this difference is not readily apparent; however, there appeared to be more patients with severe disease in the Leeds series. Alternatively, the threshold for surgical intervention at the time may have been different between the two centers. Overall, death or colectomy was the end-result in 36% of patients with severe disease, 31 % of patients over the age of 60, and 37% of patients with pancolitis. Jalan and colleagues reviewed the prognosis of 184 patients presenting with their first attack of UC at the Western General Hospital in Edinburgh between 1950 and 1967 [8]. Eighteen of the 19 deaths occurred in patients classified as having severe disease. The mortality rate in the group classified as severe was 26.4%. Ten deaths occurred while the patient was under medical supervision and the other nine fatalities followed emergency surgery. Overall, there were significant decreases in mortality rate over time. Population-based studies There are only a few population-based studies that deal with prognosis following the first attack of UC [10, 32, 49]. The most detailed of these studies is that of Sinclair and colleagues, who reviewed all 537 patients newly diagnosed with UC in the northeast of Scotland between 1967 and 1976 [10]. Patients' disease severity was mild in 68%, moderate in 26%, and severe in 6% by Truelove and Witts classification. The extent of disease was distal in 74% (rectum and sigmoid), pancolitis in 11%, and substantial involvement in 15% (proximal to sigmoid but not proximal to hepatic flexure). The mortality rate during the

Remo Panaccione and Lloyd R. Sutherland initial attack was 3% with 11 of the 17 deaths occurring in patients over the age of 70. Patient outcome was related to disease extent and severity. In patients with severe disease, 25% died, 31% underwent colectomy, 4% had continuous symptoms, and only 40% entered clinical remission. Outcomes for patients with mild disease were much better: 92% entered clinical remission, 8% had continuous symptoms, and no deaths were reported. In moderate disease 85% entered remission, 8% also had continuous symptoms, and 3%o died. Fig. 5 demonstrates the prognosis for patients characterized as having a severe first attack or pancolitis at presentation. Two additional population-based studies have been published more recently [32, 49]. Moum and colleagues [49] reported the clinical course of 496 newly diagnosed patients with UC in southeastern Norway between 1990 and 1993. The majority of patients (88%) underwent complete colonoscopy and were followed at 1-year intervals (98%) with a mean follow-up of 16.2 months. The overall mortality rate in this population was 2.2% (11 /496) but only two deaths were directly attributable to UC. Four percent of patients underwent colectomy. The colectomy rate was higher in patients with extensive colitis compared to patients with disease limited to the left colon or rectum. During the follow-up period 47% of patients remained in chnical remission. In the study from Copenhagen, involving 1161 patients diagnosed between 1962 and 1987, the overall colectomy rate in the year of diagnosis was 9% [32]. The relative risk of death in the year of diagnosis compared to age- and sex-matched population was 2.4.

Prognosis for subsequent attacks The majority of patients who enter clinical remission will have a recurrence of their disease over time. There is a small subset of patients who have a single episode of colitis and then enter a prolonged remission. It is difficult to tell whether this truly represents a subset of UC patients or simply reflects patients with presumed infectious etiology who were misdiagnosed as having UC. The most interesting unanswered question is whether the widespread introduction of 5-ASA agents for maintenance of remission purposes will have an impact on long-term prognosis. Intuitively, one would expect that patients on 5-ASA therapy would spend more time in remission and less time in

277 periods of relapse. Similarly one might expect that progression of disease extension would be reduced. In the Oxford series [4], 64% of patients had a chronic intermittent course and 7% of patients had a chronic continuous course. Eighteen percent of patients had only one attack over the study period. A proportion of these patients probably represented colitis of an infectious etiology. Stonnington and colleagues [14] reported on 182 patients in Olmsted County, Minnesota, followed between 1935 and 1979. They reported that 28% had a transient cohtis, 65%o had intermittent course, and 5% had a continuous course.

Extension of disease A major concern of patients, especially those with localized disease, is the risk of disease extension. This is a difficult area to analyze, for a variety of reasons. First, most of the available studies were performed prior to the advent of fibreoptic technology. Initial extent of disease was based on rigid sigmoidoscopy, which may underestimate disease extent. This is well described in the Cleveland Clinic experience with proctosigmoiditis [28]. In this series of 276 patients the upper limit of disease extent could be identified in only 46% of patients (127 of 276 patients). In the remaining 149 patients the examination failed to define disease beyond 15 cm due either to patient discomfort or lack of proper visualization. Second, the importance of microscopic involvement of endoscopically normal mucosa remains unknown. Many physicians do not biopsy normal mucosa, which histologically may be abnormal and therefore may underestimate the initial extent of disease. Third, the extent of disease tends to be underestimated with barium enema. This is demonstrated by the difference in the reported rates of pancolitis in population-based series in which flexible sigmoidoscopy and barium enema were used to define disease extent which range from 1 l-18%o [10,16] and those in which colonoscopy was utilized in which the rates of pancolitis ranges from 28% to 36% [15, 48, 50, 51]. Fourth, patients may be reassessed only at the time of dramatic change in disease behavior. There are very few studies in which patients undergo procedures to re-evaluate their disease endoscopically and/or radiologically on a regular basis. Finally, various studies, which have reported the proportion of patients with disease extension, did not use actuarial methods, which would deal with the issue of dropouts in analyzing the data.

Ulcerative colitis and ulcerative proctitis

278 Table 4. Clinical outcome after the first 10 years of ulcerative colitis and proctitis (%)

Involved area Rectum Sigmoid colon Descending to transverse colon Ascending colon

Surgery at 5 years

Descending colon

Transverse colon

2 + 2% 6 ± 3% 3 ±3% 30 ± 8%

5 + 2% 18 + 5%

3 ±2% Not stated

-

In the Leeds study [9], 50 of the 204 patients seen within the first year of symptom onset had disease limited to the rectum. During a mean follow-up of 3.2 years (range 6 months to 12 years), 18 of these 50 patients (36%) had proximal extension of their disease. Eleven of the 18 had substantial involvement, while seven developed pancolitis. The figure of 36% may be an underestimate, as actuarial methods were not used. In the largest series, from St. Mark's, 269 patients diagnosed with UC between 1966 and 1975 were reviewed for progression of disease [5]. All patients were initially assessed for disease extent within 6 months of symptom onset. The extent of disease was defined by rigid sigmoidoscopy and the majority of patients also underwent barium enema. Seventy-nine percent of patients had histologically confirmed disease and all but five patients were included in the follow-up. At initial presentation 28% had proctitis, 29% had proctosigmoiditis, 12% had disease extending from the sigmoid colon to the hepatic flexure, and 14% had disease extending proximal to the hepatic flexure. In 17% the exact extent of disease could not be evaluated as the barium enema failed to reveal abnormalities. The follow-up in this group ranged from 1 to 11 years. Clinical outcome data are shown in Table 4. Of the 76 patients initially diagnosed with proctitis, five were noted to have disease extension (three to the descending colon, two to the hepatic flexure). The cumulative probability after 5 years of extension into the descending colon was 5% and into the transverse colon was 3%. In the 79 patients presenting with proctosigmoiditis progression was noted in 17 (nine to the descending colon, eight to the ascending colon). The cumulative rate of progression at 5 years was 18% to the descending colon and 7% to the right colon. Thirty-three patients had left-sided disease distal to the hepatic flexure with extension occurring

Ascending colon 7 + 3% 21 + 8%

Mortality 0 (0%) 1 (2.6%) 0 (0%) 2 (5.4%)

in six. The cumulative probability of extension in this group was 21 % at 5 years. These data suggest that the cumulative 5-year probability of disease extension is approximately 10% in patients with ulcerative proctitis and 20% in patients with more extensive disease of the left colon. Farmer reported higher rates of progression in the Cleveland Clinic series [30]. Although this represents data from a tertiary center the data support a high rate of progression. Farmer and colleagues studied the natural history of 1116 patients with UC seen at the Cleveland Clinic between 1960 and 1983. Of these 1116 patients, 46% had proctosigmoiditis, 17% had left-sided colitis (disease distal to the splenic flexure), and 37% had pancolitis. During a mean follow-up of 13 years the disease had progressed in 54% of the patients. In many patients disease progression was evident without a change in symptoms, suggesting that subclinical extension is common. Seventy percent of patients with left-sided colitis and 34% of patients with proctosigmoiditis extended their disease to develop pancolitis. Factors associated with disease extension included severe disease at presentation, disease of the left colon, young age at diagnosis, joint symptoms at diagnosis, and bleeding at the time of diagnosis (most likely related to severity). Population-based studies also suggest that distal disease may not be as benign as originally thought. In one study 20-30% of patients with initial distal disease came to proctocolectomy over a 20-year follow-up period [12]. Using actuarial techniques, Sinclair and colleagues reported extension of distal disease in 12% of patients at 5 years and 30% of patients at 10 years [10]. The details of exact disease extent are not given. The colectomy rate in the group with distal disease was less than 5% at 10 years of follow-up.

Remo Panaccione and Lloyd R. Sutherland In the largest population-based study to date, using actuarial methods, Langholz and colleagues evaluated progression in 1161 patients in Copenhagen County diagnosed with UC between 1962 and 1987 [32]. Disease extent was evaluated by sigmoidoscopy (rigid or flexible) and barium enema. At the time of initial diagnosis 44% had proctosigmoiditis (mucosal changes on sigmoidoscopy combined with normal-appearing mucosa on barium enema proximal to the rectum), 36% had substantial colitis (which includes both left-sided and extensive colitis; abnormal-appearing mucosa on barium enema proximal to the rectum but not including the cecum), and 18%o had pancohtis (signs of colitis throughout the colon including the cecum). In 1.7%o of patients the initial disease was not known; therefore these patients were excluded from the analysis. Two hundred (39%) of the patients with proctosigmoiditis had extension of their disease. In 35% the extension was to substantial colitis and in 4% the extension was to pancolitis. The cumulative probability of extension at 10 years in the patients with proctosigmoiditis was 41%. The colectomy rate in these patients was 12%. Interestingly, of the 422 patients with substantial colitis only 9% progressed to pancolitis, while 44%) of patients experienced a radiographic regression of their disease. Radiographic regression was also seen in 39%) of patients diagnosed initially with pancolitis. This study suggests that inflammation of the colon is a dynamic state and that both progression and regression can be seen. The previous studies have relied on radiological assessment to define the extent of disease both at diagnosis and in follow-up. In a review of all patients seen in a district health unit in the United Kingdom [13], 48% of patients had more extensive disease at follow-up colonoscopy compared to presentation. Of the patients initially classified as having 'mild disease' (macroscopic changes at sigmoidoscopy but normal-appearing barium enema), 43% undergoing colonoscopy later in their evaluation were found to have colitis extending into the mid-transverse colon. This once again underscores the difficulty of interpreting studies when extent is based on radiological assessment.

Relapse In educating patients with respect to their disease course it is evident that the disease course will be characterized by periods of remission and relapse.

279 For most patients the time spent in remission is far greater than the time spent with a relapse or active disease. This is once again illustrated in the population-based study from Copenhagen [32]. Once again older non-population-based studies initially gave a different impression. DeDombal and colleagues [52] reported that the majority of patients seen in Leeds spent more time in clinical remission than in relapses. Relapses were usually mild by Truelove and Witts criteria. In Leeds the risk of relapse did not alter over time [53], providing little support for the concept that the will disease burn out. The report from Edwards and Truelove was not as encouraging. In their series of 250 newly referred patients with UC, 80% had a second attack within 1 year of entering remission [54]. Seven percent of patients had continuous disease. The disease activity and extent of disease at presentation did not alter the probability of relapse. However, one must remember that this report predated the widespread introduction of sulfasalazine. In the Scottish study [10], 28% of patients did not relapse after 5 years of follow-up. The authors speculated that perhaps the high proportion of patients with mild disease localized to the distal colon may have explained this, but there are no other studies to support this. The authors did introduce the concept of the attack and remission year. An attack year was noted if, at any time during the year, the patient had active symptoms suggesting disease activity. A remission year was noted if the patient was free of disease for the entire year. During followup 2674 patient-years were accumulated with 68%o of these years spent in clinical remission. There were 864 (32%) attack-years of which 84% were characterized by mild disease activity. The most complete data set relating to the prognosis of subsequent attacks of UC comes from the Copenhagen study by Langholz and colleagues evaluating their 25-year experience in 1161 patients [55]. Patients were treated according to defined protocols using sulfasalazine or 5-ASA for induction or maintenance of remission purposes, corticosteroids for more acute attacks, and proctocolectomy for patients who failed high-dose corticosteroids. Patients had easy access to subspecialty care ensuring rapid and uniform treatment. Among the 1161 patients from Copenhagen who did not require colectomy most patients had subsequent flares. The cumulative probability of a relapsing course was 90%o over the 25-year study period. The cumulative prob-

Ulcerative colitis and ulcerative proctitis

280 ability of a completely relapse free course was 10.6% at 25 years, whereas the probability of a continuous active course was 0.1%. Fig. 4 demonstrates the cumulative probability of a completely relapse-free course, the cumulative probability of a course with continuous activity, and the cumulative probability of an intermittent course. The authors also calculated the probability of remaining in remission by using Markov modeling. Fig. 5 in Chapter 20 shows the probability of remaining in remission from one year to the next. The probability of a patient in clinical remission remaining in remission the following year was 80-90%. The probability was influenced by the previous disease course in that patients with numerous years of remission were more likely to remain in remission. By contrast patients with clinically active disease had a probability of 70% of having a relapse in the subsequent year. The probability of having active disease every year after a relapse decreased over time from 50% at 2 years to 10-15% at 10 years. Langholz and colleagues also analyzed a subgroup of 600 patients who had been followed for a minimum of 7 years and had not yet died or undergone colectomy. In this group of patients, who were analyzed in the 5-year period 3-7 years after diagnosis, 25% were in continuous remission, 57% had an intermittent course, and 18% had continuous disease activity. During the 5-year period predictors of disease course included initial disease course in the year of diagnosis plus the following 2 years (more disease activity in early disease led to relapse later in the course); the calendar year (patients diagnosed in the 1960s had a more active course than those diagnosed in the 1980s); and the occurrence of systemic symptoms such as fever and weight loss at the time of diagnosis. Interestingly, in this study the clinical course over the 25-year follow-up was not influenced by anatomic extent of disease at diagnosis (except in patients with more extensive disease who underwent early colectomy), age, sex, presenting symptoms, length of time between onset symptoms and diagnosis, initial treatment, or occurrence of extraintestinal manifestations. Factors which influenced the course of disease were once again the presence of systemic symptoms at diagnosis and development of disease later in the study period. The difference in time of diagnosis suggests that patient outcomes may be improving with advances in medical therapy.

A800 kDa macromolecular trypsin-sensitive protein, but not part of a panel of defined pancreatic exocrine proteins [78]. Seibold et al. studied 222 CD and 51 UC patients for PAB and found that 31% and 4% of CD and UC patients respectively were positive for PAB, which was statistically significant [79]. PAB were not found in other inflammatory controls, and were not associated with disease severity or location. Of the seven patients with concurrent CD and pancreatic disease, four were positive for PAB. A subsequent study by Seibold et al. confirmed these observations. They reported that 40% of CD patients in their study population were positive for PAB. Furthermore, pancreatic insufficiency was detected in seven of 26 PAB-positive patients but in only three of 38 negative patients [80]. These observations support PAB as a specific marker for CD and possibly for associated pancreatic insufficiency. While PAB were found to be associated with CD, anti-goblet cell antibodies (GAB) have been associated with the diagnosis of UC. GAB were found in 13 of 51 patients with UC (29%) and in none of the CD patients or inflammatory controls. However, neither antibody was associated with disease behavior, nor suggested an antigen in the etiopathogenesis of IBD [79]. When the presence of PAB and pANCA was examined in first-degree relatives with UC and CD, neither marker wsd elevated, suggesting that they do not represent genetic markers of disease [81,82].

IL-1/IL-1ra ratio IL-1 is a proinflammatory cytokine that is inhibited by its naturally occurring receptor (IL-lra). In IBD, it has been hypothesized that an imbalance in the production of IL-1 and IL-lra exists in the gut mucosa as a result of a dysregulated immune system. The quantitative level of IL-lra detected in peripheral blood from patients with UC, CD, and inflammatory controls has been observed to be significantly

416 elevated in active CD and UC compared to control. These levels were also higher in active CD versus active UC [83]. When the ratio of IL-1 ra to IL-1 (ILlra/IL-1) was examined, the ratio was significantly decreased in the mucosa of UC and CD patients, compared to healthy controls or patients with selflimited colitis [84]. The decreased IL-lra/IL-1 ratio also correlated with disease severity. These data suggest that the measurement of IL-lra or the ILlra/IL-1 ratio may be a useful diagnostic test and indicator of disease activity. The polymorphism on intron 2 of the IL-lra gene has been previously characterized [85]. Carriage of allele 2 of this gene has been associated with 60-70% of patients with UC. Tountas etal. also showed that a decreased level of IL-lra production in PBMC was correlated with the carriage of allele 2 (IL-1RN*2) [86]. In this study, when patients were stratified according to ethnicity, carriage of allele 2 of the ILlra polymorphism was significantly increased in the Hispanic and Jewish subgroups, but not in the subgroup of patients of Italian descent [86]. Similar studies in other ethnic subgroups have also not been able to confirm the increased carriage of allele 2 of the IL-lra polymorphism in their UC populations from England, Spain, and Germany [87-89]. However, Roussomoustakaki et ai did confirm that, in their UC subgroup from England, extensive disease was significantly associated with increased carriage of allele 2 of the IL-lra gene [87]. Heresbach et al. also showed that, in severe UC patients who required colectomy, there was a higher carriage of the IL-lra allele 1-2 than non-operated UC patients. Furthermore, non-surgically treated UC patients had a significantly lower frequency of IL-lra allele 2 compared to surgically treated UC patients or controls [90]. These observations suggest that the genetic polymorphism for the IL-lra gene in UC may be associated with genetic heterogeneity among different ethnic groups, accounting for the variable carriage rates in different populations. The carriage of allele 2 of the IL-lra gene may also be associated with more extensive and severe disease in UC. The IL-ip gene polymorphism was studied in a group of Dutch UC and CD patients, but did not show any significant associations. However, when carriage of the IL-ip polymorphism and allele 2 of the IL-lra gene polymorphism were combined, UC as well as CD patients showed an increased carriage of IL-lra allele 2, but a decreased frequency of the IL-ip polymorphism. These observations provide genetic support for the hypothesis that an imbalance

New diagnostic approaches in inflammatory bowel disease in the IL-lra/ IL-1 ratio may be important in the pathogenesis of IBD [91]. However, in a similar study by Hacker et al., neither the IL-1 P gene polymorphism nor the combination with the IL-lra gene polymorphism showed any associations within this population of IBD patients. Neither could any association between the allele 2 polymorphism and UC patients be identified or associated with severity of disease. These disparate results may again reflect the genetic heterogeneity between various ethnic subgroups within UC patients [89]. Papo and co-workers studied a subgroup of Spanish UC patients for ANCA and allelic frequencies of IL-lra gene polymorphism, and polymorphisms of TNFoc and TNF-p. Overall there was no association between these polymorphisms in UC patients. However, when this study population was stratified according to ANCA status, pANCA-positive patients had a significantly higher frequency of the allele 1-2 of the IL-ra polymorphism. This suggests that pANCA and allele 2 of the IL-1 ra gene may also identify a distinct genetic subgroup of UC patients [88]. The frequency of other polymorphisms in the genes for IL-1 and IL-lra have been studied (Mwol, MspAIl, Alul, Taql, BsoFl), but have not been shown to be associated with UC or CD compared to controls [92].

Other genetic markers as novel markers of IBD Genetic markers have been associated with the presence of ANCA in IBD, and may further identify distinct subgroups of IBD patients in combination with ANCA serology. In a cohort of 53 IBD patients, Hirv et al. reported a significantly higher frequency of the HLA-DRB1*15 allele in UC, and a significantly lower frequency of this allele in CD. When UC patients were substratified according to ANCA status the increased HLA-DRB1*15 association was due to a significantly increased frequency of the allelic subtype HLA-DRB1*1501 in the ANCApositive subgroup compared to controls [93]. Although larger populations of IBD patients need to be studied, this observation supports the role of ANCA in combination with genetic markers in identifying more homogeneous IBD subgroups. Other genetic markers have been observed to be associated with the severity and prognosis of IBD. For example, in a study population of UC patients

Lori Kam and Eric A. Vasiliauskas the presence of HLA-DR2 was associated with poor prognosis and medically resistant disease [94]. HLADRB1*15 and HLA-DRB1*0103 alleles have also been reported to be associated with more extensive and severe disease in ulcerative colitis [95]. In CD HLA-DRB1*03 was found to be significantly less frequent in a selected population of fistulizing CD patients compared to control [96]. Taken together, these data suggest that a combination of genetic and serologic markers may identify specific phenotypic subgroups characterized by common patterns in disease severity and prognosis.

Clinical applications Serum immune and genetic markers can potentially be utilized at several levels including: diagnosis (IBD vs non-IBD), differentiation (UC vs CD), stratification (phenotyping), or prognostication, as well as potential therapeutic and research implications.

Diagnostic utility: differentiation of IBD from non-IBD Early recognition of IBD may allow for therapeutic interventions at an earlier point in the inflammatory process, thereby potentially avoiding or minimizing subsequent disease- and therapy-related complications. The current gold standard for the diagnosis of IBD is based on a combination of estabhshed clinical, endoscopic, histopathologic, radiologic, and serologic criteria [1]. The sensitivity and specificity of these combined tests enable clinicians to differentiate IBD for other disease processes. Establishing a definitive diagnosis of IBD, and knowing in whom and to what extent to pursue more extensive diagnostic evaluations, can be challenging. The presenting symptoms of IBD, in particular CD, can be subtle and non-specific. Initial symptoms may be nongastrointestinal in nature and may include: arthritis, arthralgias, menstrual irregularities, unexplained fevers, anorexia, or fatigue. Indeed, once a diagnosis of IBD has been established, it is not uncommon for patients and family members to relate how, in retrospect, symptoms associated with unrecognized IBD predated the definitive diagnosis by months or years, having been attributed to 'irritable bowel syndrome', 'chronic recurrent abdominal pain', or other conditions [97, 98]. Despite exhibiting minimal or no gastrointestinal symptoms, predominant manifestations of IBD in children and adolescents may include

417 delayed pubertal maturation or failure to gain weight and height appropriately [97, 99, 100]. Under these circumstances IBD may go unrecognized, or the degree of involvement may go under-appreciated for an extended period of time. Underscoring this point is that growth retardation is already present in 3050% of pediatric CD patients, including prepubertal children, at the time of presentation, with decreased height velocity preceding the onset of gastrointestinal symptoms by as much as 11 years [101-103]. Failure to recognize and address 'silent' active inflammatory disease in a timely fashion ultimately manifests permanent deficits and adult height [97, 104, 105]. Thus for pediatric patients in particular, early diagnosis, adequate nutrition, and successful control of disease activity set the stage for optimizing appropriate linear growth, normal sexual maturation, and restoration of well-being. Delays in diagnosis are common and may in part be related to patient or parental resistance and anxiety due to the invasive nature of the work-up. Radiographic and endoscopic procedures require cooperation and are unpleasant and invasive, especially from the perspective of the child or the young adolescent. Less invasive traditional blood tests are non-specific, primarily looking for evidence of inflammation, including: elevated white blood cells with a left shift and elevated band counts, thrombocytosis, elevated sedimentation rate and/or C-reactive protein, decreased serum albumin levels, and decreased iron stores. Stool markers, such as stool oci-antitrypsin, have also been used as diagnostic aids [97]. Of the described IBD-specific serum immune markers, pANCA and ASCA are clinically the best characterized and studied. As previously described, these serum immune markers are present in the majority of UC and CD patients, respectively, and are infrequently present in non-IBD controls, including patients with irritable bowel syndrome (IBS). These tests are thus specific for IBD and the falsepositive rate is low for each. Non-invasive and less expensive diagnostic screening in the form of an 'IBD serum immune marker panel' could thus provide the evidence to direct further appropriate evaluation, a strategy similar to that currently applied in patients with suspected systemic lupus erythematosus [106]. Ideally patients identified with UC and CD should have the appropriate traditional work-up to characterize their disease, and expensive and invasive testing to rule out disease should be avoided in patients who do not have IBD despite symptoms compatible

418 with these disorders. Such a strategy was recently tested in a pediatric population presenting with nonspecific symptoms suggestive of IBD, yet a normal physical examination [60]. For the purposes of testing the assays as a screening tool, an ELISA-based modification of the pANCA and ASCA assays without immunofluorescence or DNase pANCA confirmation was used. Cut-off' values were recalculated for the purpose of optimizing diagnostic sensitivity (understanding that this also lowers specificity). In this study all patients underwent simultaneous invasive and non-invasive modified serodiagnostic testing. Applying this strategy the authors found that 95% of non-1 BD patients were accurately identified by way of the modified pANCA and ASCA panel, demonstrating that had this 'IBD serum immune marker paneF been the only diagnostic tool utilized, these non-IBD patients could have avoided an unnecessary invasive evaluation. On the other hand, the third of patients with non-specific symptoms whose diagnosis would have been delayed with this approach would likely have returned with persistent or progressive symptoms more suggestive of IBD, which would have then mandated the needed further work-up. The addition of OmpC, 12 and other serum immune markers to such an 'IBD paneT could further improve sensitivity and specificity. In the appropriate clinical setting, a negative 'IBD panel' can provide reassurance and avoid unnecessarily extensive work-up, unneeded discomfort, or anxiety [60, 107]. On the other hand, a positive test may provide the evidence to pursue further appropriate evaluation. It must be emphasized, however, that the absence of serum pANCA or ASCA expression does not rule out IBD. The IBD panel is an excellent noninvasive adjunct to a good history, and clinical, endoscopic, radiologic, and histopathologic findings.

Differentiation of uicerative colitis from Crohn's disease Both pharmacologic and surgical management decisions are often heavily influenced by whether the patient has a diagnosis of CD or UC. Serum immune markers have been used in an attempt to differentiate UC from CD. While there is a strong tendency for patients who express ASCA or OmpC to have CD and pANCA to have UC, it is not surprising that overlap is seen in serum immune marker expression, just as an overlap between CD

New diagnostic approaches in inflammatory bowel disease and UC has been demonstrated at multiple levels including: clinical, endoscopic, radiographic, histopathologic, epidemiologic, genetic, immunopathologic, and response therapies [9, 13, 64, 108-115]. In general up to 30% of CD patients are pANCA"^ and up to 14% of UC patients are ASCA"". Selective marker antibody expression combining ASCA and pANCA improves the ability to differentiate between CD and UC [10, 38]. In addition to the absolute presence or absence of immune marker expression, the clinical relevance of taking into account the level of marker antibodies has recently been demonstrated [10]. An inverse relationship between ANCA and ASCA levels within CD has been observed (i.e. the higher the ASCA level, the lower the ANCA level and vice-versa) [10]. A similar inverse relationship is seen in UC (personal observations). This inverse relationship suggests that the presence and levels of a specific marker antibody could reflect distinct immunologic reactivity that leads to diff'erences in clinical expression among CD patients. The vast majority of patients who are positive for both IgA and IgG ASCA, and/or express very high levels of either ASCA subtype, have CD (personal observations). Expression of both IgG and IgA ASCA subtypes is unusual in UC, and levels of ASCA tend to be low. Interestingly, ASCA is coexpressed in only about half of pANCA"*" CD patients, generally at low levels [10]. Of further clinical relevance is that many patients with indeterminate colitis are not surprisingly pANCA"^, which is consistent with their 'UClike' presentation, and as such the ANCA and ASCA profiles are of limited usefulness in determining whether these patients will truly manifest as UC or CD [116]. 12 is detected more commonly in CD than UC and non-inflammatory controls (51-54%, 10%, and 4%, respectively) [73, 117]. Interestingly, 12 is present in 44% of pANCA"^ CD patients and additionally was detected in a third of those CD patients negative for ASCA and pANCA [117]. OmpC is present in nearly half of CD patients lacking the above two markers. The value of collective serum marker profiles, combined with the presence of specific genetic markers, may ultimately prove to be the most beneficial strategy.

Stratification Overview Serum immune and genetic markers may serve as subclinical indicators of specific patterns of disease

Lori Kam and Eric A. Vasiliauskas expression. Numerous attempts have been made to characterize CD and UC patients into uniform subgroups to better understand and predict cHnical courses and responses to medical and surgical interventions. Most stratification analyses of CD, particularly in therapeutic trials, continue to focus predominantly on anatomic location of inflammation. Behavior characteristics of inflammation, aggressiveness of disease, and responsiveness to medical interventions also need to be considered [1, 5]. Patients with UC are generally characterized by location of disease as having either proctitis, proctosigmoiditis, left-sided colitis, or pancolitis. While this description of extent of disease may guide initial approaches to treatment and future dysplasia screening recommendations, it is important to reaUze that this does not correlate with or predict aggressiveness of disease. Such classifications have generally been descriptive and not reflective of specific pathogenic mechanisms. Inflammatory bowel diseases can now be further stratified into distinct subgroups by pairing clinical features with the presence and levels of serum antibody markers [10, 41, 67]. Serum immune marker expression in IBD permits stratification at the mucosal, clinical, immunologic, genetic, and therapeutic levels. To date these associations have been best characterized with pANCA and ASCA. In IBD, pANCA and ASCA production appear to reflect distinct and divergent mucosal inflammatory processes and thereby are reflective of immunologically distinct subpopulations. Both have been associated with distinct disease subgroups [10, 31, 35, 37, 41,44,118-120].

Ulcerative colitides In UC pANCA has been associated with the clinical features of: (1) treatment-resistant left-sided disease [119]; (2) aggressive disease course [31, 120]; (3) surgery early in the disease course [35]; and (4) development of pouchitis following ileal pouch-anal anastomosis (IPAA) [41, 118, 121]. While many studies have found no association of pANCA expression with disease activity [18, 21, 22, 24-27, 31], and persistence of pANCA positivity after coletomy [19, 21, 24, 33, 34, 87, 122] others have observed higher levels of antibody expression in patients with more active disease [20, 28, 29, 123] and lower or absent levels in patients following colectomy with active disease [32, 39, 123, 124]. Some studies have shown an increased incidence of ANCA expression in

419 family members of ANCA"^ UC patients [12, 14, 34]. Genetic studies have suggested that subpopulations of UC can be defined by ANCA expression in association with specific HLA markers [13, 14, 87, 115, 125, 126] and it has recently been suggested that certain genotypes may be associated with clinical phenotypes, such as disease behavior or an increased likelihood of developing chronic pouchitis [67, 87, 127, 128].

Crohn's diseases ThQpresence of the immunologic markers ASCA and pANCA expression in CD have been found to correlate with different disease locations [14, 26, 30, 37-39, 61, 129]. Most studies evaluating ASCA expression in CD have found an association with small bowel involvement. Giaffer et al. observed that patients with isolated small bowel involvement had significantly higher ASCA IgG titers against two strains o^S. cerevisiae than those with colonic disease [129]. In a study examining ASCA expression in monozygotic twins, Lindberg et al. found higher IgG ASCA levels (to whole yeast) in patients with CD limited to the small bowel compared to those with both small bowel and colonic involvement [61]. Vasiliauskas et al. also observed that the presence of ASCA was associated with small bowel involvement (alone or in combination with colonic disease), noting minimal ileal involvement in all [10]. In contrast, several investigators have maintained that in CD pANCA expression is related to colonic involvement disease, which at minimum is left-sided and 'UC-like' [10, 14, 25, 26, 37, 39, 130]. 'UC-like' behavioral attributes are the predominant behavior pattern common to the subgroup of CD patients expressing serum pANCA, providing additional evidence that pANCA is a marker of a distinct mucosal inflammatory process. In addition to features of 'classic CD' (such as fistula, fibrostenosis, and anal or small bowel involvement) pANCA"^ CD patients have at some point in the course of their disease a minimal left-side coHtis with endoscopic and/or histopathologic features of UC [37]. Review of the clinical histories of pANCA"^ CD patients reveals that nearly a third are initially diagnosed as having UC, prior to the development of characteristics diagnostic of CD (i.e. small bowel fibrostenosing or penetrating disease, multiple fistulas, or the finding of granuloma, which by definition convert the diagnosis to CD (unpublished personal observations). Using the same methods of pANCA detection this association of

New diagnostic approaches in inflammatory bowel disease

420

Table 2. Clinical features in Crohn's disease: results of multivariate analysis Disease location

Fibrostenosing disease

Internal penetrating disease

Perianal penetrating disease

UC-like features

-

0.32 0.03 0.022

-

n.s.

0.33 0.13 0.0001

n.s.

-0.29 0.02 0.063

-

-

-

-

n.s.

n.s.

n.s.

n.s.

ASCA level

P

R" j[>Value ANCA level

P f^

p-Value

0.67 0.15 0.0001

p, regression coefficient; F^, proportion of the total variation of a dependent variable which can be explained by the variation of each independent variable Modified from ref. 10.

'UC-like' features has been confirmed in several CD populations by investigators in England and Canada [14,39, 130]. It has recently been suggested that not only the type, but also the magnitude of the immune marker expression (as measured by level of host marker antibody production) reflects divergent immune responses, that manifest as specific disease behaviors (phenotypes) in IBD. A recent study by Vasiliauskas et al. demonstrated that the associations with behavioral manifestations in CD become more evident at higher levels of both ASCA and pANCA expression (particularly if they expressed one, but not the other, immune marker) [10]. Higher levels of ASCA and pANCA expression identify CD subgroups which are immunologically and clinically progressively more homogeneous, with distinct and divergent clinical characteristics, suggesting that ASCA and pANCA are serum markers for different mucosal inflammatory mechanisms that influence disease expression [10]. Higher levels of ASCA are independently associated with earlier age of onset of CD and a tendency toward developing 'classic' fibrostenosing and internal penetrating small bowel complications, and less often UC-like features (Table 2) [10]. In contrast, higher levels of pANCA in CD are associated with later age of onset and with a 'UC-Hke' inflammatory response, as well as a relative lack of fibrostenosis and penetrating disease (Table 2). The combination of selective pANCA or ASCA expression with the added variable of the magnitude of expression (those expressing high levels of a single marker antibody and lacking the other) further defines subgroups of CD patients into more

immunologically and clinically homogeneous subgroups, including a group with a more aggressive disease course [10]. While fibrostenosing and internal penetrating behaviors have been considered by some to be indicators of a more aggressive form of small bowel CD, the need for small bowel surgery or recurrent surgeries has been suggested as a further measure of disease severity. In the study by Vasiliauskas et al. the subgroup of CD patients expressing high levels of IgG and IgA ASCA and lacking pANCA were characterized by the universal occurrence of fibrostenosis and the frequent development of internal penetrating complications (Fig. 4) [10]. Small bowel surgery was not only required by a larger percentage of CD patients expressing very high levels of ASCA and not pANCA (Fig. 5A), but the mean number of small bowel surgeries per patient with small bowel disease was also higher (Fig. 5B). Similarly, the incidence of'UC-like' behavior is greatest at higher ELISA levels of pANCA expression and by exclusion of ASCA expression. The frequency of UC-like features in low-level expression of pANCA is no different from that in the overall population, stressing the importance of taking into account the magnitude of the immune response. These data strengthen the assertion that CD and UC pANCA subgroups share a similar mucosal immune response. The associations and findings seen in the subgroup of CD patients expressing high levels of ASCA and pANCA are provocative, and suggest that the presence of serum immune markers (in this case ASCA and pANCA), and the magnitude of the host immune response (level of serum immune marker expression), correlate with

Lori Kam and Eric A. Vasiliauskas 100% n

421

^100%

m IgA & IgG ASCA* (>50) / ANCA" (n«14) • pANCA (>40) / ASCA" (n=14) All Others (n=128) 86%

>

"i

75%

m tn 0 0)

58%

g

50% H

c I cr P

25%

_j

0%

Internal Penetrating p< 0.001

Fibrostenosing p< 0.001

UC-like p< 0.001

Disease Behavior Figure 4. Substratification of Crohn's disease using selected expression of immune markers. Disease behavior characteristics were examined in more immunologically homogeneous CD subgroups (those expressing high levels of a single marker antibody) and compared to all other CD study patients. Overall differences in proportions were evaluated using the x^ test for trend {p< 0.001 for each of the disease behavior characteristics). Higher ASCA levels are associated with both fibrostenosing and internal penetrating disease behaviors. In contrast, higher ANCA levels are associated with an 'UC-like' inflammatory response (Modified from ref. 10). D High IgA & IgG ASCA* (>50) / ANCA" • High pANCA (>40) / ASCA' D AH Others 1

.S M 1.8 i

1.71

86%

i

3 Q

wl. 00 (O

_ « '1.4

75% i 57%

-B = 50% ^ E 42 «>

2 O OflO DQ ^

1

1.02

si 1 *^£

0.8

29%

I ^ H H

1.2 ( it

t 1,

CD Subgro up

CD Subgroup

B Figure 5. Surgery as a measure of disease aggressiveness In Crohn's disease. Substratification of the subset of patients with small bowel disease (CD involving the small bowel alone or in combination with the colon) using selective immune marker expression: A: percentage of patients requiring small bowel surgery and B: total number of small bowel surgeries per patient with small bowel involvement Overall differences in proportions were evaluated using the x^ test for trend. (A: p < 0.0001 \%:p< 0.005). Small bowel surgery was not only required by a larger percentage of CD patients expressing very high levels of ASCA and not pANCA, but the mean number of small bowel surgeries per patient with small bowel disease was higher as well, suggesting that very high levels of ASCA without ANCA may reflect more aggressive small intestinal disease (Modified from ref. 10).

422 distinct, and indeed divergent, clinical characteristics, suggesting that ASCA and pANCA are serum markers for different mucosal inflammatory mechanisms that influence disease expression. These marker antibodies have also provided insight into the perforating behaviors characteristic of CD, as manifested by abscess or fistula formation or perforation. While internal penetrating disease is positively correlated with high levels of ASCA expression, and negatively correlated with high levels of ANCA expression, no significant correlation was seen with these markers and perianal fistulizing disease (Table 2) [10]. This suggests that these two penetrating forms of CD involve different immunologic mechanisms, and might be considered not only cHnically, but also immunologically, distinct. Of note is that, while perforating and non-perforating behaviors are cHnically (and likely in some ways immunologically) distinct also, the observation that higher levels of ASCA expression are associated with both of these traditional 'classic' fibrostenosing and internal penetrating features of CD, suggests an overlapping mucosal mechanism, rather than solely distinct mechanisms [10]. The association of ANCA and ASCA expression with age of onset of CD is intriguing. The relationship between age of disease onset and both disease behavior and anatomic location has long been appreciated. The majority of epidemiologic studies examining age of presentation suggest a bimodal age distribution in CD [131-134]. It has been observed that childhood-onset CD is characterized by a greater prevalence of small bowel disease and stricturing/penetrating complications, compared to lateonset CD which is associated with more colonic disease and 'UC-like' features and a lower incidence of fibrostenosis or internal perforating complications [10, 135, 136]. Some studies suggest that childhood-onset CD has a more aggressive course [136138], while older-onset CD is associated with a comparatively favorable prognosis, better response to medical therapies, and lower risk of recurrence in the minority requiring surgery [134, 135, 139, 140 140]. Clinical differences between childhood- and adult-onset CD in disease distribution and clinical course may be associated with diff'erent immune responses [135-142]. The relationship between age of onset of CD and expression of ASCA and pANCA has been examined. ASCA"^ CD patients have been shown to experience onset of IBD symptoms earlier in life than patients not expressing ASCA [10, 38]. The study by Vasiliauskas et al, compared patients

New diagnostic approaches in inflammatory bowel disease 1

73% 59%

< u

29%

(0


6 months) with prednisone or antimetabolites as outpatients, and severe disease - those patients who required admission to hospital for treatment. A total of 607 eligible patients generated average charges of $12417 per patient. Charges for the mild, moderate, and severe disease groups were $6277, $10033 and $37 135, respectively. Twenty-five percent of the patients generated 80% of the total charges; 57% of the charges occurred as a consequence of caring for inpatients. Finally, Bernstein and colleagues [17] estimated the direct costs of hospitalization for patients with CD and UC admitted to a tertiary-care hospital in Manitoba, Canada, during 1994 and 1995. Based on a sample of 275 admissions the mean cost of CD was $3149 compared with $3726 for UC. Surgery

accounted for 49.8% of all admissions and 60.5% of the total inpatient costs. What conclusions can be drawn from these three studies? First, two of the studies show that the direct cost of care for CD, in the setting of a US HMO during the first half of the 1990s, was approximately $66008800 per annum (The first estimate is from Hay and Hay [13]; the latter is calculated from Feagan et al. [16] following adjustment for: (1) a cost-to-charge ratio of 0.65 and (2) an annual rate of inflation of 3%). Second, the studies demonstrate that provision of inpatient care to a minority of patients - 20-25% is responsible for the majority of costs. Third the data of Hay and Hay indicate that UC likely has less overall economic impact than CD. Finally, the relatively low cost estimates obtained by the Manitoba study underscore the importance of obtaining jurisdiction-specific cost estimates. Although all three of these studies provide valuable information they also suffer from an important limitation. Since only highly selected individuals who sought medical treatment were assessed, it is likely that the cost of treatment was overestimated. Ideally, cost data should be obtained from a defined cohort of patients (1) who are residents of a region where health care is readily accessible and (2) where an administrative database exists which has the capacity to reliably measure the total costs of care. Two such studies have been reported; one from Minnesota, the other from Sweden. The former study concentrated on the direct cost of CD whereas the latter provides us with the only published information regarding indirect costs for both disorders.

Brian G. Feagan

Poulation-based cost estimates Olmstead County Silverstein et ai [18] performed a retrospective evaluation of the direct costs of CD based on a cohort of Olmstead County Minnesota residents who were diagnosed between 1970 and 1993. Patients were classified into one of eight unique health states. A Markov model estimated the duration of time spent in each heakh state over the follow-up period for the entire cohort of patients. The transition probabilities between each of these states were estimated based on 2-month periods. The direct costs of CD were assigned to the specific health states using 1987-1996 charge data from the Olmstead County Utilization and Expediture Database. Charges were converted to cost estimates in 1995 US dollars using the Medical Price Index. The average lifetime cost of CD was estimated at $125 404 per patient. Expressed on an annual basis this translates into a cost of $4308 per patient per year. Surgery accounted for 44% of the total costs, treatment with 5-aminosalicyclic acid (5-ASA) accounted for 29% of all costs.

Sweden A second population-based study, performed in Sweden, provides unique indirect cost data. Blomquist and Ekbom [19] identified all cases of CD in Sweden during 1994. Because access to health care is universal in Sweden this study is a comprehensive, cross-sectional evaluation of the direct and indirect economic burden of the disease. However, the latter analysis was restricted to economic loss due to disease-related absence from work. The total direct cost of CD, expressed in 1994 US dollars, was estimated to be $27.5 million. Admission to hospital was responsible for 58%o of the cost. The estimate of indirect costs was even greater at $58.4 million total. These large indirect costs were attributable to: (1) a small group of patients who were permanently disabled and (2) a significant rate of absenteeism due to illness (the average sick leave was greater than 40 days). Taken collectively these population-based studies suggest that the average cost estimates derived from HMO data likely overstate the true direct costs of CD. This discordance probably reflects an overrepresentation of more severely ill patients in the claimsbased HMO data sets. The Swedish study also under-

475 scores the large potential effects of the disease on indirect costs. These costs are important from a societal perspective and should be considered when decisions are made regarding payment for new drugs. However, it should be emphasized that social programs diff'er greatly among countries. Thus further indirect costs studies from a broad sample of jurisdictions should be performed. Although these population-based studies may not reflect more recent trends in therapy which include: (1) the use of newer, more expensive forms of medical t r e a t m e n t s (aminosalicylates, antibiotics, 6mercaptopurine, azathioprine, methotrexate, infliximab); (2) increased reliance on outpatient care; and (3) increased surgery rates for patients with CD, a consistent message is identifiable. Treatments which reduce the need for admission to hospital, surgery, or absenteeism may result in important direct and indirect cost saving

Measuring outcomes: quality of life As described previously, pharmacoeconomic analyses assess the incremented cost required to achieve clinically meaningful improvement in outcomes such as disease activity, hospitalization, surgery, complications, death, or HRQL. In patients with CD, H R Q L is perhaps the most relevant outcome because: (1) hospitalization, surgery, disease-related complications, and death occur infrequently; and (2) the burden of symptoms and social and psychological morbidity are lifelong. HRQL is not simply the absence of disease, but comprises four major components: physical function, emotional and social well-being, ability to work productively, and freedom from disease-related symptoms [20, 21]. Although physical complaints of pain, diarrhoea, and fatigue are important to patients with IBD, other components of HRQL, such as the presence of disease-related emotional stress, are also relevant [ 22]. Patients with CD and UC have lower HRQL than healthy individuals. The greatest impact on HRQL occurs during disease exacerbations, but patients are also affected while in remission. In a Danish study [23], a majority of patients reported that IBD had a negative influence on their lives. This may be related to concerns about disease recurrence, the need for surgery, and the effect of chronic illness on family or professional relationships.

476

Pharmacoeconomics and inflammatory bowel disease

In an attempt to measure the relative importance of the factors which affect quahty of hfe, HRQL instruments have been developed. The two major forms of HRQL assessments are psychometric questionnaires and utility measures.

placed on a given health state by an individual or society. The concept of utility analysis incorporates the evaluation of preference. For example, given a choice, most individuals would prefer to have mild myopia (high utility) than to be blind (low utility).

Psychometric questionnaires HRQL questionnaires range widely in complexity from asking the patient to rate pain on a 10-point scale to completing a computerized instrument containing more than 100 individual questions. Quality-of-life instruments are either generic or disease-specific. Generic instruments evaluate most aspects of physical, mental, and social health, and are especially useful comparing differences in health status between diseases. Many such instruments exist, including the Sickness Impact Profile [24], the Short Form-36 [25], and the McMaster Health Status Questionnaire [26]. Disease-specific HRQL instruments address only the problems commonly encountered by individuals with a given disease, and generally respond better than generic instruments to small changes in the well-being of patients.

Estimating utility

The IBD questionnaire The IBD questionnaire (IBDQ) [27] is a diseasespecific measurement of quality of life which measures four dimensional scores, namely bowel function (e.g. loose stool, abdominal pain), systemic function (e.g. fatigue, altered sleep pattern), social function (e.g. work attendance, need to cancel social events), and emotional function (e.g. anger, depression, irritability). Each item is scored on a sevenpoint Likert scale where a score of 1 indicates a very severe problem, while a score of 7 denotes no problem at all. Thus, the aggregate score ranges from 32 (very poor HRQL) to 224 (very good HRQL). The IBDQ has been validated [28, 29] and has excellent reliability and responsiveness.

Utility analysis Valid and reliable HRQL questionnaires may help the clinician to assess the psychological and social needs of the patient and improve patient management [29]. Unfortunately, these scores are unsuitable for use in a cost-effectiveness analyses, i.e. analysing the marginal cost per unit of IBDQ score gained is not a meaningful metric. An alternative method of evaluating HRQL is utility analysis [9, 30]. Utility refers to the value

Three generally accepted methods exist for estimating utility: the Standard Gamble, the Time Trade-off, and the Visual Analogue Scale (VAS) [31]. The gold standard for utility estimation is the Standard Gamble. This procedure asks patients to consider their current state of health, and then choose between two alternatives: Choice A - to remain in that health state; or Choice B - to opt for a hypothetical gamble with two possible outcomes. The possible outcomes with Choice B are: (1) a probability (P) of returning to perfect health or; (2) a complementary probability (1 - P) of immediate death. In a simplified theoretical situation, if a patient with chronically active CD was offered a 99% chance of returning to perfect health with an associated 1% risk of immediate death, most patients would accept the gamble. Conversely, most people in this situation would not choose to gamble based on the complementary probability of a 99% chance of sudden death against a 1 % risk of returning to health. As the probability P is varied the patient should reach a point at which it is difficult to decide whether or not to gamble. At this point of equipoise the utility score is identified. In the Time Trade-off method an individual identifies the number of years of life in his/ her current health state that he/she would be willing to trade in exchange for perfect health. The number of years is varied until the patient is ambivalent about the trade. The utility score is calculated by expressing the number of years traded as a proportion of the patient's calculated life expectancy (estimated from an actuarial table). For example, if a 40-year-old malet with chronically active CD is willing to trade 6 of his estimated 38 years of life expectancy, the utility score is 38 - 6/38 = 0.84. In the visual analog scale (VAS) method, the subject chooses a point on a 100-mm visual aid that best expresses the current health state where 0 signifies death and 1.0 means perfect health. Utility measures assess patient preference. Although the use of methods such as the Standard Gamble have been criticized as abstract it can be pointed out that clinical medicine routinely requires doctors and patients to make choices under uncer-

Brian G. Feagan

All This observation indicates that utilities may be less useful than other measurements for detecting small, but clinically important, differences in HRQL. Validity was demonstrated by: (1) the observation that all three measures of utility ordered the hypothetical health states according to an a-priori prediction (patients with more severe disease scored poorer than those whose disease was in remission and (2) that the Time Trade-off and Standard Gamble scores correlated with two other widely accepted measures of disease severity: the IBDQ, which assesses HRQL, and the CDAI, which assesses disease activity. Reliability (reproducibility) was also demonstrated since utility scores in stable patients were remarkably similar at the first visit and 8 weeks thereafter.

tainty. Usually these choices are associated with quantifiable risks to the patient.

Utility In IBD Few studies have measured utility scores in patients with IBD. A 1997 publication [32] collected data from 180 patients with CD using the three accepted measures for utility estimation (Standard Gamble, Time Trade-off, and VAS). Patients were stratified into one of four health states (Table 3). Utility scores were demonstrated to be valid and reliable measures of HRQL. However, utility determination was less responsive to changes in disease severity than the CD Activity Index (CDAI). Table 3. Evaluation of utility measurement In Crohn's disease [32] Health State

Definition

Chronically active therapy-responsive disease

CDAI score < 1 5 0 with continuous prednisone therapy ( ^ 10 mg/day) and/or continuous treatment with methotrexate or purine antimetabolites for > 6 months

Chronically active therapy-resistant disease

CDAI score ^ 150 despite continuous prednisone therapy ( ^ 1 0 mg/day) and/or continuous treatment with methotrexate or purine antimetabolites for > 6 months

Acute disease exacerbation

A recent flare of disease (CDAI ^ 150) and no steroid or immunosuppressive therapy in the 12 weeks preceding the initiation of current course of treatment

Remission

Inactive Crohn's disease defined by presence of CDAI score < 1 5 G for a minimum of 6 months in the absence of steroid or immunosuppressive therapy

CDAI = Crohn's Disease Activity Index. Reprinted with permission from ref. 32.

Table 4. Mean utility scores (95% confidence interval) for the tiiree hypothetical disease-severity states [32] Model

Mild disease Visit 1 Visit 2

Moderate disease Visit 1 Visit 2

Severe disease Visit 1 Visit 2

Time Trade-off

0.95 (0.94-0.97)

0.96 (0.95-0.97)

0.88 (0.87-0.91)

0.88 (0.86-0.91)

0.73 (0.69-0.76)

0.71 (0.68-0.75)

Standard Gamble

0.81 (0.79-0.83)

0.82 (0.80-0.85)

0.72 (0.69-0.75)

0.73 (0.71-0.77)

0.50 (0.46-0.54)

0.54 (0.50-0.59)

Visual Analogue Scale

0.80 (0.78-0.82)

0.82 (0.80-0.84)

0.57 (0.55-0.59)

0.61 (0.58-0.63)

0.27 (0.23-0.30)

0.31 (0.28-0.33)

Estimated CDAI

134 (122-164)

266 (215-318)

483 (404-562)

Estimated IBDQ

174 (167-1^30)

154 (140-169)

117 (107-128)

Reprinted with permission from ref. 32.

478

Pharmacoeconomics and inflammatory bowel disease

Additional analyses of these data showed that utility estimates are highly dependent upon methodology (Table 4). The VAS yielded the lowest utility estimates. This result was attributed to the wellrecognized phenomenon of 'response-spreading', a bias associated with visual scales which systematically lowers scores. Time Trade-off estimates were significantly higher than the Standard Gamble scores since approximately one-third of respondents refused to trade life expectancy. This finding was observed consistently irrespective of the health state of the patient. For these reasons the Standard Gamble was considered a more accurate reflection of utility than the other two measures.

Cost-utility models in Crolm's disease

Application of utility measurements Utility scores can compare the HRQL of patients with chronic diseases. In a study of symptomatic patients with angina a utility score of approximately 0.9 was reported [33]. In contrast patients with chronic renal failure have a utility of 0.4 - meaning very poor quality of life. In the previously described study a Standard Gamble utility score of 0.82 was shown in patients with mild CD. This suggests that the HRQL of patients with CD is substantially worse than those with heart disease. In summary, these utility data demonstrate that CD has a profoundly negative effect on the well-being of patients. How then can this information be used? One application is in cost-utility analysis.

Cost-utility analysis In a cost-utility analysis, treatment alternative regimens are compared in terms of the incremental cost per QALY [34]. As noted previously, society is generally willing to pay for health-care interventions that cost less than US$50 000 per QALY [11]. Cost-utility analyses are suitable for evaluating new treatments for the following reasons: (1) therapy now focuses on improving patient well-being rather than changing life expectancy; (2) a single measurement can assess both the positive and negative aspects of a treatment; and (3) society must make critical decisions regarding reimbursement for effective yet relatively costly new treatments

Only two cost-utility analyses have been described in IBD. Trallori and Messori [35] evaluated the incremental cost utility of 5-ASA maintenance therapy in comparison to a strategy of no-maintenance therapy. This model evaluated two hypothetical groups of 100 patients. The natural history of the disease was estimated using data from a large-scale survey of Italian gastroenterologists [36]. The efficacy of 5ASA maintenance therapy was derived from a meta-analysis performed by one of the authors [37]. Cost estimates were based on the Hay and Hay data [13] using a 5% annual discount rate. Both lifetime and 2-year time horizons were evaluated. Utility estimates were derived from 10 expert gastroenterologists who assigned a utility score to five health states (remission in non-operated patients, remission in operated patients, relapse not requiring hospitalization, relapse requiring hospitalization without surgical intervention, and relapse requiring hospitalization and surgical intervention). The lifetime cost per QALY of maintenance therapy was estimated to be US$5015. The authors concluded that chronic therapy with 5-ASA was associated with a small incremental benefit and costs. Sensitivity analyses indicated that the result was insensitive to differences in utility estimates but highly sensitive to variances in the cost of illness. If the cost of CD was decreased by 20% the cost per QALY gained was increased to US$26436. The effect of varying efficacy estimates for 5-ASA was not examined in this study. This landmark study is the first attempt to examine economic and quality-of-life outcomes for CD. It should be emphasized that the authors were limited by the quality of the available data at the time of publication. Since this time, additional information suggests that the marginal cost-utility estimate obtained in this study likely overstates the value of 5-ASA maintenance therapy. First, the efficacy of maintenance therapy is probably less than the 12% annual absolute risk reduction used in the model. A subsequent meta-analysis by Camma etal. [38] could not identify a statistically significant effect of maintenance therapy following a medically induced remission. A modest benefit (9% absolute risk reduction per year) was observed following surgery. The authors also assumed that 5-ASA therapy prevented surgery and hospitalization. No data are available to support these assumptions. Second, as noted

Brian G. Feagan previously, the cost estimates of Hay and Hay likely overestimate the economic burden of CD in the population. Since this model is highly sensitive to cost these inputs favor maintenance therapy. Twoway sensitivity analyses, which simultaneously examine the consequences of varying costs and efficacy estimates, might have provided useful information. Finally the utility and natural history estimates used by Trallori and Messori were derived by expert opinion and are almost certainly weighted towards a tertiary-care hospital experience. Again this bias would favor 5-ASA treatment. Notwithstanding these criticisms this study is an important contribution to the literature which illustrates the potential value of cost-utility models. The second cost-utility model, by Wong and colleagues [39], examined the role of infliximab therapy for the treatment of refractory CD. Data from a randomized controlled trial were used to estimate the efficacy of infliximab. The Olmstead County Markov model served as the basis for natural history and cost estimates. Utility estimates were adapted from the study of Gregor and colleagues [32]. A marginal cost-effectiveness ratio of $14 20040000 per QALY gained was estimated for infliximab therapy: and the investigators concluded that infliximab therapy is likely cost-effective depending on the durability of the treatment benefit. Since these data have been published only as an abstract it is difficult to critique the assumptions used in the model. However, a significant limitation of this analysis is the lack of long-term data regarding the efficacy of infliximab therapy.

479 of new technologies. Consideration of data from cost-utility models is likely to become an important part of the decision-making process.

References 1. 2. 3. 4. 5. 6.

7. 8.

9. 10. 11.

12. 13. 14.

Conclusion Over the past decade preliminary data regarding the pharmacoeconomics of inflammmatory bowel disease have emerged. The synthesis of this information has culminated in the first models which evaluate the marginal cost-utility of competing therapies. Since the validity of these comparisons is critically dependent upon accurate efficacy, natural history, cost and utility inputs investigators must identify areas where important data are lacking. Through this process, more sophisticated and valid models will be developed. Highly effective, yet costly, new treatments for IBD will soon be in widespread use. Because societal resources are limited, health-care providers must make intelligent choices regarding the introduction

15. 16.

17.

18.

19.

Andres PG, Friedman LS. Epidemiology and the natural course of inflammatory bowel disease. Gastroenterol Clin N Am 1999; 28: 255-75. Hanauer SB, Meyers S. Management of Crohn's disease in aduhs. Am J Gastroenterol 1997; 92: 559-66. Binder V. Prognosis and quality of hfe in patients with ulcerative colitis and Crohn's disease. Int Disabil Stud 1988;10:172-4. Torrance GW. Preferences for health outcomes and costutility analysis. Am J Manag Care 1997; 3: S8-S20. Spilker B. Quality of Ufe and pharmacoeconomics in clinical trials, 2nd edn. Philadelphia: Lippencott-Raven, 1996. Marra FO, Frighetto LO, Marra CA et al. Cost-minimization analysis of piperacillin/tazobactam versus imipenem/ cilastatin for the treatment of serious infections: a Canadian hospital perspective. Ann Pharmacother 1999; 33: 156-62. Teutsch SM, Murray JF. Dissecting cost-effectiveness analysis for preventive interventions: a guide for decision makers. Am J Manag Care 1999; 5: 301-5. Jimenez FJ, Guallar-Castillon P, Rubio TC, Guallar E. Cost-benefit analysis of Haemophilus lifluenzae type B vaccination in children in Spain. PharmacoEconomics 1999; 15: 75-83. Gerard K, Smoker I, Seymour J. Raising the quality of costutility analyses: lessons learnt and still to learn. Health Policy 1999; 46: 217-38. Weinstein MC, Stason WB. Foundations of cost-effectiveness analysis for health and medical practices. N Engl J Med 1977;296:716-21. Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. Can Med Assoc J 1992; 146:473-81. Brooten D. Methodological issues linking costs and outcomes. Med Care 1997; 35: 87-95. Hay JW, Hay AR. Inflammatory bowel disease: cost-ofillness. J Clin Gastroenterol 1992; 14: 309-17. Schulman KA, Ohishi A, Park J, Glick HA, Eisenberg JM. CHnical economics in clinical trials: the measurement of cost and outcomes in the assessment of clinical services through clinical trials. Keio J Med 1999; 48: 1-11. Jacobs P, Fassbender K. The measurement of indirect costs in the health economics evaluation literature. A review. Int J Technol Assess Health Care 1998; 14: 799-808. Feagan BG, Vreeland MG, Larson LR, Bala MV. Annual cost of care for Crohn's disease patients. A claims-based cost of illness evaluation. Am J Gastroenterol 2000; 95: 1955-60. Bernstein CN, Papineau N, Zajaczkowski J, Rawsthorne P, Okrusko G, Blanchard JF. Direct hospital costs for patients with inflammatory bowel disease in a Canadian tertiary care university hospital. Am J Gastroenterol 2000; 95: 677-83. Silver stein MD, Loftus EV, Sandborn WJ et al Clinical course and costs of care for Crohn's disease: Markov model analysis of a population-based cohort. Gastroenterology 1999; 117:49-57. Blomqvist P, Ekbom A. Inflammatory bowel disease: health care and costs in Sweden in 1994. Scand J Gastroenterol 1997; 32: 1134-9.

Pharmacoeconomics and inflammatory bowel disease

480 20. 21.

22. 23. 24. 25. 26.

27. 28.

29. 30.

Fitzpatrick R. Quality of life measures in health care. Br Med J 1992; 305: 74^7 (Abstract). Turnbull GK, Vallis TM. Quality of life in inflammatory bowel disease: the interaction of disease activity with psychosocial function. Am J Gastroenterol 1995; 90: 14504. Mitchell A, Guyatt G, Singer J et al. Quality of life in patients with IBD. J CHn Gastroenterol 1988; 10: 306-10. Munkholm P, Langholz E, Davidsen M, Binder V. Disease activity courses in a regional cohort of Crohn's disease patients. Scand J Gastroenterol 1995; 30: 699-706. Bergner M, Bobbitt RA, Carter WB, Gilson BS. The Sickness Impact Profile: development and final revision of a health status measure. Med Care 1981; 19: 787-805. Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992; 30: 473 83. Chambers LW, MacDonald LA, Tugwell P, Buchanan WW, Kraag G. The McMaster Health Index Questionnaire as a measure of quality of life for patients with rheumatoid disease. J Rheumatol 1982; 9: 780 4. Guyatt G, Mitchell A, Irvine EJ et al. A new measure of health status for clinical trials in inflammatorybowel disease. Gastroenterology 1989; 96: 804 10. Irvine EJ, Feagan B, Rochon J ct al. Quality of life: a valid and reliable measure of therapeutic efficacy in the treatment of inflammalory bowel disease. Gastroenterology 1994; 106: 287 96. Irvine EJ. Quality of life measurement of inflammatory bowel disease. Scand J Gastroenterol 1993; 199: 36 9. Hellinger FJ. Expected utility theory a risky choice with health outcomes. Med Care 1989; 27: 273 9.

31. 32. 33. 34. 35. 36. 37.

38.

39. 40.

Read JL, Quinn RJ, Berwick DM, Fineberg HV, Weinstein MC. Preferences for health outcomes. Comparison of assessment methods. Med Decis Making 1984; 4: 315-29. Gregor JC, McDonald JWD, Klar N et al. An evaluation of the utility measurement in Crohn's disease. Inflam Bowel Dis 1997;3:265-76. Nease RF, Kneeland T, O'Connor GT et al. Variations in patient utilities for outcomes of the management of chronic stable angina. J Am Med Assoc 1995; 273: 1185-90. Drummond MF, Heyse J, Cook J, McGuire A. Selection of end points in economic evaluations of coronary-heart-disease interventions. Med Decis Making 1993; 13: 184-90. Trallori G, Messori A. Drug Treatments for maintaining remission in Crohn's disease. A lifetime cost-utility analysis. PharmacoEconomics 1997; 11:444-53. Pera A, Rocca B. Le malattie infiammatorie croniche intestinali: profili di diagnosi e terapia. EDRA srl 1995; Messori A, Brignola C, Trallori G et al. Effectiveness of 5aminosalicylic acid for maintaining remission in patients with Crohn's disease: a meta-analysis. Am J Gastroenterol 1994:89:692 8. Camma C, Giunta M, Rosselli M, Cottone M. Mesalamine in the maintenance treatment of Crohn's disease: a metaanalysis adjusted for confounding variables. Gastroenterology 1997; 113: 1465 73. Wong JB, Loflus EV, Sandborn J, Feagan BG. Estimating the cost-effectiveness of infliximab for Crohn's disease. Gastroenterology 1999; 116: G0451 (Abstract). Hay AR, Hay JW. Inflammatory bowel disease: medical cost algorithms. J Clin Gastroenterol 1992; 14: 318 27.

24 Measuring quality of life in inflammatory bowel disease E. JAN IRVINE

Introduction Most patients with inflammatory bowel disease (IBD) are not significantly at risk of dying from their disease [1]. However, other serious outcomes, such as hospitalization and the need for surgery, are clearly related to disease severity [2] In contrast, daily functioning, health perception, life satisfaction, and ambulatory health-care use correlate best with health-related quality of life (HRQL) status. Thus, clinicians and researchers must assess both disease severity and H R Q L , using well-validated instruments. These applications allow increased patient participation in disease management, fuller assessment of the disease impact, and the ability to implement the most cost-effective therapies [3-7]. The HRQL instruments, which appear to play a critical role in IBD assessment, will be discussed in this chapter.

Disease activity measurement Traditionally, disease severity has been quantified in IBD by assessing symptoms, endoscopic or radiologic bowel appearance, and tissue or serum markers that reflect the degree of inflammation. These features have been assessed with three major objectives: (1) to stratify patients for optimal management; (2) to predict disease outcome; and (3) to measure response to therapy. Several activity indices are available and their pros and cons are discussed in the previous chapter. In general there are sufficient differences in the clinical manifestations and pathophysiology of Crohn's disease (CD) and ulcerative colitis (UC) to justify the development and application of distinct severity indices. These differences are, however, less important when assessing HRQL. Although there may be good reasons to use a unique instrument to answer a

particular research question, there is a strong correlation (r>0.5) between disease severity and the physical domain of HRQL and considerable overlap in the psychosocial HRQL features of CD and UC. Generally, HRQL instruments developed for use in IBD research have been applied in both diseases.

Defining HRQL HRQL measurement assesses both function and attitude in three broad categories: physical, social, and emotional domains, as they relate to health or chronic illness. HRQL encompasses physical symptoms, mobility, body image, pain and discomfort, ability to engage in recreation or leisure activities, social and family relationships, sexual function, enjoyment of food and drink, satisfaction in and performance of work or school activities, and emotional function [3-6]. In chronic illnesses such as IBD, HRQL scores are partly predicted by disease severity, but are also influenced by psychological and social factors. These include such features as the patient's cognitive function, knowledge concerning the disease, socioeconomic status, educational level, personality, appropriate use of coping strategies, social support network, culture, beliefs, and other less clear determinants [7]. These predictors of HRQL may help explain why no two IBD patients are exactly alike, even in the face of comparable disease severity. For example, one patient might be fully employed with a spouse and children, while another is socially withdrawn, depressed, and receiving long-term disability. HRQL features can also help explain why different patients choose different treatment pathways. For example, one patient with refractory UC might consider temporary cyclosporine therapy, to delay surgery for work or family reasons, while another might opt for immediate surgery. Other potential applications of HRQL mea-

Stephan R. Targan, Fergus Shanahan and Lor en C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 481-494. © 2003 Kluwer Academic Publishers. Printed in Great Britain

482 surement include assessing an individual patient's needs, monitoring health-care delivery, assessing or predicting disease outcome or efficacy of treatments, and evaluating the cost-utility of different treatment interventions [4, 6].

Measuring HRQL In most studies which have adequately assessed HRQL, standardized questionnaires are administered to obtain data from different populations, while providing useful information for patient treatment. The three kinds of HRQL instruments that are used include the global assessment, generic instrument, and disease-specific instrument [8, 9]. Advantages and disadvantages of each type of questionnaire are shown in Table 1 and are briefly discussed here, but are well reviewed in the literature [2-4, 6-9]. The global assessment provides a summary of HRQL, often using a graded scale such as good, fair, or poor or a percent score based on a 10-cm visual analog scale (VAS). Such scores permit rapid comparisons with other conditions but cannot identify the determinants of a particular health state. Generic instruments are derived in and for general populations and consist of multi-item problem lists that are meant to be gender-, age-, and disease-independent. Items can be clustered in subscores, using statistical techniques to describe a series of domains, such as 'physical function', 'somatic sensation', or 'mental health', or may also be collapsed into summary scores to compare different populations or diseases.

Measuring quality of life in inflammatory bowel disease Generic instruments are the most likely to detect unexpected factors affecting HRQL. Examples of these include the Sickness Impact Profile [10], the Short Form-36 [11], or the EuroQoL [12]. Diseasespecific instruments are also multidimensional inventories, that have been derived in and for patients with a single condition [8, 9]. Such instruments are used to detect important treatment effects or change over time. To optimize HRQL assessment, most studies now combine generic and diseasespecific instruments. Some of the questionnaires that have been applied in IBD HRQL research are shown in Table 2.

Factors affecting the portrayal of HRQL Traditionally, HRQL assessment has been assessed from the patient's viewpoint and is supported by data in IBD and other conditions showing that spouses, family members, and physicians impose biases that underestimate the impact of IBD on HRQL [2, 13]. Clinicians and researchers selecting a HRQL instrument must consider the strength of the psychometric properties validity, reliability, and responsiveness of the instrument [8, 9]. Validity is a comparison of the new index score against a reference score (convergent validity) or a construct of what the new index is measuring, such as a prediction that patients with less severe disease will have better HRQL scores (construct validity). Reliability is a statistical summary trait describing the degree of measurement

Table 1. HRQL assessment in IBD Advantages

Disadvantages

Global assessment Simple summary Easily administered Easily scored

Specific domains of dysfunction difficult to discern May not detect small but important changes

Generic instrument Permits comparisons among groups or populations Can detect unanticipated effects

Complex administration and scoring Insensitive to important clinical changes

Disease-specific instrument Reflects most important problems to a specific population Can detect important clinical changes with time or treatment

Complex administration and scoring May miss unexpected effects

Utility Necessary for economic analysis

Insensitive to detect changes over time

483

E. Jan Irvine Table 2. HRQL instruments used in Inflammatory bowel disease Global assessments Visual analog scale (10-cm line) Graded scale (excellent, very good, good, poor, extremely poor) Utilities (Standard Gamble or Time Trade-off; 0.0 death to 1.0 perfect health) Generic instruments Sickness Impact Profile (136 items, 12 subscores; higher score = worse HRQL) 1. 2. Short Form-36 (36 items, 8 subscores; score 0, worst-lOO, best) 3. Grogono and Woodgate (20 items, 10 subscales) 4. Psychological General Well-being (22 items; 6 subscales (anxiety depression, well-being, self-control, health, vitality); reliability 0.61-0.89; score 22-132, lower score better) Cleveland Clinic Questionnaire (47 items, 4 subscores) Euro-QL (5 items - utility) Disease-specific instruments 1. Inflammatory Bowel Disease Questionnaire (IBDQ; 32 items, 4 subscores: 32-224) 2. Rating Form of IBD Patient Concerns (RFIPC) (25 items, subscores, 0-100) 3. Cleveland Clinic Questionnaire (18 items, 4 subscores; 0-100) 4. Pentasa Study instrument (7 items 0-100)

error, as in test-retest reliability, or the correlations among individual items or subscores [14]. Responsiveness provides an estimate of the signal-to-noise ratio of an instrument and allows the user to understand the degree of change that is considered to be clinically important [15]. Psychometric properties of some disease-specific instruments are illustrated in Table 3. The application of non-validated instruments and unsanctioned modification of well-validated instruments by investigators who have not followed rigorous questionnaire development methods are among the most common problems facing HRQL research. Investigators not familiar with questionnaire development methods must understand the pitfalls of inappropriate and unjustified alteration of questionnaires. Selecting an instrument for a particular study should also be based upon the research question, population to be assessed, study design, type of response options needed, questionnaire length, whether it is interviewer-administered, and how it is scored and interpreted.

Historical descriptive studies One of the earliest studies assessing HRQL was a British study, employing a global assessment to describe the 'good' outcome for IBD patients. Sixtynine percent of a cohort of 501 Oxford patients who had survived their UC were described as having 'normal function' (requiring some ambulatory visits)

while 19.3% had 'essentially normal function' (with frequent visits) [16]. In examining the predictors of HRQL in American adolescents with UC, Michener et al. observed that better HRQL occurred when the time to diagnosis was short [17]. In a series of adults with CD, postsurgical patients were less likely to describe their HRQL as 'good' compared to those who had been treated only medically (40% versus 53%; p - 0.01), suggesting that more severe disease was associated with poorer HRQL [18]. Another series of patients with CD unanimously recalled poor function preoperatively, with 37-84% describing an impact on symptoms, personal relations, school and employment, recreation, sexuality, or body image [19]. At postoperative follow-up, 6-29%) reported dysfunction and dissatisfaction with their outcome, particularly in the presence of an ileostomy. These studies collectively suggested that more severe disease was associated with poorer HRQL and that most aspects of a person's life were affected by IBD. However, further work was needed to identify more specific problems. Mallett et al. interviewed 84 British UC patients and observed a limited social life in 25%o and reduced earning capacity in 21%, even when disease was well controlled [20]. Almost 75% described impaired domestic or sex lives, while approximately 30%o claimed to have better family relations as a result of their disease. In a similar study of 85 ambulatory patients with CD, 11%) felt their marriage had

484

Measuring quality of life in inflammatory bowel disease

Table 3. Disease-specific HRQL instruments IBDQ (Inflammatory Bowel Disease Questionnaire) 32 items, total score; 4 subscores (bowel, systemic, emotional, social) Total score 32-224; some users report average score per item (1 to 7) Higher score connotesbetter HRQL Face validity: obtained from literature review, patients, caregivers Discriminative validity: poorer scores with more severe disease Convergent validity: correlates with CDAI, global assessments. Rand Test-retest reliability: ICC = 0.90; Cronbach alpha: 0.83 Clinical change over time: within patients: 0.5-1.0 per item (16-32 total score) Several modifications of IBDQ published based on face validity only Stiort IBDQ 10 items, 1-7 Discriminative validity and sensitivity to change over time shown Reliability, internal consistency, responsiveness similar to IBDQ RFIPC 25 items, 4 factors (impact of disease, sexual intimacy, complications of disease, body stigma), score 0-100; test-retest reliability Individual item score 10 cm VAS Higher score = more worries and concerns Discriminative validity, more worries and concerns in hospitalized or anxious patients, Crohn's disease after surgery, or stoma Test-retest reliability: ICC = 0 .87 Responsiveness to change: surgical patients Cleveland Clinic Questionnaire 18 items, 4 subscales (functional/economic, social/recreational, affect/life in general, medical) Items score 1-5 - strongly agree to strongly disagree Reliability: Spearman r = 0.75-0.95 Higher score, poorer HRQL Responsiveness not demonstrated

suffered due to their disease, 33% reported less frequent sexual intercourse, and 14% had completely stopped all sexual activity [21]. Thirty-three percent had received treatment for depression, and many with impaired well-being had signs of neurosis. Joachim et al. in a similar study, reported 42% of 80 IBD patients had decreased overall life satisfaction [22]. In studies from McMaster University, Mitchell et al. identified 150 potential problems described by IBD patients and their caregivers [23]. Bowel symptoms such as frequent or loose stools and abdominal pain, feeling unwell or seriously fatigued, were the commonest and most severe symptoms. Emotional difficulties such as frustration and depression also played an important role, while social problems, such as canceling or missing social events, were somewhat less important. Surprisingly, emotional and social problems were elicited only after prompting subjects with a list of problems, suggesting that these problems were overlooked or underrated by patients.

families, or health-care givers. On average, patients had each experienced up to 50 of the problems listed. In a more recent study in children, the impact of IBD on physical and emotional domains was also greater than social impairment (/70.90). A short version, the SIBDQ, consisting of 10 of the original 32 questions, explained over 90% of the variance of the IBDQ in both UC and CD [82]. The SIBDQ was able to discriminate between groups with inactive versus active disease {p - 0.0006) and an average score deterioration of 0.93 + 0.55 (p 0.001) occurred in patients at the time of relapse. Han et al. [83] in testing for differences between UC and CD, further validated the SIBDQ in over 300 British UC patients. They confirmed the excellent reliability (>0.95) of the SIBDQ and convergent validity of the instrument and saw no advantage in adopting separate instruments for UC and CD. This questionnaire, which can be completed in 3 minutes, may therefore be useful in clinical trials or in clinical practice to identify patients who are coping poorly and who would benefit from adjuvant counseling or more aggressive therapy.

Psychological disturbance In IBD North et al. undertook two overviews examining psychological relationships and noted that early studies which reported psychological abnormalities were methodologically more flawed than those with negative findings [84, 85]. Studies in UC failed to find a direct association, while in CD there was an increased incidence of psychopathology, although less than that observed in irritable bowel syndrome. The most prevalent findings were those of anxiety and depression, and the degree of disturbance appeared to correlate with disease severity. Turnbull and Vallis noted that both psychological function and disease severity were needed to predict HRQL scores [86]. In 150 outpatients Kinash et al. found that younger patients ( ^ 3 5 years), females, those with a depressive mood or affective coping style (prayer, substance abuse) had the worst HRQL. Patients who used problem-solving techniques to reduce stress tended to do better [87]. Drossman et al. also noted that higher rates of concern about

Measuring quality of life in inflammatory bowel disease Table 6. Factors associated with poor HRQL in IBD Disease Severe Extensive Treatment IVIedical Surgical Surgery Koch pouch Brook ileostomy lleorectal anastomosis For selected patients

More medication Differing problems such as incontinence For selected patients For older/elderly patients

Complicated (abscess or leak) Handsewn anastomosis Elderly Female

surgery, being treated as dilTerent and feeling out of control, occurred in patients who used unconventional therapies [2, 25]. It is possible that these subjects perceived stress more than others and that this might play a role in depression and illness behavior. Levenstein et al. described seven of 29 patients who had no symptoms but did have endoscopic evidence of rectal inflammation, their perceived stress levels were higher than in patients who had no rectal inflammation [88]. A later study by the same group suggested that patients with quiescent UC, who scored in the upper tertile of the Perceived Stress Questionnaire, had a significantly higher risk of exacerbation over long-term follow-up than those with lower scores [89]. This may have been related to poorer sleep pattern, shorter remission, persistent histological inflammation, or the use of drugs that might trigger exacerbation. However, no studies have adequately shown a temporal relationship between depression or anxiety and disease occurrence. Katon summarized the impact of major depression on chronic medical illness as leading to increased visits and medical costs, increased functional impairment and decreased HRQL, increased somatization, difficulty adhering to selfcare regimens (medication compliance, diet or quitting smoking), and increased mortality [90].

Other factors affecting HRQL Demographic and lifestyle factors also appear to have an effect on HRQL. A list of factors predicting poor HRQL is shown in Table 6. These include gender, age, and smoking status. Several studies have

E. Jan Irvine observed poorer HRQL in women than in men [2, 8, 22]. In a Dutch study smoking was associated with poorer HRQL in females with CD, particularly young females, while males with UC who smoked moderately fared better than non-smoking UC patients [91]. However, no adjustment was made for disease severity, which explains approximately 40% of the IBDQ variance [62]. In patients taking HPN for CD, older patients seemed to do more poorly, likely due to the presence of comorbid disease [48]. Two surveys using the RFIPC and SF-36, respectively, observed that African-American patients had a significantly poorer quality of life (/?< 0.002) [92] and greater disease burden than non-African-Americans, with more frequent and longer hospitalizations, more work loss, and a greater likelihood of having a stoma than non-African-Americans, despite a similar clinical course [93]. However, an adjusted analysis suggested that differences were due to social and economic factors rather than disease features [94]. Thus, age, gender, socioeconomic status, and lifestyle practices may impact on HRQL. Clearly further work is needed to elicit the specific reasons for these differences.

491 prepared to advise and address individual worries and concerns. This includes providing letters of documentation for patients when disability arises or the need for retraining occurs. Membership in local societies should be encouraged for some patients, particularly early in the disease course. Future directions in HRQL assessment could consider the use of modular questionnaires and will need to further address crosscultural translation and validation. This is already occurring with instruments such as SIP, SF-36, PGWB, RFIPC, and IBDQ but further refinement of instruments will still be needed for broader applications. More efficient data collection and analysis techniques are also needed, a problem which might be addressed by the use of computerized questionnaires [96]. Finally, cost-effectiveness and cost-utility assessments are beginning to appear in the literature. However, some of these are premature, while others are assessing outdated or ineffective treatments. These will likely be the most important applications when a wider array of effective treatment options is available.

References 1.

Does HRQL really matter? HRQL is what matters most to patients with IBD. While the majority of patients have reasonable functional status and HRQL, some with particularly aggressive disease, depression or anxiety, a poor understanding of the disease, or maladaptive coping strategies tend to do less well. Appropriate tools can be used to identify these patients, to optimize management. Most clinicians do elicit data regarding symptoms; social and emotional issues; support network; family interactions; and ability to carry on with education, domestic, or job activities. Responses are then used to develop clear objectives, tailored to the individual, with patient involvement in decision-making (perhaps selecting a less efficacious drug with fewer adverse effects). Already, many patients self-medicate and appreciate the ability to do so. This is in keeping with an observed improvement in HRQL of UC patients who used a patient-developed guidebook for self-management [95]. The use of alternative medicines in IBD often occurs due to frustration with lack of efficacy of conventional treatments. As a new generation of immunologicals is being tested, we must still be

2.

3. 4.

5.

6. 7. 8. 9. 10.

Travis SPL. Review article: insurance risks for patients with ulcerative colitis or Crohn's disease. Aliment Pharmacol Ther 1997; 11:51-9. Drossman DA, Patrick DL, Mitchell CM et al. Healthrelated quality of life in inflammatory bowel disease: functional status and patient worries and concerns. Dig Dis Sci 1989; 34: 1379-86. Irvine EJ. Activity scores and quality of life indices in inflammatory bowel disease. Curr Opin Gastroenterol 1995; 11:331-6. Irvine EJ. Quality of life - rationale and methods for developing a disease-specific instrument for inflammatory bowel disease. Scand J Gastroenterol 1993; 28(Suppl. 199): 22-7. Maunder RG, Cohen Z, McLeod RS et al. Eff'ect of intervention in inflammatory bowel disease on health-related quality of life: a critical review. Dis Colon Rectum 1995; 38: 1147-61. Fitzpatrick R, Fletcher A, Gore S et al. Quality of life measures in health care. I: Applications and issues in assessment. Br Med J 1992; 305: 1074-7. Garrett JW, Drossman DA. Health status in inflammatory bowel disease: biological and behavioral considerations. Gastroenterology 1990; 99: 90-6. Guyatt GH, Feeny DH, Patrick DL. Measuring health related quality of life in chronic disease. Ann Intern Med 1993; 118:622-9. Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care 1989; 27(Suppl.):S217-32. Bergner M, Bobbitt RA, Carter V^B et al The Sickness Impact Profile: development and final revision of a health status measure. Med Care 1981; 19: 787-805.

492 11. 12. 13. 14.

15. 16. 17. 18. 19.

20. 21. 22. 23. 24. 25. 26.

27. 28. 29.

30.

31. 32. 33.

Measuring quality of life in inflammatory bowel disease Ware JE, Sherbourne CD. The MOS 36 item Short Form Health Survey (SF-36). Conceptual framework and item selection. Med Care 1992: 30: 473-83. The EuroQol Group. EuroQol - a new facility for the measurement of health-related quality of life. Health Policy 1990;16: 199-208. Guyatt GH, Mitchell A, Irvine EJ et al. A new measure of health status for clinical trials in IBD. Gastroenterology 1989;96: 804-10. Deyo RA, Diehr P, Patrick DL. Reproducability and responsiveness of health status measures: statistics and strategies for evaluation. Control Clin Trials 1991; 12: 14258S. Guyatt GH, Walter S, Norman G. Measuring change over time: assessing the usefulness of evaluative instruments. J ChronDis 1987; 40: 171-8. Edwards FC, Truelove SC. The course and prognosis of ulcerative colitis. Gut 1963; 4: 299 315. Michencr WM, Farmer RG, Mortimer EA. Long-term prognosis of ulcerative colitis with onset in childhood or adolescence. J Clin Gastroenterol 1979; 1: 301 5. Farmer RG, Whclan G, Fazio VW. Long-term follow-up of patients with Crohn's disease. Gastroenterology 1985; 88: 1818 25. Meyers S, Walfish JS, Sachar DB, Grccnstcin AJ, Hill AG, Janowitz HD. Quality of life after surgery for Crohn's disease: a psychological survey. Gastroenterology 1980; 78: 1 6. Mallctt SJ, Lennard-Joncs JE, Binglcy J, Gilon E. Colitis. Lancet 1978; 2: 619 21. Gazzard BG, Price HL, Libby GW, Dawson AM. The social toll of Crohn's disease. Br Med J 1978; 2: 117 19. Joachim G, Milne B. The effects of inflammatory bowel disease on lifestyle. Can Nurse 1985; 81: 39 40. Mitchell A, Guyatt G, Singer J ct al. Quality o'i life in patients with IBD. J Clin Gastroenterol 1988; 10: 306 10. Forget S, Issenman RM, Gold N, Irvine EJ. Assessment of functional status in children and adolescents with inflammatory bowel disease. Gastroenterology 1997; 112: A974. Drossman DA, Leserman J, Li Z et al. The rating form of IBD patient concerns: a new measure of health status. Psychosom Med 1991; 53: 701 12. Drossman DA, Leserman J, Mitchell CM et al. Health status and health care use in persons with inflammatory bowel disease: a national sample. Dig Dis Sci 1991; 36: 1746-55. Farmer RG, Easley KA, Farmer JM. Quality of life assessment of patients with inflammatory bowel disease. Cleve Clin J Med 1992;59:35-42. Grogono AW, Woodgate DJ. Index for measuring health. Lancet 1971: 1: 1024-6. Shivananda S, van Blankenstein M, Schouten WR, Themans B, vsDoes E. Quality of life in Crohn's disease: results of a case-control study. Eur J Gastroenterol Hepatol 1993; 5: 919-25. Hendriksen C, Kreiner S, Binder V. Long term prognosis in ulcerative colitis - based on results from a regional patient group from the county of Copenhagen. Gut 1985; 26: 15863. Sorensen VZ, Olsen BG, Binder V. Life prospects and quality of life in patients with Crohn's disease. Gut 1987; 28: 382-5. Guassora AD, Kruuse C, Thomsen OO, Binder V. Quality of life study in a regional group of patients with Crohn's disease. Scand J Gastroenterol 2000; 35: 1068-74. Munkholm P, Langholz E, Davidsen M et al. Disease activity courses in a regional cohort of Crohn's disease patients. Scand J Gastroenterol 1995; 30: 699-706.

34. 35. 36.

37. 38. 39.

40. 41.

42.

43.

44.

45. 46. 47.

48. 49. 50. 51. 52. 53.

Langholz E, Munkholm P, Davidsen M, Binder V. Course of ulcerative colitis. Analysis of changes in disease activity overyears. Gastroenterology 1994; 107: 3-11. Sonnenberg A. Disability and need for rehabilitation among patients with inflammatory bowel disease. Digestion 1992; 51: 168-78. Mayberry MK, Probert C, Srivastava E, Rhodes J, Mayberry A J F Perceived discrimination in education and employment by people with Crohn's disease: a case control study of educational achievement and employment. Gut 1992; 33: 312-14. Eckardt VF, Lesshafift C, Kanzler G et al. Disability and health care use in patients with Crohn's disease: A spouse control study. Am J Gastroenterol 1994; 89: 2157-62. Ferguson A, Sedgwick DM, Drummond J. Morbidity of juvenile onset inflammatory bowel disease: effects on education and employment in early adult life. Gut 1994; 35: 665-8. Sagar PM, Lewis W, Holdsworth PJ et al. Quality of life after restorative proctocolectomy with a pelvic ileal reservoir compares favorably with that of patients with medically treated colitis. Dis Colon Rectum 1993; 36: 584-92. Martin A, Dinca M, Leone L et al. Quality of life after proctocolectomy and ileo-anal anastomosis for severe ulcerative colitis. Am J Gastroenterol. 1998 Feb; 93: 166 9. Cohen RD, Brodsky AL, Hanauer SB. A comparison of the quality of life in patients with severe ulcerative colitis after total colectomy versus medical treatment with intravenous cyclosporin. Inflam Bowel Dis 1999; 5: 1 10. Kohler LW, Pcmberton JH, Zinsmeister AR et al. Quality of life after proctocolectomy. A comparison of Brooke ileostomy, Kock pouch, and ileal pouch anal anastomosis. Gastroenterology 1991; 101: 679 84. Tjandra JJ, Fazio VW, Church JM, Oakley JR, Milsom JW, Lavery IC. Similar functional results after restorative proctocolectomy in patients with familial adenomatous polyposis and ulcerative colitis. Am J Surg 1993; 165: 322 5. Scaglia M, Bronsino E, Canino V et al. Impatto della proctocoleclomia convenzionale sulla funzione sessuale (Influence of conventional proctocolectomy on sexual function). Minerva Chir 1993; 48: 903 10. Moody G, Probert CS, Srivastava EM, Rhodes J, Mayberry J F Sexual dysfunction amongst women with Crohn's disease: a hidden problem. Digestion 1992; 52: 179-83. Moody GA, Mayberry JF. Perceived sexual dysfunction amongst patients with inflammatory bowel disease. Digestion 1993; 54: 256-60. Colombel JF, Yazdanpanah Y, Laurent F, Houcke P, Delas N, Marquis P. Qualite de vie dans les maladies inflammatoires chroniques de I'intestin. Gastroenterol Clin Biol 1996; 20: 10771-77. Richards DM, Irving, MH. Assessing the quality of Hfe of patients with intestinal failure on home parenteral nutrition. Gut 1997;40:218-22. Mughal M, Irving M. Home parenteral nutrition in the United Kingdom and Ireland. Lancet 1986; 2: 383-7. Galandiuk S, O'Neill M, McDonald P, Fazio VW, Steiger E. A century of home parenteral nutrition for Crohn's disease. Am J Surg 1990; 159: 540-4; discussion 544-5. Detsky AS, McLaughlin JR, Abrams HB et al. Quality of life of patients on long-term total parenteral nutrition at home. J Gen Intern Med 1986; 1: 26-33. Nissan A, Zamir O, Spira R, Seror D, Alweiss T et al. A more liberal approach to the surgical treatment of Crohn's disease. Am J Surg 1997; 174: 339-41. McLeod RS, Churchill DN, Lock AN, Vanderburgh S, Cohen Z. Quality of life with ulcerative colitis preoperatively and postoperatively. Gastroenterology 1991; 101: 1307-13.

E. Jan Irvine 54.

55.

56.

57.

58. 59. 60. 61. 62.

63. 64.

65. 66. 67.

68.

69.

70. 71.

72. 73.

Provenzale D, Phillips-Bute B, Shearin M et al. Clinical predictors of post-colectomy health status in patients with a history of ulcerative colitis. Gastroenterology 1995; 108: A30 (Abstract). O'Bichere A, Wilkinson K, Rumbles S et al. Functional outcome after restorative Panproctocolectomy for ulcerative colitis decreases an otherwise enhances quahty of life. Br J Surg 2000; 87: 8802-7. Tiainen J, Matikainen M. Functional outcome of conversion of ileorectal anastomosis to ileal pouch anal anastomosis in patients with familial adenomatous polyposis and ulcerative colitis. Scand J Gastroenterol 1999; 34: 601-5. Irvine EJ, Feagan B, Rochon J et al. Quahty of hfe: a valid and reliable measure of outcome for clinical trials in inflammatory bowel disease. Gastroenterology 1994; 106: 287-96. Irvine EJ, Donnelly M. Quality of life is comparably impaired in patients with ulcerative colitis and irritable bowel syndrome. Gastroenterology 1994; 106: A517. Love JR, Irvine EJ, Fedorak RN. Quality of life in inflammatory bowel disease. J Clin Gastroenterol 1992; 14: 15-19. Martin A, Naccarato R, Fries W, Leone L. Quality of hfe in inflammatory bowel disease. Ital J Gastroenterol 1995; 27: 450-4. Casellas F, Lopez-Vivancos J, Badia X, Vilaseca J, Malagelada JR. Impact of surgery for Crohn's disease on healthrelated quahty of life. Am J Gastroenterol 2000; 95: 177-82. Irvine EJ, Bolin T, Grace E et al. and the International Quality of Life Study Group. Geographic differences in health related quality of life in inflammatory bowel disease. Gastroenterology 1997; 112: A1003. Farrokhyar F, Marshall JK, Cawdron R, Irvine EJ. Mood disorders worsen health-related quality of life in inflammatory bowel disease. Gastroenterology 2001; 119A (In press). Almeida RT, Hjortsway H, Strom M, Aimer S, Persson J. Technology assessment using the association between outcome measures and patterns of illness severity. Med Biol Eng Comput 1997; 35: 386-90. Moser G, Tillinger W, Sachs G et al. Disease-related worries and concerns: a study on out-patients with inflammatory bowel disease. Eur J Gastroenterol Hepatol 1995; 7: 853-8. Lange A, Haslbeck E, Bregenzer N, Gross V, Scholmerich J, Lamparter-Lang R. Ambulatory education of Crohn's disease/ulcerative colitis. Z Gastroenterol 1996; 34: 411-15. Maunder RG, de Rooey EC, Toner BB et al. Health-related concerns of people who receive psychological support for inflammatory bowel disease. Can J Gastroenterol 1997; 11: 681-5. Provenzale D, Shearin M, Phillips-Bute BG et al. Health related quality of life after ileoanal pull-through: evaluation and assessment of new health status measures. Gastroenterology 1997; 113:7-14. Tilhnger W, Mittermaier C, Lochs H, Moser G Healthrelated quality of life in patients with Crohn's disease. Influence of surgical operation-a prospective trial. Dig Dis Sci 1999; 44: 932-8. Robinson M, Hanauer S, Hoop R et al. Mesalamine capsules enhance the quality of life for patients with ulcerative colitis. Aliment Pharmacol Ther 1994; 8: 27-34. Singleton JW, Hanauer S, Robinson M. Quality of life results of a double-blind placebo controlled trial of mesalamine in patients with Crohn's disease. Dig Dis Sci 1995; 40: 931-5. Gasche C, Dejaco C, Waldhoer T et al. Intravenous iron and erythropoietin for anemia associated with Crohn's disease. Ann Intern Med 1997; 126: 782-7. Thomsen OO, Cortot A, Jewell D et al. A comparison of budesonide and mesalamine for active Crohn's disease. N Engl J Med. 1998; 339: 3 7 0 ^ .

493 74.

75.

76. 77. 78.

79.

80. 81. 82.

83. 84. 85. 86. 87. 88.

89.

90. 91.

92.

93.

Dupuy HJ. The Psychological General Weil-Being (PGWB) Index. In: Wenger NK, Masson ME, Furberg CD, Elinson J, eds. Assessment of Quality of Life in Clinical Trials of Cardiovascular Therapies. New York: Le Lacq, 1984: 17083. Martin F, Sutherland L, Beck IT et al. Oral 5ASA versus prednisone in short term treatment of Crohn's disease: a multicentre controUed trial. Can J Gastroenterol 1990; 4: 452-58. Greenberg GR, Feagan BG, Martin FM et al. Oral budesonide for active Crohn's disease. N Engl J Med 1994; 331: 836-41. Feagan BG, Rochon J, Fedorak RN et al. Methotrexate for the treatment of Crohn's disease. N Engl J Med 1995; 332: 292-7. Targan S, Hanauer SB, Van Deventer SJH et al. A shortterm study of chimeric monoclonal antibody cA2 to tumor necrosis factor a for Crohn's disease. N Engl J Med 1997; 337: 1029-35. Rutgeerts P, D'Haens G, Targan S et al. Efficacy and safety of retreatment with anti-tumour necrosis factor antibody (infliximab) to maintain remission in Crohn's disease. Gastroenterology 1999; 117: 761-9. Feagan B, McDonald JWD, Rochon J et al. Chronic low dose cyclosporine in the treatment of Crohn's disease. N Engl J Med 1994: 330: 1846-51. Irvine EJ, Feagan B, Wong C. Does self-administration of a quality of life instrument change the results? J Clin Epidemiol 1996; 49: 1177-85. Irvine EJ, Zhou Q, Thompson A and the CCRPT Investigators. The short IBDQ: a quality of life instrument for community physicians managing inflammatory bowel disease. Am J Gastroenterol 1996; 91: 1571-8. Han SW, Gregory W, Nylander D et al. The SIBDQ: Further validation in ulcerative colitis patients. Am J Gastroenterol 2000; 95: 145-51. North CS, Alpers DH. A review of studies of psychiatric factors in Crohn's disease. Ann Clin Psychiatry 1994; 6: 117-24. North CS, Clouse RE, Spitznagel EL, Alpers DH. The relation of ulcerative colitis to psychiatric factors: a review of findings and methods. Am J Psychiatry 1990; 147: 974-8. Turnbull GK, Vallis TM. Quality of life in inflammatory bowel disease: the interaction of disease activity with psychosocial function. Am J Gatroenterol 1995; 90: 1450-4. Kinash RG, Fischer DG, Lukie BE, Carr TL. Inflammatory bowel disease impact and patient characteristics. Gastroenterol Nurs 1993; Feb: 147-55. Levenstein S, Prantera C, Varvo V et al. Psychological stress and disease activity in ulcerative colitis: a multidimensional cross-sectional study. Am J Gastroenterol 1994; 89: 121925. Levenstein S, Prantera C, Varvo V et al. Stress and exacerbation in ulcerative colitis: a prospective study of patients enrofled in remission. Am J Gastroenterol 2000; 95: 121320. Katon W Editorial: The impact of major depression on chronic medical illness. Gen Hosp Psychiatry 1996; 18: 215-19. Russel MG, Nieman FH, Bergers JM, Stockbrugger RW Cigarette smoking and quality of life in patients with inflammatory bowel disease. Eur J Gastroenterol Hepatol 1996; 8: 1075-81. Eisen GM, Straus WL, Sandler RS et al. Health status in Crohn's disease: comparison of the rating form of IBD patient concerns in African and non-African Americans. Gastroenterology 1996; 110: A15. Straus WL, Eisen GM, Sandler RS et al. Race and Crohn's disease: clinical and therapeutic comparison of African

Measuring quality of life in inflammatory bowel disease

494

94. 95.

Americans and non-African Americans. Report of a multicenter survey. Gastroenterology 1996; 110: A1021 Straus WL, Eisen GM, Sandler RS, Murray SC, Sessions JT. Crohn's disease: does race matter? Am J Gastroenterol 2000; 95: 479-83. Drummond HE, Ghosh S, Ferguson A, Brackenridge D, Tiplady B. Electronic quality of life questionnaires: a

96.

comparison of pen-based electronic questionnaires with conventional paper in a gastrointestinal study. Qual Life Res 1995; 4: 21-6. Kennedy AP, Thompson DG, Robinson AJ, Wilkin D. Randomized controlled trial of a patient-centred guidebook; effect on quality of life in ulcerative colitis. Gastroenterology 2000; 118(Suppl.2): A213.

25 Clinical pliarmacology in inflammatory bowel disease: optimizing current medical therapy LAURENCE J. EGAN AND WILLIAM J. SANDBORN

Introduction Effective therapy of inflammatory bowel disease (IBD) often requires the implementation of diverse pharmacologic strategies (Fig. 1). Unlike, for example, the treatment of an acid-peptic disease where blockade of the proton pump, a specific target enzyme, is highly effective in inhibiting the pathogenic process, investigators have been unable to identify a specific unifying etiologic or pathogenic factor in IBD to pharmacologically target. For this reason, current therapy of ulcerative colitis (UC), Crohn's disease (CD) and pouchitis often requires the use of more than one drug that belong to different therapeutic classes, and that are non-specific in action. In this chapter the clinical pharmacology of drugs that are effective in IBD is reviewed. The pharmacokinetics and mechanisms of action of these agents are discussed, with particular reference to their use in IBD patients. Although the list of medicines available to treat IBD is expanding, none is universally effective or safe. Therefore, we also explore various strategies to individualize and optimize drug therapy in IBD, including therapeutic drug monitoring and the potential use of pharmacogenetic data to guide Systemic inhibitors of inflammation Glucocorticoids

therapeutic decision-making. Several agents recently introduced for use in IBD, such as mycophenolate mofetil, interleukin (IL)-10 and thalidomide are not discussed because of a paucity of information on their use.

Aminosalicylates Chemistry Drugs that contain 5-aminosalicylic acid (5-ASA) have been used successfully in the treatment of IBD for decades. The aminosalicylates have evolved to become a cornerstone of modern IBD therapy since the observation that the 5-ASA-containing agent sulfasalazine, designed to provide both anti-inflammatory and antibacterial activities for rheumatoid arthritis therapy, produced symptomatic improvement of bowel disease in patients with coexisting colitis [1]. Sulfasalazine consists of one molecule of 5-ASA azo-bound to a molecule of the sulfa antibiotic sulfapyridine. A number of studies have suggested that the therapeutic action of sulfasalazine in IBD resides with the 5-ASA moiety, rather than with sulfapyridine [2-4]. Olsalazine is an azo-conjugated dimer of 5-ASA, and balsalazide consists of a molecule of 5-ASA azo-bound to 4-aminobenzoylalanine, an inert carrier (Fig. 2).

I mmunosuppressants Azathioprine, methotrexate cyclosporine

Cytokine modulators Anti-TNFa, IL-10

Topical inhibitors of inflammation Aminosalicylates

Figure 1. Effective therapy of IBD often requires a combination of diverse pliarmacological approaches.

Mechanisms of action 5-ASA differs from salicylic acid only by the addition of an amino group at the 5 (meta) position. However, this change produces a chemical entity with pharmacologic properties different from conventional salicylates, and appears to be necessary for activity in IBD. The primary anti-inflammatory, antipyretic and analgesic actions of salicylates such as aspirin are due to blockade of prostaglandin

Stephan R. Targan, Fergus Shanahan and Lor en C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 495-521. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Clinical pharmacology in inflammatory bowel disease

496

COOH

COOH

^^-•^S_ 6 )

Distal jejunum, proximal ileum

Colon 15 mM [30], small bowel 0.3 mM [33]

40% [34], 54% [30]

1.5-4 g

Mesalamine

Pentasa

Ethylcellulose-coated microgranules (time and pH dependant release)

Entire small bowel, colon

Small bowel 0.3 mM [35], colon 13 mM [30]

30% [36], 36% [30]

2-4 g

Sulfasalazine

Azulfidine

5-ASA azo bound to sulfapyridine

Colon

Colonic lumen 7 mM [37]

11% [31]

1-4 g

Olsalazine

Dipentum

5-ASA dimer, linked by azo-bond

Colon

Colonic lumen 14-23 mM [37], colon 24 mM [30]

10% [31], 17% [32], 22% [30]

1-3 g

Balsalazide

Colazide, Colazal

5-ASA azo-bound to inert carrier

Colon

Sites of delivery

Absorption^ [Ref.]

Daily dose range 1-4 g

2-6.75 g

^Luminal 5-ASA concentrations achieved in these studies of the various formulations should not be directly contrasted because different doses of drug were administered in each study. ^Absorption is estimated from cumulative urinary excretion of 5-ASA + acetyl-5-ASA, expressed as a percentage of total administered dose.

498

Clinical pharmacology in inflammatory bowel disease

of Pentasa allows 5-ASA to dissolve and diffuse into the gut lumen over several hours. Three azo-conjugated preparations of 5-ASA are available: sulfasalazine, balsalazide and olsalazine. Compared to free 5-ASA, oral administration of azoconjugated 5-ASA results in greatly reduced systemic absorption of the active drug, and therefore greater delivery of drug to the site of presumed pharmacologic activity [30-38], Intestinal bacteria contain azo-reductase activity, and are most abundant in the terminal ileum and colon. Azo-conjugated aminosalicylates reach the distal bowel largely unabsorbed, and there are effectively metabolized to yield free 5-ASA and the carrier molecule. In the case of olsalazine, two molecules of 5-ASA are released. These preparations thus deliver high concentrations of active drug to the terminal ileum and colon. Compared to the delayed-release preparations, azoconjugated aminosalicylates are less subject to absorption in the small bowel and consequently, less drug is delivered systemically (Table 1). A number of factors may affect the delivery of orally administered 5-ASA to the site of disease in IBD patients. For example, rapid intestinal transit may decrease exposure of the small intestine to delayed-release preparations of 5-ASA, and alterations in bacterial flora due to antibiotic use or small bowel bacterial overgrowth could affect liberation of 5-ASA from the azo-bound aminosalicylates. Clinically significant differences between the available oral 5-ASA preparations have been diflftcult to detect and, compared to sulfasalazine, the newer products have not been shown to have superior clinical efficacy. However, aminosalicylates lacking sulfapyridine are tolerated better than sulfasalazine [39]. Comparisons between doses of azo-bound aminosalicylates and free aminosalicylates should be made on the basis of molar content of 5-ASA; thus, 1 g of sulfasalazine contains approximately 0.4 g of 5ASA. Rectal administration of 5-ASA as a suppository or enema provides a high concentration of drug at the site of disease in distal UC. Suppositories deliver 5-ASA to the rectum only. Enemas contain an aqueous suspension of 5-ASA that gradually dissolves into the luminal free water to deliver active drug to the rectum and left colon [40]. The effectiveness and absorption of rectally administered 5-ASA depend on the patient's ability to retain the drug, which can be difficult for individuals with active proctitis. The absorption of rectal 5-ASA averages around 20% of the administered dose.

Adverse effects Aminosalicylates are well tolerated, especially the newer drugs lacking the sulfapyridine moiety of sulfasalazine [39, 41]. Sulfasalazine, but not 5-ASA, has been associated with oligospermia [42] and folate deficiency [43]. Adverse reactions to 5-ASA are generally mild and reversible. In most of the efficacy trials the frequency of drug reactions sufficient to warrant discontinuation was not significantly greater in the active treatment group (3-4%) than in the placebo group. The most frequently reported side effects of 5-ASA include dizziness, fever, headache, abdominal pain, nausea and rash [44-46]. Certain rare and potentially serious adverse events are occasionally associated with 5-ASA use. These include pneumonitis and pericarditis [47, 48], aplastic anemia [49], pancreatitis [50] and rarely, exacerbation of existing colitis [51]. High doses of aminosalicylates that are associated with higher salicylate blood levels have generated concern about potential nephrotoxicity [52, 53]. Careful studies of renal function and urinary sediment have documented subtle changes indicative of tubular epithelial cell damage in IBD patients receiving high doses of 5-ASA drugs [54]. However, despite this concern, clinically significant renal dysfunction appears to occur very infrequently in IBD patients treated with 5-ASA drugs.

Optimizing therapy The effectiveness of aminosalicylate therapy of IBD depends on getting a high concentration of the drug to the site of disease. The different preparations of 5ASA listed in Table 1 offer unique characteristics that can theoretically be exploited to deliver the greatest amount of drug to the site of disease in individual patients. Thus, many clinicians use 5-ASA suppositories and enemas in patients with proctitis or proctosigmoiditis. Indeed, even in the presence of pancolitis, the use of rectal 5-ASA may be very useful to control distal disease, the source of the most troublesome symptoms. In more extensive UC, both the azo-conjugated and delayed-release formulations of oral 5-ASA are suitable. No individual preparation has been conclusively shown to have an efficacy advantage over another in UC. For the treatment of small bowel CD, the timed-release Pentasa preparation has the theoretical, but not clinically proven, advantage over the other oral aminosalicylates.

Laurence J. Egan and William J. Sandborn

499

Table 2. Comparison of the potency, bioavailability and common doses of oral and rectal glucocorticoid preparations for use in IBD Drug

Anti-inflammatory potency

Bioavailability

Usual maximum dose, mg/day 100-200

Hydrocortisone

1

Oral: > 50%; rectal: 16-30%

Prednisone

4

Oral: > 50%

40-80

Prednisolone

4

Oral: > 50%; rectal, as metasulfobenzoate: < 10%

40-80

Budesonide, controlled-ileal-release

200

9-15

Oral: 10%; rectal: 16%

A consistent theme to emerge from clinical trials of aminosalicylates in IBD is that more is better. For active UC, doses of 5-ASA in the 2.4-4.8 g per range were more efficacious than 1.6 g or placebo [44, 46]. Similarly, in the treatment of active CD, 5-ASA doses of 3-4 g per day [55] are more effective than 1-2 g per day [56, 57]. For maintenance of medically or surgically induced remission in CD, 5-ASA doses of 2.4 g or greater are effective, but lower doses are not [5860]. Furthermore, in studies which have compared oral and rectal 5-ASA for distal UC, combination therapy seems to be superior to either alone. Higher doses of 5-ASA produce higher mucosal drug levels [61], and the occurrence of ileal relapse after resection in CD patients on prophylactic 5-ASA, 2.4 g per day, was found to be associated with lower ileal mucosa drug levels [62]. Taken together, these data support the concept that the optimum use of aminosalicylates in active IBD demands the highest tolerated dose. In quiescent disease, lower doses may be more tolerable to the patient and are less costly. 5-ASA is primarily metabolized by A^-acetylation to a product that does not appear to possess significant anti-inflammatory activity. Of the two A^acetyltransferase isoenzymes, A^-acetyltransferase-1 has much greater affinity for aminosalicylates than A^-acetyltransferase-2 [63]. A^-acetyltransferase-1 is highly expressed in the epithelium of the small bowel and colon [64]. The NATl gene is polymorphic, and inactivating mutations are associated with a phenotype of decreased enzymatic activity [65]. Conversion of 5-ASA to acetyl-5-ASA by A^-acetyltransferase-1 takes place in the gut epithelium [66, 67] and in the liver, where phase II conjugation with glucuronic acid also takes place. After oral or rectal administration of an aminosalicylate, acetyl-5-ASA is the chief

species present in the serum and urine [32, 33, 68] and in the intestine [36]. One study has examined the potential role of NATl polymorphisms and A^-acetyltransferase-1 activity in colonic mucosa as a determinant of clinical response to 5-ASA [69]. In this study IBD patients classified on clinical grounds as normal or poor responders to 5-ASA were compared with respect to NATl polymorphisms and colonic Nacetyltransferase-1 activity. Interestingly, poor responders had a lower rate of 5-ASA acetylation than normal responders, and three out of five poor responders, but none of the normal responders, were heterozygous for NATl polymorphisms. These intriguing but preliminary data run contrary to the notion that acetylation inactivates 5-ASA, but require confirmation.

Glucocorticoids Chemistry The glucocorticoids that are useful in the therapy of IBD are synthetic molecules that bear a number of modifications or substitutions of the naturally occurring steroid, Cortisol (hydrocortisone). These alterations to the steroid molecule result in dramatic differences in its absorption, transport, metabolism and molecular actions. The synthetic analogs of Cortisol have been developed with the primary goal of maximizing glucocorticoid activity while minimizing mineralocorticoid effects, and decreasing the protean adverse effects of high quantities of systemically active glucocorticoid. Prednisone and prednisolone are the steroids most commonly used to treat IBD, and have an intermediate duration of action

500

Clinical pharmacology in inflammatory bowel disease

that allows once-daily dosing. A variety of high glucocorticoid potency, low-bioavailability synthetic molecules have been developed. Delayed-release oral preparations of these molecules, for example controlled-ileal-release budesonide, allow topical delivery of the drug to the site of disease in IBD (Table 2).

Mechanisms of action Glucocorticoids exert a diverse range of effects on most tissues of the body. When present at supraphysiological amounts after exogenous administration, the glucocorticoids have pronounced anti-inflammatory and immunosuppressive actions. The majority of the physiologic and pharmacologic etfects of glucocorticoids are dependent on diffusion of the steroid across the cell membrane to the cytoplasm where they bind to the glucocorticoid receptor, a member of the superfamily of steroid hormone receptors (Fig. 3). The glucocorticoid receptor exists in the cytosol bound to the 90 kDa heat-shock proteins that act as chaperone molecules and prevent nuclear translocation. After steroid binding, the glucocorticoid-glucocorticoid receptor-heat-shock protein complex undergoes conformational changes that allow dissociation of the steroid and its receptor from the heat-shock proteins. The steroid-receptor complex is then transported to the nucleus where it can interact via its zinc finger domains with glucocorticoid response elements that are found in the 5' promotor regions of up to 100 genes. Binding of the steroid-receptor complex to glucocorticoid response elements changes the rate of transcription of target genes. In some situations this favors the assembly of the transcription initiation complex and up-regulates the rate of gene transcription. In contrast, when the steroid-receptor complex binds to negatively acting glucocorticoid response elements, repression of gene transcription results. In addition to these direct effects on the rate of gene transcription, activated glucocorticoid receptors also physically associate with other transcription factors in the nucleus via leucine zipper interactions, and reciprocally modulate each others' activities. One important example of this mechanism is the inhibition of N F - K B and activator protein-1 binding to DNA by activated glucocorticoid receptors [70, 71]. Glucocorticoids also inhibit the activity of N F - K B by inducing transcription of its natural inhibitor, IKBOC [72, 73]. These direct effects, and numerous indirect effects resulting from the modulation of gene expression,

Figure 3. Mechanism of action of glucocorticoids (GC). After dissociation from corticosteroid-binding globulin (CBG), GC diffuse across tfie plasma membrane. In the cytoplasm the steroid binds to the glucocorticoid receptor (GCR), inducing conformational changes that cause dissociation of two chaperone molecules, heat-shock protein-90 (hsp90). The GC-GCR complex migrates to the nucleus and binds to glucocorticoid response elements (GRE) in the promoter regions of various target genes, regulating the rate of gene transcription. In addition, the GC-GCR complex interacts with important peptide transcription factors in the nucleus such as NF-KB, and can regulate their activity (not illustrated).

are the major molecular mechanisms of action of glucocorticoids. The dramatic capacity of glucocorticoids to inhibit inflammation relates to the potent inhibitory actions that these drugs exert on molecules that promote the activity of immune and inflammatory processes (Table 3). Effects of glucocorticoids that are well documented include inhibition of the production of proinflammatory cytokines such asTNF-a and IL-1, and chemokines such as IL8; repression of the transcription of the genes for certain enzymes such as inducible nitric oxide synthase, phospholipase A2 and cyclo-oxygenase II; and blockade of adhesion molecule expression. These molecular actions of glucocorticoids result in a blockade of leukocyte migration and function, and inhibit the effects of numerous important peptideand lipid-derived mediators of inflammation. A number of studies have examined the antiinflammatory or immune-modulating effects of glucocorticoids in patients with IBD. Nitric oxide is found in elevated quantity in the inflamed mucosa in IBD [79], where it is produced by the inducible form of nitric oxide synthase. Glucocorticoids inhibited in-vitro production of nitric oxide by cultured mucosal biopsy specimens, but glucocorticoid treatment of severe UC was not shown to down-regulate immunoreactive nitric oxide synthase [80]. In

Laurence J. Egan and William J. Sandborn

501

Table 3. Gene targets of glucocorticoids that are important in mediating their anti-inflammatory actions Target substance

Effect of glucocorticoid

Target cell

Mechanism [ref.]

Result

Cytokines and chemokines IL-1JNFa,IL-8

Decrease

Many

Inhibition of NF-KB [15]

Decrease

Lymphocytes

Inhibition of AP-1 [74]

Decreased inflammatory signaling Decreased lymphocyte activation

Decrease

Epithelial and endothelial

Inhibition of NF-KB [15]

Decreased vascular permeability

Phospholipase A2

Decrease

Many

Inhibition of NF-KB, induction of lipocortin gene expression [75]

Decreased eicosanoid production

Cyclo-oxygenase II

Decrease

Many

Inhibition of NF-KB [15]

Decreased prostaglandin production

Receptors IL-1 receptor type II

Increase

Many

Direct induction of gene expression [76]

Decreased availability of IL-1 for biological activity

Adhesion molecules E-selectin

Decrease

Endothelial

Inhibition of NF-KB [77]

Decrease

Epithelial, monocytic

Inhibition of NF-KB [78]

Decreased leukocyte extravasation Decreased leukocyte migration

IL-2 Enzymes Inducible nitric oxide synthase

Intracellular adhesion molecule-1

patients with active CD prednisolone, but not budesonide, has been demonstrated to inhibit neutrophil and peripheral blood lymphocyte expression of activation markers and the proinflammatory cytokines, IL-1 and TNF-a [81]. Similarly, expression of the adhesion molecules intracellular adhesion molecule3 and P2-integrin by peripheral blood lymphocytes in CD and UC was lower in patients receiving prednisone [82]. At the level of the inflamed mucosa in UC, prednisolone enemas and oral tablets have been shown to decrease the luminal concentrations of the arachidonic acid metabolites prostaglandin-E2 and leukotriene-B4 [22]. At the mechanistic level prednisolone therapy inhibits the activation of N F - K B in the gut mucosa [11, 14], supporting an important role for this transcription factor in mediating the beneficial effects of glucocorticoids in IBD. Overall, these human studies provide evidence that the effects of glucocorticoids on expression of the genes listed in Table 3 underlie the in-vivo anti-inflammatory effects of these drugs in IBD.

Pharmaceutics and pharmacokinetics The conventional synthetic corticosteroids prednisone and prednisolone are the drugs of this class most commonly used in IBD. Prednisone is a prodrug that requires metabolic activation in the liver to prednisolone. There is some evidence that chronic liver disease may impair this process [83], and for this reason prednisolone may be preferable in patients with advanced liver disease. Both of these drugs are well absorbed after oral administration and bioavailability is high, averaging over 70%. There is some evidence suggesting decreased absorption of prednisolone in patients with CD of the small bowel [84, 85], but it is not clear if this difference is cHnically significant. After absorption, glucocorticoids circulate in the plasma partially bound to corticosteroidbinding globulin and albumin, but it is only the free form that is available for biologic activity. The affinity of different corticosteroids for these carrier proteins varies considerably and in part determines anti-inflammatory potency. Although various physiologic and pathologic conditions alter the plasma concentration of the steroid-binding pro-

502

Clinical pharmacology in inflammatory bowel disease

teins, the resulting changes that occur in the concentration of free glucocorticoid are transient and are not likely to be significant. There has been a considerable effort to develop glucocorticoids for delivery to the site of disease in IBD, with the goal of providing high disease-tissue drug concentrations, with less systemic absorption than conventional steroids. One approach is rectal administration of the drug. Hydrocortisone enemas are used in ulcerative proctosigmoiditis to provide topical delivery of glucocorticoid directly to the diseased tissue, with good clinical results. Absorption of hydrocortisone enemas is less than after oral administration; nevertheless, bioavailability ranges from about 15% to 30% [86, 87], sufficient to suppress adrenal steroid production. A derivative of prednisolone with a metasulfobenzoate substitution is also used rectally in distal UC. Compared to prednisolone this agent is less well absorbed and should therefore provide a therapeutic advantage by acting locally and causing less systemic exposure to glucocorticoid activity [88, 89]. However, careful studies of adrenal function and bone turnover demonstrate that prednisolone metasulfobenzoate enemas do result in significant systemic glucocorticoid activity when used to treat distal UC [90, 91]. Budesonide is a synthetic analog of prednisolone that has proved to be therapeutically effective as oral delayed-release capsules in CD of the ileum and right colon, and as an enema in distal UC. This drug has high affinity for the glucocorticoid receptor resulting in very high potency, and is subject to rapid inactivation in the liver leading to low systemic bioavailability [92, 93]. A controlled-ileal-release formulation of budesonide is used orally, in which the drug is formulated as capsules containing microgranules of budesonide coated with Eudragit LI00-55, an acrylic resin that dissolves above pH 5.5 to release the drug in the distal ileum. The systemic bioavailabihty of oral budesonide capsules is approximately 10% [94], and the majority of absorption appears to occur in the distal small bowel. However, even this degree of systemic exposure is sufficient to suppress adrenal steroid production, but is less than occurs with equivalent doses of prednisolone [95-97]. For rectal delivery, budesonide is formulated as a suspension enema. The bioavailability of budesonide administered by this route, mean 16%, is greater than found with the oral controlled-ileal-release formulation, probably reflecting absorption into the distal rectal veins that drain directly into the systemic venous system [98]. Rectal administration of budesonide

produces dose-dependent suppression of adrenal function [99]. Thus, with both the oral and rectal formulations of budesonide, some degree of adrenal suppression occurs, but this appears it to be less than seen with conventional steroids. A number of other high glucocorticoid potency, low systemic bioavailability corticosteroids have been developed and tested for use in IBD. Tixocortol pivalate is a potent, well-absorbed and rapidly metabolized steroid molecule that has been used as an enema for distal UC. Fluticasone is a high potency fluorinated corticosteroid that is poorly absorbed from the intestine and undergoes extensive first-pass inactivation, leading to systemic bioavailability of less than 1%. Fluticasone has been evaluated for oral use in CD and UC, but in preliminary studies was not found to efficacious [100, 101]. Beclomethasone is a highly potent glucocorticoid with low systemic bioavailability. Preliminary studies of beclomethasone enemas in active distal UC have revealed efficacy and little adrenal suppression, but further studies are needed [102-104].

Adverse effects The use of glucocorticoids for prolonged periods in IBD patients is associated with serious side-effects. These are not different from those that occur when these drugs are given chronically for other indications. In the National Cooperative Crohn's Disease Study, adverse drug effects that occurred more frequently in prednisone-treated patients than in those receiving placebo were petechiae and bruising of skin, striae, moon face, acne, hypertension and infection [105]. About 30% of prednisone-treated patients in this study experienced a side-effect of this therapy, and the likelihood of occurrence of most of the reported adverse events increased with duration of exposure to the drug. Physicians who care for IBD patients requiring frequent use of glucocorticoids also recognize that the full gamut of steroid-induced problems develop with prolonged consumption of these drugs. Certain adverse effects of glucocorticoids are of particular relevance in IBD. Hyperglycemia and non-alcoholic steatohepatitis are frequent metabolic complications of chronic glucocorticoid therapy. Abnormal liver chemistry test results in IBD patients can also be due to primary sclerosing cholangitis, or short-bowel syndrome, so one must consider steroid-induced steatohepatitis in the differential diagnosis of abnormal liver tests in IBD. Individuals with IBD are at

Laurence J. Egan and William J. Sandborn greatly increased risk for the development of osteoporosis [106-108]. The use of glucocorticoids may increase this risk, as in one study 58% of steroidtreated patients had lumbar spine osteopenia, compared to 28% of steroid non-users [108]. In these studies the use of glucocorticoids may simply be a marker of more active disease, itself a risk factor for osteoporosis. However, in diseases other than IBD that are not strongly associated with osteoporosis, glucocorticoid use induces bone loss and fractures [109]. Aseptic necrosis of the hip is an unpredictable, devastating complication of glucocorticoid use. The newer high-potency, low-bioavailability steroids such as budesonide may be associated with less risk of inducing bone disease. Options for treating and preventing bone disease due to glucocorticoids include calcium and vitamin D, estrogen for postmenopausal women, and bisphosphonates. IBD patients who have been treated with glucocorticoids within 1 year should receive a short course of steroids perioperatively to protect against an inadequate adrenal response to the stress of surgery [110].

Optimizing ttierapy It is paradoxical that glucocorticoids, the first therapy proven effective for IBD [111,112] have not been as carefully studied as most other classes of drugs. The optimum doses of glucocorticoids to induce remission in active UC and CD are not known. Few studies have addressed the question of the optimum rate at which to taper the dose of steroids after a patient responds, so individual clinical judgement and experience, rather than data, direct practice. Nevertheless, there are some well-established principles to guide the effective and safe use of glucocorticoids in IBD. Almost all IBD patients with moderately to severely active disease should be treated with a corticosteroid, as these drugs are highly efficacious for induction of remission and are probably responsible for reducing the high death rate first reported for severe UC [111]. Although oral steroids are well absorbed, most authorities recommend intravenous administration in the sickest patients, because alterations in gut motility and mucosal function may compromise absorption [84, 85]. However, after a patient responds, the dose should be tapered gradually to minimize development of corticosteroidrelated side-effects. A typical rate of withdrawal for oral prednisone is by 5 mg per day each week, but this is only a guideline. Medications other than glucocorticoids, such as aminosalicylates and immunosup-

503 pressives, should be used to maintain remission because of superior efficacy and toxicity profiles. Rectally administered steroids have lower systemic absorption than oral, and therefore may provide a therapeutic advantage by acting locally in distal UC. Similarly, in CD of the terminal ileum and right colon, controUed-ileal-release budesonide appears to be almost as effective as prednisolone, but has significantly less systemic effect [97]. Intravenous, oral and rectal steroids are highly efl&cacious in most patients with IBD. However, up to 20% of the IBD population do not respond clinically to oral prednisolone after 30 days [113], and on clinical grounds are resistant to steroids. The molecular basis for resistance to the anti-inflammatory effects of glucocorticoids is an important issue in IBD therapy. A syndrome of familial steroid resistance has been described not only for glucocorticoids, but also for sex hormones, mineralocorticoids and vitamin D [114]. Familial glucocorticoid resistance is a rare disorder that appears to be due to a mutation in the glucocorticoid receptor. More common is isolated resistance to the anti-inflammatory effects of glucocorticoids, which has been best described in asthma [115]. In asthma, steroid resistance is associated with defects in the suppression of peripheral blood mononuclear cell function by glucocorticoids [116]. This may be related to alterations in the number of glucocorticoid receptors, and to altered affinity for ligand in target cells [117-119]. The molecular basis for these abnormalities has been linked to increased expression of the transcription factor c-fos in the peripheral blood mononuclear cells of glucocorticoid-resistant asthmatics [120]. In CD, peripheral blood leukocytes have reduced sensitivity to the inhibitory effects of dexamethasone on cytokine secretion after ex-vivo stimulation, compared to controls [121]. Among a group of patients with severely active UC, a therapeutic response to intravenous hydrocortisone was significantly more likely in those with the greatest sensitivity of peripheral T cells to dexamethasone-induced inhibition of prohferation after ex-vivo stimulation [122]. In IBD, more study is clearly needed to delineate the molecular mechanisms that underlie sensitivity and resistance to glucocorticoids.

504

Azathioprine and 6mercaptopurine Chemistry Azathioprine and 6-mercaptopurine are purine antimetabolite drugs that were found to have immunesuppressing properties when used for cancer chemotherapy, leading to their investigation and ultimate use as immunosuppressive drugs in a variety of clinical settings. Azathioprine is a pro-drug that is rapidly converted to 6-mercaptopurine in a nonenzymatic reaction by sulfhydryl-containing substances such as glutathione. These two drugs possess identical actions with respect to their use in IBD. Considering mass (azathioprine is 55% 6-mercaptopurine by weight) and the efficiency of conversion of azathioprine to 6-mercaptopurine (88%), one must administer approximately twice as much azathioprine as 6-mercaptopurine to achieve similar amounts of bioavailable active drug. 6-Mercaptopurine can enter three known metabolic pathways (Fig. 4). Two of these routes, leading to the inactive metabolites 6methylmercaptopurine and 6-thiouric acid, are catalyzed by thiopurine methyltransferase and xanthine oxidase, respectively. The action of hypoxanthine phosphoribosyl transferase commits 6-mercaptopurine to conversion to the 6-thioguanine nucleotides, the putative active metabolites [123].

Mechanism of action After administration of azathioprine or 6-mercaptopurine, the 6-thioguanine nucleotides accumulate intracellularly, and are believed to mediate the biologic actions of these drugs. Despite over 50 years of investigation the precise molecular basis for the therapeutic effects of the purine analogs is not known. However, intracellular accumulation of 6thioguanine nucleotides causes inhibition of the pathways of purine nucleotide metabolism and DNA synthesis and repair, resulting in inhibition of cell division and proHferation. It is plausible, but not proven, that antiproliferative or functionally inhibitory actions on cells of the immune system, such as lymphocytes, underlie the immunosuppressive actions of azathioprine and 6-mercaptopurine.

Pharmacolcinetics and pharmacogenetics The bioavailability of azathioprine, estimated as 6mercaptopurine, ranges from 27%o to 83%, and of 6-

Clinical pharmacology in inflammatory bowel disease 6-methylmercaptopurine TPMT Azathioprine —• 6-mercaptopurine HPRT Thioinosinic—^ 6-thioguanine dLc'id^^^"'^ nucleotides XO 6-thiouric acid

Figure 4. Metabolic pathways of azathioprine and 6mercaptopurlne. TPIVIT, thiopurine methyltransferase; HPRT, hypoxanthine phosphoribosyltransferase; XO, xanthine oxidase. Reproduced with permission from Chan GL et al., J Clin Pharmacol 1990; 30: 358-63.

mercaptopurine ranges from 5% to 37%. An oral preparation of azathioprine coated with Eudragit-S, designed for release above pH 7 in the distal ileum, had a mean bioavailability of 7%. Rectal administration of azathioprine in hydrophilic or hydrophobic enemas consistently resulted in less than 10% systemic bioavailability [124]. At present there are no data on the clinical efficacy or toxicity of rectal or delayed-release oral azathioprine in IBD. Many studies have examined the kinetics of the active metabolites of azathioprine and 6-mercaptopurine, the 6-thioguanine nucleotides. Although azathioprine and 6-mercaptopurine are rapidly cleared from plasma, 6-thioguanine nucleotides are concentrated in cells resulting in a half-life of elimination of many days [125]. In patients with CD and other conditions, erythrocyte 6-thioguanine nucleotide concentration gradually rises to a plateau after 2-4 weeks of oral azathioprine [126-128] (Fig. 5). The rather delayed time to reach steady-state 6thioguanine nucleotide kinetics parallels the slow onset of therapeutic benefit of oral azathioprine or 6-mercaptopurine therapy in IBD, which has been estimated to average 17 weeks [129]. This observation prompted a pilot study [130], followed by a placebo-controlled trial of an intravenous loadingdose of azathioprine to hasten the onset of clinical benefit in CD [126]. In this study, despite initially higher erythrocyte 6-thioguanine nucleotide concentration in those who received active intravenous drug, a clinical response was seen no sooner than in placebo-treated patients. However, the median time to steady-state erythrocyte 6-thioguanine nucleotide concentration and first clinical improvement was only 4 weeks, suggesting that azathioprine may act more quickly than had previously been realized.

Laurence J. Egan and William J. Sandborn

505

O Placebo A IV azathioprine

0} E

0

2

4

6

8

10

12

14

16

Weeks of treatment

Figure 5. Mean erythrocyte 6-thioguanine nucleotide concentration in Crohn's disease patients after beginning azathioprine 2 mg/kg per day by mouth. A , received an intravenous loading dose of azathioprine; 0 » received a placebo intravenous loading dose. Reproduced with permission from Sandborn WJ et al. Gastroenterology 1999; 117:527-35.

One of the pathways of biotransformation that is available to 6-mercaptopurine is metabolized by a polymorphically expressed enzyme, thiopurine methyltransferase [131]. Erythrocyte thiopurine methyltransferase activity correlates well with the enzyme's activity in other cells and tissues. The majority of the Caucasian population, ^ 9 0 % , have an erythrocyte thiopurine methyltransferase activity of about 14-25 U/ml erythrocytes, and these individuals are homozygous for two wild-type TPMT alleles. Approximately 10% of this population have an enzyme activity that is detectable but below 14 U/ ml erythrocytes, and are heterozygous for one active and one inactive copy of TMPT. One percent or less of individuals have no detectable thiopurine methyltransferase activity and have 2 inactive TTMralleles [131, 132]. The erythrocyte 6-thioguanine nucleotide concentration varies considerably among patients being treated with azathioprine or 6-mercaptopurine, and appears to be dependent on the individuals' inherent, genetically determined thiopurine methyltransferase activity [132, 133]. Thus, patients with low thiopurine methyltransferase activity have greater 6-thioguanine nucleotide concentrations, presumably because more 6-mercaptopurine is shunted into the hypoxanthine phosphoribosyl transferase pathway.

occurring early during treatment; and dose-related, generally occurring later. Allergy to these drugs can manifest as pancreatitis, fever, rash, nausea, diarrhea, and hepatitis. Approximately 5% of the IBD population beginning treatment with azathioprine or 6-mercaptopurine will experience an allergic reaction to the drug [134], which can be confirmed by rechallenge if necessary. Dose-related toxicities of azathioprine or 6-mercaptopurine include bone marrow depression leading to leukopenia, anemia or thrombocytopenia. If severe, leukopenia can potentially cause profound immune suppression, and thus predispose to opportunistic infections or neoplasms. Cytopenias have been reported to develop from 2 weeks to 11 years after beginning therapy, at a cumulative frequency of about 10% [126, 135, 136]. Although the numbers are insufficient for statistical analysis, there are enough reports of fatal or life-threatening opportunistic infections in leukopenic IBD patients treated with azathioprine or 6-mercaptopurine to dispel any tendency towards complacency [134, 135]. If leukopenia and a serious infection develop in a patient receiving azathioprine or 6-mercaptopurine, reversal of bone marrow suppression is usually possible with granulocyte colony-stimulating factor. It is not completely clear if azathioprine or 6mercaptopurine therapy confers an excess risk of malignancy. In the New York series, a wide spectrum of malignant neoplasms developed in IBD patients who had at some point received 6-mercaptopurine, but the overall rate of 3.1 % was low and probably not greater than the general population [134]. In the British series, of 755 patients treated with azathioprine for a median duration of 12.5 years at a dose of 2 mg/kg per day, 31 developed a cancer compared to 24.3 expected cases [137]. The difference was significant only when considering colorectal adenocarcinomas, a recognized complication of IBD. Thus, the data are quite reassuring, but isolated and anecdotal reports of unusual lymphomas arising in IBD patients during or after azathioprine or 6-mercaptopurine therapy continue to generate some concern [136, 138]. Nevertheless, if there is an increased cancer risk a t t r i b u t a b l e to these drugs, the magnitude of the increased risk is small.

Optimizing ttierapy Adverse effects Two categories of adverse effects to azathioprine and 6-mercaptopurine have been described: allergic.

In some studies of patients being treated with azathioprine or 6-mercaptopurine for diseases other than IBD, low thiopurine methyltransferase activity.

506

Clinical pharmacology in inflammatory bowel disease

resulting in greater 6-thioguanine nucleotide concentration, has been associated with the development of leukopenia [123, 139]. Conversely, patients with higher thiopurine methyltransferase activity may be undertreated with standard doses of azathioprine or 6-mercaptopurine. For example, in childhood acute leukemia the risk of relapse during therapy with 6mercaptopurine is greatest in those patients with the highest erythrocyte thiopurine methyltransferase activity, because they inactivate the drug more efficiently and have lower concentrations of active 6thioguanine nucleotides [140, 141]. In IBD patients it is currently uncertain if significant associations exist between azathioprine or 6mercaptopurine metabolism and the occurrence of clinical response or toxicity. A number of studies that were performed in CD patients selected without regard to thiopurine methyltransferase activity have suggested that 6-mercaptopurine pharmacogenetics may in fact be clinically relevant. In a group of pediatric CD patients receiving 6-mercaptopurine therapy, one study found that disease severity was least in those with the highest erythrocyte 6-thioguanine nucleotide concentration [142]. Another found that 65% of patients with inactive disease had 6thioguanine nucleotide levels greater than 230 pmol/8x 10^ erythrocytes, compared to 28% of patients with active CD [143]. Similar studies in adults with CD and UC have documented a greater mean 6-thioguanine nucleotide concentration among responders to azathioprine or 6-mercaptopurine than among patients with persistent active disease [144, 145]. In contrast, two prospective studies of azathioprine in CD arrived at different conclusions. These clinical trials examined the relationships between thiopurine methyltransferase activity and erythrocyte 6-thioguanine nucleotide concentration, and the clinical outcomes of leukocyte count and efficacy [126, 130]. To minimize the potential for toxicity in these studies, individuals were excluded if thiopurine methyltransferase activity was in the heterozygous or homozygous low range (approximately 20% of the screened population, unpublished observation, W.J. Sandborn). No significant correlations were found between thiopurine methyltransferase activity or erythrocyte 6-thioguanine nucleotide concentration, and leukocyte count or clinical efficacy. The exclusion of individuals with intermediate or low thiopurine methyltransferase activity probably decreased the power of these studies to detect pharmacogenetic differences. Furthermore, a cross-sectional study of

IBD patients that included individuals with intermediate-range thiopurine methyltransferase activity who were receiving azathioprine or 6-mercaptopurine also found no significant associations between thiopurine methyltransferase activity or erythrocyte 6-thioguanine nucleotide concentration and disease severity [146]. Based on these data the benefit of monitoring erythrocyte 6-thioguanine nucleotide levels during azathioprine or 6-mercaptopurine therapy of IBD remains uncertain. Prospective studies that examine if a clinical benefit derives from using erythrocyte 6thioguanine nucleotide concentrations to guide therapy in IBD are needed. It is our practice to measure thiopurine methyltransferase activity before beginning therapy with one of these drugs, to exclude the rare patient that is completely deficient in this enzyme. Individuals who have intermediate thiopurine methyltransferase activity can probably be treated safely with azathioprine or 6-mercaptopurine, but at half the standard dose. We currently do not routinely use erythrocyte 6-thioguanine nucleotide concentration for therapeutic drug monitoring, but this practice is commonplace at some centers. Therapeutic trials have demonstrated that azathioprine should be given at a dose of 2-2.5 mg/ kg per day, and 6-mercaptopurine at 1.5 mg/kg per day. There is no role for an intravenous loading dose to hasten the onset of clinical response. We counsel patients about the risk of leukopenia, and check a complete blood count every month during therapy. The results of one retrospective study suggested that the development of mild leukopenia, an indicator of intensity of treatment with 6-mercaptopurine, identified CD patients with a greater likelihood of a favorable response [147]. Based on this observation, one group of investigators has tried dose escalation in CD patients refractory to standard-dose azathioprine. In this uncontrolled study, after increasing the dose of azathioprine from a mean of 2 mg/kg per day to 2.7 mg/kg per day, 14 of 18 patients improved and were able to discontinue steroids [148]. Although no significant leukopenia was observed, two patients developed serious cytomegalovirus infection. Other studies that have examined the relationship between leukocyte count and clinical response during azathioprine and 6mercaptopurine therapy of IBD have not supported the concept that leukopenia is associated with efficacy [126, 146]. For these reasons we do not recommend use of greater than standard doses of these agents.

Laurence J. Egan and William J. Sandborn

Methotrexate Chemistry and metabolism Methotrexate is an antimetabolite anticancer drug that, like azathioprine, has been found to have beneficial effects in a number of chronic inflammatory diseases. Structurally, methotrexate is the 4amino, N-10 methyl analog of folic acid. Like folic acid, methotrexate is a substrate for intracellular enzyme folylpolyglutamate synthase, which adds up to five glutamic acid residues to these compounds [149]. This modification promotes intracellular retention of methotrexate and increases the drug's affinity for a number of target enzymes [150, 151]. Over 90% of a dose of parenteral methotrexate is excreted unchanged in the urine [152]. In the liver a minor metabolic pathway of methotrexate produces small quantities of the 7-hydroxy derivative.

IVIechanism of action The principal biochemical action of methotrexate, and the mechanism that is believed to be responsible for the drug's cytotoxic activity, is inhibition of dihydrofolate reductase. This enzyme is critical for regenerating the fully reduced folate co-factors that are required for reactions involving transfer of onecarbon fragments, such as the production of thymidylate and purines. Methotrexate acts as a reversible, competitive inhibitor of dihydrofolate reductase. The polyglutamated metabolites of methotrexate are important not only because they are retained intracellularly, but also because they are potent inhibitors of enzymes that lie downstream of dihydrofolate reductase in the metabolic pathway of folic acid [151, 153]. Inhibition of these enzymes, including thymidylate synthase and 5-aminoimidazole-4-carboxamide ribotide transformylase, further impairs folate-dependent synthetic reactions. At the high doses used in cancer chemotherapy, methotrexate blocks de-novo synthesis of DNA precursors leading to inhibition of cellular proliferation and cytotoxicity. Although methotrexate is beneficial in a number of chronic inflammatory diseases, cytotoxic or antiproliferative effects of folic acid analog therapy are difficult to invoke as the anti-inffammatory mechanism of action because clinically efficacious doses of methotrexate do not decrease circulating neutrophil, lymphocyte or platelet numbers [154, 155]. The occurrence of cytopenias or mucositis, clinical evidence of cytotoxicity, are in fact indica-

507 tions to lower the dose of methotrexate in chronic inflammatory diseases. Also, administration of folic acid ameliorates some of the toxicities of methotrexate that are attributable to its antiprohferative eff*ects, without impairing its anti-inflammatory action [156]. In efforts to identify the mechanism of action of methotrexate in inflammatory diseases the effects of this drug on a number of mediators of inffammation have been examined, principally in patients with rheumatoid arthritis. Methotrexate has been found to decrease neutrophil production of leukotriene B4 [157], to decrease IL-1 concentration in synovial fluid [158] and to lower plasma IL-6 and soluble IL2 receptor concentrations [159]. However, other agents such as glucocorticoids produce similar effects, so these changes may reffect general improvement rather than methotrexate-specific phenomena. It has alternatively been proposed that the antiinflammatory action of methotrexate may be mediated by increased release of the endogenous anti-inflammatory autocoid, adenosine. Experiments in tissue culture and animal models of acute inflammation have demonstrated that methotrexate can promote extracellular release of adenosine, which inhibits neutrophil adherence to endothelial cells and accumulation at sites of inflammation [160, 161]. In animal experiments, methotrexate has also been shown to decrease leukocyte-endothelial cell interactions through release of adenosine and occupancy of adenosine A2 receptors [162]. To determine if the methotrexate-induced elevations of extracellular adenosine concentration that were demonstrated in tissue culture and animal models of acute inflammation were relevant to the use of this drug in humans with IBD, the effect of methotrexate on rectal and plasma adenosine was studied. In 10 IBD patients with varying degrees of rectal involvement, adenosine was measured in the rectal lumen using in-vivo rectal dialysis [163]. The concentration of adenosine in the rectum tended to correlate with the severity of proctitis. After a subcutaneous injection of methotrexate 15 or 25 mg the mean rectal adenosine concentration did not change. Furthermore, the plasma adenosine concentration did not differ significantly from baseline at any time during the 24 h after the methotrexate injection. These data suggest that the in-vivo anti-inffammatory actions of methotrexate are not related to alterations of adenosine metabohsm. Activated T lymphocytes form an important part of the inflammatory infiltrate in IBD, and the major

508 contribution of T lymphocytes in the pathogenesis of mucosal inflammation is highlighted by the many animal models of IBD that occur with various perturbations of T lymphocyte function [164]. The effect of methotrexate on T lymphocyte survival has been studied [165]. Methotrexate-treated T lymphocytes were found to be primed to die by apoptosis when stimulated with mitogens, by a Fas-independent mechanism. Furthermore, the inhibitory effect of methotrexate was confined to cells that were activated at the time of drug exposure. This process was inhibited by fully reduced folate and thymidine, indicating that methotrexate-induced lymphocyte apoptosis was dependent on inhibition of dihydrofolate reductase and thymidylate synthase. These results demonstrate that methotrexate inhibits the activation of T lymphocytes, a process that is likely to be important to this drug's in-vivo anti-inflammatory actions.

Clinical pharmacology in inflammatory bowel disease

o E

UJ 2

n(15 mg) 10 n(25 mg) 5

6

8

10

12

14

10 5

10 5

Figure 6. Mean erythrocyte methotrexate concentrations after starting methotrexate 15 mg/week ( # ) or 25 mg/week ( O ) by subcutaneous injection. Reproduced with permission from Egan U et ai, Aliment Pharmacol Ther 1999; 13:1597-604.

300 n

Pharmacokinetics For the treatment of chronic inflammatory diseases, such as psoriasis and rheumatoid arthritis, methotrexate has traditionally been given once weekly, because the incidence of hepatotoxicity appeared to be less with this approach than with daily administration [166]. The doses that have been used range from 7.5 to 25 mg per week, usually given orally or by intramuscular injection. When methotrexate was first tried for IBD, this therapeutic regime was empirically adopted [167]. The bioavailability of oral methotrexate is dosedependent, because the pathways of absorption are saturable, and at higher doses methotrexate induces malabsorption through its antiproliferative effects on epithelia [168]. Reported oral bioavailability at the low doses used in chronic inflammatory diseases range from 50% to 90% [169-171]. Intramuscular and subcutaneous methotrexate exhibit near-complete bioavailability [152, 172]; thus, considering ease of administration, the subcutaneous route is preferred. After administration, methotrexate is widely distributed and rapidly cleared, mostly in unchanged form by the kidney. However, a small fraction of the drug is concentrated in many cell types as polyglutamated methotrexate. In IBD patients receiving weekly injections of methotrexate, the drug has been shown to gradually accumulate in erythrocytes, reaching a plateau after about 8 weeks (Fig. 6) [173]. Methotrexate also accumulates in many other

4

Weeks after starting methotrexate

300

• Pre-methotrexate D Post-methotrexate

I

200 H

-S o

200

100

Dialysate

M.

o E

MOO 0)

Biopsy

Figure 7. In IBD patients receiving weekly injections of methotrexate the drug can be detected in the rectal lumen and mucosa 7 days after the last dose. The concentration is higher in biopsy material than in dialysate, indicating intracellular concentration of methotrexate. Reproduced with permission from Egan LJ et a/., Clin Pharmacol Ther 1999; 65: 29-39.

tissues, including the intestinal mucosa. Methotrexate was easily identified in rectal dialysate and biopsies 7 days after a dose, at a time when plasma methotrexate was barely detectable (Fig. 7). Comparison of rectal lumen and rectal mucosa methotrexate concentrations reveals that the drug is also concentrated intracellularly at this site. The trough rectal mucosa methotrexate concentration was found to be sufficient to exert pharmacologic activity, and this observation thus provides a potential explanation for the continuous activity of methotrexate throughout a 7-day dosing interval [152].

Laurence J. Egan and William J. Sandborn

Adverse effects The long-term use of methotrexate for chronic inflammatory diseases has the potential for serious toxicity. In reported series of IBD patients treated with methotrexate about 10% have discontinued the drug because of suspected drug-induced toxicity [167, 173-179]. The adverse effects of methotrexate can be considered in three categories: predictable dose-related anti-proliferative eff'ects on bone marrow, gut epithelia and hair follicles; idiosyncratic allergic/hypersensitive-type reactions; and hepatic damage due to the cumulative effects of chronic exposure. Methotrexate inhibits cellular proliferation in a dose-dependent manner, by depleting cells of the fully reduced folate co-factors needed for thymidylate and purine biosynthesis. At the high doses used for cancer chemotherapy, bone marrow depression, gastrointestinal ulceration and stomatitis, and alopecia predictably occur, necessitating 'rescue' with tetrahydrofolate. For eff'ective therapy of chronic inflammatory diseases with methotrexate, it is not clear that overt inhibition of cellular proliferation is required. Consequently, most clinicians regard the occurrence of cytopenias, intestinal distress or alopecia as indicators of unacceptable toxicity, and reduce the dose of methotrexate or coadminister folic acid [156]. Idiosyncratic allergic-type reactions to methotrexate include rash and pneumonitis. Methotrexateinduced lung damage is important, because in the rheumatologic literature this is reported to occur in 3-11% of patients, and has occasionally been fatal [180]. Clinically, methotrexate-induced pneumonitis is characterized by cough, dyspnea, fever, hypoxemia, a restrictive ventilatory defect with impaired gas exchange, and a radiographic picture of a diffuse alveolar/interstitial process. Discontinuation of methotrexate usually results in resolution. Long-term use of weekly methotrexate is associated with abnormal serum transaminase activity in up to 30% of patients [177, 181]. However, hypertransaminasemia does not correlate well with histologic liver damage. The spectrum of histologic lesions attributed to chronic methotrexate use includes macrovesicular steatosis, hepatocyte necrosis and portal inflammation, progressing to fibrosis and cirrhosis [182]. It has been difficult to determine the true incidence of significant methotrexateinduced liver disease. In studies of psoriasis patients treated with methotrexate, progressive histologic liver damage has been well documented [183, 184].

509 In one study of 364 liver biopsies from 168 patients with psoriasis, only 50% of patients were without hepatic fibrosis after a cumulative dose of 3 g methotrexate [185]. However, in rheumatoid arthritis, the risk of cirrhosis over 5 years of methotrexate therapy appears to be only 1 in 1000 [186, 187]. Thus, there may be a disease-specific risk for methotrexate liver damage. In addition to the cumulative dose of methotrexate, other identified risk factors for liver damage include heavy alcohol ingestion, obesity and diabetes [185].

Optimizing therapy The effective and safe use of methotrexate in IBD demands a trade-off* between efficacy and toxicity, both of which may be dose-dependent. Studies in rheumatoid arthritis have demonstrated that higher doses of methotrexate are clinically more effective than lower doses [188], and some forms of methotrexate-induced toxicity are clearly dose-dependent. In patients with refractory IBD a 16-week comparison of methotrexate 15 or 25 mg per week did not reveal different rates of efficacy or toxicity between the two doses [173]. Until more data become available the clinical approach is necessarily somewhat empiric. Typically, patients are started on methotrexate 15 or 25 mg per week by subcutaneous injection. Once a satisfactory clinical response occurs the dose is reduced to 15 mg per week. Our proposed guidelines for liver monitoring in methotrexate-treated IBD patients are outlined in Table 4. We advise patients to refrain from alcohol consumption during methotrexate therapy. Erythrocyte methotrexate concentration reflects intensity of exposure to the drug (Fig. 6). To determine if an association exists between the occurrence of clinical response or drug toxicity and circulating methotrexate concentrations, the mean trough concentrations of erythrocyte and plasma methotrexate, and plasma 7-hydroxymethotrexate were compared in IBD patients who responded or did not respond to weekly methotrexate injections [173]. No significant differences were found, suggesting that this approach to therapeutic drug monitoring of methotrexate in IBD is not helpful. In cancer chemotherapy, poor outcomes are associated with the development of resistance to methotrexate. Two primary mechanisms have been identified: increased expression of the main target enzyme, dihydrofolate reductase, and increased drug efflux from cells, mediated by expression of the multi-

Clinical pharmacology in inflammatory bowel disease

510

Table 4. Guidelines for preventing and monitoring hepatotoxicity in IBD patients receiving long-term methotrexate Baseline

Withhold methotrexate

Liver tests

AST, ALT, ALP, bilirubin Hepatitis B and C tests

If > 2 X normal If actively infected

Liver biopsy

If abnormal AST, ALT, ALP, bilirubin; if clinical suspicion of liver disease

If hepatitis, fibrosis or cirrhosis is present Stop or reduce methotrexate

During treatment

AST

Every 6 weeks

If > 2 X baseline

Liver biopsy

Every cumulative 1.5 g methotrexate; if progressive elevation of AST; if > 50% of AST tests are high

If hepatitis, fibrosis or cirrhosis is present

AST, aspartate aminotransferase; ALT alanine aminotransferase, ALP alkaline phosphatase.

drug resistance pump, p-glycoprotein [189]. It is not known if either of these mechanisms operates in IBD patients who do not respond to methotrexate.

Monoclonal antibody anti-TNF-oc therapy Several observations suggest that TNF-oc is an important factor in the pathogenesis of mucosal inflammation [190]. This cytokine is capable of triggering intracellular signaling cascades, such as the N F - K B pathway, that result in an inflammatory phenotype in epithelial cells and cells of the lamina propria. Stool and mucosal levels of TNF-oc are elevated in IBD [9, 191-193]. Recently, specific monoclonal antibody therapies to antagonize TNFa have demonstrated clinical efficacy in CD [194196].

Structure of monoclonal antibodies and fusion proteins against TNF-oc There are three currently available synthetic antiTNF-a molecules: two monoclonal antibodies and one fusion protein (Fig. 8). These molecules have been developed with the goal of binding free and membrane-bound TNF-oc and thereby preventing the cytokine from binding to its cell-surface receptor and exerting biologic activity. Linkage of the TNF-oc binding portions of these molecules to immunoglobulin-like structures slows their clearance from plasma and allows for a variety of effector functions related to the immunoglobulin constant regions,

such as complement fixation and antibody-mediated cellular cytotoxicity. Infliximab (cA2) is a chimeric mouse-human monoclonal antibody to human TNF-oc. This molecule consists of human IgGj constant regions, human kappa light chains and transplanted monoclonal murine variable regions that recognize TNF-oc with very high affinity [197]. CDP571 is a 'humanized' monoclonal antibody to TNF-oc. This molecule consists of human IgG4 constant regions, human kappa light chains and transplanted complementarity determining regions from a murine monoclonal antibody that recognizes TNF-oc, also with very high affinity [198]. Etanercept is a fusion protein that is generated by linking the human p75 TNF-oc receptor to the Fc portion of IgGi [199].

Pharmacokinetics Infliximab and CDP571 are administered as intravenous infusions, and etanercept is given by subcutaneous injections. Pharmacokinetic studies of infliximab and CDP571 demonstrate that the plasma concentrations of these monoclonal antibodies are linearly proportional to dose, and that their elimination follows first-order kinetics [198, 200, 201]. At therapeutic doses of 5-10 mg/kg the half-lives of infliximab and CDP571 are in the range of 8-10 days. This results in plasma disappearance of these antibodies after approximately 8-12 weeks. In contrast, the receptor fusion protein etanercept has a shorter plasma existence. After subcutaneous injections of 25 mg etanercept, the median half-life of elimination was approximately 5 days [202].

Laurence J. Egan and William J. Sandborn

Chimerio Monoclonal Antibody

Schematio Diagram of an Antibody Ught« chain gene

511

Heavyohain gene

\

HymanJzed Monoclonal Antibody

Human Recombinant Reoeptor/Fc Fusion Protein

COR

/

Constant 2 Fc < Variabie Constant Constant 1 W^ Constant 2 5 Constant 3 EO Fab fli Hinge ^ Fc a

76% human • I iMlouse O

Human

9 5 % human • i Mlouse a Human COR « Complementarity de^mfilnlng regions

Constant 3

100% human •

Human

Figure 8. Schematic comparjson of the available antl-TNF-a monoclonal antibodies Infliximab and CDP571, and the TNF-a receptorFc fusion protein, etanercept. Reproduced with permission from Sandborn WJ et al., Inflam Bowel DIs 1999; 5:119-33.

Mechanism of action The biologic activity of TNF-a is mediated by the binding of soluble trimers of the cytokine, or membrane-bound trimers, to the cell-surface TNF-a receptors, types I and II. Infliximab, CDP571 and etanercept all bind to soluble TNF-a trimers with high affinity (Table 5). As a result serum TNF-a concentrations may actually increase because binding to the long-residing TNF-a antagonist molecules slows clearance [200, 202]. However, the binding of circulating TNF-a by the antagonist molecules prevents the cytokine from binding to its receptors, and thus the biologic activity of TNF-a is effectively neutralized [197, 199, 203, 204]. Membrane-bound TNF-a also exhibits biologic activity. Infliximab, CDP571 and etanercept can bind to this form of TNF-a and probably neutralize its activity. Immunoglobulin isotype is an important determinant of antibody function. For example, antibodies of the IgGi, but not the IgG4 isotype, can stimulate secondary effector mechanisms mediated by the Fc portion of the molecule. Fc receptor binding can result in complement activation and antibody-mediated cellular cytotoxicity [205]. In this regard infliximab but not CDP571 or etanercept can generate Fc receptor-dependent effector functions

Table 5. Comparison of the anti-TNF-a activities of infliximab, CDP571 and etanercept

Binding to soluble TNF-a Infliximab CDP571 Etanercept

Yes Yes Yes

Binding to membranebound TNF-a

Yes Probably

Yes

Lysis of cells expressing membranebound TNF-a

Yes No No

resulting in cytotoxicity [206]. This function may in part underlie the effects of infliximab in vivo, and thus may represent an important difference from CDP571 and etanercept, which do not appear to cause complement-mediated or antibody-mediated cytotoxicity [202, 205]. In vivo, treatment with infliximab has been demonstrated to down-regulate Th-1 type cytokine production in CD mucosa [10], to decrease the activity of pathways of coagulation and fibrinolysis [207], and to decrease circulating levels of phospholipase A2. Further, in rheumatoid arthritis patients, infliximab has been shown to decrease circulating IL-1, IL-6, E-

512

Clinical pharmacology in inflammatory bowel disease

selectin, intracellular adhesion molecule-1 and matrix metalloproteinases [208-210]. CDP571 has been shown to decrease IL-1 production in an animal model [211]. Etanercept has been shown to decrease levels of IL-1, IL-6, E-selectin, intracellular adhesion molecule-1 and matrix metalloproteinases [202].

detectable acquired immune response to infliximab can be retarded by immune suppression. Higher doses of infliximab have been associated with lower rates of development of anti-infliximab antibodies [212]. Antibodies to CDP571 were detected in 35% of CD patients after an infusion [196]. Anti-etanercept antibodies developed in 16% of rheumatoid arthritis patients after receiving this drug. It is presently unclear if these antibodies are clinically significant. The development of autoantibodies has been reported to occur after treatment with infliximab [201] and etanercept [202]. There are currently no data available on autoantibodies in CDP571-treated patients. After treatment with infliximab, 34% of CD patients developed antinuclear antibodies, and 9% developed antidouble-stranded DNA antibodies. A small minority of these patients developed a lupuslike syndrome that resolved gradually. A variety of malignant neoplasms have developed in patients who were treated with the anti-TNF-oc agents, infliximab and etanercept. However, the numbers have been too small to determine if the risk of cancer is greater than in appropriate control groups.

Adverse effects The most important adverse effects of anti-TNF-oc monoclonal antibody and fusion protein therapies are the development of infusion reactions, and induction of treatment-related antibodies. Initial infusions of infliximab have resulted in adverse reactions in approximately 17% of patients, but in only 2% was the reaction severe enough to warrant discontinuation of the infusion [201]. Four percent of patients experienced fever or chills, 1% experienced pruritis or urticaria and 1% experienced cardiopulmonary symptoms such as chest pain, dyspnea or hemodynamic instability. Infusion reactions are more common with the second or subsequent infusions, than with the first. In CD patients who were retreated with infliximab after a break of 2-4 years, 25% experienced adverse reaction that occurred 3-12 days after the infusion. These reactions consisted of myalgia; arthralgia; fever; rash; urticaria; and facial, hand and lip edema. In all cases of infusion reactions to infliximab, symptomatic treatment with antihistamines or corticosteroids resulted in resolution. The development of a reaction to infliximab was more common in CD patients who subsequently had detectable circulating anti-infliximab antibodies (36%) than in those without the antibodies (see below). The incidence of adverse reactions to infusions of CDP571 or to injections of etanercept appears to be less than occurs with infliximab. Thus, the risk of adverse infusion reactions seems to be proportional to the amount of murine protein in the antibody. Administration of a foreign protein can stimulate development of an immune response directed at exogenous peptide antigens. Thus, it is not surprising that human antibodies that recognize epitopes on the foreign protein have been reported to develop in patients treated with infliximab, CDP571 and etanercept. Anti-infliximab antibodies developed in 10% of CD patients who were receiving concomitant immunosuppressive therapy with azathioprine or 6mercaptopurine, and in 23%o of those who were not [201]. A similar trend was seen in rheumatoid arthritis patients, suggesting that the development of a

Optimizing therapy Currently, infliximab is the preferred anti-TNF-oc agent for use in CD. Further studies of CDP571 and etanercept in IBD are awaited. For induction of remission in active inflammatory or fistulizing CD, an initial dose of infliximab, 5 mg/kg is indicated [194, 195]. Because favorable clinical responses to infliximab tend to last only 2-4 months it is our practice to initiate immunosuppressive therapy with azathioprine or 6-mercaptopurine at this time, to act as a long-acting maintenance drug. Methotrexate is probably a good alternative for patients not tolerating azathioprine or 6-mercaptopurine. It is reasonable to retreat patients who do not respond to an initial dose of infliximab after about 4 weeks, because in the fistula study some patients responded only after the second or third dose [195]. Long-term retreatment with infliximab should be reserved for patients who relapse despite concomitant use of a long-acting immunosuppressive agent. The optimum dose and frequency of infliximab infusions for maintenance of remission in CD are not known. However, a recent study of every 8 week doses of infliximab suggested that retreatment may be a reasonable approach in patients who have failed other options.

Laurence J. Egan and William J. Sandborn

Cyclosporine and tacrolimus (FK 506) Chemistry and mechanism of action Cyclosporine is a cyclic polypeptide of fungal origin, and tacrolimus is a macrolide antibiotic. Although these entities are chemically unrelated their mechanisms of immunosuppression are very similar. Both drugs bind to cytoplasmic proteins; cyclosporine to cyclophilin, and tacrolimus to FK-binding protein. These drug-protein complexes inhibit the cytoplasmic phosphatase calcineurin, an enzyme essential for activation of the T lymphocyte-specific nuclear factor of activated T cells. This transcription factor regulates the rate of transcription of cytokines important in T lymphocyte activation, notably IL-2 and IFN-y [213]. Through this mechanism cyclosporine and tacrolimus act as powerful and specific inhibitors of T lymphocyte activation.

Pharmacol(inetics Standard cyclosporine is poorly and erratically absorbed, with oral bioavailabihty of only 20-50%. In patients with small bowel CD absorption may be even lower [214, 215]. The newer microemulsion formulation of cyclosporine is better absorbed by a factor of about two [216]. Tacrohmus is much more potent than cyclosporine. Although oral tacrolimus is also not highly available systemically, it is less dependent than cyclosporine on the presence of bile for absorption.

Adverse effects Both cyclosporine and tacrolimus have low therapeutic indices. With chronic use of either agent, toxicity is likely to develop. The most frequently reported adverse effects of cyclosporine in IBD patients are paresthesias (26%), hypertrichosis (13%)), hypertension (ll%o), tremor (7%), reversible decline of renal function (6%)), headache (5%), and nausea (6%)) [217]. The experience with tacrolimus in IBD is much more limited than with cyclosporine, but in addition to the side-effects reported with cyclosporine, hyperglycemia can occur. Rare but potentially life-threatening adverse effects of cyclosporine and tacrolimus include opportunistic infections and malignancies [218, 219].

513

Optimizing therapy Optimum use of these powerful immunosuppressives requires a careful consideration of the risk to benefit ratio. There is a significant risk of long-term toxicity with cyclosporine and tacrolimus, and most clinical trials with oral cyclosporine in CD have been negative [220-222]. However, the short-term use of intravenous cyclosporine and tacrolimus has been highly effective in very sick IBD patients [223-226]. For these reasons we rarely initiate cyclosporine or tacrolimus in IBD patients with the anticipation of long-term use. The key principles to effective and safe cyclosporine and tacrolimus therapy of IBD are use only in highly selected patients, concomitant initiation of safer forms of long-term immune suppression, the use of blood levels to guide dosing, and careful toxicity monitoring. Intravenous cyclosporine has a role in the management of patients with severely active UC who have failed to respond to high-dose intravenous steroids, who do not have colonic dilation or impending perforation, and who are not immediate surgical candidates [223]. In CD patients there are several potential situations in which cyclosporine or tacrolimus may be beneficial. Severely active inflammatory or fistulizing disease without a stricture, that cannot be controlled with intravenous steroids or infliximab, may respond to intravenous cyclosporine or tacrolimus [224, 225]. There is no role for oral cyclosporine or tacrolimus in the longterm management of UC or CD. The use of these rapidly acting drugs should be considered a 'bridge' to the onset of action of a safer, slow-acting immunosuppressive such as azathioprine, 6-mercaptopurine or methotrexate. It is our practice to start one of these agents along with cyclosporine or tacrolimus, and to overlap them for a sufficient duration to allow the slow-acting drug to work. Monitoring of blood levels is essential during cyclosporine and tacrolimus therapy, because of inter- and intra-individual variation in absorption and metaboHsm. During intravenous therapy the drug blood levels should be checked daily, at least until a stable dose is achieved. During oral therapy the levels should be checked twice weekly initially and once every 1-2 weeks after a stable dose is achieved. Target blood levels of cyclosporine that correlate with clinical efficacy range from 150 to 300 ng/ml (high-pressure liquid chromatography assay) [217, 224]. Effective tacrolimus concentrations range from 3 to 8 ng/ml [225, 226]. Both cyclosporine and tacrolimus are metabolized by the 3A4 isoform of

514

Clinical pharmacology in inflammatory bowel disease

cytochrome P450. This enzyme is inhibited by grapefruit juice, so patients should be cautioned against drinking this juice, unless they consume a constant amount on a regular basis, when its use may allow a lower dose of drug to be effective [227]. Toxicity to cyclosporine and tacrolimus often occur at these levels, and dose adjustment may be necessary. Hypocholesterolemia should be excluded before initiating cyclosporine [228]. Blood pressure should be checked twice weekly, and serum creatinine, potassium and glucose should be monitored every 1-2 weeks.

A delayed-release formulation of nicotine had a mean oral bioavailability of 40% [235].

Nicotine Chemistry and mechanism of action

Adverse effects In therapeutic trials for UC, transdermal nicotine has caused considerable side effects [229, 236, 237]. Adverse events occurring more frequently with nicotine than with placebo included dermatitis, dizziness and nausea. The occurrence of side-effects appears to be dose-related, but significant associations with plasma nicotine levels have not been found [229, 234, 236].

Optimizing therapy

Nicotine is a naturally occurring drug that is an agonist at ganglionic cholinergic receptors. This drug evokes a wide range of pharmacologic effects mediated by binding to receptors in autonomic ganglia, the adrenal medulla, the neuromuscular junction and in the central nervous system. In the gut, nicotine stimulates parasympathetic ganglia to cause increased cholinergic outflow, resulting in elevated smooth muscle tone and motor activity. In UC the therapeutic mechanism of action of nicotine is not known. This drug undoubtedly has profound effects of gastrointestinal motility that may be relevant to its action in UC. However, endoscopic inflammation is reduced by nicotine [229], suggesting an anti-inflammatory effect. In patients with UC, cigarette smoking is associated with decreased colonic production of IL-1, IL-8 and TNF-oc [230]. One study has found that colon mucosal expression of IL-8 decreases in UC patients who respond to nicotine patches, but not in those who fail to respond, or who were treated with placebo [231]. Based on our current level of knowledge no firm conclusions can be drawn about the mechanism of nicotine's beneficial action in UC, and more study is needed.

Nicotine therapy of UC is hampered by only modest efficacy in the face of considerable dose-limiting toxicity. Ex-smokers seem to be somewhat resistant to the nausea and light-headedness caused by nicotine [236]. Patients who develop UC for the first time, or who experience a disease flare after quitting smoking, may therefore represent a subpopulation that would be more likely to benefit from nicotine. When nicotine patches are prescribed for UC it is wise to start with a low dose, such as 7 mg/day, and gradually titrate the dose upwards according to response. Nicotine doses in the 22-25 mg/day range appear to be most effective. Nicotine is metabolized by the polymorphic cytochrome P450 2A6 enzyme. Individuals with decreased activity of this enzyme metabolize nicotine more slowly than normal, are less likely to become smokers, and when they do smoke they consume less cigarettes [238]. It is not known if nicotine metabolizer status is predictive of clinical outcome in the therapy of UC, but serum nicotine levels did not correlate with response [229]. Thus, there is no role for therapeutic drug monitoring of nicotine in IBD at present.

Pharmacolcinetics

References

The systemic bioavailability of nicotine from inhalation and transdermal patch is high, ranging from 75% to 90% [232, 233]. For use with therapeutic intent in UC, high blood levels of nicotine are not desirable because of side-effects. One study has evaluated the bioavailability of nicotine after oral and rectal administration [234]. Both oral and various rectal formulations of nicotine were found to have approximately 20%o systemic bioavailability.

1. 2. 3.

4.

Svartz N. Salazopyrine, a new sulfanilamide preparation. Acta Med Scand 1942; 110: 577. Azad Khan AK, Piris J, Truelove SC. An experiment to determine the active therapeutic moiety of sulfasalazine. Lancet 1977; 2: 892-5. van Hees PA, Bakker JH, van Tongeren JH. Effect of sulphapyridine, 5-aminosalicylic acid, and placebo in patients with idiopathic proctitis: a study to determine the active therapeutic moiety of sulphasalazine. Gut 1980; 21: 632-5. Klotz U, Maier K, Fischer C, Heinkel K. Therapeutic efficacy of sulfasalazine and its metabolites in patients with

Laurence J. Egan and William J. Sandborn

5.

6.

7.

8. 9.

10. 11. 12. 13.

14.

15. 16. 17. 18.

19.

20. 21.

22.

ulcerative colitis and Crohn's disease. N Engl J Med 1980; 303: 1499-502. Sharon P, Ligumsky M, Rachmilewitz D, Zor U. Role of prostaglandins in ulcerative colitis. Enhanced production during active disease and inhibition by sulfasalazine. Gastroenterology 1978; 75: 638-40. Ligumsky M, Karmeli F, Sharon P, Zor U, Cohen F, Rachmilewitz D. Enhanced thromboxane A2 and prostacyclin production by cultured rectal mucosa in ulcerative colitis and its inhibition by steroids and sulfasalazine. Gastroenterology 1981; 81: 444-9. Punchard NA, Boswell DJ, Greenfield SM, Thompson RP. The effects of sulphasalazine and its metabolites on prostaglandin production by human mononuclear cells. Biochem Pharmacol 1992; 43: 2369-76. Hawkey CJ, Broughton-Smith NK, Whittle BJR. Modulation of human colonic arachadonic acid metabolism by sulphasalazine. Dig Dis Sci 1985; 30: 1161-5. Isaacs KL, Sartor RB, Haskill S. Cytokine messenger RNA profiles in inflammatory bowel disease mucosa detected by polymerase chain reaction amplification. Gastroenterology 1992; 103: 1587-95. Plevy SE, Landers CJ, Prehn J et al A role for TNF-alpha and mucosal T helper-1 cytokines in the pathogenesis of Crohn's disease. J Immunol 1997; 159: 6276-82. Schreiber S, Nikolaus S, Hampe J. Activation of nuclear factor kappa B inflammatory bowel disease. Gut 1998; 42: 477-84. Rogler G, Brand K, Vogl D et al. Nuclear factor kappa B is activated in macrophages and epithelial cells of inflamed intestinal mucosa. Gastroenterology 1998; 115: 357-69. Neurath MF, Pettersson S, Meyer zum Buschenfelde KH, Strober W. Local administration of antisense phosphorothioate oligonucleotides to the p65 subunit of NF-kappa B abrogates established experimental colitis in mice. Nat Med 1996; 2: 998-1004. Thiele K, Bierhaus A, Autschbach F et al. Cell specific effects of glucocorticoid treatment on the NF-kappaBp65/ IkappaBalpha system in patients with Crohn's disease. Gut 1999; 45: 693-704. Barnes PJ, Karin M. Nuclear factor kappa B - a pivotal transcription factor in chronic inflammatory diseases. N Engl J Med 1997; 336: 1066-71. Kopp E, Ghosh S. Inhibition of NF-kB by sodium salicylate and aspirin. Science 1994; 265: 956-8. Wahl C, Liptay S, Adler G, Schmid RM. Sulfasalazine: a potent and specific inhibitor of nuclear factor kappa B. J Clin Invest 1998; 101: 1163-74. Egan LJ, Mays DC, Huntoon CJ et al. Inhibition of interleukin-1-stimulated NF-kappaB RelA/p65 phosphorylation by mesalamine is accompanied by decreased transcriptional activity. J Biol Chem 1999; 274: 26448-53. Kaiser GC, Yan F, Polk DB. Mesalamine blocks tumor necrosis factor growth inhibition and nuclear factor kappaB activation in mouse colonocytes. Gastroenterology 1999; 116:602-9. Shanahan F, Niederlehner A, Carramanzana N, Anton P. Sulfasalazine inhibits the binding of TNF alpha to its receptor. Immunopharmacology 1990; 20: 217-24. Crotty B, Rosenberg WM, Aronson JK, Jewell DP. Inhibition of binding of interferon-gamma to its receptor by saHcylates used in inflammatory bowel disease. Gut 1992; 33: 1353-7. Lauritsen K, Laursen LS, Bukhave K, Rask-Madsen J. Effects of topical 5-aminosalicylic acid and prednisolone on prostaglandin E2 and leukotriene B4 levels determined by equilibrium in vivo dialysis of rectum in relapsing ulcerative colitis. Gastroenterology 1986; 91: 837-44.

515 23.

24.

25.

26.

27. 28.

29. 30.

31.

32.

33. 34.

35.

36. 37.

38.

39. 40.

Roberts WG, Simon TJ, Berlin RG et al. Leukotrienes in ulcerative colitis: results of a multicenter trial of a leukotriene biosynthesis inhibitor, MK-591. Gastroenterology 1997; 112:725-32. Dallegri F, Ottonello L, Ballestrero A, Bogliolo F, Ferrando F, Patrone F. Cytoprotection against neutrophil derived hypochlorous acid: a potential mechanism for the therapeutic action of 5-aminosalicylic acid in ulcerative colitis. Gut 1990; 31: 184-6. Sandoval M, Liu X, Mannick EE, Clark DA, Miller MJ. Peroxynitrite-induced apoptosis in human intestinal epithelial cells is attenuated by mesalamine. Gastroenterology 1997;113: 1480-8. Burress GC, Musch MW, Jurivich DA, Welk J, Chang EB. Effects of mesalamine on the hsp72 stress response in rat IEC-18 intestinal epithelial cells. Gastroenterology 1997; 113:1474-9. Yu DK, Elvin AT, Morrill B et al. Effect of food coadministration on 5-aminosalicylic acid oral suspension bioavailability. Clin Pharmacol Ther 1990; 48: 26-33. Aimer S, Norlander B, Strom M, Osterwald H. Steady-state pharmacokinetics of a new 4-g 5-aminosahcylic acid retention enema in patients with ulcerative colitis in remission. Scand J Gastroenterol 1991; 26: 327-35. Klotz U. Clinical pharmacokinetics of sulphasalazine, its metabolites and other prodrugs of 5-aminosalicyhc acid. Clin Pharmacokinet 1985; 10: 285-302. Staerk Laursen L, Stokholm M, Bukhave K, Rask-Madsen J, Lauritsen K. Disposition of 5-aminosalicylic acid by olsalazine and three mesalazine preparations in patients with ulcerative colitis: comparison of intraluminal colonic concentrations, serum values, and urinary excretion. Gut 1990; 31: 1271-6. Stretch GL, Campbell BJ, Dwarakanath AD et al. 5-Amino salicylic acid absorption and metabolism in ulcerative colitis patients receiving maintenance sulphasalazine, olsalazine or mesalazine. Aliment Pharmacol Ther 1996; 10: 941-7. Gionchetti P, Campieri M, Venturi A et al. Systemic availability of 5-aminosalicylic acid: comparison of delayed release and an azo-bond preparation. Aliment Pharmacol Ther 1996; 10: 601-5. Goebell H, Klotz U, Nehlsen B, Layer P. Oroileal transit of slow release 5-aminosalicylic acid. Gut 1993; 34: 669-75. Norlander B, Gotthard R, Strom M. Pharmacokinetics of a 5-aminosalicylic acid enteric-coated tablet in patients with Crohn's disease or ulcerative colitis and in healthy volunteers. Aliment Pharmacol Ther 1990; 4: 497-505. Layer PH, Goebell H, Keller J, Dignass A, Klotz U. Delivery and fate of oral mesalamine microgranules within the human small intestine. Gastroenterology 1995; 108: 1427-33. Yu DK, Morrill B, Eichmeier LS et al. Pharmacokinetics of 5-aminosalicylic acid from controlled-release capsules in man. Eur J Clin Pharmacol 1995; 48: 273-7. Lauritsen K, Hansen J, Ryde M, Rask-Madsen J. Colonic azodisalicylate metabolism determined by in vivo dialysis in healthy volunteers and patients with ulcerative colitis. Gastroenterology 1984; 86: 1496-500. Dew MJ, Ebden P, Kidwai NS, Lee G, Evans BK, Rhodes J. Comparison of the absorption and metabolism of sulphasalazine and acrylic-coated 5-amino salicyhc acid in normal subjects and patients with colitis. Br J CHn Pharmacol 1984; 17: 474-6. Sutherland SR, May GR, Schaffer GA. Sulfasalazine revisited: a meta-analysis of 5-aminosalicylic acid in the treatment of ulcerative colitis. Ann Intern Med 1993; 118: 540. Gionchetti P, Venturi A, Rizzello F et al. Retrograde colonic spread of a new mesalazine rectal enema in patients with

Clinical pharmacology in inflammatory bowel disease

516

41. 42. 43. 44.

45. 46. 47.

48. 49. 50. 51. 52. 53. 54. 55.

56. 57.

58.

59.

60.

distal ulcerative colitis. Aliment Pharmacol Ther 1997; 11: 679-84. Sutherland LR, Roth DE, Beck PL. Alternatives to sulfasalazine: a meta-analysis of 5-ASA in the treatment of ulcerative colitis. Inflam Bowel Dis 1997; 3: 65-78. O'Morain C, Smethurst P, Dore CJ, Levi AJ. Reversible male infertility due to sulphasalazine: studies in man and rat. Gut 1984; 25: 1078-84. Franklin JL, Rosenberg HH. Impaired folic acid absorption in inflammatory bowel disease: effects of salicylazosulfapyridine (Azulfidine). Gastroenterology 1973; 64: 517-25. Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral 5aminosalicylic acid therapy for mildly to moderately active ulcerative colitis. A randomized study. N Engl J Med 1987; 317:1625-9. The Mesalamine Study Group. An oral preparation of mesalaminc as long-term maintenance therapy for ulcerative colitis. Ann Intern Med 1996; 124: 204-11. Sninsky CA, Cort DH, Shanahan F et al. Oral mesalamine (Asacol) for mildly to moderately active ulcerative colitis. A multicenter study. Ann Intern Med 1991; 115: 350 5. Sentongo TA, Piccoli DA. Recurrent pericarditis due to mesalaminc hypersensitivity: a pediatric case report and review of the literature. J Pcdiatr Gastroenterol Nutr 1998; 27: 344 7. Lazaro MT, Garcia-Tcjcro MT, Diaz-Lobato S. Mcsalamine-induccd lung disease. Arch Intern Med 1997; 157: 462. Abboudi ZH, Marsh JC, Smith-Laing G, Gordon-Smith EC. Fatal aplastic anaemia after mcsalazinc. Lancet 1994; 343: 542. Block MB, Genant HK, Kirsner JB. Pancreatitis as an adverse reaction to salicylazosulfapyridine. N Engl J Med 1970; 282: 380 2. Sturgeon JB, Bhatia P, Hermens D, Miner PB, Jr. Exacerbation of chronic ulcerative colitis with mesalamine. Gastroenterology 1995; 108: 1889 93. Novis BH, Korzets Z, Chen P, Bcrnhcim J. Nephrotic syndrome after treatment with 5-aminosalicylic acid. Br Med J (Clin Res Ed) 1988; 296: 1442. harbour VM, Williams PF. Nephrotic syndrome associated with sulphasalazine. Br Med J 1990; 301: 818. Schreiber S, Hamling J, Zehnter E ct al. Renal tubular dysfunction in patients with inflammatory bowel disease treated with aminosalicylate. Gut 1997; 40: 761 6. Singleton JW, Hanauer SB, Gitnick GL ct al. Mesalamine capsules for the treatment of active Crohn's disease: results of a 16-week trial. Pentasa Crohn's Disease Study Group. Gastroenterology 1993; 104: 1293-301. Mahida YR, Jewell DP. Slow-release 5-amino-salicylic acid (Pentasa) for the treatment of active Crohn's disease. Digestion 1990;45:88-92. Rasmussen SN, Lauritsen K, Tage-Jensen U et al. 5-Aminosalicylic acid in the treatment of Crohn's disease. A 16-week double-blind, placebo-controlled, multicentre study with Pentasa. Scand J Gastroenterol 1987; 22: 877-83. Steinhart AH, Hemphill D, Greenberg GR. Sulfasalazine and mesalazine for the maintenance therapy of Crohn's disease: a meta-analysis. Am J Gastroenterol 1994; 89: 2116-24. Messori A, Brignola C, Trallori G et al. Effectiveness of 5aminosalicylic acid for maintaining remission in patients with Crohn's disease: a meta-analysis. Am J Gastroenterol 1994;89:692-8. Gendre JP, Mary JY, Florent C et al. Oral mesalamine (Pentasa) as maintenance treatment in Crohn's disease: a multicenter placebo-controlled study. Groupe d'Etudes Therapeutiques des Affections Inflammatoires Digestives (GETAID). Gastroenterology 1993; 104: 435-9.

61.

62.

63.

64. 65. 66. 67. 68.

69.

70.

71.

72.

73.

74.

75. 76. 77.

78.

Frieri G, Pimpo MT, Palumbo GC et al. Rectal and colonic mesalazine concentration in ulcerative colitis: oral vs. oral plus topical treatment. Aliment Pharmacol Ther 1999; 13: 1413-17. Frieri G, Pimpo MT, Andreoli A et al. Prevention of postoperative recurrence of Crohn's disease requires adequate mucosal concentration of mesalazine. Gruppo Italiano per lo Studio del Colon e del Retto. Aliment Pharmacol Ther 1999; 13:577-82. Hein DW, Doll MA, Rustan TD et al. Metabolic activation and deactivation of arylamine carcinogens by recombinant human NATl and polymorphic NAT2 acetyltransferases. Carcinogenesis 1993; 14: 1633-8. Hickman D, Pope J, Patil SD et al. Expression of arylamine N-acetyltransferase in human intestine. Gut 1998; 42: 4029. Grant DM, Hughes NC, Janezic SA et al. Human acetyltransferase polymorphisms. Mutat Res 1997; 376: 61 70. Allgayer H, Ahnfelt NO, Kruis W et al. Colonic N-acetylation of 5-aminosalicylic acid in inflammatory bowel disease. Gastroenterology 1989; 97: 38 41. Ireland A, Priddle JD, Jewell DP. Acetylation of 5-aminosalicylic acid by isolated human colonic epithelial cells. Clin Sci(Colch) 1990; 78: 105 11. Karamanolis DG, Papatheodoridis GV, Xourgias V. Systemic absorption of 5-aminosalicylic acid in patients with inactive ulcerative colitis treated with olsalazine and mesalazine. Eur J Gastroenterol Hepatol 1996; 8: 1083 8. Easterbrook J, Armstrong G, Bayless TM, Harris ML, Montrose MH. Pharmacokinetics of 5-ASA metabolism in IBD patients: preliminary report of variation in N-acctyltransferase 1 gene polymorphisms and 5-ASA metabolism rates in mucosal biopsies. Gastroenterology 1998; 114: G3792. Scheinman RI, Gualbcrto A, Jewell CM, Cidlowski JA, Baldwin AS, Jr. Characterization of mechanisms involved in Iransrcpression of NF-kappa B by activated glucocorticoid receptors. Mol Cell Biol 1995; 15: 943 53. Ray A, Prefontaine KE. Physical association and functional antagonism between the p65 subunit of transcription factor NF-kappa B and the glucocorticoid receptor. Proc Natl AcadSci USA 1994; 91: 752 6. Scheinman RI, Cogswell PC, Lofquist AK, Baldwin AS, Jr. Role oi' transcriptional activation of I kappa B alpha in mediation of immunosuppression by glucocorticoids. Science 1995;270:283-6. Auphan N, DiDonato JA, Rosette C, Helmberg A, Karin M. Immunosuppression by glucocorticoids: inhibition of NF-kappa B activity through induction of I kappa B synthesis. Science 1995; 270: 286-90. Paliogianni F, Raptis A, Ahuja SS, Najjar SM, Boumpas DT. Negative transcriptional regulation of human interleukin 2 (IL-2) gene by glucocorticoids through interference with nuclear transcription factors AP-1 and NF-AT. J Clin Invest 1993; 91: 1481-9. Flower RJ, Rothwell NJ. Lipocortin-1: cellular mechanisms and clinical relevance. Trends Pharmacol Sci 1994; 15:71-6. Colotta F, Re F, Muzio M et al. Interleukin-1 type II receptor: a decoy target for IL-1 that is regulated by IL-4. Science 1993;261:472-5. Brostjan C, Anrather J, Csizmadia V, Natarajan G, Winkler H. Glucocorticoids inhibit E-selectin expression by targeting NF-kappaB and not ATF/c-Jun. J Immunol 1997; 158: 3836^4. Wissink S, van Heerde EC, vand der Burg B, van der Saag P T A dual mechanism mediates repression of NF-kappaB activity by glucocorticoids. Mol Endocrinol 1998; 12: 35563.

Laurence J. Egan and William J. Sandborn 79.

80.

81.

82.

83. 84. 85. 86. 87. 88. 89. 90.

91. 92.

93. 94.

95.

96. 97. 98.

Rachmilewitz D, Stamler JS, Bachwich D, Karmeli F, Ackerman Z, Podolsky DK. Enhanced colonic nitric oxide generation and nitric oxide synthase activity in ulcerative colitis and Crohn's disease. Gut 1995; 36: 718-723. Leonard N, Bishop AE, Polak JM, Talbot IC. Expression of nitric oxide synthase in inflammatory bowel disease is not affected by corticosteroid treatment. J Clin Pathol 1998; 51: 750-3. Tillinger W, Gasche C, Reinisch W et al. Influence of topically and systemically active steroids on circulating leukocytes in Crohn's disease. Am J Gastroenterol 1998; 93: 1848-53. Bernstein CN, Sargent M, Rawsthorne P, Rector E. Peripheral blood lymphocyte beta 2 integrin and ICAM expression in inflammatory bowel disease. Dig Dis Sci 1997; 42: 233849. Powell LW, Axelsen E. Corticosteroids in liver disease: studies on the biological conversion of prednisone to prednisolone and plasma protein binding. Gut 1972; 13: 690-6. Tanner AR, Halliday JW, Powell LW. Serum prednisolone levels in Crohn's disease and coeliac disease following oral prednisolone administration. Digestion 1981; 21: 310-15. Shaffer JA, Williams SE, Turnberg LA, Houston JB, Rowland M. Absorption of prednisolone in patients with Crohn's disease. Gut 1983; 24: 182-6. Petitjean O, Wendling JL, Tod M et al. Pharmacokinetics and absolute rectal bioavailability of hydrocortisone acetate in distal colitis. Aliment Pharmacol Ther 1992; 6: 351-7. Cann PA, Holdsworth CD. Systemic absorption from hydrocortisone foam enema in ulcerative colitis. Lancet 1987; 1:922-3. Lee DA, Taylor M, James VH, Walker G. Rectally administered prednisolone - evidence for a predominantly local action. Gut 1980; 21: 215-18. Mclntyre PB, Macrae FA, Berghouse L, English J, LennardJones JE. Therapeutic benefits from a poorly absorbed prednisolone enema in distal colitis. Gut 1985; 26: 822-24. Luman W, Gray RS, Pendek R, Palmer KR. Prednisolone metasulphobenzoate foam retention enemas suppress the hypothalamo-pituitary-adrenal axis. Aliment Pharmacol Ther 1994; 8: 255-8. Robinson RJ, Iqbal SJ, Whitaker RP, Abrams K, Mayberry JF. Rectal steroids suppress bone formation in patients with colitis. AHment Pharmacol Ther 1997; 11: 201-4. Dahlberg E, Thalen A, Brattsand R et al. Correlation between chemical structure, receptor binding, and biological activity of some novel, highly active, 16 alpha, 17 alphaacetal-substituted glucocorticoids. Mol Pharmacol 1984; 25: 70-8. Ryrfeldt A, Edsbacker S, Pauwels R. Kinetics of the epimeric glucocorticoid budesonide. Clin Pharmacol Ther 1984; 35: 525-30. Edsbacker S, Wollmer P, Nilsson A, Nilsson M. Pharmacokinetics and gastrointestinal transit of budesonide controlled ileal release capsules. Gastroenterology 1983; 104: A695. Edsbacker S, Nilsson M, Larsson P. A Cortisol suppression dose-response comparison of budesonide in controlled ileal release capsules with prednisolone. Aliment Pharmacol Ther 1999; 13: 219-24. Greenberg GR, Feagan BG, Martin F et al. Oral budesonide for active Crohn's disease. Canadian Inflammatory Bowel Disease Study Group. N Engl J Med 1994; 331: 836-41. Rutgeerts P, Lofberg R, Malchow H et al. A comparison of budesonide with prednisolone for active Crohn's disease. N Engl J Med 1994; 331: 842-5. Dahlstrom K, Edsbacker S, Kallen A. Rectal pharmacokinetics of budesonide. Eur J Clin Pharmacol 1996; 49: 293-8.

517 99.

100.

101. 102.

103.

104.

105.

106.

107.

108. 109. 110.

111. 112. 113. 114. 115. 116. 117.

118.

Hanauer SB, Robinson M, Pruitt R et al. Budesonide enema for the treatment of active, distal ulcerative colitis and proctitis: a dose-ranging study. U.S. Budesonide enema study group. Gastroenterology 1998; 115: 525-32. Hawthorne AB, Record CO, Holdsworth CD et al. Double blind trial of oral fluticasone propionate v prednisolone in the treatment of active ulcerative coHtis. Gut 1993; 34: 1258. Angus P, Snook JA, Reid M, Jewell DP. Oral fluticasone propionate in active distal ulcerative colitis. Gut 1992; 33: 711-14. D'Arienzo A, Manguso F, Castiglione GN et al. Beclomethasone dipropionate (3 mg) enemas combined with oral 5-ASA (2.4 g) in the treatment of ulcerative colitis not responsive to oral 5-ASA alone. Ital J Gastroenterol Hepatol 1998; 30: 254-7. Campieri M, Cottone M, Miglio F et al. Beclomethasone dipropionate enemas versus prednisolone sodium phosphate enemas in the treatment of distal ulcerative colitis. Aliment Pharmacol Ther 1998; 12: 361-6. Halpern Z, Sold O, Baratz M, Konikoff"F, Halak A, Gilat T. A controlled trial of beclomethasone versus betamethasone enemas in distal ulcerative colitis. J Clin Gastroenterol 1991; 13:38-41. Singleton JW, Law DH, Kefley ML, Jr, Mekhjian HS, Sturdevant RA. National Cooperative Crohn's Disease Study: adverse reactions to study drugs. Gastroenterology 1979; 77: Pt 2): 870-82. Bjarnason I, Macpherson A, Mackintosh C, Buxton-Thomas M, Forgacs I, Moniz C. Reduced bone density in patients with inflammatory bowel disease. Gut 1997; 40: 228-33. Silvennoinen JA, Karttunen TJ, Niemela SE, Manelius JJ, Lehtola JK. A controlled study of bone mineral density in patients with inflammatory bowel disease. Gut 1995; 37: 7 1 6. Abitbol V, Roux C, Chaussade S et al. Metabolic bone assessment in patients with inflammatory bowel disease. Gastroenterology 1995; 108: 417-22. Adinoff" AD, Hollister JR. Steroid-induced fractures and bone loss in patients with asthma. N Engl J Med 1983; 309: 265-8. Schlaghecke R, Kornely E, Santen RT, Ridderskamp P. The effect of long-term glucocorticoid therapy on pituitaryadrenal responses to exogenous corticotropin-releasing hormone. N Engl J Med 1992; 326: 226-30. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Final report on a therapeutic trial. Br Med J 1955; 4947: 1041-8. Truelove SC. Treatment of ulcerative colitis with local hemisuccinate sodium. A report on a controlled therapeutic trial. Br Med J 1958; 2: 1072-7. Munkholm P, Langholz E, Davidsen M, Binder V. Frequency of glucocorticoid resistance and dependency in Crohn's disease. Gut 1994; 35: 360-2. Lamberts SW, Koper JW, de Jong FH. Familial and iatrogenic Cortisol receptor resistance. J Steroid Biochem Mol Biol 1992; 43: 385-8. Barnes PJ, Adcock IM. Steroid resistance in asthma. Q J Med 1995; 88: 455-68. Lane SJ, Lee TH. Mononuclear cells in corticosteroidresistant asthma. Am J Respir Crit Care Med 1996; 154: S49-51; discussion S52. Sher ER, Leung DYM, Surs W et al. Steroid resistant asthma. Cellular mechanisms contributing to inadequate response to glucocorticoid therapy. J Clin Invest 1994; 93: 33-9. Lane SJ, Lee TH. Glucocorticoid receptor characteristics in monocytes of patients with corticosteroid-resistant bronchial asthma. Am Rev Respir Dis 1991; 143: 1020-4.

518 119. Corrigan CJ, Brown PH, Barnes NC, Tsai JJ, Frew AJ, Kay AB, Glucocorticoid resistance in chronic asthma. Peripheral blood T lymphocyte activation and comparison of the T lymphocyte inhibitory effects of glucocorticoids and cyclosporin A. Am Rev Respir Dis 1991; 144: 1026-32. 120. Lane SJ, Adcock IM, Richards D, Hawrylowicz C, Barnes PJ, Lee TH. Corticosteroid resistant bronchial asthma is associated with increased c-fos expression in monocutes and T lymphocytes. J Clin Invest 1998; 102: 2156-64. 121. Franchimont D, Louis E, Dupont P et al. Decreased corticosensitivity in quiescent Crohn's disease: an ex vivo study using whole blood cell cultures. Dig Dis Sci 1999; 44: 1208-15. 122. Hearing SD, Norman M, Probert CS, Haslam N, Dayan CM. Predicting therapeutic outcome in severe ulcerative colitis by measuring in vitro steroid sensitivity of proliferating peripheral blood lymphocytes. Gut 1999; 45: 382 8. 123. Lennard L. The clinical pharmacology of 6-mercaptopurine. Eur J Clin Pharmacol 1992; 43: 329 39. 124. Van Os EC, Zins BJ, Sandborn WJ ct al. Azathioprine pharmacokinetics after intravenous, oral, delayed release oral and rectal foam administration. Gut 1996; 39: 63 8. 125. Lennard L, Brown CB, Fox M, Maddocks JL. Azathioprine metabolism in kidney transplant recipients. Br J Clin Pharmacol 1984; 18:693 700. 126. Sandborn WJ, Tremainc WJ, Wolf DC ct al. Lack of effect of intravenous administration on time to respond to azathioprine for steroid-treated Crohn's disease. North American Azathioprine Study Group. Gastroenterology 1999; 117: 527 35. 127. Lennard L. Clinical implications of thiopurine methyltransferasc optimization of drug dosage and potential drug interactions. Ther Drug Monit 1998; 20: 527 31. 128. Bergan S, Rugstad HE, Bentdal O, Stokkc O. Monitoring of azathioprine treatment by determination of 6-thioguanine nucleotide concentrations in erythrocytes. Transplantation 1994; 58: 803 8. 129. Pearson DC, May G R , Fick G H , Sutherland LR. Azathioprine and 6-mercaptopurine in Crohn disease. A meta-analysis. Ann Intern Med 1995; 122: 132 42. 130. Sandborn WJ, Van O, EC, Zins BJ, Tremaine WJ, Mays DC, Lipsky JJ. An intravenous loading dose of azathioprine decreases the time to response in patients with Crohn's disease. Gastroenterology 1995; 109: 1808 17. 131. Weinshilboum RM, Sladek SL. Mercaptopurine pharmacogenetics: monogenic inheritance of erythrocyte thiopurine methyltransferase activity. Am J Hum Genet 1980; 32: 65162. 132. Weinshilboum R. Methyltransferase pharmacogenetics. Pharmacol Ther 1989; 43: 77-90. 133. Lennard L, Van Loon JA, Lilleyman JS, Weinshilboum RM. Thiopurine pharmacogenetics in leukemia: correlation of erythrocyte thiopurine methyltransferase activity and 6thioguanine nucleotide concentrations. Clin Pharmacol Ther 1987; 41: 18-25. 134. Present DH, Meltzer SJ, Krumholz MP, Wolke A, Korelitz BI. 6-Mercaptopurine in the management of inflammatory bowel disease: short- and long-term toxicity. Ann Intern Med 1989; 111:641-9. 135. Connell WR, Kamm MA, Ritchie JK, Lennard-Jones JE. Bone marrow toxicity caused by azathioprine in inflammatory bowel disease: 27 years of experience. Gut 1993; 34: 1081-5. 136. Bouhnik Y, Lemann M, Mary JY et al. Long-term follow-up of patients with Crohn's disease treated with azathioprine or 6-mercaptopurine. Lancet 1996; 347: 215-19. 137. Connell WR, Kamm MA, Dickson M, Balkwill AM, Ritchie JK, Lennard-Jones JE. Long-term neoplasia risk

Clinical pharmacology in inflammatory bowel disease

138. 139.

140.

141. 142. 143.

144. 145. 146.

147. 148.

149.

150.

151.

152.

153.

154.

after azathioprine treatment in inflammatory bowel disease. Lancet 1994; 343: 1249-52. Larvol L, Soule JC, Tourneau AL. Reversible lymphoma in the setting of azathioprine therapy for Crohn's disease. N Engl J Med 1994;331:883-4. Krynetski EY, Tai HL, Yates CR et al. Genetic polymorphism of thiopurine S'-methyltransferase: clinical importance and molecular mechanisms. Pharmacogenetics 1996; 6: 279-90. Lennard L, Lilleyman JS, Van Loon J, Weinshilboum RM. Genetic variation in response to 6-mercaptopurine for childhood acute lymphoblastic leukaemia. Lancet 1990; 336: 225-9. Lilleyman JS, Lennard L. Mercaptopurine metabolism and risk of relapse in childhood lymphoblastic leukaemia. Lancet 1994; 343: 1188 90. Cuffari C, Theoret Y, Latour S, Seidman G. 6-Mercaptopurine metabolism in Crohn's disease: correlation with efficacy and toxicity. Gut 1996; 39: 401 6. Dubinsky MC, Lamolhe S, Yang HY et al. Optimizing and individualizing 6-MP therapy in IBD: the role of 6-MP metabolite levels and TPMT genotyping. Gastroenterology 1999; 116: A702. Dubinsky M, Hassard P, Yang HY et al. 6-MP metabolite levels correlate with clinical response and drug toxicity in adult IBD. Am J Gastroenterol 1999; 94: 2641. Cuffari C, Sharma S. 6-Mcrcaptopurinc metabolite levels tailored to achieve clinical responsiveness in pediatric Crohn's disease. Gastroenterology 1999; 116: A694. Lowry PW, Franklin CL, Weaver AL et al. Measurement of thiopurine methyltransferase activity and azathioprine metabolites in patients with inflammatory bowel disease. Gut 2001; 49: 665 70. Colonna T, Korelitz BI. The role of leukopenia in the 6mercaptopurine-induced remission of refractory Crohn's disease. Am J Gastroenterol 1994; 89: 362 6. Barbc L, Matycau P, Lemann M et al. Dose raising of azathioprine beyond 2.5 mg/kg/day in Crohn's disease patients who fail to improve with a standard dose. Gastroenterology 1998; 114: A925. Jolivel J, Schilsky RL, Bailey BD, Drake JC, Chabner BA. Synthesis, retention, and biological activity of methotrexate polyglutamates in cultured human breast cancer cells. J Clin Invest 1982; 70: 351 60. Allegra CJ, Drake JC, Jolivet J, Chabner BA. Inhibition of phosphoribosylaminoimidazolecarboxamide transformylase by methotrexate and dihydrofolic acid polyglutamates. Proc Natl Acad Sci USA 1985; 82: 4881-5. Baggott JE, Vaughn WH, Hudson BB. Inhibition of 5aminoimidazole-4-carboxamide ribotide transformylase, adenosine deaminase and 5'-adenylate deaminase by polyglutamates of methotrexate and oxidized folates and by 5aminoimidazole-4-carboxamide riboside and ribotide. BiochemJ 1986; 236: 193-200. Egan LJ, Sandborn WJ, Mays DC, Tremaine WJ, Fauq AH, Lipsky JJ. Systemic and intestinal pharmacokinetics of methotrexate in patients with inflammatory bowel disease. Clin Pharmacol Ther 1999; 65: 29-39. Jolivet J, Chabner BA. Intracellular pharmacokinetics of methotrexate polyglutamates in human breast cancer cells. Selective retention and less dissociable binding of 4-NH210-CH3-pteroylglutamate4 and 4-NH2- 10-CH3-pteroylglutamate5 to dihydrofolate reductase. J Clin Invest 1983; 72: 773-8. Weinstein GD, Jeffes E, McCullough JL. Cytotoxic and immunologic effects of methotrexate in psoriasis. J Invest Dermatol 1990; 95: 49-52S.

Laurence J. Egan and William J. Sandborn 155. Olsen NJ, Murray LM. Antiproliferative effects of methotrexate on peripheral blood mononuclear cells. Arthritis Rheum 1989; 32: 378-85. 156. Morgan SL, Baggott JE, Vaughn WH et al. Supplementation with folic acid during methotrexate therapy for rheumatoid arthritis. A double-blind, placebo-controlled trial. Ann Intern Med 1994; 121: 833-41. 157. Sperling RI, Benincaso AI, Anderson RJ, Coblyn JS, Austen KF, Weinblatt ME. Acute and chronic suppression of leukotriene B4 synthesis ex vivo in neutrophils from patients with rheumatoid arthritis beginning treatment with methotrexate. Arthritis Rheum 1992; 35: 376-84. 158. Thomas R, Carroll GJ. Reduction of leukocyte and interleukin-1 beta concentrations in the synovial fluid of rheumatoid arthritis patients treated with methotrexate. Arthritis Rheum 1993; 36: 1244-52. 159. Crilly A, Mclnness IB, McDonald AG, Watson J, Capell HA, Madhok R. Interleukin 6 (IL-6) and soluble IL-2 receptor levels in patients with rheumatoid arthritis treated with low dose oral methotrexate. J Rheumatol 1995; 22: 224-6. 160. Cronstein BN, Eberle MA, Gruber HE, Levin RI. Methotrexate inhibits neutrophil function by stimulating adenosine release from connective tissue cells. Proc Natl Acad Sci USA 1991; 88: 2441-5. 161. Cronstein BN, Naime D, Ostad E. The antiinflammatory mechanism of methotrexate. Increased adenosine release at inflamed sites diminishes leukocyte accumulation in an in vivo model of inflammation. J Chn Invest 1993; 92: 267582. 162. Asako H, Wolf RE, Granger DN. Leukocyte adherence in rat mesenteric venules: effects of adenosine and methotrexate. Gastroenterology 1993; 104: 31-7. 163. Egan LJ, Sandborn WJ, Mays DC, Tremaine WJ, Lipsky JJ. Plasma and rectal adenosine in inflammatory bowel disease: eff'ect of methotrexate. Inflam Bowel Dis 1999; 5: 167-73. 164. Elson CO, Sartor RB, Tennyson GS, Riddell RH. Experimental models of inflammatory bowel disease. Gastroenterology 1995;109: 1344-67. 165. Genestier L, Paillot R, Fournel S, Ferraro C, Miossec P, Revillard J-P. Immunosuppressive properties of methotrexate: apoptosis and clonal deletion of activated peripheral T cells. J Clin Invest 1998; 102: 322-8. 166. Dahl MGG, Gregory MM, Scheurer PJ. Methotrexate hepatotoxicity in psoriasis. Comparison of different dose regimens. Br Med J 1972: 654-56. 167. Kozarek RA, Patterson DJ, Gelfand MD, Botoman VA, Ball TJ, Wilske KR. Methotrexate induces clinical and histologic remission in patients with refractory inflammatory bowel disease. Ann Intern Med 1989; 110: 353-6. 168. Phelan MJ, Taylor W, van Heyningen C, Williams E, Thompson RN. Intestinal absorption in patients with rheumatoid arthritis treated with methotrexate. Clin Rheumatol 1993; 12: 223-5. 169. Teresi ME, Crom WR, Choi KE, Mirro J, Evans WE. Methotrexate bioavailability after oral and intramuscular administration in children. J Pediatr 1987; 110: 788-92. 170. Jundt JW, Browne BA, Fiocco GP, Steele AD, Mock D. A comparison of low dose methotrexate bioavailability: oral solution, oral tablet, subcutaneous and intramuscular dosing. J Rheumatol 1993; 20: 1845-9. 171. Oguey D, KoUiker F, Gerber NJ, Reichen J. Effect of food on the bioavailability of low-dose methotrexate in patients with rheumatoid arthritis. Arthritis Rheum 1992; 35: 61114. 172. Seideman P, Beck O, Eksborg S, Wennberg M. The pharmacokinetics of methotrexate and its 7-hydroxy metabolite in patients with rheumatoid arthritis. Br J Clin Pharmacol 1993; 35: 409-12.

519 173. Egan LJ, Sandborn WJ, Tremaine WJ et al. A randomized dose-response and pharmacokinetic study of methotrexate for refractory inflammatory Crohn's disease and ulcerative colitis. Aliment Pharmacol Ther 1999; 13: 1597-604. 174. Baron TH, Truss CD, Elson CO. Low-dose oral methotrexate in refractory inflammatory bowel disease. Dig Dis Sci 1993; 38: 1851-6. 175. Lemann M, Chamiot-Prieur C, Mesnard B etal. Methotrexate for the treatment of refractory Crohn's disease. Aliment Pharmacol Ther 1996; 10: 309-14. 176. Arora S, Katkov WN, Cooley J et al. A double-blind, randomized, placebo-controlled trial of methotrexate in Crohn's disease. Gastroenterology 1992; 102: A591. 177. Feagan BG, Rochon J, Fedorak RN et al. Methotrexate for the treatment of Crohn's disease. The North American Crohn's Study Group Investigators. N Engl J Med 1995; 332: 292-7. 178. Oren R, Arber N, Odes S et al. Methotrexate for chronic ulcerative colitis: a double-blind, randomized, Israeli multicenter trial. Gastroenterology 1996; 110: 1416-21. 179. Oren R, Moshkowitz M, Odes S et al. Methotrexate in chronic active Crohn's disease: a double-blind, randomized, Israeli multicenter trial. Am J Gastroenterol 1997; 92: 2203-9. 180. Goodman TA, Polisson RP. Methotrexate: adverse reactions and major toxicities.. Rheum Dis Clin N Am 1994; 20: 513-28. 181. Kremer JM, Alarcon GS, Lightfoot RW, Jr. etal. Methotrexate for rheumatoid arthritis. Suggested guidelines for monitoring liver toxicity. American College of Rheumatology. Arthritis Rheum 1994; 37: 316-28. 182. Podurgiel BJ, McGill DB, Ludwig J, Taylor WF, Muller SA. Liver injury associated with methotrexate therapy for psoriasis. Mayo Clin Proc 1973; 48: 787-92. 183. Zachariae H, Kragballe K, Thestrup-Pedersen K, Kissmeyer-Nielsen F. HLA antigens in methotrexate-induced Hver cirrhosis. Acta Dermato-Venereol 1980; 60: 165-6. 184. Nyfors A, Poulsen H. Liver biopsies from psoriatics related to methotrexate therapy. 2. Findings before and after methotexate therapy in 88 patients. A blind study. Acta Pathol Microbiol Scand Pathol 1976; 84: 262-70. 185. Newman M, Auerbach R, Feiner H et al. The role of liver biopsies in psoriatic patients receiving long- term methotrexate treatment. Improvement in liver abnormalities after cessation of treatment. Arch Dermatol 1989; 125: 1218-24. 186. Lewis JH, Schiff" E. Methotrexate-induced chronic liver injury: guidelines for detection and prevention. The ACG Committee on FDA-related matters. American College of Gastroenterology. Am J Gastroenterol 1988; 83: 1337-45. 187. Walker AM, Funch D, Dreyer NA et al. Determinants of serious liver disease among patients receiving low-dose methotrexate for rheumatoid arthritis. Arthritis Rheum 1993; 36: 329-35. 188. Seideman P. Methotrexate - the relationship between dose and clinical effect. Br J Rheumatol 1993; 32: 751-3. 189. Gorlick R, Goker E, Trippett T, Waltham M, Banerjee D, Bertino JR. Intrinsic and acquired resistance to methotrexate in acute leukemia. N Engl J Med 1996; 335: 1041-8. 190. Van Deventer SJ. Tumour necrosis factor and Crohn's disease. Gut 1997; 40: 443-8. 191. Braegger CP, Nicholls S, Murch SH, Stephens S, MacDonald TT. Tumour necrosis factor alpha in stool as a marker of intestinal inflammation. Lancet 1992; 339: 89-91. 192. Nicholls S, Stephens S, Braegger CP, Walker-Smith JA, MacDonald TT. Cytokines in stools of children with inflammatory bowel disease or infective diarrhoea. J Clin Pathol 1993; 46: 757-60. 193. Cappeflo M, Keshav S, Prince C, Jewell DP, Gordon S. Detection of mRNAs for macrophage products in inflam-

Clinical pharmacology in inflammatory bowel disease

520

194.

195. 196. 197. 198.

199.

200.

201. 202. 203.

204.

205.

206.

207.

208.

209.

210.

matory bowel disease by in situ hybridisation. Gut 1992; 33: 1214-19. Targan SR, Hanauer SB, van Deventer SJH et al. A shortterm study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn's disease. N Engl J Med 1997; 337: 1029-35. Present DH, Rutgeerts P, Targan S et al. Infliximab for the treatment of fistulas in patients with Crohn's disease. N Engl J Med 1999; 340: 1398^05. Stack WA, Mann SD, Roy AJ et al. Randomised controlled trial of CDP571 antibody to tumour necrosis factor-alpha in Crohn's disease. Lancet 1997; 349: 521-4. Knight DM, Trinh H, Le J et al. Construction and initial characterization of a mouse-human chimeric anti-TNF antibody. Mol Immunol 1993; 30: 1443-53. Stephens S, Emtage S, Vetterlein O et al. Comprehensive pharmacokinetics of a humanized antibody and analysis of residual anti-idiotypic responses. Immunology 1995; 85: 668 74. Mohler KM, Torrance DS, Smith CA et al. Soluble tumor necrosis factor (TNF) receptors are effective therapeutic agents in lethal cndotoxemia and function simultaneously as both TNF carriers and TNF antagonists. J Immunol 1993; 151: 1548 61. Wagner C, Mace K, Dc Woody K et al. Infliximab treatment benefits correlate with pharmacodynamic parameters in Crohn's disease patients. Digestion 1998; 59 (Suppl. 3): 124 5. Anon. Rcmicadc (infliximab). Prescribing information. Physicians Desk Reference, 1999. Anon. Enbrel (ctanercept). Prescribing information. Physicians Desk Reference, 1999. Scallon BJ, Moore MA, Trinh H, Knight DM, Ghraycb J. Chimeric anti-TNF-alpha monoclonal antibody cA2 binds recombinant transmembrane TNF-alpha and activates immune effector functions. Cytokine 1995; 7: 251 9. Siegel SA, Shcaly DJ, Nakada MT et al. The mouse/human chimeric monoclonal antibody cA2 neutralizes TNF in vitro and protects transgenic mice from cachexia and T N F lethality in vivo. Cytokine 1995; 7: 15 25. Suitters AJ, Foulkes R, Opal SM et al. Differential effect of isotype on efficacy of anti-tumor necrosis factor alpha chimeric antibodies in experimental septic shock. J Exp Med 1994; 179: 849 56. Scallon BJ, Trinh H, Nedelman M, Brennan FM, Feldmann M, Ghrayeb J. Functional comparisons of different tumour necrosis factor receptor/IgG fusion proteins. Cytokine 1995;7:759-70. Hommes DW, van Dullemen HM, Levi M et al. Beneficial effect of treatment with a monoclonal anti-tumor necrosis factor-alpha antibody on markers of coagulation and fibrinolysis in patients with active Crohn's disease. Haemostasis 1998;27:269-77. Lorenz HM, Antoni C, Valerius T et al. In vivo blockade of TNF-alpha by intravenous infusion of a chimeric monoclonal TNF-alpha antibody in patients with rheumatoid arthritis. Short term cellular and molecular effects. J Immunol 1996; 156: 1646-53. Paleolog EM, Hunt M, Elliott MJ, Feldmann M, Maini RN, Woody JN. Deactivation of vascular endothelium by monoclonal anti-tumor necrosis factor alpha antibody in rheumatoid arthritis. Arthritis Rheum 1996; 39: 1082-91. Brennan FM, Browne KA, Green PA, Jaspar JM, Maini RN, Feldmann M. Reduction of serum matrix metalloproteinase 1 and matrix metalloproteinase 3 in rheumatoid arthritis patients following anti-tumour necrosis factoralpha (cA2) therapy. Br J Rheumatol 1997; 36: 643-50.

211. Redl H, Schlag G, Paul E et al. Endogenous modulators of TNF and IL-1 response are under partial control of TNF in baboon bacteremia. Am J Physiol 1996; 271: Rl 193-8. 212. Maini RN, Breedveld FC, Kalden JR et al. Therapeutic eflficacy of multiple intravenous infusions of anti-tumor necrosis factor alpha monoclonal antibody combined with low-dose weekly methotrexate in rheumatoid arthritis. Arthritis Rheum 1998; 41: 1552-63. 213. Flanagan WM, Corthesy B, Bram RJ, Crabtree GR. Nuclear association of a T-cell transcription factor blocked by FK-506 and cyclosporin A. Nature 1991; 352: 803-7. 214. Brynskov J, Freund L, Campanini MC, Kampmann JP. Cyclosporin pharmacokinetics after intravenous and oral administration in patients with Crohn's disease. Scand J Gastroenterol 1992; 27: 961-7. 215. Fluckiger SS, Schmidt C, Meyer A, Kallay Z, Johnston A, Kutz K. Pharmacokinetics of orally administered cyclosporine in patients with Crohn's disease. J Clin Pharmacol 1995; 35:681-7. 216. Mueller EA, Kovarik JM, van Bree JB,Tetzloff W, Grevel J, Kutz K. Improved dose linearity of cyclosporine pharmacokinetics from a microemulsion formulation. Pharmaceut Res 1994; 11: 301 4. 217. Sandborn WJ. A critical review of cyclosporine therapy in inflammatory bowel disease. Inflam Bowel Dis 1995; 1: 48 63. 218. Abrcu-Martin MT, Vasiliauskas EA, Gaiennic J, Voigt B, Targan SR. Continuous infusion cyclosporine in effective for severe acute Crohn's disease ... but for how long? Gastroenterology 1996; 110: A851. 219. Scalzini A, Barni C, Stcllini R, Sucri L. Fatal invasive aspergillosis during cyclosporine and steroids treatment for Crohn's disease. Dig Dis Sci 1995; 40: 528. 220. Fcagan BG, McDonald JW, Rochon J et al. Low-dose cyclosporine for the treatment of Crohn's disease. Canadian Crohn's Relapse Prevention Trial Investigators. N Engl J Med 1994; 330: 1846 51. 221. Jewell DP, Lcnnard-Jones JE. Oral cyclosporine for chronic active Crohn's disease: a multicentre controlled trial. Eur J Gastroenterol Hepatol 1994; 6: 499 505. 222. Stange EF, Modigliani R, Pena AS, Wood AJ, Feutren G, Smith PR. European trial of cyclosporine in chronic active Crohn's disease: a 12-month study. European Study Group. Gastroenterology 1995; 109: 774 82. 223. Lichtiger S, Present DH, Kornbluth A et al. Cyclosporine in severe ulcerative colitis refractory to steroid therapy. N Engl J Med 1994; 330: 1841-5. 224. Egan LJ, Sandborn WJ, Tremaine WJ. Clinical outcome following treatment of refractory inflammatory and fistulizing Crohn's disease with intravenous cyclosporine. Am J Gastroenterol 1998; 93: 442-8. 225. Sandborn WJ. Preliminary report on the use of oral tacrolimus (FK506) in the treatment of complicated proximal small bowel and fistulizing Crohn's disease. Am J Gastroenterol 1997; 92: 876-9. 226. Lowry PW, Weaver AL, Tremaine WJ, Sandborn WJ, Combination therapy with oral tacrolimus (FK506) and azathioprine or 6-mercaptopurine for treatment-refractory Crohn's disease perianal fistulae. Inflam Bowel Dis 1999; 5: 239-45. 227. Yee GC, Stanley DL, Pessa LJ et al. Effect of grapefruit juice on blood cyclosporin concentration. Lancet 1995; 345: 9556. 228. de Groen PC, Aksamit AJ, Rakela J, Forbes GS, Krom RA. Central nervous system toxicity after liver transplantation. The role of cyclosporine and cholesterol. N Engl J Med 1987;317:861-6. 229. Sandborn WJ, Tremaine WJ, Offord KP et al. Transdermal nicotine for mildly to moderately active ulcerative colitis. A

Laurence J. Egan and William J. Sandborn

230. 231.

232.

233.

randomized, double-blind, placebo-controlled trial. Ann Intern Med 1997; 126: 364^71. Sher ME, Bank S, Greenberg R et al. The influence of cigarette smoking on cytokine levels in patients with inflammatory bowel disease. Inflam Bowel Dis 1999; 5: 73-8. Louvet B, Buisine MP, Desreumaux P et al. Transdermal nicotine decreases mucosal IL-8 expression but has no effect on mucin gene expression in ulcerative coHtis. Inflam Bowel Dis 1999; 5: 174-81. Bannon YB, Corish J, Corrigan OI, Devane JG, Kavanagh M, Mulligan S. Transdermal delivery of nicotine in normal human volunteers: a single dose and multiple dose study. Eur J Clin Pharmacol 1989; 37: 285-90. Mulligan SC, Masterson JG, Devane JG, Kelly JG. Clinical and pharmacokinetic properties of a transdermal nicotine patch. Clin Pharmacol Ther 1990; 47: 331-7.

521 234. Zins BJ, Sandborn WJ, Mays DC et al. Pharmacokinetics of nicotine tartrate after single-dose liquid enema, oral, and intravenous administration. J Clin Pharmacol 1997; 37: 426-36. 235. Compton RF, Sandborn WJ, Lawson GM et al. A doseranging pharmacokinetic study of nicotine tartrate following single-dose delayed-release oral and intravenous administration. AHment Pharmacol Ther 1997; 11: 865-74. 236. Pullan RD, Rhodes J, Ganesh S et al. Transdermal nicotine for active ulcerative colitis. N Engl J Med 1994; 330: 81115. 237. Thomas GA, Rhodes J, Mani N et al. Transdermal nicotine as maintenance therapy for ulcerative colitis. N Engl J Med 1995; 332: 988-92. 238. Pianezza ML, Sellers EM, Tyndale RF. Nicotine metabolism defect reduces smoking. Nature 1998; 393: 750.

26 Multi-site tlierapeutic modaiities for infiammatory bowel diseases - meclianisms of action GERHARD ROGLER

Introduction Genetic susceptibilities such as the recently identified polymorphisms in the N0D2 gene, as well as environmental factors such as certain bacteria, play a role in the etiology of inflammatory bowel diseases (IBD). Over the years many factors have been identified that contribute to the pathogenesis and the process of inflammation in Crohn's disease (CD) and ulcerative colitis (UC), leading to new therapeutic concepts. Among these contributing factors are cytokines and chemokines, as well as adhesion molecules, which are relevant for the emigration of immune cells into the intestinal mucosa. Mucosal and systemic concentrations of many proinflammatory cytokines are elevated in IBD. An inadequate, and/or prolonged activation of the intestinal immune system plays an important role in the pathophysiology of chronic mucosal inflammation. An 'imbalance' between proinflammatory and anti-inflammatory cytokines has been described in the inflamed mucosa of patients with CD and UC. Many, if not all, of the involved cytokine-mediated pathways have back-up systems. Therefore a therapeutic intervention at a particular, singular, very specific point in the complex network of cytokine and chemokine interactions with each other or their receptors is frequently less likely to be successful than a multi-site targeted anti-inflammatory strategy [1]. Even when the etiology of IBD is completely elucidated a causative therapy might not be possible, and the multi-site anti-inflammatory strategy could still be favorable. Therefore the improvement of 'classic' multi-site targeted anti-inflammatory therapies, as well as the development of new concepts for this approach, is of great importance for the future management of patients with IBD. The improvement of 'classic' concepts and therapies is necessary, as so far no therapeutic strategy has proved successful in

all patients. Clinically we observe a 'resistance' of a certain percentage of patients to any particular therapy. An important goal for the future must be a better understanding of mechanisms leading to a relative resistance to classic multi-site anti-inflammatory strategies. This would allow the early identification of patients who are not likely to respond to a treatment modality, would avoid frustrating treatments for the patient and the physician, and could allow more specific concepts for the individual patient. To understand the mechanisms of resistance to a particular multi-site targeted anti-inflammatory therapy we first have to understand the molecular and cellular mechanisms involved in an effective therapy. Important insights into a number of those mechanisms of drug action have been made in the last two decades. A classic example for a multi-site targeting antiinflammatory treatment is therapy with glucocorticoids. Therefore the principles of glucocorticoid therapy, the structure of the effect-mediating receptor, the molecular mechanisms of action and the effects on the cellular levels will be highlighted first in this chapter. The glucocorticoid receptor (GR) is a member of a family of receptors, the so-called nuclear receptor superfamily, which shares structural and functional similarities. Other members of this family are the peroxisome proliferation-activated receptors (PPAR). Ligands to PPARy have recently been shown to be effective in animal models of IBD whereas potential ligands of PPAR-oc were beneficial in a clinical study in active IBD. The inhibition of the proinflammatory transcription factor nuclear factor kappa B ( N F - K B ) is likely to be the most important target of glucocorticoid therapy as well as PPAR-mediated effects. Therefore, finally the mechanisms and pathways of N F - K B activation, as well as the consequences of N F - K B inhibition, will be explained.

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.J, Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 523-551. © 2003 Kluwer Academic Publishers. Printed in Great Britain

524

Multi-site therapeutic modalities for inflammatory bowel diseases

Glucocorticoids Glucocorticoids are used for the suppression or reduction of inflammation in a wide variety of diseases such as rheumatoid diseases, allergic diseases, IBD and in general autoimmune diseases [2-8]. In many of these cases they are still the standard or first-line therapy due to their high efficiency; however, their use is limited by systemic side-effects. The understanding of the mechanisms by which glucocorticoids suppress or reduce inflammation has increased dramatically during recent years [9-14]. Glucocorticoids have been proven to be the first choice in the treatment of acute flares of IBD in several major studies [15-23]. The systemic administration of glucocorticoids (orally or intravenously) during the acute exacerbation of CD or UC is followed by a multitude of different effects in different body cells. One of the intended eff'ects is the down-regulation of proinflammatory cytokines [24], This mechanism is part of the feedback system between inflammation-derived cytokines and the central nervous system-adrenal axis regulating corticosteroid synthesis with the physiological relevance to balance host defense and anti-inflammatory systems of the body [25, 26]. Among the molecules down-regulated by GR action are multiple cytokines and their receptors, chemokines and their receptors, kinins and their receptors, adhesion molecules and inflammation-associated enzymes such as inducible nitric oxide synthase (iNOS) and the inducible cyclooxygenase (COX-2) [3, 27].

Principal meclianisms of giucocorticoid action Most, if not all, effects on cells of naturally occurring glucocorticoids such as Cortisol or synthetic corticosteroids such as prednisolone and its methylated or acetylated derivates (triamcinolone, dexamethasone or beclomethasone) are mediated by binding to cytosolic GR. GR are present in almost all body cells in concentrations between 2000 and 30 000 binding sites/cell [28]. The GR is a member of the nuclear receptor superfamily, which also includes PPARoc, PPARy and PPAR5 (see below). GR consists of 777 amino acids (AA) and was cloned in 1985 [29-31]. There is only a single GR binding glucocorticoids, with no evidence for subtypes of differing affinity in different tissues [3]. A splice variant of GRoc, termed GRP, has been identified that is not able to bind glucocorticoids (see below).

Normally the interaction of glucocorticoids with GRa is followed by an activation and dissociation of GR from its inhibitory protein complex and by a translocation of the receptor into the nucleus, where the complex of steroid and receptor interacts with promoter regions of different genes, finally leading to an increase or decrease of gene transcription (Fig. 1). This process is very similar within the family of nuclear receptors, which show a high degree of genetic similarity with 40-90% of identical AA sequences [32-34]. An important problem for the therapy of patients with IBD and in general chronic inflammatory diseases is the occurrence of glucocorticoid responders and non-responders. Non-response to glucocorticoids may be mediated by mutations of the receptor, a reduced number of GR or a down-regulation of the receptor [35]. Studies demonstrated that primary (hereditary) abnormalities in the GR gene make only 2.3% of patients with asthma relatively 'resistant'. 'Resistance' to the beneficial clinical effects of glucocorticoid therapy in patients with IBD, therefore, is probably rarely related to generalized primary (hereditary) glucocorticoid resistance. In the majority of patients with rheumatoid arthritis or asthma the glucocorticoid resistance seems to be acquired and localized to the sites of inflammation, where it reflects high local cytokine production, which interferes with glucocorticoid action [36]. One of the basic mechanisms that could be responsible for glucocorticoid resistance, the competition for co-factors, will be explained in detail later. Glucocorticoid levels (^max) and binding affinities (A^d) vary among patients and have been correlated to patient response. A certain threshold level of GR is necessary for glucocorticoid responsiveness [37]. In patients with rheumatoid arthritis a decrease of systemic GR in leukocytes has been found [38]. However, the GR density did not correlate with inflammatory disease activity.

Ttie molecular structure of glucocorticoid receptor The mechanisms mediating the glucocorticoid response via GR are complex. Regulation of gene expression and GR actions occurs at several levels in the body and the single cell. However, the glucocorticoid-mediated anti-inflammatory mechanisms are a classic example for the transmission of basic research to the bedside. It is important to first understand the structure of GR and then the molecular

Gerhard Rogler

525

i l l l l i i i l l l cell membrane nuclear membrane DNA

^

translation

iiiiiiiilii

new protein, altered cell function

Figure 1. Cellular mechanisms of glucocorticoid receptor action. After binding of glucocorticoid(s) to its receptor hsp90 is released from the complex. This is followed by an exposure of the nuclear localization signal, homodimerlzation and a rapid translocation of the activated GR/glucocorticoid complex to the nucleus where it can bind its response elements (glucocorticoid response elements, GRE) and mediate fra/7s-activation or frans-repression of gene transcription.

mechanisms of glucocorticoid action. GRa normally is a phosphorylated 92 kDa protein, which in its inactivated form is bound into a complex with other proteins. Like most members of the nuclear receptor superfamily GR contains a DNA-binding site (DBD) consisting of two zinc-finger domains [10]. This area contains an invariant pattern of eight cysteins arranged in two groups of four, so as to coordinate the binding of two zinc atoms [39-41]. The DBD is located in the middle of the GR molecule (Fig. 2). The two zinc-finger motifs in the DBD induce the formation of a tertiary structure containing helices that interact with specific DNA sequences organized in 'GR response elements' [42-47]. Between the DBD and the carboxyl-terminal ligand-binding domain (LBD) is a so-called 'hinge region'. This region of the GR protein contains a nuclear localization signal, a binding region for heat-shock protein 90 (hsp90) and a second ^ra^^-activating domain (activation function 2, AF2) [3]. The carboxyl-terminal LBD not only represents the specific binding sites for glucocorticoids but also serves as a h o m o d i m e r l z a t i o n d o m a i n [10]. Additionally it interacts with other proteins regulating GR activity.

The N-terminal part of GR contains another trans-2iCii\2iiion domain (activation function 1, AFl) that plays an important role in gene regulation [48-50]. This structure may be especially important during trans-a.ctiv3.tion but not ^ra^^'-repression activity [51-54]. It is not yet clear how AFl interacts with other proteins to induce transcription. AF-1 has been reported to interact with the basal transcriptional machinery [55, 56] but also with other adaptors or co-activators [57-60]. The inactivated GR is bound to a protein complex of approximately 300 kDa. It includes two molecules of hsp90, a 59 kDa immunophilin protein and other inhibitory proteins [3] (Fig. 2). The hsp90 proteins cover the nuclear localization site, preventing the unliganded GR from localizing to the nuclear compartment. After binding of the specific ligand to the LBD the tertiary structure of the molecule changes, followed by a release of hsp90 from the complex. This leads to an exposure of the nuclear localization signals and a rapid translocation of the activated GR-glucocorticoid complex to the nucleus where it can bind its response elements (glucocorticoid response elements, GRE).

Multi-site therapeutic modalities for inflammatory bowel diseases

526

COOH gene activating domain

inhibitory protein

hormone binding site

hinge region DNA-binding domian steroid hormone

I

.i

/ "

DNA'binding site expos^

COOH

1 Steroid hormone Figure 2. Molecular structure of the glucocorticoid receptor. The N-terminal part of GR contains a fra/7s-activation domain that plays an Important role in gene regulation. The DNA binding site (DBD) consists of two zinc-finger domains and contains an invariant pattern of eight cysteins arranged in two groups of four. Between the DBD and the carboxyl-termlnal ligand binding domain (LBD) is a so-called 'hinge region'. This region of the GR protein contains a nuclear localization signal, a binding region for heat-shock protein 90 (hsp90) and a second frans-activating domain. The C-terminal LBD not only represents the specific binding sites for glucocorticoids but also serves as a homodimerlzatlon domain. Additionally it Interacts with other proteins regulating GR activity. The inactivated GR is bound to a protein complex of approximately 300 kDa Including two molecules of hspQO. The hspQO proteins cover the nuclear localization site preventing the unliganded GR from localizing to the nuclear compartment. After binding of the specific ligand to the LBD the tertiary structure of the molecule changes, followed by a release of hspQO from the complex.

Molecular mechanisms of glucocorticoid receptor action Activated and nuclear-translocated GR can act in two principal ways. It can mediate /ra^i^-activation of gene transcription leading to an increased mRNA expression or /ra«^-repression followed by a downregulation of shut-off of a certain gene product.

Transactivation In the nucleus GR can bind to classical GRE to activate transcription of the response gene [61-63] (Fig. 3A). The GRE has a palindromic motif (consensus sequence: GGT ACA N N N TGT TCT). Usually GR binds this or similar DNA sequences cooperatively as a homodimer (Fig. 3A). For homodimerization interaction of a group of five AA known as the dimerization or D loop is needed. They are located within the DBD of the GR molecule and are essential for dimerization and transcriptional activation.

A direct /mA?.v-activation of gene transcription by GR has been described for IL-I type II receptor that binds IL-1 without induction of signal transduction, thus preventing cells from activation and inflammatory reaction [64-67] (Fig. 3A). Another important protein whose transcription is /ra^^-activated by GR is iKBa, the inhibitory protein for the proinflammatory transcription factor N F - K B (see below). For other genes such as serine protease inhibitor 3 or arginase a cooperative /ra«^-activation involving GR and another transcription factor such as C/EBP or AP-1 has been described (Fig. 3B). This means that binding of activated and translocated GR is necessary but not sufficient for the increased transcription of these genes. However, as soon as both, GR and the cooperating transcription factor (C/EBP or AP-1) are bound to the promoter, increased genetranscription takes place.

Gerhard Rogler

527

Figure 3. Frans-activation of gene transcription by glucocorticoid receptor. A: GR binds as a homodimer to classical GRE to activate transcription of response genes. The GRE has a palindromic motif (consensus sequence: GGT ACA NNN TGT TCT). A direct transactivation of gene transcription by GR has been described for IL-1 type II, I K B , IL-1 RA or lipocortin I. B: For other genes such as serine protease inhibitor 3 or arginase a cooperative fra/]s-activation involving GR and other transcription factors like C/EBP has been described.

GR-mediated rra/t^-activation of genes is also involved in the induction of apoptosis in T cells by dexamethasone [68-78]. The detailed mechanisms mediating the induction of apoptosis are not yet elucidated. In addition to these trans-?iCi\\diimg actions described, which usually take several hours, GR have more rapid ^ra^i^-repressing activities. 7A5/7S-repression via DNA binding After characterization of positive transcriptional regulating GRE the existence of negative GRE sites (nGRE) mediating a negative regulation of transcription (/raw^-repression) via glucocorticoids was postulated [63]. However, the concept of a nGRE site is still a matter of discussion, as the consensus binding site is variable and described for only a few genes [13]. A binding to a promoter sequence and subsequent ^ra^^'-repression by GR has been shown for the prO'Opiomelanocortin (POMC) gene, an ACTH precursor allowing a negative feedback circle [79, 80] (Fig. 4A). Another promoter containing a negative GR binding site (at -278 to -249 of the

promoter) is the corticotropin-releasing hormone (CRH) promoter [81, 82]. Thus it seems that direct ^ra«^-repression activity is reserved for the negative feedback circle in the hypothalamic-pituitary-adrenal axis. GR and /raw^-activating transcription factors may also compete for binding sites in promoters [83]. In this case the presence of GR blocks ^ra^^-'activation by another factor and does not actively rra^^-repress transcription itself (Fig. 4B). In the osteocalcin promoter GR overlaps the TATA box. Activation and nuclear translocation of GR can therefore prevent the binding of the basal transcription factor, TATAbinding protein (TBP), which is necessary for the recruitment of RNA polymerase II and initiation of transcription [84]. This may be one of the reasons for the occurrence of osteoporosis during long-term glucocorticoid therapy. 7?'5/?s-repression without DNA binding Besides the competition for binding sites in promoters a rra^^-repression mechanism for GR has been described that does not involve DNA binding (Fig.

528

Multi-site therapeutic modalities for inflammatory bowel diseases

B

Figure 4. frans-represslon of gene transcription by glucocorticoid receptor. A: A binding to a negative glucocorticoid response element (nGRE) and subsequent fra/7S-repression by GR has been shown for the pro-opiomelanocortin (POMC) gene, an ACTH precursor and CRH both allowing a negative feedback circle. B: GR and frans-activating transcription factors can compete for binding sites in promoters. In the osteocalcin promoter GR overlaps the TATA box. Activation and nuclear translocation of GR can therefore prevent the binding of the basal transcription factor, TATA binding protein (TBP), which is necessary for the recruitment of RNA polymerase II and initiation of transcription. C: For the p65 subunit of NF-KB and GR a direct physical interaction has been demonstrated, leading to a frans-repression of transcription of the NF-KB dependent gene without DNA binding of GR. This interaction effect does not impair the DNA binding ability of NF-KB.

4C). The promoters of most proinflammatory genes contain binding sites for the proinflammatory transcription factor N F - K B or AP-1. In the genes regulated by both factors nGRE are not found. However, glucocorticoids are able to rapidly /ra«^-repress AP1 or NF-KB-induced transcription of these genes. This might in fact be the most important mechanisms triggering their therapeutic efficacy in the treatment of inflammatory diseases. It is not completely clear whether direct proteinprotein interaction between GR and AP-1 or N F - K B are sufficient for /ra/i^-repression [85-87] or whether other mechanisms are involved. However, GR association to AP-1 does not reduce AP-1-DNA-binding, as demonstrated by electrophoretic mobility shift assays (EMSA) [86, 88]. Similar observations were made for the interaction of N F - K B and GR. For the p65 subunit of N F - K B

and GR a direct or semi-direct physical interaction has been demonstrated [14, 89-94]. This interaction

occurs without impairing the DNA-binding ability of both transcription factors [90-92, 95-97]. These data indicate that co-activating or co-repressing proteins may play a major role for that function. Proinflammatory transcription factors such as AP-1 and N F - K B on one hand, and GR on the other

hand, could compete for a coactivator molecule called CREB-binding protein (CBP) (for more details see ref. 93). CBP plays an essential role in the activation of transcription by numerous transcription factors/transcriptional activators [12 ,93, 98103]. CBP binds and is necessary for coactivation of CREB, AP-1, signal transducer and activator of transcription (STAT) proteins and N F - K B , as well as nuclear receptors such as as GR, PPAR, progesterone receptor and retinoid receptors [12, 103]. It is likely that CBP and the related p300 are not acting as single coactivator proteins but are present in the nucleus in a complex of proteins [12, 103]. CPB and p300 acetylate free histones in vitro, which is neces-

Gerhard Rogler sary for successful transcription [12, 103]. Other coactivators are pl60/SRC, the TRAP/DRIP/ARC complex and the P/CAF complex (for a review see ref. 12).

529 from the elucidation of these basic mechanisms for the steroid treatment of patients with IBD can be expected.

Factors determiningfra^s-activationor fr^^s-repression

Regulation of glucocorticoid receptor expression

The factors determining whether glucocorticoid binding to its receptor induces more rra^^-activation or more ^raw^-repression activity on genes are not completely understood. Clearly specific DNA-binding sites play a role. However, the question still arises as to why some ligands are able to induce mainly /ra^^-repression actions while others may mainly induce ^ra/t^'-activation. In experiments modeling the GR ligand-binding domain it could be shown that tyrosine 735 may interact with the D ring of dexamethasone and that the substitution of D ring functional groups results in partial agonist steroids with reduced ability to direct ^ra«^-activation [104]. A substitution of Tyr735 by phenylalanine (Tyr735Phe) did not reduce ligandbinding affinity and did not alter ^ra/i^'-repression of N F - K B , but reduced ^raw^'-activation. These data suggest that tyrosine 735 is important for ligand interpretation and ^ra«^-activation [104]. In addition, recent data suggest that GR does not necessarily have to form homodimers to modulate gene expression. Negative recognition elements (nGRE) have been identified in keratin gene promoters that bind GR as four monomers [105]. Thereby a specific set of co-repressors is bound, including histone acetyltransferase and CBP but not SRC- 1 and GRIP-1. In the regulation of GR action so-called glucocorticoid-modulating elements (GME) in promoters and proteins binding to them (GMEB) have gained attention in recent years. A GME in the gene tyrosine aminotransferase (TAT) was identified in 1992 [106]. The GME was identified as a 21-basepair sequence of the rat TAT gene; it is located at -3.6 kb and 1 kb upstream of the GRE [107]. It modulates both the dose-response curve of agonists bound to the GR and the residual agonist activity of GR-bound antisteroids. The expression of GME activity involves the binding of two novel proteins (GMEB-1 and GMEB-2) [107-112]. Taken together it seems that four important factors determine the dose-response curve and the preference of ^ran^-activation or rm«5'-repression of glucocorticoids: the ligand structure, the GR concentration, the co-activator or co-repressor concentration/availability and the GME. Important effects

To further complicate the system of ^ra«5'-activation, ^mn^'-repression with DNA-binding and transrepression without DNA-binding involving co-activators and co-repressor the expression of the GR gene itself is regulated by activated GR. The GR (GRoc) represses its own synthesis in a hormonedependent manner [113]. The reduction in cellular receptor levels is followed by insensitivity of the cells to glucocorticoids. It is dose- and time-dependent and reversible upon hormone withdrawal. For this down-regulation of GR expression by GR itself, DNA-binding seems to be crucial [114-118]. However, this is not the end of the story. The phosphorylation status of GR may be important for this feedback function. Deletion of phosphorylation sites from the GR protein by site-directed mutagenesis resulted in a GR that cannot repress its own transcription upon binding of glucocorticoids [119, 120].

glucocorticoid receptor p In contrast to G R a the shorter GRP isoform that is generated by alternative splicing lacks a LBD. GRp is identical to GRa through the first 727 AA but differs in the carboxyl-terminus, where it lacks the last 50 AA found in GRa but contains additional 15 non-homologous AA [14]. As GRa, GRp is widely expressed in adult human tissues but is primarily localized to the cell nucleus independent of the presence of ligand [14]. GRp still binds to DNA and may therefore potentially interfere with the action of G R a [121-123]. It has been speculated that GRp might be an antagonist of GRa action as it blocks DNA-binding sites without suppressing gene transcription [30, 121-127]. However, this inhibition seems to require increased GRp expression relative to GRa. It has been studied mainly in transfection assays with overexpression the P-isoform. A clear functional role has not yet been confirmed [13]. Glucocorticoid resistance in asthma patients has been described to be associated with elevated in-vivo expression of the GR P-isoform [128-130]. However, conflicting data have also been obtained showing no overexpression of GRp in resistant patients [131133].

530

Multi-site therapeutic modalities for inflammatory bowel diseases

Regulation of glucocorticoid receptor in IBD The regulation of GR action and the mechanisms involved in its function are complex. Investigations on possible mechanisms of steroid-refractory IBD are still at early stages. The few studies raise more questions than they answer. Hsp90 has some features of an inhibitory protein of GR and is bound to its inactive form (see above). The expression of human hsp90 in patients with CD and UC was studied; however, no differences between patients and controls were found, making a role of hsp90 in glucocorticoid-refractory IBD unlikely [134]. The number and dissociation constant (A^j) of GR in peripheral blood mononuclear cells of six nonresponders of glucocorticoid treatment with UC, five responders and 10 healthy controls was determined in another study [135]. A significant increase in the number of binding sites and the dissociation constant in non-responders compared to responders was found. Surprisingly the number of binding sites was highest in non-responders. When GR levels were determined via dexamethasone binding only in the cytosolic fraction to ensure that only free receptor and not already steroidassociated, translocated receptor molecules could bind [ H]dexamethasone, the situation seemed to be different. The dexamethasone binding in cytosol isolated from peripheral blood mononuclear cells (PBMNC) of corticosteroid-treated IBD patients was significantly lower compared to controls and IBD patients not treated with glucocorticoid [136]. Systemic GR levels in untreated IBD patients did not differ significantly from controls [136]. There was no difference in the binding affinity of patients and control, with an obvious lower binding maximum indicating a reduced receptor number in the steroidtreated patients group, which is in contrast to the results reported by Shimada et al. In contrast to the findings in PBMNC mucosal GR levels of IBD patients were significantly decreased in both steroidtreated and untreated patients compared to controls [136]. The reduced binding of [^H]dexamethasone in cytosol from steroid-treated IBD patients is most likely due to a feedback regulation of GR in the cells by the ligand. This assumption is supported by a number of studies [137-139]. The data indicate that, in contrast to patients with rheumatoid arthritis or other connective tissue diseases in IBD, there is no difference in systemic GR levels between patients not treated with glucocorticoids and controls. This could mean that a localized inflammation in the intestinal

mucosa is not followed by a systemic depression of GR levels in leukocytes. Honda and co-workers studied the expression of GRa and GRp in PBMNC of patients with UC and controls [140]. They found expression of GR(3 in only 9.1% of patients with steroid-sensitive disease whereas it was present in 83.3% of steroid-resistant patients as detected by polymerase chain reaction (PCR). The authors conclude that the determination of GRp expression could provide a tool to predict steroid responsiveness of UC patients [140]. However, other laboratories have reported that GRP transcripts could be amplified from all control patients investigated, making a role of GRp for glucocorticoid refractivity unlikely (personal communication). Despite these studies most of the questions regarding the mechanisms of steroid-refractory disease are still unanswered [141]. It is not clear why some patients express GRP and others do not. It is not clear whether changes in glucocorticoid binding are just an epiphenomenon or a cause of different disease courses. It is not clear whether GRp expression or decreased GRa levels are really crucial for treatment success. Future studies on GR expression in IBD need to answer several important questions: are GR levels at the onset of the disease predictive for the success of glucocorticoid therapy? Are low levels correlated with the development of a steroid-refractory disease? The question of whether measurement of GR in the mucosa can be predictive for therapy success needs to be answered for the future management of patients with IBD. Patients with low GR levels could then be primarily treated with other drugs such as azathioprine. Prospective studies investigating GRoc and GRP transcripts, as well as glucocorticoid binding in PMNC and in the mucosa at the onset of disease before any treatment, will need to be performed to clarify whether both mechanisms are related to steroid-refractory IBD.

Cellular mechanisms of glucocorticoid action in IBD Glucocorticoid receptors that can be activated by ligand binding mediate glucocorticoid action and induce ?ra«5'-activation or /ra/t^-repression of genes as well as an inhibition of proinflammatory transcription factors such as N F - K B and AP-1. The question arises as to which effects the ligand-induced

Gerhard Rogler

531

Glycocortlcoids/Glucocorticold-Receptor

translation - — ! •

new protein. vrelevant for apoptosis

r

antagonism to NF-KB

mediated transactivation of pro-inflammatory genes^

translation

M

Induction of apoptosis

Inhibition of N F - K B mediated gene induction

Figure 5. Glucocorticold-induced anti-inflammatory mechanisms. Two major principles mediate the anti-inflammatory effect of glucocorticoids. After penetrating the cell membrane and binding to the GR nuclear translocation occurs. The complex can then fraA7s-activate genes that are involved in the induction of apoptosis. The induction of apoptosis in activated lymphocytes, eosinophils, basophils and other cell types reduces the overall amount of circulating cytokines and inflammatory mediators and allows reconstitution of the intestinal mucosa. On the other hand the heterodimer of GR interacts with translocated NF-KB and prevents transcriptional activation by this proinflammatory transcription factor As NF-KB activation has anti-apoptotic effects in a number of cell types antagonism to its action also has pro-apoptotic effects, indicating the interconnection of both mechanisms of glucocorticoid action.

activation of GR have at the cellular level, or more precisely which eflfects can be observed in different cell types? A simplified view could focus on two major principles (Fig. 5): 1. The cellular effects induced by a down-regulation of proinflammatory transcription factors. As mentioned this is facilitated by an incompletely understood antagonism to transcription factors such as N F - K B and AP-1 (Fig. 5). 2. The induction of apoptosis of activated immune cells, which limits the immune response and as a consequence is also followed by reduced levels of circulating proinflammatory factors (Fig. 5).

In this chapter only a few of the effects on some of the relevant cell populations in the intestinal immune system can be highlighted.

Lymphocytes Glucocorticoids affect the growth, differentiation and function of lymphocytes, the distribution of cellular subsets, and the production of cytokines [142]. In the chronic inflamed mucosa the number of lymphocytes is greatly increased. Most of these cells are activated T-helper cells. Glucocorticoids reduce lymphocyte proHferation and induce apoptosis in these cells [143-146]. Studies indicate that mainly transcriptional ^ra«^-activation functions are required for this glucocorticoidmediated apoptosis [147]. Interestingly intestinal intraepithelial lymphocytes (lEL) may be resistant

532

Multi-site therapeutic modalities for inflammatory bowel diseases

to steroid-induced apoptosis, which could be due to the expression of high levels of the anti-apoptotic protein Bcl-2 and Bcl-x [148]. Recent data indicate that glucocorticoid-mediated inhibition of cytokine secretion in inflammatory diseases may be mediated not only by direct action on lymphocyte and monocyte/macrophage N F - K B but in addition indirectly through promotion of aThelper cell type 2 (Th2) induction [143] with increased levels of Th2 cytokine (IL-4) and reduced levels of Thl cytokine (IL-12) secretion [149]. In addition dexamethasone can inhibit IL-12-induced phosphorylation of STAT-4 without altering IL-4induced STAT-6 phosphorylation [149]. Both would result in a blockade of proinflammatory T-helper cell type 1 (Thl) cytokine expression. A further example of the multi-site mechanisms involved in glucocorticoid action is given by another monocyte/macrophage-lymphocyte interaction. Glucocorticoids have been shown to down-regulate T-cell co-stimulatory molecules B7-1 and B7-2 on macrophages which are essential for clonal T-cell expansion in reaction to antigen-presenting cells [150]. On the other hand co-stimulatory molecules can prevent cells from glucocorticoid-induced apoptosis [151 ]. It is interesting to note in this context that normal intestinal macrophages express no co-stimulatory molecules, whereas there is a clear up-regulationinIBD[152, 153]. Macrophages Macrophages are known to play an important role during inflammation in many diff'erent tissues [154]. Intestinal macrophages represent one of the largest compartments of the mononuclear phagocyte system in the body [155]. They are localized preferentially in the subepithelial region and constitute 10-20% of mononuclear cells in the intestinal lamina propria [156]. Macrophages are able to secrete proinflammatory cytokines which are known to be regulated by N F - K B such as IL-ip,TNF-oc, IL-6, IL-8, MCP-1. In normal mucosa only very few macrophages express activation-associated markers such as CD14, CD16, HLA-DR, GDI lb, and G D l l c [152, 153, 157], supporting a concept of anergy in the normal mucosa. Several findings indicate a phenotypic change of the intestinal macrophage population in IBD. Mahida etal. demonstrated the presence of GDI6 (Fcylll receptor) in IBD by immunohistochemical methods and in isolated cells [158]. GD54 (IGAMl) expression increased from 7% to 70% in UG and to 46% in GD [159]. Inflammation-

associated intestinal macrophages express LPS receptors, Fc receptors and co-stimulatory molecules [152, 153, 157], which enable them to stimulate T cells and partially prevent them from glucocorticoid-induced apoptosis (see above). The number of macrophages is clearly increased in both GD and UG mucosa [160-162]. The activated macrophage population secretes a multitude of inflammatory mediators such as prostaglandins; leukotrienes; cytokines such as TNF, IL-1, IL-12; chemokines such as MGP-1 or IL-8 as well as tissuedamaging reactive radicals and tissue-degrading enzymes [163-176]. The transcription, translation and secretion of almost all molecules mentioned above can be reduced or inhibited by the administration of glucocorticoids. In the promoters of most of those gene N F - K B binding sites can be found [177187], indicating that the glucocorticoid-mediated inhibition of N F - K B activation may be the most important mechanism for their anti-inflammatory potential. Intestinal epithelial cells Intestinal epithelial cells (lEG) play an active role in the intestinal immune system [188-190]. After stimulation with IFN-y they express MHG molecules [191-195]. In addition there is evidence that epithelial cells are able to respond to damage or bacterial invasion by secreting cytokines and chemokines [196-205]. Among the cytokines secreted by intestinal epithelial cells are inflammatory mediators such as IL-6 and IL-8. As already discussed for the other cell types, it could be shown that, for example, the induction of IL-6 production in human intestinal epithelial cells following stimulation with IL-1P was associated with activation of the transcription factor N F - K B [206]. Jobin and co-workers demonstrated that adenovirus-mediated transfection of HT-29 and Gaco-2 cells with a N F - K B superrepressor caused a reduced induction of inducible nitric oxide synthase (iNOS), IL-1 (3 and IL-8 by IL-1 or TNF, indicating involvement of the N F - K B system in regulation of these genes [207]. In the inflamed mucosa activation of N F - K B has been demonstrated in intestinal epithelial cells, whereas activation was absent in non-inflamed mucosa [208, 209]. Glucocorticoids down-regulate the expression of IL-6, IL-8, iNOS and class II molecules in intestinal epithelial cells and restore epithelial cell physiology [210, 211]. This is again likely to be mediated by the inhibition of N F - K B activation.

Gerhard Rogler

Other cell types As GR are expressed in most if not all cell types, effects of glucocorticoid therapy on all cell types involved in intestinal inflammation have been found. During the course of IBD increased numbers of eosinophils and mast cells are present in the intestinal mucosa [212-216]. These cells are activated and secrete different tissue-damaging proteins as well as cytokines and chemokines [217-224]. Again N F - K B has been found to be a major factor in activation of eosinophils [225], and glucocorticoids have been shown to down-regulate the production and secretion of the proinflammatory proteins [226, 227]. In addition the number of mucosal and circulating eosinophils is decreased by glucocorticoids. This is probably mediated by an induction of apoptosis in these cells [228-233]. As well as the number of eosinophils and mast cells, the number of basophils is increased in IBD mucosa [212]. Glucocorticoids reduce the number of basophils similar to the reduction described for eosinophils. Neutrophils are a major component in active lesions in UC and to a lesser extent in CD [234-240]. They are recruited under the influence of the neutrophil chemoattractant IL-8 [237]. An important feature of these neutrophils is their ability to induce the so-called 'oxidative burst' reaction involving the NADPH oxidase system leading to a secretion of oxygen radicals that not only kill surrounding bacteria but also damage surrounding tissue [238]. Glucocorticoids act in different ways in neutrophils. They down-regulate the oxidative burst reaction and reduce IL-8 secretion, leading to a reduced immigration of neutrophils into the mucosa. On the other hand an increased release from the bone marrow, followed by a leukocytosis, is induced [241]. Another important feature of the treatment with glucocorticoids is down-regulation of the expression of adhesion molecules on endothelial cells, mononuclear cells and epithelial cells [89, 241-247], which again may be mainly mediated by antagonistic effects to N F - K B .

Glucocorticoid ttierapy in combination witti other anti-inflammatory drugs in IBD Due to this multitude of desired effects the standard therapy for acute flares of IBD is still the systemic application of glucocorticoids. The effectiveness of a glucocorticoid regimen has been shown in numerous

533 multicenter trials [23, 248]. Initial remission rates in patients with acute flares of CD under a standard therapy vary from 60% to 80%, which is higher than under treatment with sulfasalazine or 5-aminosalicylic acid (40-50%). The combination of glucocorticoids and anti-inflammatory drugs, e.g. sulfasalazine, shows no additive effect and no higher remission rates than therapy with prednisolone alone [23]. This may indicate that glucocorticoids and drugs such as sulfasalazine may both act in a common signal transduction cascade. Recently evidence has been found that some of the antiinflammatory effects of 5-ASA and sulfasalazine in IBD patients may be mediated by the inhibition of the proinflammatory transcription factor N F - K B . Acetylsalicylic acid, 5-ASA and sodium salicylate inhibit activation of N F - K B by blocking iKB-kinases (IKK), which are key factors in N F - K B activation [249-251]. Similar results were found for sulfasalazine [252,253].The inhibition of bothlKKocandlKKp has been shown [254]. The lack of additional effects of a combination of glucocorticoid and salicylate therapy indicates that the glucocorticoid effect on N F - K B inhibition may be superior. However, long-term studies show that, despite the high initial response rates, only 44% of patients initially treated with glucocorticoids show long-term remission; 25-35% of patients become 'steroid dependent', indicating that steroid treatment cannot be completely tapered and omitted. About 20% of glucocorticoid-treated patients prove to be primarily 'glucocorticoid resistant' [255, 256].

Peroxisome proliferator-activated receptors (PPAR) The GR that mediates glucocorticoid effects in the treatment of IBD is a member of the so-called steroid or nuclear receptor superfamily. Other members of this family have gained increasing attention in the recent years. Particularly PPARy and PPARa have been shown to have NF-KB-inhibiting activities, and could be possible tools for the treatment of IBD in the future. They are further examples for a multisite treatment approach. PPARs - like GR - are ligand-activated receptors. They have been discovered to be regulators of lipid and lipoprotein metabolism [257]. However, in recent years it was shown that they also regulate cellular proliferation, differentiation and apoptosis [257], features that may be very important during

534

Multi-site therapeutic modalities for inflammatory bowel diseases

repair processes after or during intestinal inflammation. Three family members are known. Besides PPARoc and PPARy, PPAR5 is encoded by a separate gene and expressed in most tissues. PPARs form heterodimers with the retinoid X receptor (RXR) and bind to PPAR response elements (PPRE) in the promoters of target genes.

PPARy As mentioned, PPARy is a member of the nuclear hormone receptor family and a ligand-activated transcription factor. High expression is found in adipose tissue, the adrenal gland, the spleen and, interestingly, in the colon [257-262]. PPARy is involved in the induction of adipocyte differentiation and glucose-homeostasis [262]. PPARy is also expressed in diff'erentiated macrophages, whereas there is no expression in monocytes [263]. Thiazolidinediones were identified as synthetic ligands of PPARy [262, 2 6 4 - 2 7 3 ] . Besides pharmacologically developed ligands natural ligands such as free fatty acids and the prostaglandin D2 metabolite 15-deoxy-A12,14 prostaglandin J2 bind to the PPARy protein and stimulate the transcription of target genes [274-279]. Prostaglandin D2 metabolites cannot be detected in adipose tissue; however, they are important intermediate products of arachidonic acid metabolism in macrophages and antigen-presenting cells [280]. In activated macrophages a significant-upregulation of PPARy was found [281-283]. An activation of PPARy by its ligands 15A-PGJ2 or thiazolidinediones inhibited iNOS, gelatinase B and scavenger receptor A genes. This inhibition is partially mediated by an antagonism to the transcription factors AP-1, STAT-1 and N F - K B [283]. Furthermore it could be demonstrated that PPARy inhibits the expression of a number of proinflammatory cytokines in monocytes [283, 284]. Besides macrophages and adipocytes, colonic epithelial cells express high levels of PPARy mRNA and protein [258, 285-290]. The physiological function of PPARy in intestinal epithelial cells is not well understood. A recent study by Su and co-workers showed that the expression of proinflammatory cytokines was reduced by incubation with PPARy ligands, which was mediated by inhibition of N F - K B activation [291]. The administration of thiazolidinediones during the recovery phase in an acute model of DSS (dextran sulphate sodium)-induced colitis in mice was followed by a dramatic improvement in

histological signs of inflammation [291]. Clinical trials to test the efficacy of thiazolidinediones in IBD are under way. When activated, PPARy molecules form heterodimers with another transcription factor of the nuclear receptor superfamily, the retinoid-X-receptor (RXR). Heterozygous PPARy and RXRot knockout mice display a significantly enhanced susceptibility to 2,4,6-trinitrobenzene sulfonic acid (TNBS)induced colitis compared with their wild-type littermates, indicating a role for the RXR/PPARy heterodimer in protection against colon inffammation [292]. The administration of both PPARy and RXR agonists also reduced TNBS-induced colitis, reflected by a decrease in TNF and IL-ip mRNA levels and a reduction of N F - K B DNA-binding activity [292]. A synergistic eff'ect of PPARy and RXR ligands was observed. In addition to their N F - K B antagonistic properties ligands for PPARy, similar to GR have been proven to induce apoptosis in a number of different cell types and cell lines [257, 293-303]. However, it is not clear whether the induction of apoptosis is simply mediated by the inhibition of the antiapoptotic N F KB or by a yet-incompletely elucidated specific mechanism. In addition it is necessary to clarify whether the induction of apoptosis by PPARy ligands plays a role in the regulation of the human immune system. It is important to note that PPARy activators have also been discussed as inductors of colon polyp formation [289], which could be a major problem in the therapy of IBD patients, increasing their risk of developing colon cancer. Future studies are necessary to clarify this problem.

PPARa PPARa is another member of the PPAR family. Recently it has been shown that dehydroepiandrosterone (DHEA) could be a natural ligand for PPARa. DHEA and its sulphated metabolite (DHEAS) are the most abundant steroid hormones in the body. They are predominantly synthesized in the adrenal glands in response to adrenocorticotrophin (ACTH) from its precursors cholesterol and pregnenolone, and can be further metabolized via androstenedione into androsterone and via testosterone into 17Pestradiol [304]. DHEA inhibits activation of N F - K B and secretion of IL-6 and IL-12 via the activation of PPARa [305-307]. Furthermore DHEA stimulates the production of IL-10 in murine spleen cells [308].

Gerhard Rogler Treatment with 50-200 mg/day DHEA was shown to be eflfective in patients with lupus erythematosus [309-312]. On the other hand DHEAS concentrations are decreased in patients with IBD [313, 314]. In a pilot trial treatment with DHEA was safe and effective in patients with refractory active CD or UC [315]. In this study a dose of 200 mg/day DHEA in all patients was used. Ligands for PPARoc are at present being developed by several major pharmaceutical companies. Future studies are necessary to clarify whether these ligands are favorable compared to the endogenous ligand DHEA, and whether PPARa ligation is followed by antiapoptotic or proapoptotic stimuli [316]. The data for both members of the PPAR family are so far incomplete and controlled randomized clinical trials are needed. However, they show clearly that possibilities for multi-site anti-inflammatory therapies are not limited to GR. PPAR ligands could prove to have great potential for the treatment of IBD.

535 phosphorylation of IKB at two conserved aminoterminal serine residues by a multiprotein I K B kinase complex containing iKB-kinase alpha (IKKoc), beta (IKKP) and gamma (IKKy) (Fig. 6) [330-332]. This is followed by a poly-ubiquination of the IKB proteins by a specialized E3 ubiquitin ligase complex ( E 3 I K B ) [333] which makes them accessible to proteolytic degradation by the 26S proteasome (Fig. 6) [328]. The removal of IKB proteins exposes nuclear localization signals (NLS) followed by translocation of the activated N F - K B into the nucleus [330]. There, the activated N F - K B dimer interacts with regulatory N F - K B elements in promoters and enhancers [324, 325, 329, 330] (Fig. 6). Among the genes transactivated by the N F - K B dimer, interestingly, is IKBOC

[328]. This usually limits N F - K B action. If IKB does not find a binding partner in the cytosol it translocates to the nucleus, binds activated N F - K B and mediates re-shutthng of the complex to the cytosol.

NF-KB

Inhibition of NF-KB-mediated fra/75-activation as a centrai target of IBD tiierapy The discussion of GR functions showed that one of the most important anti-inflammatory mechanisms mediated by glucocorticoid therapy is the inhibition of the proinflammatory transcription factor N F - K B . Furthermore, the first genetic mutations associated with susceptibility to CD were found in NOD2 [317, 318], which is a NF-KB-activating protein [319-321]. Therefore the N F - K B system needs more detailed consideration.

Molecular mechanisms of N F - K B activation Transcription factors of the N F - K B / R C I family form dimeric complexes which control the expression of a variety of inducible genes involved in inflammation and proUferation [322, 323]. The prototypic heterodimeric complex N F - K B consists of the subunits p50 and p65 (RelA) [324, 325]. The inactive N F - K B dimer is present in the cytosol bound to inhibitory proteins, termed IKB [326, 327] (Fig. 6). iKBa, IKBP and IKBS

have similar functions in inhibiting the translocation of the N F - K B dimer [328]. The activation of N F - K B

by inflammatory cytokines and microorganisms requires the release of IKB from the complex [324327, 329]. The release of I K B is induced by

activation in IBD

There is evidence that N F - K B transcription factors might play an important role in the inflammatory process of IBD. It was shown that the administration of antisense phosphothioate oligonucleotides to the p65 subunit of N F - K B abrogates colitis in IL-10deficient mice and in the TNBS-colitis model [334]. Inhibition of the proteasome complex which is responsible for the rapid degradation of I K B prevented N F - K B activation and attenuated the colonic and splenic injury and inflammation in the PG/PS model [335]. In the DSS model of chronic colitis the DSS-induced intestinal inflammation was characterized by an increase of N F - K B activity [336]. Blocking N F - K B activation by administering gliotoxin, a fungal product, was accompanied by a significant suppression of intestinal inflammation and mRNA expression of TNF-oc and IL-la in vivo [336]. N F - K B activation was demonstrated in cultured rat intestinal epithelial cefls (IEC-6) by IL-ip and TNF-a [207, 337]. Dexamethasone prevented epithelial cells from being activated by these mediators [207, 337]. In mucosal biopsies from patients suffering from IBD, activation of N F - K B was demonstrated to be mainly localized in two cell types in the intestinal mucosa by double-labeling techniques: (1) in lamina propria macrophages and (2) in epithelial cells [208].

Multi-site therapeutic modalities for inflammatory bowel diseases

536

cell membrane TNF-RI

nuclear membrane activator (e.g. TNF)

pro raptor

DNA

i

p65 '{R«'A)

transcription

(RelA)

inslocation NF-KB translocation

licB-degradation after ubiquinatlon the 26S-proteasome

c ^z^-"^'^ translation

mm Mini

new protein, altered cell function

Figure 6. Principles of NF-KB activation: In the cytoplasm the NF-KB complex is associated with the Inhibitory protein IKB. Activation of NF-KB IS Induced by a number of different signals, as for example IL-1 and TNF-a. These signals activate an IKBklnase complex that contains the kB-kinases IKK-a (IKK1) and IKK-jJ (IKK2). IKK-a and IKK-P are phosphorylated, leading to a recruitment of IKB and phosphorylation of IKB at serine32 and serine36. This Is followed by release of IKB from the complex. Consecutively a rapid proteolytic degradation of IKB and a translocation of the activated NF-KB Into the nucleus takes place. In the nucleus the activated NF-KB dimer Interacts with regulatory NF-KB elements In promoters and enhancers, leading to alteration in transcription rates and altered cell function. When the degradation of IKB is blocked by Inhibitors of the proteasome complex IKB reassociates with the NF-KB dimer and NF-KB activation is inhibited.

Furthermore, mutations in N0D2, a gene which is mainly expressed in monocytes/macrophages and activates N F - K B , has been shown to be associated with susceptibiHty to CD [317, 318]. A variety of genes are induced in the inflamed mucosa which have been shown to be regulated by N F - K B , including the genes encoding TNF-<x [171, 329, 338-341], IL-lp [171, 184, 339, 341-345], IL-6 [346, 347], IL-8 [348, 349], macrophage colony-stimulating factor (M-CSF) [329], macrophage granulocyte colony-stimulating factor (GM-CSF) [329], monocyte chemotactic protein-1 (MCP-1) [325, 350], vascular cell adhesion molecule-1 (VCAMl) [351, 352] and intercellular adhesion molecule-1 (ICAMl) [325]. Some of these gene products such as TNF-oc and IL1, are also able to activate N F - K B [324, 325], leading to a positive autoregulatory loop [184].

NF-KB and co-factors As mentioned above, one of the most important eff'ects of glucocorticoids (and ligands for other nuclear receptors) in the treatment of inflammatory diseases is the suppression of the /raw^'-activating ability of N F - K B . Nuclear receptors (GR and PPAR), as well as N F - K B depend on the co-activators CREP-binding protein (CBP) and steroid receptor coactivator-1 (SRC-1) for their transcriptional activity [99, 100]. It has been suggested that some of the inhibitory affect of GR action is mediated by a nuclear competition for limited amounts of the coactivators CBP and SRC-1 [100]. This means that the presence of activated GR would induce binding of CBP and/or SRC-1 and that as a consequence activated N F - K B (p65) does not find a sufficient

Gerhard Rogler number of co-activator molecules to be transcriptionally active. A heterozygous mutation in CBP is sufficient to induce Rubinstein-Taybi syndrome, illustrating the critical importance of a certain amount of CBP protein in the nucleus [353]. It is not clear whether direct protein-protein interactions between N F - K B and the GR occur or whether both proteins could be bound to a common interacting protein such as CBP. Glucocorticoids are also known to directly induce iKBa gene transcription [97]. However, it has been reported that in many cases this is not sufficient to reduce or abolish N F - K B DNA-binding activity. In addition, glucocorticoids are able to inhibit the expression of N F - K B target genes in the absence of protein synthesis, indicating that IKB induction may contribute to the inhibition of N F - K B activation but the interaction described above is more important [97].

Pathways leading to N F - K B activation A number of different pathways have been described that are able to activate N F - K B and induce its transactivating activity. In many of these pathways the IKK complex and subsequent degradation of IKB is involved. Extracellular signals can induce IKK activation by specific cell surface receptors. However, alternative pathways have been recently described, in which the signal transduction cascade is initiated by intracellular receptors or detection systems [319-321]. TNF signaling as a prototype of cell-surface receptorinitiated NF-KB activation TNF is an important mediator of mucosal inflammation and is thought to play a key role in the pathophysiology of IBD [354-356]. The TNF system is complex and has a number of regulatory mechanisms. Besides the induction of other proinflammatory cytokines TNF has pleiotropic functions mediating induction of apoptosis, cell proliferation, modulating viral replication, allergy, and inducing insulin resistance [357, 358]. The TNF homotrimer consisting of three 17 kDa units transduces these cellular responses through two distinct receptors: type I (p60 or p55), which are expressed on all cell types, and type II (p80 or p75), which are expressed only on cells of the immune system and endothelial cells. Both receptors can induce N F - K B and AP-1 as well as mitogen-activated protein (MAP) kinase

537 activation [357-362]. The activation of N F - K B by ligated TNF receptors is mediated first by recruitment of an adaptor protein (TRADD), activation of a seronine-threonine kinase (RIP) followed by activation of TNF receptor associated factor (TRAF) 2 (Fig. 7). TRAF-2 recruits the iKBa-kinase complex (IKK) to the TNF receptor [357]. It also associates with N F - K B inducing kinase (NIK), which can be a component of the IKK complex, at least when it is overexpressed (Fig. 7). NIK then activates IKKa by phosphorylation of Ser 176 [357]. However, IKKoc phosphorylation after T N F receptor ligation can also occur withoutthe presence ofNIK. BesidesIKKa the IKK complex contains IKKp, which has a very similar structure to IKKa and the unrelated IKKy. Binding of TNF to its receptors induces phosphorylation of all three kinases. The presence of IKKp and IKKy seems to bee essential for TNF-mediated N F KB activation [357]. The activated IKK complex then phosphorylates IKB, leading to its ubiquination and degradation. TheTNF-receptor-IKK-NF-KB cascade is prototypic for a cell-surface-receptor-initiated N F - K B activation. It has been speculated the inhibition of the interaction of the T N F homotrimer with its receptors may prevent N F - K B activation, and may be beneficial in inflammatory diseases. As this point of intervention is different to glucocorticoids, which take action in the nucleus, it was assumed early that anti-TNF treatment could prevent NF-KB-mediated effects in steroid-resistant refractory patients. Several methods of intervention are possible and have found their way into clinical trials. The synthesis of TNF itself can be reduced, neutralizing antibodies to TNF can be administered and recombinant soluble T N F receptor can also block T N F signaling cascades. In 1993, for the first time, a mouse/human chimeric antibody was described, which was generated in mice and had a high affinity to human TNF. This chimeric antibody (cA2) was initially used in clinical studies with patients suffering from rheumatoid arthritis, and showed very positive effects [363-369]. A pilot study in patients with severe, steroid-dependent CD showed a similar rapid improvement of the patients' symptoms [370]. A number of studies have since proved the efficacy of the anti-TNF treatment, and it can be regarded as an estabhshed therapy for patients with a glucocorticoid-dependent or -resistant disease course [371-388]. One of the most likely mechanisms by which the administration of anti-TNF antibodies reduces joint

Multi-site therapeutic modalities for inflammatory bowel diseases

538

IL1-RI

RICK {RIP''2, CARDIAK)

Ubiquination and subsequent IKBdegradation by the 26S-proteasome

p50

/v.

p65 (RelA)

p50

p65 (RelA)

NF-KB-translocation

Figure 7. Pathways of NF-KB activation. TNF-RI (binding TNF) and the IL-1 receptor family (binding IL-1a or IL-1 p to IL-1RI and extracellular LPS toTLRs) utilize distinct but analogous signaling cascades resulting in IKK activation, IKB degradation and finally in NF-KB translocation and activation. Both bind adaptor proteins (TRADD associates to TNF-Rl, MyD88 associates with IL-1 Rl and TLRs). This Is followed by a recruitment of a seronine-threonine kinase (RIP and IRAK respectively) that are phosphorylated. Both pathways then utilize TRAF factors (TRAF-2 and TRAF-6) leading to IKK activation. A new pathway leading to NF-KB activation has gained attention by the demonstration of an association of polymorphisms in the N0D2 gene with the susceptibility for Crohn's disease. N0D1 and N0D2 are potential intracellular receptors for LPS, binding LPS In a leucin-rich domain. After binding their CARD domain interacts with the CARD domain of RICK, followed by IKK-complex activation. Potential targets for therapeutic intervention may be defined in the future by further elucidation of the pathways shown.

and mucosal inflammation is reduced activation of N F - K B . The antibody prevents circulating TNF from binding to its receptor, subsequent IKK activation, IKB phosphorylation and N F - K B activation. However, this may not be the whole story, as a positive effect of the anti-TNF antibody administration is still observed up to 1 year after the antibody has been cleared from the circulation [355]. It has been suggested that anti-TNF therapy may also induce apoptosis in activated lymphocytes and macrophages. On the other hand, in glucocorticoid responders, longlasting remissions can be achieved when no therapeutic glucocorticoids are administered, indicating that the effector needs not be present for a longlasting change of immune mechanisms.

N F - K B activation induced by the IL-1 receptor family Like TNF, IL-1 is a central regulator of immune functions. It plays a major role in both systemic and local immune responses to bacteria and bacterial wall products such as LPS [359, 389-391]. For IL-1 two receptors (IL-IRI and IL-IRII) are known that mediate cellular responses. These receptors belong to a family of receptors based on homologous molecule structures within the intracellular signaling domain. Other family members include the IL-18 receptor and, perhaps more importantly, the recently discovered Toll-like receptors (TLR), which recognize different bacterial products (TLR4: lipopolysaccharides (LPS),TLR2: Gram-positive bacteria, bacterial proteins, TLR9: bacterial DNA) [392-400].

Gerhard Rogler

539

Glucocorticoids/Glucocorticoid-receptor 5-ASA, Sulasalazine

TNF-I DNA

pro mot

\ transcription

^

=

translation . » « ^ ^ -

^

anti-TNF antibodies

transcrlp^n

protein, " - p - " relevant for apoptosis

PPARy PPARa

Induction of apoptosis

Inhibition of N F - K B mediated gene induction

Figure 8. Summary: Multi target therapies for Inflammatory bowel disease act on two major pathways they induce apoptosis and/or inhibit the activation of NF-KB.

The type I IL-1 receptor (IL-IRI) is an 80 kDa high-affinity receptor present on most cells of the immune system, including monocytes/macrophages and lymphocytes. The receptor consists of a 319 AA aminoterminal extracellular domain, a 219 A A carboxyl-terminal intracellular domain, and a transmembrane domain containing a short region of uncharged AA [390]. The cytoplasmic domain of IL-IRI shows homology to the IL-18 receptor and the TLR family. The region between AA 513 and 529 of the IL-IRI molecule appears to be crucial for NFKB activation. Both biologically active isoforms of IL-1, oc and P, can bind to IL-IRI. A so-called IL-1 receptor associated protein (IL-lRAcP) is necessary for IL-1 receptor function. IL-lRAcP does not bind either IL-1 a or IL-ip, but it increases their binding affinity to the IL-lRI and, more importantly, it activates additional signal transduction pathways that are not activated in cells expressing IL-IRI alone [390]. The type II IL-1 receptor (IL-IRII) is a 60 kDa receptor with a very short cytoplasmic domain lacking signal transduction capabilities even though it can bind IL-1 with high affinity. Mammalian TLR are homologs of Drosophila Toll receptor

and are also members of the IL-1 receptor family with the ability to activate N F - K B . TO date, 10 TLRs (TLR 1-10) have been described. The first event in IL-1 signal transduction following IL-1 binding to IL-IRI is the formation of the IL-IR and IL-1 R-AcP complex [390] (Fig. 7). This induces a rapid and stable association of an adapter protein called MyD88 with the receptor complex. MyD88 allows the interaction of IL-1 receptor associated kinase (IRAK) with the complex (Fig. 7). The association of MyD88 with IRAK is rapid and transient, followed by a phosphorylation of IRAK resulting in reduced affinity for MyD88 and its release from the IL-1RI complex [390]. To date three IRAK molecules (IRAK, IRAK-2 and IRAK-M) are known. Phosphorylated IRAK and IRAK-2 subsequently associate with the downstream signaling protein T N F receptor-associated factor 6 (TRAF6), a member of the TNF receptor-associated factor family. Activation of TRAF6 is followed by activation of NIK (Fig. 7). As mentioned (see above) N I K may form a complex with IKKoc, IKKP, and IKKy. Ligands that bind to TLR proteins activate N F - K B by using similar signal transduction pathways.

540

Multi-site therapeutic modalities for inflammatory bowel diseases

Because of the importance of IL-1 receptor family members for N F - K B activation, future therapy strategies may well target this system. Interesting targets for a therapeutic intervention are not only the IL-1 /IL-1 R system but also the LPS recognizing TLR4 and the bacterial DNA-bindingTLR9. Targeting IRAK or MyD88 may also turn out to be an interesting strategy. NF-KB activation by NODI and N0D2 Recent data indicate an association of the susceptibility for CD with polymorphisms or mutations in the NOD2 gene [317,318]. For the first time a genetic mutation involved in the pathogenesis of IBD could be demonstrated. It is intriguing that that mutated protein is expressed mainly in monocytes, binds LPS and is capable of N F - K B activation. NODI and NOD2 as well as Apafl are members of a family of intracellular proteins that contain an N-terminal caspase recruitment domain (CARD) [319-321]. NODI and NOD2 (now termed CARD 15) both respond to LPS with N F - K B activation independent of TLR4 and MyD88 function [320] (Fig. 7). However, their preferences for different LPS forms and different bacteria seem to differ [320]. After binding of LPS to a leucin-rich domain in the NODI or NOD2 (CARD 15) protein the CARD domain of NOD interacts with the corresponding CARD domain of a protein called RICK (also called RIP2 and CARDIAK). RICK is a protein kinase that is able to activate N F - K B in a IKK-dependent pathway (Fig. 7). lKBa,IKKa,IKK(3 as well as IKKy seem to be necessary for NODRICK mediated N F - K B activation [319]. Based upon these results NODI and NOD2 (CARD 15) are likely to be members of an intracellular recognition system for LPS regulating N F - K B activation.

Side-effects of N F - K B blockade We all are aware of the immunosuppressive sideeffects of a systemic glucocorticoid therapy that can lead to fatal consequences in the case of infections. One of the reasons for the observed immunosuppression is the induction of apoptosis in activated lymphocyte and monocyte populations. As the inhibition of N F - K B is one of the most important mechanisms of glucocorticoid (and PPAR-ligand) mediated effects, it is clear also that N F - K B inhibition may contribute to the side-effects. The disruption of the p65 gene in mice is followed by embryonic lethality at 15-16 days of gestation and

a massive degeneration of the liver by programmed cell death or apoptosis at this early time point [401]. TNF-mediated gene induction is absent in those mice, whereas a basic NF-KB-dependent transcription of genes is conserved [401]. This indicates the p65 subunit is essential for inducible, but not basal, transcription in NF-KB-regulated pathways. Mice lacking the p50 subunit, in contrast, show no developmental abnormalities [402]. However, they exhibit multifocal defects in various forms of immune responses, e.g. a lack of proliferation of B cells in response to bacterial lipopolysaccharide and a defect in basal and specific antibody production [402]. From these observations it can be concluded that a systemic complete blockade of N F - K B activation could be a dangerous treatment approach. In a DSS model of colitis, treatment with 2 x 20 |ig/day of the N F - K B inhibitor gliotoxin was followed by reduced survival of the mice despite improved histology in the colon [336]. It may therefore be better to achieve a local (mucosal) N F - K B inhibition. Curcumin, a substance derived from the plant Curcuma longa, which is used as a spice in India, inhibits IKB degradation and thereby N F - K B activation [403-409]. As the absorption rate of Curcumin from the gut is low, a local N F KB inhibition could be potentially achieved with this drug. Like 'topical' steroids such as budesonide this 'locaP N F - K B inhibition would avoid systemic sideeffects. Another possibility that has been discussed would be the development of tissue-specific IKK inhibitors; however, so far there is no clear evidence for a clear tissue-specificity of IKK activation pathways. Until a local or tissue specific direct N F - K B inhibition becomes possible indirect and systemic approaches such as therapy with anti-TNF antibodies must be evaluated. Taken together, N F - K B seems to play a key role for the treatment of chronic and acute inflammatory diseases and is the final target of a number of successful multi-site therapeutic approaches. Inhibition of N F - K B at different cellular levels has provided therapeutic alternatives in the past and will provide new alternatives to glucocorticoid therapy in the future. This may finally lead to an optimized therapy for individual patients with their specific course of IBD. The understanding of the mechanisms involved in N F - K B activation on the one hand, and GR and PPAR-mediated suppression of N F - K B activation on the other, is essential for this future development.

Gerhard Rogler

References 1. 2. 3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Schleimer RP. An overview of glucocorticoid anti-inflammatory actions. Eur J Clin Pharmacol 1993; 45: S3-7; discussion S43—4. Barnes PJ, Adcock I. Anti-inflammatory actions of steroids: molecular mechanisms. Trends Pharmacol Sci 1993; 14: 436-41. Barnes PJ. Anti-inflammatory actions of glucocorticoids: molecular mechanisms. Clin Sci (Colch) 1998; 94: 557-72. Kirwan JR. Effects of long-term glucocorticoid therapy in rheumatoid arthritis. Z Rheumatol 2000; 59: II: 85-9. van der Velden VH. Glucocorticoids: mechanisms of action and anti-inflammatory potential in asthma. Mediators Inflamm 1998; 7: 229-37. Spahn JD, Leung DY. The role of glucocorticoids in the management of asthma. Allergy Asthma Proc 1996; 17: 341-50. Lombardino JG. Mechanism of action of drugs for treating inflammation and arthritis. Eur J Rheumatol Inflamm 1983; 6: 24-35. Friend DR. Review article: Issues in oral administration of locally acting glucocorticosteroids for treatment of inflammatory bowel disease. Aliment Pharmacol Ther 1998; 12: 591-603. Adcock IM, Ito K. Molecular mechanisms of corticosteroid actions. Monaldi Arch Chest Dis 2000; 55: 256-66. Kumar R, Thompson EB. The structure of the nuclear hormone receptors. Steroids 1999; 64: 310-19. Webster JC, Cidlowski JA. Mechanisms of glucocorticoidreceptor-mediated repression of gene expression. Trends Endocrinol Metab 1999; 10: 396^02. Glass CK, Rosenfeld MG. The coregulator exchange in transcriptional functions of nuclear receptors. Genes Dev 2000; 14: 121-41. Newton R. Molecular mechanisms of glucocorticoid action: what is important? Thorax 2000; 55: 603-13. McKay LI, Cidlowski JA. Molecular control of immune/ inflammatory responses: interactions between nuclear factor-kappa B and steroid receptor-signaling pathways. Endocrinol Rev 1999; 20: 435-59. Truelove S, Witts L. Cortisone in ulcerative coHtis: preliminary report on a therapeutic trial. Br Med J 1954; 2: 375-8. Truelove S, Witts L. Cortisone and corticotrophin in ulcerative colitis. Br Med J 1959; 10: 387-94. Jones J, Lennard-Jones J. Corticosteroids and corticotrophin in the treatment of Crohn's disease. Gut 1966; 7: 1817. Roberts G, Naish J. Corticosteroids in Crohn's disease. Gut 1968; 9: 736. Lennard-Jones J. Toward optimal use of corticosteroids in ulcerative colitis and Crohn's disease. Gut 1983; 24: 177-81. Jewell D. Corticosteroids for the management of ulcerative colitis and Crohn's disease. Gastroenterol Clin N Am 1989; 18:21-34. Routes J, Claman H. Corticosteroids in inflammatory bowel disease. A review. J Clin Gastroenterol 1987; 9: 529-35. Hanauer S, Baert F. Medical therapy of inflammatory bowel disease. Med Clin N Am 1994; 78: 1413-26. Malchow H, Ewe K, Brandes J et al. European Cooperative Crohn's Disease Study (ECCDS): results of drug treatment. Gastroenterology 1984; 86: 249-66. Brattsand R, Linden M. Cytokine modulation by glucocorticoids: mechanisms and actions in cellular studies. Aliment Pharmacol Ther 1996; 10: 81-92. Arzt E, Paez Pereda M, Costas M et al. Cytokine expression and molecular mechanisms of their auto/paracrine regulation of anterior pituitary function and growth. Ann NY Acad Sci 1998; 840: 525-31.

541 26. 27. 28. 29.

30. 31. 32. 33.

34. 35.

36. 37.

38.

39. 40.

41.

42. 43.

44. 45.

46.

Arzt E, Kovalovsky D, Igaz LM et al. Functional cross-talk among cytokines, T-cell receptor, and GR transcriptional activity and action. Ann NY Acad Sci 2000; 917: 672-7. Barnes PJ. Novel approaches and targets for treatment of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999; 160: S72-9. Venkatesh VC, Ballard PL. Glucocorticoids and gene expression. Am J Respir Cell Mol Biol 1991; 4: 301-3. Weinberger C, Hollenberg SM, Rosenfeld MG, Evans RM. Domain structure of human glucocorticoid receptor and its relationship to the v-erb-A oncogene product. Nature 1985; 318:670-2. Hollenberg SM, Weinberger C, Ong ES et al. Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 1985; 318: 635-41. Weinberger C, Hollenberg SM, Ong ES et al. Identification of human glucocorticoid receptor complementary DNA clones by epitope selection. Science 1985; 228: 740-2. Evans R. The steroid and thyroid hormone receptor superfamily. Science 1988; 240: 889-95. Arriza JL, Weinberger C, CerelH G et al. Cloning of human mineralocorticoid receptor complementary DNA: structural and functional kinship with the glucocorticoid receptor. Science 1987; 237: 268-75. Evans R. Molecular characterization of the glucocorticoid receptor. Recent Prog Horm Res 1989; 45: 1-22. Lamberts S, Koper J, Biemond P, den-Holder F, de-Jong F. Cortisol receptor resistance: the variability of its clinical presentation and response to treatment. J Clin Endocrinol Metab 1992; 74: 313-21. Lamberts SW, Huizenga AT, de Lange P, de Jong FH, Koper JW. Clinical aspects of glucocorticoid sensitivity. Steroids 1996; 61: 157-60. Okret S, Dong Y, Tanaka H, Cairns B, Gustafsson J. The mechanism for glucocorticoid-resistance in a rat hepatoma cell variant that contains functional glucocorticoid receptor. J Steroid Biochem Mol Biol 1991; 40: 353-61. Schlaghecke R, Beuscher D, Kornely E, Specker C. Effects of glucocorticoids in rheumatoid arthritis. Diminished glucocorticoid receptors do not result in glucocorticoid resistance. Arthritis Rheum 1994; 37: 1127-31. Muller M, Renkawitz R. The glucocorticoid receptor. Biochim Biophys Acta 1991; 1088: 171-82. Muller M, Baniahmad C, Kaltschmidt C, Renkawitz R. Multiple domains of the glucocorticoid receptor involved in synergism with the CACCC box factor(s). Mol Endocrinol 1991; 5: 1498-503. Baniahmad C, Muller M, Altschmied J, Renkawitz R. Cooperative binding of the glucocorticoid receptor DNA binding domain is one of at least two mechanisms for synergism. J Mol Biol 1991; 222: 155-65. Baumann H, Paulsen K, Kovacs H et al. Refined solution structure of the glucocorticoid receptor DNA-binding domain. Biochemistry 1993; 32: 13463-71. Luisi BE, Xu WX, Otwinowski Z, Freedman LP, Yamamoto KR, Sigler PB. Crystallographic analysis of the interaction of the glucocorticoid receptor with DNA. Nature 1991; 352: 497-505. Pan T, Freedman LP, Coleman JE. Cadmium-113 NMR studies of the DNA binding domain of the mammalian glucocorticoid receptor. Biochemistry 1990; 29: 9218-25. Freedman LP, Luisi BF, Korszun ZR, Basavappa R, Sigler PB, Yamamoto KR. The function and structure of the metal coordination sites within the glucocorticoid receptor DNA binding domain. Nature 1988; 334: 543-6. Freedman LP, Yamamoto KR, Luisi BF, Sigler PB. More fingers in hand. Cell 1988; 54: 444.

542 47. 48. 49.

50.

51.

52.

53.

54.

55.

56.

57. 58.

59. 60. 61. 62. 63. 64.

65.

Multi-site therapeutic modalities for inflammatory bowel diseases Luisi BF, Schwabe JW, Freedman LP. The steroid/nuclear receptors: from three-dimensional structure to complex function. Vitam Horm 1994; 49: 1-47. Giguere V, Hollenberg SM, Rosenfeld MG, Evans RM. Functional domains of the human glucocorticoid receptor. Cell 1986;46:645-52. Hollenberg SM, Giguere V, Evans RM. Identification of two regions o^ the human glucocorticoid receptor hormone binding domain that block activation. Cancer Res 1989; 49: 2292^. Hollenberg SM, Giguere V, Segui P, Evans RM. Colocalization of DNA-binding and transcriptional activation functions in the human glucocorticoid receptor. Cell 1987; 49: 39 46. Dahlman-Wright K, Baumann H, McEwan IJ et al. Structural characterization of a minimal functional transactivation domain from the human glucocorticoid receptor. Proc Natl Acad Sci USA 1995; 92: 1699-703. Dahlman-Wright K, McEwan IJ. Structural studies of mutant glucocorticoid receptor transactivation domains establish a link between transactivation activity in vivo and alpha-helix-forming potential in vitro. Biochemistry 1996; 35: 1323 7. McEwan IJ, Wright AP, Dahlman-Wright K, CarlstedtDukc J, Guslafsson JA. Direct interaction of the tau 1 transactivation domain of the human glucocorticoid receptor with the basal transcriptional machinery. Mol Cell Biol 1993; 13: 399 407. McEwan I J, Dahlman-Wright K, Amlof T, Ford J, Wright AP, Gustafsson JA. Mechanisms of transcription activation by nuclear receptors: studies on the human glucocorticoid receptor tau 1 transactivation domain. Mutat Res 1995; 333: 15 22. Henriksson A, Almlof T, Ford J, McEwan IJ, Guslafsson JA, Wright AP. Role of the Ada adaptor complex in gene activation by the glucocorticoid receptor. Mol Cell Biol 1997; 17: 3065 73. Ford J, McEwan IJ, Wright AP, Guslafsson JA. Involvement of the transcription factor I ID protein complex in gene activation by the N-terminal transactivation domain of the glucocorticoid receptor in vitro. Mol Endocrinol 1997; 11: 1467 75. Guarente L. Transcriptional coactivators in yeast and beyond. Trends Biochem Sci 1995; 20: 517 21. Zeiner M, Gehring U. A protein that interacts with members of the nuclear hormone receptor family: identification and cDNA cloning. Proc Natl Acad Sci USA 1995; 92: 11465-9. Onate SA, Tsai SY, Tsai MJ, O'Malley BW. Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 1995; 270: 1354-7. Chakravarti D, LaMorte VJ, Nelson MC et al. Role of CBP/ P300 in nuclear receptor signalling. Nature 1996; 383: 99103. Jantzen HM, Strahle U, Gloss B et al. Cooperativity of glucocorticoid response elements located far upstream of the tyrosine aminotransferase gene. Cell 1987; 49: 29-38. Beato M, Truss M, Chavez S. Control of transcription by steroid hormones. Ann NY Acad Sci 1996; 784: 93-123. Beato M. Gene regulation by steroid hormones. Cell 1989; 56: 335-44. Abbinante N, Simpson L, Leikauf G. Corticosteroids increase secretory leukocyte protease inhibitor transcript levels in airway epithelial cells. Am J Physiol 1995; 268: L601-6. Colotta F, Dower SK, Sims JE, Mantovani A. The type II 'decoy' receptor: a novel regulatory pathway for interleukin 1. Immunol Today 1994; 15: 562-6.

66. 67. 68.

69. 70.

71.

72.

73.

74. 75.

76.

77.

78. 79. 80. 81.

82.

83.

84.

Colotta F, Mantovani A. Induction of the interleukin-1 decoy receptor by glucocorticoids. Trends Pharmacol Sci 1994; 15: 138-9. Colotta F, Re F, Muzio M et al. Interleukin-1 type II receptor: a decoy target for IL-1 that is regulated by IL-4. Science 1993;261:472-5. Purton JF, Boyd RL, Cole TJ, Godfrey DI. Intrathymic T cell development and selection proceeds normally in the absence of glucocorticoid receptor signaling. Immunity 2000; 13: 179-86. Kofler R. The molecular basis of glucocorticoid-induced apoptosis of lymphoblastic leukemia cells. Histochem Cell Biol 2000; 114: 1-7. Hulkko SM, Wakui H, Zilliacus J. The pro-apoptotic protein death-associated protein 3 (DAP3) interacts with the glucocorticoid receptor and affects the receptor function. Biochem J 2000;349: 885-93. Jamieson CA, Yamamoto KR. Crosstalk pathway for inhibition of glucocorticoid-induced apoptosis by T cell receptor signaling. Proc Natl Acad Sci USA 2000; 97: 7319 24. Tolosa E, King LB, Ashwell JD. Thymocyte glucocorticoid resistance alters positive selection and inhibits autoimmunity and lymphoproliferative disease in MRL-lpr/lpr mice. Immunity 1998; 8: 67 76. Cidlowski JA, King KL, Evans-Storms RB, Montague JW, Bortner CD, Hughes FM, Jr. The biochemistry and molecular biology of glucocorticoid-induced apoptosis in the immune system. Recent Prog Horm Res 1996; 51: 457 90. Nicto MA, Gonzalez A, Gambon F, Diaz-Espada F, LopezRivas A. Apoptosis in human thymocytes after treatment with glucocorticoids. Clin Exp Immunol 1992; 88: 341 4. Wyllic AH, Morris RG. Hormone-induced cell death. Purification and properties of thymocytes undergoing apoptosis after glucocorticoid treatment. Am J Pathol 1982; 109: 78 87. Wyllie AH, Morris RG, Smith AL, Dunlop D. Chromatin cleavage in apoptosis: association with condensed chromatin morphology and dependence on macromolecular synthesis. J Pathol 1984; 142:67 77. Ramdas J, Liu W, Harmon JM. Glucocorticoid-induced cell death requires autoinduction of glucocorticoid receptor expression in human leukemic T cells. Cancer Res 1999; 59: 1378-85. Ramdas J, Harmon JM. Glucocorticoid-induced apoptosis and regulation of NF-kappaB activity in human leukemic T cells. Endocrinology 1998; 139: 3813-21. Drouin J, Sun YL, Chamberland M et al. Novel glucocorticoid receptor complex with DN A element of the hormonerepressed POMCgene. EMBO J 1993; 12: 145-56. Charron J, Drouin J. Glucocorticoid inhibition of transcription from episomal proopiomelanocortin gene promotor. Proc Natl Acad Sci USA 1986; 83: 8903-7. Malkoski SP, Handanos CM, Dorin RI. Localization of a negative glucocorticoid response element of the human corticotropin releasing hormone gene. Mol Cell Endocrinol 1997;127: 189-99. Malkoski SP, Dorin RI. Composite glucocorticoid regulation at a functionally defined negative glucocorticoid response element of the human corticotropin-releasing hormone gene. Mol Endocrinol 1999; 13: 1629-44. Murphy EP, Conneely OM. Neuroendocrine regulation of the hypothalamic pituitary adrenal axis by the nurrl /nur77 subfamily of nuclear receptors. Mol Endocrinol 1997; 11: 39-47. Newton R, Barnes P, Adcock I. Transcription factors. In: Barnes PJ, Rodger IW, Thomson NC, eds. Asthma: Basic mechanisms and clinical management. London: Academic Press, 1998: 459-74.

Gerhard Rogler 85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98. 99. 100.

101. 102.

Schule R, Rangarajan P, Kliewer S, Ransone LJ, Bolado J, Yang N, Verma IM, Evans RM. Functional antagonism between oncoprotein c-Jun and the glucocorticoid receptor. Cell 1990; 62: 1217-26. Jonat C, Rahmsdorf HJ, Park KK et al. Antitumor promotion and antiinflammation: down-modulation of AP-1 (Fos/ Jun) activity by glucocorticoid hormone. Cell 1990; 62: 1189-204. Yang-Yen HF, Chambard JC, Sun YL et al. Transcriptional interference between c-Jun and the glucocorticoid receptor: mutual inhibition of DNA binding due to direct proteinprotein interaction. Cell 1990; 62: 1205-15. Konig H, Ponta H, Rahmsdorf HJ, Herrlich P. Interference between pathway-specific transcription factors: glucocorticoids antagonize phorbol ester-induced AP-1 activity without altering AP-1 site occupation in vivo. EMBO J 1992; 11: 2241-6. Caldenhoven E, Liden J, Wissink S et al. Negative cross-talk between RelA and the glucocorticoid receptor: a possible mechanism for the antiinflammatory action of glucocorticoids. Mol Endocrinol 1995; 9: 401-12. Ray A, Prefontaine KE. Physical association and functional antagonism between the p65 subunit of transcription factor NF-kappa B and the glucocorticoid receptor. Proc Natl Acad Sci USA 1994; 91: 752-6. Ray A, Siegel MD, Prefontaine KE, Ray P. Anti-inflammation: direct physical association and functional antagonism between transcription factor NF-KB and the glucocorticoid receptor. Chest. 1995; 107: 139S. Scheinman RI, Gualberto A, Jewell CM, Cidlowski JA, Baldwin AS, Jr. Characterization of mechanisms involved in transrepression of NF-kappa B by activated glucocorticoid receptors. Mol Cell Biol 1995; 15: 943-53. McKay LI, Cidlowski JA. CBP (CREB binding protein) integrates NF-kappaB (nuclear factor-kappaB) and glucocorticoid receptor physical interactions and antagonism. Mol Endocrinol 2000; 14: 1222-34. McKay LI, Cidlowski JA. Cross-talk between nuclear factor-kappa B and the steroid hormone receptors: mechanisms of mutual antagonism. Mol Endocrinol 1998; 12: 4 5 56. Brostjan C, Anrather J, Csizmadia V et al. Glucocorticoidmediated repression of NFkappaB activity in endothelial cells does not involve induction of IkappaBalpha synthesis. J Biol Chem 1996; 271: 19612-16. Wissink S, van Heerde EC, Schmitz ML et al. Distinct domains of the RelA NF-kappaB subunit are required for negative cross-talk and direct interaction with the glucocorticoid receptor. J Biol Chem 1997; 272: 22278-84. Wissink S, van Heerde EC, van der Burg B, van der Saag PT. A dual mechanism mediates repression of NF-kappaB activity by glucocorticoids. Mol Endocrinol 1998; 12: 35563. Kamei Y, Xu L, Heinzel T et al. A CBP integrator complex mediates transcriptional activation and AP-1 inhibition by nuclear receptors. Cell 1996; 85: 403-14. Sheppard KA, Rose DW, Haque ZK et al. Transcriptional activation by NF-kappaB requires multiple coactivators. Mol Cell Biol 1999; 19: 6367-78. Sheppard KA, Phelps KM, Williams AJ et al. Nuclear integration of glucocorticoid receptor and nuclear factorkappaB signaling by CREB-binding protein and steroid receptor coactivator-1. J Biol Chem 1998; 273: 29291-4. Kurokawa R, Kalafus D, Oghastro MH et al. Differential use of CREB binding protein-coactivator complexes. Science 1998; 279: 700-3. Horvai AE, Xu L, Korzus E et al. Nuclear integration of JAK/STATand Ras/AP-1 signahng by CBP and p300. Proc Natl Acad Sci USA 1997; 94: 1074-9.

543 103. Glass CK, Rose DW, Rosenfeld MG. Nuclear receptor coactivators. Curr Opin Cell Biol 1997; 9: 222-32. 104. Ray DW, Suen CS, Brass A, Soden J, White A. Structure/ function of the human glucocorticoid receptor: tyrosine 735 is important for transactivation. Mol Endocrinol 1999; 13: 1855-63. 105. Radoja N, Komine M, Jho SH, Blumenberg M, TomicCanic M. Novel mechanism of steroid action in skin through glucocorticoid receptor monomers. Mol Cell Biol 2000; 20: 4328-39. 106. Oshima H, Simons SS, Jr. Modulation of transcription factor activity by a distant steroid modulatory element. Mol Endocrinol 1992; 6: 416-28. 107. Zeng H, Plisov SY, Simons SS, Jr. Abihty of the glucocorticoid modulatory element to modify glucocorticoid receptor transactivation indicates parallel pathways for the expression of glucocorticoid modulatory element and glucocorticoid response element activities. Mol Cell Endocrinol 2000; 162: 221-34. 108. Kaul S, Blackford JA, Jr., Chen J, Ogryzko VV, Simons SS, Jr. Properties of the glucocorticoid modulatory element binding proteins GMEB-1 and -2: potential new modifiers of glucocorticoid receptor transactivation and members of the family of KDWK proteins. Mol Endocrinol 2000; 14: 1010-27. 109. Jimenez-Lara AM, Heine MJ, Gronemeyer H. Cloning of a mouse glucocorticoid modulatory element binding protein, a new member of the KDWK family. FEBS Lett 2000; 468: 203-10. 110. Theriault JR, Charette SJ, Lambert H, Landry L Cloning and characterization of hGMEBl, a novel glucocorticoid modulatory element binding protein. FEBS Lett 1999; 452: 170-6. 111. Zeng H, Jackson DA, Oshima H, Simons SS, Jr. Cloning and characterization of a novel binding factor (GMEB-2) of the glucocorticoid modulatory element. J Biol Chem 1998; 273: 17756-62. 112. Zeng H, Kaul S, Simons SS, Jr. Genomic organization of human GMEB-1 and rat GMEB-2: structural conservation of two multifunctional proteins. Nucleic Acids Res 2000; 28: 1819-29. 113. Oakley RH, Cidlowski JA. Homologous down regulation of the glucocorticoid receptor: the molecular machinery. Crit Rev Eukaryot Gene Expr. 1993; 3: 63-88. 114. Burnstein KL, Cidlowski JA. Regulation of gene expression by glucocorticoids. Annu Rev Physiol 1989; 51: 683-99. 115. Burnstein KL, Jewell CM, Cidlowski JA. Human glucocorticoid receptor cDNA contains sequences suflicient for receptor down-regulation. J Biol Chem 1990; 265: 7284-91. 116. Burnstein KL, Bellingham DL, Jewell CM, Powell-Oliver FE, Cidlowski JA. Autoregulation of glucocorticoid receptor gene expression. Steroids 1991; 56: 52-8. 117. Burnstein KL, Cidlowski JA. The down side of glucocorticoid receptor regulation. Mol Cell Endocrinol 1992; 83: C l 8. 118. Burnstein KL, Cidlowski JA. Multiple mechanisms for regulation of steroid hormone action. J Cell Biochem 1993; 51: 130-4. 119. Webster JC, Cidlowski JA. Downregulation of the glucocorticoid receptor. A mechanism for physiological adaptation to hormones. Ann NY Acad Sci 1994; 746: 216-20. 120. Webster JC, Jewell CM, Bodwell JE, Munck A, Sar M, Cidlowski JA. Mouse glucocorticoid receptor phosphorylation status influences multiple functions of the receptor protein. J Biol Chem 1997; 272: 9287-93. 121. Bamberger CM, Bamberger AM, de Castro M, Chrousos GP. Glucocorticoid receptor beta, a potential endogenous inhibitor of glucocorticoid action in humans. J Clin Invest 1995;95:2435-41.

544

Multi-site therapeutic modalities for inflammatory bowel diseases

122. Bamberger CM, Else T, Bamberger AM, Beil FU, Schulte HM. Regulation of the human interleukin-2 gene by the alpha and beta isoforms of the glucocorticoid receptor. Mol Cell Endocrinol 1997; 136: 23-8. 123. Bamberger CM, Bamberger AM, Wald M, Chrousos GP, Schulte HM. Inhibition of mineralocorticoid activity by the beta-isoform of the human glucocorticoid receptor. J Steroid Biochem Mol Biol 1997; 60: 43-50. 124. Bamberger CM, Schulte HM, Chrousos GP. Molecular determinants of glucocorticoid receptor function and tissue sensitivity to glucocorticoids. Endocrinol Rev 1996; 17: 245-61. 125. Oakley RH, Sar M, Cidlowski JA. The human glucocorticoid receptor beta isoform. Expression, biochemical properties, and putative function. J Biol Chem 1996; 271: 9550-9. 126. Oakley RH, Webster JC, Sar M, Parker CR, Jr, Cidlowski JA. Expression and subcellular distribution of the betaisoform of the human glucocorticoid receptor. Endocrinology 1997; 138: 5028 38. 127. Oakley RH, Jewell CM, Yudt MR, Bofetiado DM, Cidlowski JA. The dominant negative activity of the human glucocorticoid receptor beta isoform. Specificity and mechanisms of action. J Biol Chem 1999; 274: 27857 66. 128. Sousa AR, Lane SJ, Cidlowski JA, Staynov DZ, Lcc TH. Glucocorticoid resistance in asthma is associated with elevated in vivo expression of the glucocorticoid receptor beta-isoform. J Allergy Clin Immunol 2000; 105: 943 50. 129. Christodoulopoulos P, Leung DY, Elliott MW ct al. Increased number of glucocorticoid rcceplor-beta-cxpressing cells in the airways in fatal asthma. J Allergy Clin Immunol 2000; 106:479 84. 130. Hamid QA, Wenzel SE, Hauk PJ ct al. Increased glucocorticoid receptor beta in airway cells of glucocorticoid-insensitivc asthma. Am J Respir Crit Care Med 1999; 159: 1600 4. 131. Gagliardo R, Chanez P, Vignola AM ct al. Glucocorticoid receptor alpha and beta in glucocorticoid dependent asthma. Am J Respir Crit Care Med 2000; 162: 7 13. 132. Carlstedt-Dukc J. Glucocorticoid receptor beta: View II. Trends Endocrinol Metab 1999; 10: 339 42. 133. Vottero A, Chrousos GP. Glucocorticoid Receptor beta: View I. Trends Endocrinol Metab 1999; 10: 333 8. 134. Stahl M, Ludwig D, Fellermann K, Stange EF. Intestinal expression of human heat shock protein 90 in patients with Crohn's disease and ulcerative colitis. Dig Dis Sci 1998; 43: 1079 87. 135. Shimada T, Hiwatashi N, Yamazaki H, Kinouchi Y, Toyota T. Relationship between glucocorticoid receptor and response to glucocorticoid therapy in ulcerative colitis. Dis Colon Rectum 1997; 40: S54-8. 136. Rogler G, Meinel A, Lingauer A ct al. Glucocorticoid receptors are down-regulated in inflamed colonic mucosa but not in peripheral blood mononuclear cells from patients with inflammatory bowel disease. Eur J Clin Invest 1999; 29: 330-6. 137. Rosewicz S, McDonald AR, Maddux BA, Goldfine ID, Miesfeld RL, Logsdon CD. Mechanism of glucocorticoid receptor down-regulation by glucocorticoids. J Biol Chem 1988;263:2581-4. 138. Okret S, Poellinger L, Dong Y, Gustafsson JA. Downregulation of glucocorticoid receptor mRNA by glucocorticoid hormones and recognition by the receptor of a specific binding sequence within a receptor cDNA clone. Proc Natl Acad Sci USA 1986; 83: 5899-903. 139. Cidlowski J A, Cidlowski NB. Regulation of glucocorticoid receptors by glucocorticoids in cultured HeLa S3 cells. Endocrinology. 1981; 109: 1975-82. 140. Honda M, Orii F, Ayabe T et al. Expression of glucocorticoid receptor beta in lymphocytes of patients with gluco-

141. 142.

143.

144.

145. 146. 147. 148.

149.

150.

151.

152. 153.

154.

155.

156. 157. 158.

159.

corticoid-resistant ulcerative colitis. Gastroenterology 2000; 118:859-66. Stange EF. Glucocorticoid receptor activity in inflammatory bowel disease: hindsight or foresight? Eur J Clin Invest 1999;29:278-9. Boumpas DT, Paliogianni F, Anastassiou ED, Balow JE. Glucocorticosteroid action on the immune system: molecular and cellular aspects. Clin Exp Rheumatol 1991; 9: 41323. Almawi WY, Melemedjian OK, Rieder MJ. An alternate mechanism of glucocorticoid anti-proliferative effect: promotion of a Th2 cytokine-secreting profile. Clin Transplant 1999;13: 365-74. Kirsch AH, Mahmood A A, Endres J ct al. Apoptosis of human T-cells: induction by glucocorticoids or surface receptor ligation in vitro and ex vivo. J Biol Regul Homeost Agents 1999; 13: 80 9. Lanza L, Scudeletti M, Puppo F ct al. Prednisone increases apoptosis in in vitro activated human peripheral blood T lymphocytes. Clin Exp Immunol 1996; 103: 482 90. King LB, Ashwell JD. Signaling for death of lymphoid cells. Curr Opin Immunol 1993; 5: 368 73. Chapman MS, Askew DJ, Kuscuoglu U, Miesfeld RL. Transcriptional control of steroid-regulated apoptosis in murine thymoma cells. Mol Endocrinol 1996; 10: 967 78. Van Houten N, Blake SF, Li EJ ct al. Elevated expression of Bcl-2 and Bcl-x by intestinal intraepithelial lymphocytes: resistance to apoptosis by glucocorticoids and irradiation. Int Immunol 1997; 9: 945 53. Franchimont D, Galon J, Gadina M ct al. Inhibition of Thl immune response by glucocorticoids: dcxamethasone selectively inhibits IL-12-induced Stat4 phosphorylation in T lymphocytes. J Immunol 2000; 164: 1768 74. Girndl M, Scster U, Kaul H, Hunger F, Kohler H. Glucocorticoids inhibit activation-dependent expression of costimu 1 a t o r y m o l e c u l e B7-1 in h u m a n m o n o c y t e s . Transplantation 1998; 66: 370 5. Wagner DH, Jr, Hagman J, Linsley PS, Hodsdon W, Freed JH, Newell MK. Rescue of thymocytes from glucocorticoidinduced cell death mediated by CD28/CTLA-4 costimulatory interactions with B7-1/B7-2. J Exp Med 1996; 184: 1631 8. Rogler G, Hausmann M, Vogl D ct al. Isolation and phenotypic characterization of colonic macrophages. Clin Exp Immunol 1998; 1 12: 205 15. Rogler G, Hausmann M, Spottl Tct al. T-cell co-stimulatory molecules are upregulated on intestinal macrophages from inflammatory bowel disease mucosa. Eur J Gastroenterol Hepatol 1999; 11: 1105-1111. Andus T, Rogler G, Daig R, Falk W, Scholmerich J, Gross V. The role of macrophages. In: Tygat GNJ, Bartelsman JFWM, van Deventer SJH, eds. Inflammatory Bowel Disease. Dordrecht: Kluwer, 1995: 281-97 Lee SH, Starkey PM, Gordon S. Quantitative analysis of total macrophage content in adult mouse tissues. Immunochemical studies with monoclonal antibody F4/80. J Exp Med 1985; 161:475-489. Donnellan WL. The structure of the colonic mucosa. The epithelium and subepithelial reticulohistiocytic complex. Gastroenterology 1965; 49: 496-514. Rogler G, Andus T, Aschenbrenner E et al. Alterations of the phenotype of colonic macrophages in inflammatory bowel disease. Eur J Gastroenterol Hepatol 1997; 9: 893-9. Mahida YR, Patel S, Gionchetti P, Vaux D, Jewell DR Macrophage subpopulations in lamina propria of normal and inflamed colon and terminal ileum. Gut 1989; 30: 82634. Malizia G, Calabrese A, Cottone M et al. Expression of leukocyte adhesion molecules by mucosal mononuclear

Gerhard Rogler

160. 161.

162. 163.

164. 165.

166.

167. 168. 169. 170.

171.

172. 173.

174. 175. 176. 177.

178.

phagocytes in inflammatory bowel disease. Gastroenterology 1991; 100: 150-9. Tanner AR, Arthur MJ, Wright R. Macrophage activation, chronic inflammation and gastrointestinal disease. Gut 1984; 25: 760-83. Thyberg J, Graf W, Klingenstrom R Intestinal fine structure in Crohn's disease. Lysosomal inclusions in epithelial cells and macrophages. Virchows Arch A Pathol Anat Histol 1981;391: 141-52. Meuret G, Bitzi A, Hammer B. Macrophage turnover in Crohn's disease and ulcerative colitis. Gastroenterology 1978;74: 501-3. Hirata I, Murano M, Nitta M et al. Estimation of mucosal inflammatory mediators in rat DSS-induced colitis. Possible role of PGE(2) in protection against mucosal damage. Digestion 2001; 63: 73-80. Nieto N, Torres MI, Fernandez MI et al. Experimental ulcerative colitis impairs antioxidant defense system in rat intestine. Dig Dis Sci 2000; 45: 1820-7. Raab Y, Sundberg C, Hallgren R, Knutson L, Gerdin B. Mucosal synthesis and release of prostaglandin E2 from activated eosinophils and macrophages in ulcerative colitis. Am J Gastroenterol 1995; 90: 614-20. van Heeckeren AM, Rikihisa Y, Park J, Fertel R. Tumor necrosis factor alpha, interleukin-1 alpha, interleukin-6, and prostaglandin E2 production in murine peritoneal macrophages infected with Ehrlichia risticii. Infect Immun 1993; 61:4333-7. Yamashita S. Studies on changes of colonic mucosal PGE2 levels and tissue localization in experimental colitis. Gastroenterol Jpn 1993; 28: 224-35. Schreiber S, Raedler A, Stenson WF, MacDermott RP. The role of the mucosal immune system in inflammatory bowel disease. Gastroenterol Clin N Am 1992; 21: 451-502. Donowitz M. Arachidonic acid metabolites and their role in inflammatory bowel disease. An update requiring addition of a pathway. Gastroenterology 1985; 88: 580-7. Eliakim R, Karmeli F, Razin E, Rachmilewitz D. Role of platelet-activating factor in ulcerative colitis. Enhanced production during active disease and inhibition by sulfasalazine and prednisolone. Gastroenterology 1988; 95: 116772. Reinecker HC, Steff"en M, Witthoeft T et al Enhanced secretion of tumour necrosis factor-alpha, IL-6, and IL-1 beta by isolated lamina propria mononuclear cells from patients with ulcerative colitis and Crohn's disease. Clin Exp Immunol 1993; 94: 174-81. Grimm MC, Elsbury SK, Pavli P, Doe WF. Interleukin 8: cells of origin in inflammatory bowel disease. Gut 1996; 38: 90-8. Mahida YR, Wu K, Jewell DP. Enhanced production of interleukin 1-beta by mononuclear cells isolated from mucosa with active ulcerative cohtis of Crohn's disease. Gut 1989;30: 835-8. Rogler G, Andus T. Cytokines in inflammatory bowel disease. World J Surg. 1998; 22: 382-9. Gross V, Andus T, Leser HG, Roth M, Scholmerich J. Inflammatory mediators in chronic inflammatory bowel diseases.. Klin Wochenschr. 1991; 69: 981-7. Andus T, Gross V, Casar I et al. Activation of monocytes during inflammatory bowel disease. Pathobiology 1991; 59: 166-70. Styhanou E, Nie M, Ueda A, Zhao L. c-Rel and p65 transactivate the monocyte chemoattractant protein-1 gene in interleukin-1 stimulated mesangial cefls. Kidney Int 1999; 56: 873-82. Vincenti MP, Coon CI, Brinckerhoff" CE. Nuclear factor kappaB/p50 activates an element in the distal matrix metalloproteinase 1 promoter in interleukin-1 beta-stimu-

545

179.

180. 181. 182. 183.

184.

185. 186.

187. 188. 189. 190. 191. 192.

193. 194. 195. 196.

197.

198.

lated synovial fibroblasts. Arthritis Rheum 1998; 41: 198794. Martin T, Cardarelli PM, Parry GC, Felts KA, Cobb RR. Cytokine induction of monocyte chemoattractant protein-1 gene expression in human endothelial cells depends on the cooperative action of NF-kappa B and AP-1. Eur J Immunol 1997; 27: 1091-7. Mori N, Prager D. Transactivation of the interleukin-1 alpha promoter by human T-cell leukemia virus type I and type II Tax proteins. Blood 1996; 87: 3410-17. Parry GC, Mackman N. Transcriptional regulation of tissue factor expression in human endothelial cells. Arterioscler Thromb Vase Biol 1995; 15: 612-21. Kunsch C, Lang RK, Rosen CA, Shannon MF. Synergistic transcriptional activation of the IL-8 gene by NF-kappa B p65 (RelA) and NF-IL-6. J Immunol 1994; 153: 153-64. Dunn SM, Coles LS, Lang RK, Gerondakis S, Vadas MA, Shannon MF. Requirement for nuclear factor (NF)-kappa B p65 and NF-interleukin-6 binding elements in the tumor necrosis factor response region of the granulocyte colonystimulating factor promoter. Blood 1994; 83: 2469-79. Hiscott J, Marois J, Garoufahs J et al. Characterization of a functional NF-kappa B site in the human interleukin 1 beta promoter: evidence for a positive autoregulatory loop. Mol CeflBiol 1993; 13:6231-40. Kunsch C, Rosen CA. NF-kappa B subunit-specific regulation of the interleukin-8 promoter. Mol Cell Biol 1993; 13: 6137^6. Roebuck KA, Carpenter LR, Lakshminarayanan V, Page SM, Moy JN, Thomas LL. Stimulus-specific regulation of chemokine expression involves differential activation of the redox-responsive transcription factors AP-1 and NF-kappaB. J Leukoc Biol 1999; 65: 291-8. Baeuerle PA, Baltimore D. NF-kappa B: ten years after. Cell 1996;87:13-20. McKay DM, Perdue MH. Intestinal epithelial function: the case for immunophysiological regulation. Cells and mediators (1). Dig Dis Sci 1993; 38: 1377-87. McKay DM, Perdue MH. Intestinal epithelial function: the case for immunophysiological regulation. Implications for disease (2). Dig Dis Sci 1993; 38: 1735-45. Perdue MH, McKay DM. Integrative immunophysiology in the intestinal mucosa. Am J Physiol 1994; 267: G151-65. Selby WS, Janossy G, Mason DY, Jewell DP. Expression of HLA-DR antigens by colonic epithelium in inflammatory bowel disease. Clin Exp Immunol 1983; 53: 614-18. Bland PW, Whiting CV. Induction of MHC class II gene products in rat intestinal epithelium during graft-versushost disease and effects on the immune function of the epithelium. Immunology 1992; 75: 366-71. Bland PW, Warren LG. Antigen presentation by epithelial cells of the rat smafl intestine. II. Selective induction of suppressor T cells. Immunology 1986; 58: 9-14. Bland PW, Warren LG. Antigen presentation by epithelial cells of the rat small intestine. I. Kinetics, antigen specificity and blocking by anti-la antisera. Immunology 1986; 58: 1-7. Mayer L, Shhen R. Evidence for function of la molecules on gut epithehal cells in man. J Exp Med 1987; 166: 1471-83. Panja A, Barone A, Mayer L. Stimulation of lamina propria lymphocytes by intestinal epithelial cells: evidence for recognition of nonclassical restriction elements. J Exp Med 1994; 179: 943-50. Jung HC, Eckmann L, Yang SK et al. A distinct array of proinflammatory cytokines is expressed in human colon epithelial cells in response to bacterial invasion. J Clin Invest 1995; 95: 55-65. Eckmann L, Reed SL, Smith JR, Kagnoff" MF. Entamoeba histolytica trophozoites induce an inflammatory cytokine response by cultured human cells through the paracrine

546

199.

200.

201.

202. 203.

204.

205.

206.

207.

208. 209. 210.

211.

212. 213.

214. 215. 216.

Multi-site therapeutic modalities for inflammatory bowel diseases action of cytolytically released interleukin-1 alpha. J Clin Invest 1995; 96: 1269-79. Gibson P, Rosella O. Interleukin 8 secretion by colonic crypt cells in vitro: response to injury suppressed by butyrate and enhanced in inflammatory bowel disease. Gut 1995; 37: 536-43. Daig R, Rogler G, Aschenbrenner E et al. Human intestinal epithelial cells secrete interleukin-1 receptor antagonist and interleukin-8 but not interleukin-1 or interleukin-6. Gut 2000; 46: 350-8. Shirota K, LeDuy L, Yuan SY, Jothy S. Interleukin-6 and its receptor are expressed in human intestinal epithelial cells. Virchows Arch B Cell Pathol Incl Mol Pathol 1990; 58: 3038. Bocker U, Damiao A, Holt L ctal. Differential expression of interleukin 1 receptor antagonist isoforms in human intestinal epithelial cells. Gastroenterology 1998; 115: 1426 38. Warhurst AC, Hopkins SJ, Warhurst G. Interferon gamma induces differential upregulation of alpha and beta chemokine secretion in colonic epithelial cell lines. Gut 1998; 42: 208 13. Yang SK, Eckmann L, Panja A, Kagnoff MP. Differential and regulated expression of C-X-C, C-C, and C-chemokincs by human colon epithelial cells. Gastroenterology 1997; 113: 1214 23. Eckmann L, Jung HC, Schurer-Maly C, Panja A, Morzycka-Wroblcwska E, Kagnoff MF. Differential cytokine expression by human intestinal epithelial cell lines: regulated expression of interleukin 8. Gastroenterology 1993; 105: 1689 97. Parikh AA, Salzman AL, Kane CD, Fischer JE, Hasselgren PO. IL-6 production in human intestinal epithelial cells following stimulation with IL-1 beta is associated with activation of the transcription factor NF-kappa B. J Surg Res 1997; 69: 139 44. Jobin C, Panja A, Hellerbrand C et ^//.Inhibition of proinflammatory molecule production by adenovirus- mediated expression of a nuclear factor kappaB super-repressor in human intestinal epithelial cells. J Immunol 1998; 160: 410 18. Rogler G, Brand K, Vogl D et al. Nuclear factor kappaB is activated in macrophages and epithelial cells of inflamed intestinal mucosa. Gastroenterology 1998; 115: 357 69. Schreiber S, Nikolaus S, Hampe J. Activation of nuclear factor kappa B inflammatory bowel disease. Gut 1998; 42: 477-84. Cavicchi M, Whittle BJ. Regulation of induction of nitric oxide synthase and the inhibitory actions of dexamethasone in the human intestinal epithelial cell line, Caco-2: influence of cell differentiation. Br J Pharmacol 1999; 128: 705-15. Zareie M, Brattsand R, Sherman PM, McKay DM, Perdue MH. Improved effects of novel glucocorticosteroids on immune-induced epithelial pathophysiology. J Pharmacol ExpTher 1999; 289: 1245-9. Willoughby CP, Piris J, Truelove SC. Tissue eosinophils in ulcerative colitis. Scand J Gastroenterol 1979; 14: 395-9. Dvorak AM, Monahan RA, Osage JE, Dickersin GR. Crohn's disease: transmission electron microscopic studies. II. Immunologic inflammatory response. Alterations of mast cells, basophils, eosinophils, and the microvasculature. Hum Pathol 1980; 11: 606-19. Bischoff SC, Wedemeyer J, Herrmann A et al. Quantitative assessment of intestinal eosinophils and mast cells in inflammatory bowel disease. Histopathology 1996; 28: 1-13. Geboes K. From inflammation to lesion. Acta Gastroenterol Belg 1994;57:273-84. Sarin SK, Malhotra V, Sen Gupta S, Karol A, Gaur SK, Anand BS. Significance of eosinophil and mast cell counts

217. 218. 219. 220.

221.

222.

223. 224. 225. 226.

227. 228. 229.

230.

231. 232.

233. 234. 235. 236. 237.

in rectal mucosa in ulcerative colitis. A prospective controlled study. Dig Dis Sci 1987; 32: 363-7. Choy MY, Walker-Smith JA, Williams CB, MacDonald TT. Activated eosinophils in chronic inflammatory bowel disease. Lancet 1990; 336: 126-7. Seegert D, Rosenstiel P, Pfahler H, Pfefferkorn P, Nikolaus S, Schreiber S. Increased expression of IL-16 in inflammatory bowel disease. Gut 2001; 48: 326-32. Louahed J, Zhou Y, Maloy WL etal. Interleukin 9 promotes influx and local maturation of eosinophils. Blood 2001; 97: 1035-42. Jinquan T, Jing C, Jacobi HH et al. CXCR3 expression and activation of eosinophils: role of IFN-gamma-inducible protein-10 and monokine induced by IFN-gamma. J Immunol 2000; 165: 1548-56. Schwingshackl A, Duszyk M, Brown N, Moqbel R. Human eosinophils release matrix metalloproteinase-9 on stimulation with TNF-alpha. J Allergy Clin Immunol 1999; 104: 983 9. Velazquez JR, Lacy P, Mahmudi-Azer S, Moqbel R. Effects of interferon-gamma on mobilization and release of eosinophil- derived RANTES. Int Arch Allergy Immunol 1999; 118:447 9. Miyamasu M, Hirai K, Takahashi Y et al. Chemotactic agonists induce cytokine generation in eosinophils. J Immunol 1995; 154: 1339 49. Hansel TT, Braun RK, De Vrics IJ et al. Eosinophils and cytokines. Agents Actions Suppl. 1993; 43: 197 208 Yamashita N, Koizumi H, Murata M, Mano K, Ohta K. Nuclear factor kappa B mediates interlcukin-8 production in eosinophils. Int Arch Allergy Immunol 1999; 120: 230 6. Jang AS, Choi IS, Koh YI et al. Effects of prednisolone on eosinophils, IL-5, eosinophil cationic protein, EG2+ eosinophils, and nitric oxide metabolites in the sputum of patients with exacerbated asthma. J Korean Med Sci 2000; 15: 521 8. Miyamasu M, Misaki Y, Izumi S et al. Glucocorticoids inhibit chcmokine generation by human eosinophils. J Allergy Clin Immunol 1998; 101: 75 83. Shidham VB, Swami VK. Evaluation of apoptotic leukocytes in peripheral blood smears. Arch Pathol Lab Med 2000; 124: 1291-4. Arai Y, Nakamura Y, Inoue F, Yamamoto K, Saito K, Furusawa S. Glucocorticoid-induced apoptotic pathways in eosinophils: comparison with glucocorticoid-sensitive leukemia cells. Int J Hematol 2000; 71: 340-9. Nittoh T, Fujimori H, Kozumi Y, Ishihara K, Mue S, Ohuchi K. Effects of glucocorticoids on apoptosis of infiltrated eosinophils and neutrophils in rats. Eur J Pharmacol 1998;354:73-81. Walsh GM. Mechanisms of human eosinophil survival and apoptosis. Clin Exp Allergy 1997; 27: 482-7. Adachi T, Motojima S, Hirata A et al. Eosinophil apoptosis caused by theophylline, glucocorticoids, and macrolides after stimulation with IL-5. J Allergy Clin Immunol 1996; 98:S207-15. Schleimer RP, Bochner BS. The effects of glucocorticoids on human eosinophils. J Allergy Clin Immunol 1994; 94: 120213. Jaye DL, Parkos CA. Neutrophil migration across intestinal epithelium. Ann NY Acad Sci 2000; 915: 151-61. Kim B, Barnett JL, Kleer CG, Appelman HD. Endoscopic and histological patchiness in treated ulcerative colitis. Am J Gastroenterol 1999; 94: 3258-62. Bartram CI, Talbot IC. Colonoscopic biopsies in colitis. Abdom Imaging 1995; 20: 384-6. Mitsuyama K, Toyonaga A, Sasaki E et al. IL-8 as an important chemoattractant for neutrophils in ulcerative

Gerhard Rogler

238. 239.

240.

241. 242.

243. 244.

245.

246. 247.

248. 249. 250. 251.

252. 253.

254.

255. 256.

colitis and Crohn's disease. Clin Exp Immunol 1994; 96: 432-6. Grisham MB, Granger DN. Neutrophil-mediated mucosal injury. Role of reactive oxygen metabolites. Dig Dis Sci 1988;33:6S-15S. Lewis DC, Walker-Smith JA, Phillips AD. Polymorphonuclear neutrophil leucocytes in childhood Crohn's disease: a morphological study. J Pediatr Gastroenterol Nutr 1987; 6: 430-8. Kane SP, Vincenti AC. Mucosal enzymes in human inflammatory bowel disease with reference to neutrophil granulocytes as mediators of tissue injury. Clin Sci (Colch) 1979; 57: 295-303. Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: mechanisms of action and clinical considerations. Ann Intern Med 1976; 84: 304-15. van der Saag PT, Caldenhoven E, van de Stolpe A. Molecular mechanisms of steroid action: a novel type of cross-talk between glucocorticoids and NF-kappa B transcription factors. Eur Respir J Suppl. 1996; 22: 146-53s. Wheller SK, Perretti M. Dexamethasone inhibits cytokineinduced intercellular adhesion molecule-1 up-regulation on endothehal cell lines. Eur J Pharmacol 1997; 331: 65-71. Wissink S, van de Stolpe A, Caldenhoven E, Koenderman L, van der Saag PT. NF-kappa B/Rel family members regulating the ICAM-1 promoter in monocytic THP-1 cells. Immunobiology 1997; 198: 50-64. Tessier P, Audette M, Cattaruzzi P, McColl SR. Up-regulation by tumor necrosis factor alpha of intercellular adhesion molecule 1 expression and function in synovial fibroblasts and its inhibition by glucocorticoids. Arthritis Rheum 1993; 36:1528-39. Fattal-German M, Ladurie FL, Cerrina J, Lecerf F, BerrihAknin S. Modulation of ICAM-1 expression in human alveolar macrophages in vitro. Eur Respir J 1996; 9: 463-71. Atsuta J, Plitt J, Bochner BS, Schleimer RP. Inhibition of VCAM-1 expression in human bronchial epithelial cells by glucocorticoids. Am J Respir Cell Mol Biol 1999; 20: 64350. Summers RW, Switz DM, Sessions JT, Jr et al. National Cooperative Crohn's Disease Study: results of drug treatment. Gastroenterology 1979; 77: 847-69. Jue DM, Jeon KI, Jeong JY. Nuclear factor kappaB (NFkappaB) pathway as a therapeutic target in rheumatoid arthritis. J Korean Med Sci 1999; 14: 231-8. MacDermott RP. Progress in understanding the mechanisms of action of 5-aminosalicylic acid. Am J Gastroenterol 2000; 95: 3343-5. Nikolaus S, Folsch U, Schreiber S. Immunopharmacology of 5-aminosalicylic acid and of glucocorticoids in the therapy of inflammatory bowel disease. Hepatogastroenterology 2000; 47: 71-82. Wahl C, Liptay S, Adler G, Schmid RM. Sulfasalazine: a potent and specific inhibitor of nuclear factor kappa B. J Clin Invest 1998; 101: 1163-74. Liptay S, Bachem M, Hacker G, Adler G, Debatin KM, Schmid RM. Inhibition of nuclear factor kappa B and induction of apoptosis in T- lymphocytes by sulfasalazine. Br J Pharmacol 1999; 128: 1361-9. Weber CK, Liptay S, Wirth T, Adler G, Schmid RM. Suppression of NF-kappaB activity by sulfasalazine is mediated by direct inhibition of IkappaB kinases alpha and beta. Gastroenterology 2000; 119: 1209-18. Munkholm P, Langholz E, Davidsen M, Binder V. Frequency of glucocorticoid resistance and dependency in Crohn's disease. Gut 1994; 35: 360-2. Munkholm P, Langholz E, Davidsen M, Binder V. Disease activity courses in a regional cohort of Crohn's disease patients. Scand J Gastroenterol 1995; 30: 699-706.

547 257. Chinetti G, Fruchart JC, Staels B. Peroxisome proliferatoractivated receptors (PPARs): nuclear receptors at the crossroads between lipid metabolism and inflammation. Inflamm Res 2000; 49: 497-505. 258. Fajas L, Auboeuf D, Raspe E et al. The organization, promoter analysis, and expression of the human PPARgamma gene. J Biol Chem 1997; 272: 18779-89. 259. Kliewer SA, WiUson TM. The nuclear receptor PPARgamma - bigger than fat. Curr Opin Genet Dev 1998; 8: 576-81. 260. Kliewer SA, Forman BM, Blumberg B et al. Differential expression and activation of a family of murine peroxisome proliferator-activated receptors. Proc Natl Acad Sci USA 1994;91:7355-9. 261. Lemberger T, Braissant O, Juge-Aubry C et al. PPAR tissue distribution and interactions with other hormone-signaling pathways. Ann NY Acad Sci 1996; 804: 231-51. 262. Tontonoz P, Hu E, Graves RA, Budavari Al, Spiegelman BM. mPPAR gamma 2: tissue-specific regulator of an adipocyte enhancer. Genes Dev 1994; 8: 1224-34. 263. Chinetti G, Griglio S, Antonucci M et al. Activation of proliferator-activated receptors alpha and gamma induces apoptosis of human monocyte-derived macrophages. J Biol Chem 1998; 273: 25573-80. 264. Wiesenberg I, Chiesi M, Missbach M, Spanka C, Pignat W, Carlberg C. Specific activation of the nuclear receptors PPARgamma and RORA by the antidiabetic thiazolidinedione BRL 49653 and the antiarthritic thiazolidinedione derivative CGP 52608. Mol Pharmacol 1998; 53: 1131-8. 265. Murphy GJ, Holder JC. PPAR-gamma agonists: therapeutic role in diabetes, inflammation and cancer. Trends Pharmacol Sci 2000; 21: 469-74. 266. Fujiwara T, Horikoshi H. Troglitazone and related compounds: therapeutic potential beyond diabetes. Life Sci 2000; 67: 2405-16. 267. Spiegelman BM. PPAR-gamma: adipogenic regulator and thiazoHdinedione receptor. Diabetes 1998; 47: 507-14. 268. Berger J, Bailey P, Biswas C et al. Thiazolidinediones produce a conformational change in peroxisomal proliferator-activated receptor-gamma: binding and activation corr e l a t e with a n t i d i a b e t i c a c t i o n s in d b / d b mice. Endocrinology 1996; 137: 4189-95. 269. Pershadsingh HA. Pharmacological peroxisome proliferator-activated receptorgamma ligands: emerging clinical indications beyond diabetes. Expert Opin Investig Drugs 1999; 8:1859-72. 270. Willson TM, Lehmann JM, Kliewer SA. Discovery of ligands for the nuclear peroxisome proliferator-activated receptors. Ann NY Acad Sci 1996; 804: 276-83. 271. Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM, Kliewer SA. An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptor gamma (PPAR gamma). J Biol Chem 1995; 270: 12953-6. 272. Willson TM, Wahli W. Peroxisome proliferator-activated receptor agonists. Curr Opin Chem Biol 1997; 1: 235-41. 273. De Vos P, Lefebvre AM, Miller SG et al. ThiazoHdinediones repress ob gene expression in rodents via activation of peroxisome proliferator-activated receptor gamma. J Clin Invest 1996; 98: 1004-9. 274. Kliewer SA, Lenhard JM, Willson TM, Patel I, Morris DC, Lehmann JM. A prostaglandin J2 metabolite binds peroxisome proliferator-activated receptor gamma and promotes adipocyte differentiation. Cell 1995; 83: 813-19. 275. Kliewer SA, Sundseth SS, Jones SA et al. Fatty acids and eicosanoids regulate gene expression through direct interactions with peroxisome proliferator-activated receptors alpha and gamma. Proc Natl Acad Sci USA 1997; 94: 4318-23. 276. Forman BM, Tontonoz P, Chen J, Brun RP, Spiegelman BM, Evans RM. 15-Deoxy-delta 12, 14-prostaglandin J2 is

548

277. 278.

279.

280.

281.

282.

283.

284. 285. 286.

287. 288. 289. 290.

291. 292.

293.

Multi-site therapeutic modalities for inflammatory bowel diseases a ligand for the adipocyte determination factor PPAR gamma. Cell 1995; 83: 803-12. Forman BM, Chen J, Evans RM. The peroxisome proliferator-activated receptors: ligands and activators. Ann NY AcadSci 1996;804:266-75. Forman BM, Chen J, Evans RM. Hypolipidemic drugs, polyunsaturated fatty acids, and eicosanoids are ligands for peroxisome proliferator-activated receptors alpha and delta. Proc Natl Acad Sci USA 1997; 94: 4312-17. Krey G, Braissant O, L'Horset F et al. Fatty acids, eicosanoids, and hypolipidemic agents identified as ligands of peroxisome proliferator-activated receptors by coactivatordependent receptor ligand assay. Mol Endocrinol 1997; 11: 779 91. Urade Y, Ujihara M, Horiguchi Y, Ikai K, Hayaishi O. The major source of endogenous prostaglandin D2 production is likely antigen-presenting cells. Localization of glutathionerequiring prostaglandin D synthetase in histiocytes, dendritic, and Kupffcr cells in various rat tissues. J Immunol 1989; 143: 2982 9. Ricote M, Huang JT, Welch JS, Glass CK. The peroxisome proliferator-activated receptor (PPARgamma) as a regulator of monocyte/macrophage function. J Lcukoc Biol 1999; 66: 733 9. Ricote M, Huang J, Fajas L ct al. Expression of the peroxisome proliferator-activated receptor gamma (PPARgamma) in human atherosclerosis and regulation in macrophages by colony stimulating factors and oxidized low density lipoprotein. Proc Natl Acad Sci USA 1998; 95: 7614 19. Ricote M, Li AC, Willson TM, Kelly CJ, Glass CK. The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature 1998; 391:79 82. Jiang C, Ting AT, Seed B. PPAR-gamma agonists inhibit production of monocyte inflammatory cytokines. Nature 1998; 391: 82 6. Fajas L, Frucharl JC, Auwerx J. PPARgamma3 mRNA: a distinct PPARgamma mRNA subtype transcribed from an independent promoter. FEBS Lett 1998; 438: 55-60. DuBois RN, Gupta R, Brockman J, Reddy BS, Krakow SL, Lazar MA. The nuclear eicosanoid receptor, PPARgamma, is aberrantly expressed in colonic cancers. Carcinogenesis 1998; 19:49 53. Sarraf P, Mueller E, Jones D et al. Differentiation and reversal of malignant changes in colon cancer through PPARgamma. Nat Med 1998; 4: 1046-52. Sarraf P, Mueller E, Smith WM et al. Loss-of-function mutations in PPAR gamma associated with human colon cancer. Mol Cell 1999; 3: 799-804. Saez E, Tontonoz P, Nelson MC et al. Activators of the nuclear receptor PPARgamma enhance colon polyp formation. Nat Med 1998; 4: 1058-61. Lefebvre M, Paulweber B, Fajas L et al. Peroxisome proliferator-activated receptor gamma is induced during differentiation of colon epithelium cells. J Endocrinol 1999; 162: 331-40. Su CG, Wen X, Bailey ST et al. A novel therapy for colitis utilizing PPAR-gamma ligands to inhibit the epithelial inflammatory response. J Clin Invest 1999; 104: 383-9. Desreumaux P, Dubuquoy L, Nutten S et al. Attenuation of colon inflammation through activators of the retinoid X receptor (RXR)/peroxisome proliferator-activated receptor gamma (PPARgamma) heterodimer. A basis for new therapeutic strategies. J Exp Med 2001; 193: 827-38. Rohn TT, Wong SM, Cotman CW, Cribbs DH. 15-Deoxydeltal2,14-prostaglandin J2, a specific ligand for peroxisome proliferator-activated receptor-gamma, induces neuronal apoptosis. Neuroreport 2001; 12: 8 3 9 ^ 3 .

294. Marx J. Cancer research. Anti-inflammatories inhibit cancer growth - but how? Science 2001; 291: 581-2. 295. Tsubouchi Y, Sano H, Kawahito Y et al. Inhibition of human lung cancer cell growth by the peroxisome proliferator-activated receptor-gamma agonists through induction of apoptosis. Biochem Biophys Res Commun 2000; 270: 400-5. 296. Okura T, Nakamura M, Takata Y, Watanabe S, Kitami Y, Hiwada K. Troglitazone induces apoptosis via the p53 and Gadd45 pathway in vascular smooth muscle cells. Eur J Pharmacol 2000; 407: 227-35. 297. McCarty MF. Activation of PPARgamma may mediate a portion of the anticancer activity of conjugated linoleic acid. Med Hypotheses 2000; 55: 187-8. 298. Kawahito Y, Kondo M, Tsubouchi Y et al. 15-Deoxydelta(12,14)-PGJ(2) induces synoviocyte apoptosis and suppresses adjuvant-induced arthritis in rats. J Clin Invest 2000; 106:189 97. 299. Padilla J, Kaur K, Harris SG, Phipps R P PPAR-gammamediated regulation of normal and malignant B lineage cells. Ann NY Acad Sci 2000; 905: 97 109. 300. Miller RT, Anderson SP, Corton JC, Cattley RC. Apoptosis, mitosis and cyclophilin-40 expression in regressing peroxisome proliferator-induccd adenomas. Carcinogenesis 2000; 21:647 52. 301. Chang TH, Szabo E. Induction of difi'erentiation and apoptosis by ligands of peroxisome proliferator-activated receptor gamma in non-small cell lung cancer. Cancer Res 2000; 60: 1129 38. 302. Wu GD. A nuclear receptor to prevent colon cancer. N Engl J Med 2000; 342:651 3. 303. Takahashi N, Okumura T, Molomura W, Fujimoto Y, Kawabata I, Kohgo Y. Activation of PPARgamma inhibits cell growth and induces apoptosis in human gastric cancer cells. FEBS Lett 1999; 455: 135 9. 304. Krobolh PD, Salek FS, Pittenger AL, Fabian TJ, Frye R F DHEA and DHEA-S: a review. J Clin Pharmacol 1999; 39: 327 48. 305. Straub RH, Scholmerich J, Zietz B. Replacement therapy with DHEA plus corticosteroids in patients with chronic inflammatory diseases substitutes of adrenal and sex hormones. Z Rheumatol 2000; 59: 108 18. 306. Poynter ME, Daynes RA. Peroxisome proliferator-activated receptor alpha activation modulates cellular redox status, represses nuclear factor-kappaB signaling, and reduces inflammatory cytokine production in aging. J Biol Chem 1998;273:32833-41. 307. Poynter ME, Daynes RA. Age-associated alterations in splenic iNOS regulation: influence of constitutively expressed IFN-gamma and correction following supplementation with PPARa Ipha activators or vitamin E. Cell Immunol 1999; 195: 127-36. 308. Cheng GF, Tseng J. Regulation of murine interleukin-10 production by dehydroepiandrosterone. J Interferon Cytokine Res 2000; 20: 471-8. 309. van Vollenhoven RF, Morabito LM, Engleman EG, McGuire JL. Treatment of systemic lupus erythematosus with dehydroepiandrosterone: 50 patients treated up to 12 months. J Rheumatol 1998; 25: 285-9. 310. van Vollenhoven RF, Engleman EG, McGuire JL. Dehydroepiandrosterone in systemic lupus erythematosus. Results of a double-blind, placebo-controlled, randomized clinical trial. Arthritis Rheum 1995; 38: 1826-31. 311. van Voflenhoven RF, Engleman EG, McGuire JL. An open study of dehydroepiandrosterone in systemic lupus erythematosus. Arthritis Rheum 1994; 37: 1305-10. 312. van Vollenhoven RF, Park JL, Genovese MC, West JP, McGuire JL. A double-blind, placebo-controlled, clinical

Gerhard Rogler

313.

314.

315.

316.

317. 318. 319.

320. 321. 322. 323. 324. 325. 326. 327. 328. 329. 330. 331. 332.

trial of dehydroepiandrosterone in severe systemic lupus erythematosus. Lupus 1999; 8: 181-7 Straub RH, Vogl D, Gross V, Lang B, Scholmerich J, Andus T. Association of humoral markers of inflammation and dehydroepiandrosterone sulfate or Cortisol serum levels in patients with chronic inflammatory bowel disease. Am J Gastroenterol 1998; 93: 2197-202. de la Torre B, Hedman M, Befrits R. Blood and tissue dehydroepiandrosterone sulphate levels and their relationship to chronic inflammatory bowel disease. Clin Exp Rheumatol 1998; 16: 579-82. Andus T, Klebl F, Rogler G, Bregenzer N, Scholmerich J, Straub RH. Successful treatment of active Crohn's disease with oral dehydroepiandrosterone (DHEA): An open controlled pilot trial. Gastroenterology 2001; 120: A1440 Diep QN, Touyz RM, Schiffrin EL. Docosahexaenoic acid, a peroxisome proliferator-activated receptor- alpha ligand, induces apoptosis in vascular smooth muscle cells by stimulation of p38 mitogen-activated protein kinase. Hypertension 2000; 36: 851-5. Hugot JP, Chamaillard M, Zouali H et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 2001; 411: 599-603. Ogura Y, Bonen DK, Inohara N et al. A frameshift mutation in N 0 D 2 associated with susceptibility to Crohn's disease. Nature 2001; 411: 603-6. Ogura Y, Inohara N, Benito A, Chen FF, Yamaoka S, Nunez G. Nod2, a Nodl/Apaf-1 family member yhat is restricted to monocytes and activates NF-kappa B. J Biol Chem 2001; 276: 4812-18. Inohara N, Ogura Y, Chen FF, Muto A, Nunez G. Human nodi confers responsiveness to bacterial lipopolysaccharides. J Biol Chem 2001; 276: 2551-4. Inohara N, Koseki T, del Peso L et al. Nodi, an Apaf-1-like activator of caspase-9 and nuclear factor-kappaB. J Biol Chem 1999; 274: 14560-7. Lenardo MJ, Baltimore D. NF-kappa B: a pleiotropic mediator of inducible and tissue-specific gene control. Cell 1989; 58: 227-9. Baeuerle PA. The inducible transcription activator NFkappa B: regulation by distinct protein subunits. Biochim Biophys Acta 1991; 1072: 63-80. Grilli M, Chiu JJ, Lenardo MJ. NF-kappa B and Rel: participants in a multiform transcriptional regulatory system. Int Rev Cytol 1993; 143: 1-62. Baeuerle PA, Henkel T. Function and activation of NFkappa B in the immune system. Annu Rev Immunol 1994; 12: 141-79. Baeuerle PA, Baltimore D. I kappa B: a specific inhibitor of the NF-kappa B transcription factor. Science 1988; 242: 540-6. Cheng Q, Cant CA, Moll T et al. NK-kappa B subunitspecific regulation of the I kappa B alpha promoter. J Biol Chem 1994; 269: 13551-7. Huxford T, Malek S, Ghosh G. Structure and mechanism in NF-kappa B/I kappa B signaling. Cold Spring Harb Symp Quant Biol 1999; 64: 533-40. Muller JM, Ziegler-Heitbrock HW, Baeuerle PA. Nuclear factor kappa B, a mediator of lipopolysaccharide effects. Immunobiology 1993; 187: 233-56. Maniatis T. Catalysis by a multiprotein IkappaB kinase complex. Science 1997; 278: 818-19. Mercurio F, Zhu H, Murray BW et al. IKK-1 and IKK-2: cytokine-activated IkappaB kinases essential for NF-kappaB activation. Science 1997; 278: 860-6. Woronicz JD, Gao X, Cao Z, Rothe M, Goeddel DV. IkappaB kinase-beta: NF-kappaB activation and complex formation with IkappaB kinase-alpha and NIK. Science 1997; 278: 866-9.

549 333. Karin M, Delhase M. The I kappa B kinase (IKK) and NFkappa B: key elements of proinflammatory signalling. Semin Immunol 2000; 12: 85-98. 334. Neurath MF, Pettersson S, Meyer zum Buschenfelde KH, Strober W. Local administration of antisense phosphorothioate oligonucleotides to the p65 subunit of NF-kappa B abrogates established experimental colitis in mice. Nat Med 1996; 2: 998-1004. 335. Conner EM, Brand S, Davis JM et al. Proteasome inhibition attenuates nitric oxide synthase expression, VCAM- 1 transcription and the development of chronic colitis. J Pharmacol ExpTher 1997; 282: 1615-22. 336. Herfarth H, Brand K, Rath HC, Rogler G, Scholmerich J, Falk W. Nuclear factor-kappa B activity and intestinal inflammation in dextran sulphate sodium (DSS)-induced colitis in mice is suppressed by ghotoxin. Clin Exp Immunol 2000; 120: 59-65. 337. Jobin C, Hellerbrand C, Licato LL, Brenner DA, Sartor RB. Mediation by NF-kappa B of cytokine induced expression of intercellular adhesion molecule 1 (ICAM-1) in an intestinal epithelial cell line, a process blocked by proteasome inhibitors. Gut 1998; 42: 779-87. 338. Beutler B, Van Huffel C. An evolutionary and functional approach to the TNF receptor/ligand family. Ann NY Acad Sci 1994; 730: 118-33. 339. Stevens C, Walz G, Singaram C, et al. Tumor necrosis factoralpha, interleukin-1 beta, and interleukin-6 expression in inflammatory bowel disease. Dig Dis Sci 1992; 37: 818-26. 340. MacDonald TT, Hutchings P, Choy MY, Murch S, Cooke A. Tumour necrosis factor-alpha and interferon-gamma production measured at the single cell level in normal and inflamed human intestine. Clin Exp Immunol 1990; 81: 3015. 341. Pullman WE, Elsbury S, Kobayashi M, Hapel A J, Doe WF. Enhanced mucosal cytokine production in inflammatory bowel disease. Gastroenterology 1992; 102: 529-37. 342. Dinarello CA. The biological properties of interleukin-1. Eur Cytokine Netw 1994; 5: 517-31. 343. Dinarello CA. Role of pro- and anti-inflammatory cytokines during inflammation: experimental and clinical findings. J Biol Regul Homeost Agents 1997; 11: 91-103. 344. Youngman KR, Simon PL, West GA et al. LocaUzation of intestinal interleukin 1 activity and protein and gene expression to lamina propria cells. Gastroenterology 1993; 104: 749-58. 345. Andus T, Daig R, Vogl D et al. Imbalance of the interleukin 1 system in colonic mucosa - association with intestinal inflammation and interleukin 1 receptor antagonist genotype 2. Gut 1997; 41: 651-7. 346. Kishimoto T, Akira S, Narazaki M, Taga T. Interleukin-6 family of cytokines and gpl30. Blood 1995; 86: 1243-54. 347. Kusugami K, Fukatsu A, Tanimoto M et al. Elevation of interleukin-6 in inflammatory bowel disease is macrophageand epithelial cell-dependent. Dig Dis Sci 1995; 40: 949-59. 348. Daig R, Andus T, Aschenbrenner E, Falk W, Scholmerich J, Gross V. Increased interleukin 8 expression in the colon mucosa of patients with inflammatory bowel disease. Gut 1996; 38: 216-22. 349. Baggiolini M, Dewald B, Moser B. Interleukin-8 and related chemotactic cytokines - CXC and CC chemokines. Adv Immunol 1994;55:97-179. 350. Reinecker HC, Loh EY, Ringler DJ, Mehta A, Rombeau JL, MacDermott RP. Monocyte-chemoattractant protein 1 gene expression in intestinal epithelial cells and inflammatory bowel disease mucosa. Gastroenterology 1995; 108: 40-50. 351. lademarco MF, McQuillan JJ, Rosen GD, Dean DC. Characterization of the promoter for vascular cell adhesion molecule-1 (VCAM-1). J Biol Chem 1992; 267: 16323-9.

550

Multi-site therapeutic modalities for inflammatory bowel diseases

352. Neish AS, Williams AJ, Palmer HJ, Whitley MZ, Collins T. Functional analysis of the human vascular cell adhesion molecule 1 promoter. J Exp Med 1992; 176: 1583-93. 353. Murata T, Kurokawa R, Krones A et al. Defect of histone acetyltransferase activity of the nuclear transcriptional coactivator CBP in Rubinstein-Taybi syndrome. Hum Mol Genet 2001; 10: 1071-6. 354. Papadakis KA, Targan SR. Role of cytokines in the pathogenesis of inflammatory bowel disease. Annu Rev Med 2000;51:289-98. 355. Targan SR. Biology of inflammation in Crohn's disease: mechanisms of action of anti-TNF-a therapy. Can J Gastroenterol 2000; 14(Suppl. C): 13-16C. 356. Plevy SE, Landers CJ, Prehn J ct al. A role for TNF-alpha and mucosal T helper-1 cytokines in the pathogenesis of Crohn's disease. J Immunol 1997; 159: 6276 82. 357. Aggarwal BB. Tumour necrosis factors receptor associated signalling molecules and their role in activation of apoptosis, JNK and NF-kappaB. Ann Rheum Dis 2000; 59(Suppl. 1): 16 16. 358. Darnay BG, Aggarwal BB. Signal transduction by tumour necrosis factor and tumour necrosis factor related ligands and their receptors. Ann Rheum Dis 1999; 58(Suppl. I): 12 13. 359. Cao Z, Tanaka M, Regnier C cf al. NF-kappa B activation by tumor necrosis factor and intcrlcukin-1. Cold Spring Harb Symp Quant Biol 1999; 64: 473 83. 360. Chan KF, Sicgel MR, Lcnardo JM. Signaling by the TNF receptor supcrfamily and T cell homeostasis. Immunity 2000; 13:419 22. 361. Gravestein LA, Borst J. Tumor necrosis factor receptor family members in the immune system. Semin Immunol 1998; 10:423 34. 362. Malek NP, Pluempe J, Kubicka S, Manns MP,Trautwcin C. Molecular mechanisms of TNF receptor-mediated signaling. Recent Results Cancer Res 1998; 147: 97 106. 363. Lorcnz HM, Antoni C, Valerius T ct al. In vivo blockade of TNF-alpha by intravenous infusion of a chimeric monoclonal TNF-alpha antibody in patients with rheumatoid arthritis. Short term cellular and molecular effects. J Immunol 1996; 156: 1646 53. 364. Lorenz HM. TNF inhibitors in the treatment of arthritis. Curr Opin Investig Drugs 2000; 1: 188 93. 365. Maini RN, Elliott M, Brennan FM, Williams RO, Feldmann M. Targeting TNF alpha for the therapy of rheumatoid arthritis. Clin Exp Rheumatol 1994; 12(Suppl. 11): S63 6. 366. Maini RN, Elliott MJ, Brennan FM et al. Monoclonal antiTNF alpha antibody as a probe of pathogenesis and therapy of rheumatoid disease. Immunol Rev 1995; 144: 195-223. 367. Maini RN, Elliott MJ, Brennan FM, Feldmann M. Beneficial effects of tumour necrosis factor-alpha (TNF-alpha) blockade in rheumatoid arthritis (RA). Clin Exp Immunol 1995; 101:207-12. 368. Maini RN, Elliott M, Brennan FM, Williams RO, Feldmann M. TNF blockade in rheumatoid arthritis: implications for therapy and pathogenesis. Apmis 1997; 105: 257-63. 369. Maini RN, Taylor PC, Paleolog E et al. Insights into the pathogenesis of rheumatoid arthritis from application of anti-TNF therapy. Nihon Rinsho Meneki Gakkai Kaishi 2000; 23: 487-9. 370. van Dullemen HM, van Deventer SJ, Hommes DW et al. Treatment of Crohn's disease with anti-tumor necrosis factor chimeric monoclonal antibody (cA2). Gastroenterology 1995;109:129-35. 371. Lang KA, Peppercorn MA. Promising new agents for the treatment of inflammatory bowel disorders. Drugs RD 1999; 1:237-44.

372. Rutgeerts P, Baert F. New strategies in the management of inflammatory bowel disease. Acta Clin Belg 1999; 54: 27480. 373. Ricart E, Panaccione R, Loftus EV, Tremaine WJ, Sandborn WJ. Successful management of Crohn's disease of the ileoanal pouch with infliximab. Gastroenterology 1999; 117:429-32. 374. Sandborn WJ, Hanauer SB. Antitumor necrosis factor therapy for inflammatory bowel disease: a review of agents, pharmacology, clinical results, and safety. Inflamm Bowel Dis 1999; 5: 119-33. 375. Heresbach D, Semana G, Gosselin M, Bretagne MG. An immunomodulation strategy targeted towards immunocompetent cells or cytokines in inflammatory bowel diseases (IBD). Eur Cytokine Netw 1999; 10: 7-15. 376. Evans RC, Clarke L, Heath P, Stephens S, Morris Al, Rhodes JM. Treatment of ulcerative colitis with an engineered human anti-TNFalpha antibody CDP571. Aliment Pharmacol Ther 1997; 11: 1031 5. 377. van Deventer SJ, Camoglio L. Monoclonal antibody therapy of inflammatory bowel disease. Pharm World Sci 1997; 19: 55 9. 378. Schreiber S, Campicri M, Colombel JF et al. Use of antitumour necrosis factor agents in inflammatory bowel disease. European guidelines for 2001 2003. Inl J Colorectal Dis 2001; 16: 1 11; discussion 12 13. 379. Kam LY, Targan SR. TNF-alpha antagonists for the treatment of Crohn's disease. Expert Opin Pharmacother 2000; 1:615 22. 380. Mikula CA. Anti-TNF alpha: new therapy for Crohn's disease. Gastroenterol Nurs 1999; 22: 245 8. 381. Shanahan F. Anti-TNF therapy for Crohn's disease: a perspective (infliximab is not the drug we have been waiting for). Inflamm Bowel Dis 2000; 6: 137 39. 382. Rutgeerts PJ, Targan SR. Introduction: anti-TNF strategies in the treatment of Crohn's disease. Aliment Pharmacol Ther 1999; 13(Suppl. 4): 1. 383. van Deventer SJ. Anti-TNF antibody treatment of Crohn's disease. Ann Rheum Dis 1999; 58(Suppl. 1): II14 20. 384. D'Haens G, Van Deventer S, Van Hogczand R et al. Endoscopic and histological healing with infliximab antitumor necrosis factor antibodies in Crohn's disease: A European multicenter trial. Gastroenterology 1999; 116: 1029 34. 385. Baert FJ, Rutgeerts PR. Anti-TNF strategies in Crohn's disease: mechanisms, clinical effects, indications. Int J Colorectal Dis 1999; 14:47-51. 386. Hommes DW, van Dullemen HM, Levi M et al. Beneficial effect of treatment with a monoclonal anti-tumor necrosis factor-alpha antibody on markers of coagulation and fibrinolysis in patients with active Crohn's disease. Haemostasis 1997; 27: 269-77. 387. van Hogezand RA, Verspaget HW. The future role of antitumour necrosis factor-alpha products in the treatment of Crohn's disease. Drugs 1998; 56: 299-305. 388. van Hogezand RA, Verspaget HW. New therapies for inflammatory bowel disease: an update on chimeric antiTNF alpha antibodies and IL-10 therapy. Scand J Gastroenterol Suppl. 1997; 223: 105-7. 389. Kimbrell DA, Beutler B. The evolution and genetics of innate immunity. Nat Rev Genet 2001; 2: 256-67. 390. Daun JM, Fenton MJ. Interleukin-1/Toll receptor family members: receptor structure and signal transduction pathways. J Interferon Cytokine Res 2000; 20: 843-55. 391. O'Neill L. The ToU/interleukin-l receptor domain: a molecular switch for inflammation and host defence. Biochem Soc Trans 2000; 28: 557-63.

Gerhard Rogler 392. Imler JL, Hoffmann JA. Toll and Toll-like proteins: an ancient family of receptors signaling infection. Rev Immunogenet 2000; 2: 294-304 393. Kaisho T, Akira S. Critical roles of Toll-like receptors in host defense. Crit Rev Immunol 2000; 20: 393-405. 394. Beutler B, Poltorak A. Positional cloning of Lps, and the general role of toll-like receptors in the innate immune response. Eur Cytokine Netw 2000; 11: 143-52. 395. Means TK, Golenbock DT, Fenton MX The biology of ToUHke receptors. Cytokine Growth Factor Rev 2000; 11: 21932. 396. Beutler B. Endotoxin, toll-like receptor 4, and the afferent limb of innate immunity. Curr Opin Microbiol. 2000; 3: 2 3 8. 397. Beutler B. Tlr4: central component of the sole mammalian LPS sensor. Curr Opin Immunol 2000; 12: 20-6. 398. Qureshi ST, Gros P, Malo D. The Lps locus: genetic regulation of host responses to bacterial lipopolysaccharide. Inflamm Res 1999; 48: 613-20. 399. Hemmi H, Takeuchi O, Kawai T et al. A Toll-like receptor recognizes bacterial DNA. Nature 2000; 408: 740-5. 400. Du X, Poltorak A, Wei Y, Beutler B. Three novel mammalian toll-like receptors: gene structure, expression, and evolution. Eur Cytokine Netw 2000; 11: 362-71. 401. Beg AA, Sha WC, Bronson RT, Ghosh S, Baltimore D. Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature 1995; 376: 167-70. 402. Sha WC, Liou HC, Tuomanen EI, Baltimore D. Targeted disruption of the p50 subunit of NF-kappa B leads to multifocal defects in immune responses. Cell 1995; 80: 32130.

551 403. Xu YX, Pindolia KR, Janakiraman N, Chapman RA, Gautam SC. Curcumin inhibits ILl alpha and TNF-alpha induction of AP-1 and N F - K B DNA-binding activity in bone marrow stromal cells. Hematopathol Mol Hematol 1997; 11:49-62. 404. Plummer SM, Holloway KA, Manson MM et al. Inhibition of cyclo-oxygenase 2 expression in colon cells by the chemopreventive agent curcumin involves inhibition of NF-kappaB activation via the N I K / I K K signalhng complex. Oncogene 1999; 18: 6013-20. 405. Pan MH, Lin-Shiau SY, Lin JK. Comparative studies on the suppression of nitric oxide synthase by curcumin and its hydrogenated metaboHtes through down-regulation of IkappaB kinase and NFkappaB activation in macrophages. Biochem Pharmacol 2000; 60: 1665-76. 406. Kumar A, Dhawan S, Hardegen NJ, Aggarwal BB. Curcumin (Diferuloylmethane) inhibition of tumor necrosis factor (TNF)-mediated adhesion of monocytes to endothelial cells by suppression of cell surface expression of adhesion molecules and of nuclear factor-kappaB activation. Biochem Pharmacol 1998; 55: 775-83. 407. Jobin C, Bradham CA, Russo MP et al. Curcumin blocks cytokine-mediated NF-kappa B activation and proinflammatory gene expression by inhibiting inhibitory factor Ikappa B kinase activity. J Immunol 1999; 163: 3474-83. 408. Han SS, Chung ST, Robertson DA, Ranjan D, Bondada S. Curcumin causes the growth arrest and apoptosis of B cell lymphoma by downregulation of egr-1, c-myc, bcl-XL, NFkappa B, and p53. Clin Immunol 1999; 93: 152-61. 409. Brennan P, O'Neill LA. Inhibition of nuclear factor kappaB by direct modification in whole cells-mechanism of action of nordihydroguaiaritic acid, curcumin and thiol modifiers. Biochem Pharmacol 1998; 55: 965-73.

271 Targeted therapies for inflammatory bowel disease I SANDER J.H. VAN DEVENTER

Introduction Medical treatment of inflammatory bowel disease (IBD) has traditionally consisted of administration of anti-inflammatory drugs such as sulfasalazine, mesalazine, glucocorticosteroids, and more recently immunomodulating drugs including azathioprine, methotrexate, and cyclosporine. None of these drugs has been primarily designed for use in IBD, and for most the precise mechanisms of action are unknown. The development of drugs for IBD has long been hindered by the lack of suitable animal models and lack of knowledge of the regulation of (immunemediated) mucosal inflammation. In the past decade the cells involved in mucosal inflammation, and many of the proteins that lymphocytes, monocytes and antigen-presenting cells use to communicate, have been characterized. Mice that either lack or over-express these proteins often develop IBD, and this has not only greatly improved basic knowledge, but also provided elegant models for testing of therapeutic interventions. Concurrently, the rapid development of biotechnology has resulted in new classes of drugs that are now commonly known as 'biologicals'. The term 'biologicals' originally referred to a group of therapeutically administered endogenous proteins such as antibodies, cytokines, and naturally occurring cytokine-neutralizing proteins. However, most biologicals have been extensively modified to increase the therapeutic effects or to increase half-life, and compounds such as antisense oligonucleotides, that do not occur naturally, are usually also categorized as biologicals. Hence, a better operational definition for biological therapies is that these approaches all use peptides, proteins or nucleic acids in order to specifically target cell activation, or the consequences thereof. Biological therapies differ from traditional small molecules in several aspects. Because biological therapies specifically interfere with a single target, the clinical efficacy of such interventions directly reflects the pathogenic importance of such targets. As a result this

class of therapeutic agents has greatly contributed to current understanding of the pathogenesis of IBD. Current biotechnology allows rapid development of biologicals (antibodies, peptides, cytokines, antisense oligonucleotides) for clinical intervention studies, thereby greatly decreasing drug development time. Most of these compounds are highly speciesspecific and because extensive animal testing is usually not feasible, the path from bench to bedside becomes even shorter. A second line of targeted therapies has been a result of rapidly increased understanding of the pathways through which receptors on the cell membrane transmit signals to the nucleus, resulting in the transcriptional activation of specific genes (the signal transduction pathways). Specific inhibitors of components of these pathways have been generated, and some have anti-inflammatory properties in preclinical models. Finally, gene therapy aUows very specific manipulation of cell function, and although no gene therapy trials have been initiated in IBD, preclinical development has been initiated. In this chapter the fundamentals and therapeutic efficacy of targeted therapies are discussed. Although IBD has become a test bed for development of targeted therapies for chronic inflammatory diseases, and the number of approaches and clinical trials is overwhelming, it should be noted that only one of these approaches (treatment with TNF-aneutralizing antibodies) has at present been approved for use in Crohn's disease (CD).

Therapeutic targets in IBD The development of biological therapies has been facilitated by remarkable progress in the understanding of the pathogenesis of mucosal inflammatory mechanisms (Fig. 1). This has largely been a consequence of the availability of mouse models of IBD, most of which resulted from genetic targeting [1]. Many of the observations originally made in mice

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 553-571. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Targeted therapies for inflammatory bowel disease

554

antigen presenting cell

regulatory T cell

Lamina propria Figure 1. Crohn's disease is thought to result from uncontrolled immune activation following presentation of intestinal antigens by antigen-presenting cells. This requires the recognition of the antigen/HLA complex by the lymphocyte T cell receptor, the expression of membrane-bound expressed co-stimulatory molecules, and cytokines such as IL-12 and IL-18. Activated T lymphocytes produce proinflammatory cytokines that recruit and activate other immune cells. Normally, this reaction is controlled by induction of apoptosis of activated lymphocytes, and by regulatory T lymphocytes. Therapeutically, this inflammation can be controlled by blocking of cytokines (IL-12, IL-18, TNF-ot), by induction of apoptosis or by interfering with neutrophil of lymphocyte recruitment.

have subsequently been confirmed in patients with IBD. Based on these data it is now possible to construct models of the inflammatory mechanisms that define specific targets (Fig. 1). Important differences exist between the disease mechanisms in CD and ulcerative colitis (UC), and the development of biological therapies for these diseases may diverge considerably. Certain disease mechanisms in IBD are also found in other chronic inflammatory diseases, such as rheumatoid arthritis, multiple sclerosis, and psoriasis, and biologicals that are developed in parallel for these indications may prove to be effective in CD or UC. Finally, the fields of immune suppression and transplantation have identified molecular targets that are involved in cell activation or trafficking, enabling the development of therapeutic interventions that can be applied in IBD. In

general, three groups of therapeutic targets for biological therapies can be defined: neutralization of proinflammatory cytokines, application of antiinflammatory cytokines, and interference with cell activation or cell trafficking.

Neutralization of proinfiammatory cytokines In the final decades of the last century the pivotal insight emerged that inflammatory reactions are mediated by families of proteins with multiple, often redundant, autocrine, paracrine and endocrine activities, and these are now named cytokines. The two prototype pro-inflammatory cytokines are tumor necrosis factor a (TNF-a) and interleukin-ip (IL-

Sander J.H. van Deventer

Variable

Constant

555

wv

Fc

igGi

//

1/

1

•

Murine

1

1

•

Human

1

human/mouse chimeric

humanized

human receptor fusion protein

infliximab

CDP571

etanercept

Figure 2. Three different TNF-a-neutralizing molecules. CDP571 contains less mouse protein than infliximab and presumably is less immunogenic. Etanercept is not an antibody, but a fusion protein of two TNF-oc p75 receptors and an lgG1 tail.

1P) (for extensive reviews of their biological activities see refs 2 and 3). In addition, a system of T lymphocyte- and natural killer cell-activation cytokines has been discovered, which links innate and adaptive immunity, and includes the cytokines IL-12, IL-18 and interferon-y (Fig. 3) [4]. Virtually all current experimental anticytokine approaches in IBD have targeted one of these cytokines.

Interleukin-1p IL-ip, the first endogenous pyrogen to be identified, is produced by professional inflammatory cells, but also by many parenchymal cells. Generation of the mature form of IL-1 p is dependent on cleavage of the IL-ip propeptide by IL-1 converting enzyme (ICE), a caspase that is also necessary for activation of IL18 [5]. IL-ip is recognized by a specific receptor, the IL-ip type I receptor, which is widely expressed, and signal transduction is dependent on the presence of a second membrane protein, the IL-1 receptor associated protein [6]. Signal transduction proceeds through activation of TRAF6 and several other components that are shared with the endotoxin and the IL-18 signal transduction pathways [7]. IL-ip induces a wide range of secondary cytokines and

chemokines, increases eicosanoid production, and up-regulates cell adhesion molecules and tissue factor (thereby activating the coagulation cascade). The potent inflammatory effects of IL-1 P potentially can damage the host, but are tightly controlled in vivo by a naturally occurring antagonist, IL-1 receptor antagonist (IL-IRA), by a membrane-bound decoy receptor (IL-IRII) and by soluble receptors (Fig. 2). Increased concentrations of IL-1 p are detected in the mucosa of patients with active CD and UC, and it has been suggested that the latter disease is characterized by an imbalance of IL-1 p/IL-IRA production [8-10]. This imbalance may be a result of a combination of genetic polymorphisms that are associated with a high IL-ip and a low IL-lRA production [11]. However, the specificity of the increased mucosal IL-ip/IL-lRA production ratio for UC has been questioned [12]. In a neutrophildependent model of IBD (immune complex colitis in rabbits), IL-IRA showed remarkable therapeutic efficacy, but no clinical trials in IBD have been reported [13]. IL-IRA can be produced in sufficient amounts for large-scale clinical use, and was tested for efficacy in three clinical trials in patients with sepsis syndrome. Although the drug was found to be safe, no effect on the primary outcome, 28-day all-

Targeted therapies for inflammatory bowel disease

556

IL-12

IFNy

TNFa

IL-18R

Figure 3. IL-12 and IL-18 both cause Th1 differentiation of CF4"*" T lymphocytes, eventually leading to increased production of IFN-y and TNF-a. IL-12 can be blocked using specific antibodies, whereas IL-18 is blocked both by antibodies and by a naturally occurring protein named IL-18 binding protein.

cause mortality, was observed. A clinical study in rheumatoid arthritis showed long-term efficacy, with a significant effect on the progression of joint erosions. Hence, despite its short half-life, systemically administered IL-IRA can modulate compartmentalized inflammatory disease in the long term. In view of these data, IL-lRA remains an attractive candidate drug for the treatment of IBD.

Tumor necrosis factor-a The tumor necrosis factor family of cytokines and receptors is involved in development of the immune system, inflammation, and apoptosis. Members of this family that have been impHcated in the pathogenesis of IBD include lymphotoxin-(3, fas and TNFa [14]. TNF-ot is a 17 kDa cytokine that is mainly produced by monocytes and lymphocytes upon a variety of stimuli. The two most important signal transduction pathways that lead to transcription of the TNF-oc gene involve activation of N F - K B and MAP kinases (see below). Several small molecules

influence transcription and translation of TNF-a by interfering with intracellular targets. An increase of the intracellular cAMP concentration by inhibition of phosphodiesterase reduces TNF-a transcription, and this mechanism causes the reduction of TNF-a production by pentoxifyllin and by thalidomide (analogs) [15-18]. TNF-a is translated as an immature protein with an unusually long signal peptide, which retains the protein within the extracellular membrane, where it forms homotrimers. Membrane-bound homotrimeric TNF-a is biologically active (mainly through interaction with the type II T N F receptor - see below) but can only exert paracrine effects. A specific metalloproteinase, TNF-a converting enzyme (TACE), cleaves the TNF molecule extracellularly, thereby liberating homotrimeric soluble TNF-a [19-22]. TNF-a is recognized by two receptors, TNFRI, which is expressed by a wide variety of cells, and is considered responsible for signaling inflammation and apoptosis, and TNFRII, which is mainly expressed by white blood cells [23, 24]. Signaling through TNFRI acti-

Sander J.H. van Deventer vates two distinct pathways, leading to inflammatory responses and apoptosis respectively (see Fig. 5). It is of importance to realize that induction of apoptosis is both activated (through activation of the death domain) and inhibited (by N F - K B ) by TNF-oc, and the net result depends on the inflammatory context and the type of cell studied [25]. The importance of TNF-a for chronic mucosal inflammation has been clearly established and is discussed in chapters 4, 5, and 7. For this reason TNF-oc has become a major therapeutic target (using small molecules, antibodies and designed proteins) inlBD. Small molecular TNF-oc inhibitors Phosphodiesterase inhibitors Methylxanthine compounds are vasodilators, inhibit platelet aggregation and thromboxane A2 synthesis, and were initially designed to treat intermittent claudication resulting from large vessel atherosclerosis. The best-known drug in this class is pentoxifylline, a non-specific phosphodiesterase inhibitor, that was subsequently found to diminish the production of proinflammatory cytokines in vitro as well as in vivo [18, 26]. This finding led to a series of clinical trials in patients with sepsis and chronic inflammatory diseases. In patients with active CD, oxpentifylline (a pentoxifylline derivative) reduced the production of TNF-a by peripheral blood mononuclear cells, but had no eff'ect on disease activity [27]. Rolipram is a specific inhibitor of the type 4 phosphodiesterase and inhibits cAMP-dependent signaling through cAMP-response element-binding factor (CREB) [15, 28]. This not only causes a reduction of the production of proinflammatory cytokines, but also reduces T lymphocyte chemotaxis. Rolipram is eff'ective in various animal models of chronic T lymphocyte-dependent inflammatory disease, such as adjuvant arthritis and multiple sclerosis, and was shown to reduce mucosal TNF-a production in dextran sulfate-induced colitis, thereby preventing tissue damage [29, 30]. The clinical efficacy of rolipram in IBD has not been investigated. Thalidomide Thalidomide was introduced in Europe as a sedative, but its use was soon abandoned because of teratogenicity. It has long been known that thalidomide has immunomodulatory effects, and thalidomide is used in erythema nodosum leprosum, a T lymphocytemediated complication of lepra that is characterized by high circulating concentrations of TNF-a and

557 IFN-y. More recently, thalidomide has been clinically used in rheumatoid arthritis, Behcet's disease, systemic lupus erythematosus, multiple myeloma, and therapy-refractory tuberculosis [31-35]. Oral ulcerations secondary to CD and those occurring in AIDS do respond favorably to thalidomide administration [36]. Two open-label trials have explored the potential efficacy of thahdomide in CD [37, 38]. Both included patients with chronic active, steroid-unresponsive disease, and patients received daily thalidomide doses ranging from 100 to 300 mg. Both studies reported a substantial reduction of the CD activity index, and one study suggested that thalidomide could have steroid-sparing effects [37]. The results of these two uncontrolled studies suggest that thalidomide may have beneficial effects in patients with chronic active CD. It remains to be seen whether this clinical effect is related to a reduction of TNF-a production. In previous studies in HIV-infected patients and patients with tuberculosis no inhibition of TNF-a production by thalidomide treatment has been observed [33]. In addition to inhibiting TNF-a, thahdomide also affects the production of IL-10 and IL-12, and activates T lymphocytes and down-regulates the expression of T lymphocyte chemokine receptors [16, 39-41]. Thalidomide analogs with more specific phosphodiesterase-inhibiting and T lymphocyte-stimulating properties have been synthesized, allowing a more specific targeting [42]. TACE inhibitors TNF-a converting enzyme (TACE) is a metalloproteinase that cleaves membrane-bound TNF between alanine at position 76 and valine at position 77, thereby releasing soluble trimeric TNF-a [19, 21, 22]. Several metalloproteinase inhibitors more or less specifically inhibit TACE, which results in a reduction of the release of soluble TNF-a, while the number of membrane-bound TNF-a molecules is unaff'ected. TACE inhibitors indeed inhibit the release of TNF-a in various mouse models of inffammatory disease, and one of these compounds reduced inflammatory tissue damage in TNBS-induced colitis in rats [43]. However, although in the latter model the severity of colitis was attenuated, this was not associated with a reduction of TNF-a production. Therefore, it is likely that the TACE inhibitor used in this experiment reduced inflammation by a nonspecific inhibition of metalloproteinases other than TACE. Most TACE inhibitors available not only interfere with cleavage of TNF-a, but also reduce the shedding of TNF receptors that occurs in acti-

558 vated mononuclear cells. Shedding of TNF receptors dampens the biological effects of TNF-oc, because of the reduced number of potential TNF-oc binding sites, and because soluble receptors retain the ability to bind and neutralize TNF-oc. It has been suggested that the reduced shedding of TNF receptors caused a paradoxical increase of disease activity in collageninduced arthritis in mice that were treated with a TACE inhibitor [44]. In endotoxin-challenged human volunteers, pretreatment with an oral TACE inhibitor significantly reduced the release of TNF-a, but also interfered with the shedding of TNF receptors [45, 46]. Results from clinical trials with TACE inhibitors have not been published. TNF-oc-binding molecules The development of TNF-binding molecules mainly resulted from the hypothesis that the mortality of sepsis was caused by the proinflammatory effects of TNF-oc. Indeed, administration of TNF-a neutralizing antibodies reduced mortality in certain (primate) sepsis models, particularly those induced by injection of large numbers of Gram-negative bacteria or endotoxin [47]. However, the clinical efficacy of TNF-a neutralizing antibodies in sepsis has been disappointing, and with a few exceptions drug development for this indication has been discontinued [48]. Nonetheless, this line of research has led to the development of two classes of TNF neutralizing proteins that are currently used in chronic inflammatory diseases. The first class includes TNF neutralizing antibodies. The first generation of these antibodies was induced in mice that were immunized with human TNF-a, resulting in the generation of hybridomas that produced murine high-affinity antihuman T N F - a antibodies. Because (repeated) administration of mouse antibodies in humans causes immunogenic responses, all currently available antibodies have been modified in order to reduce mouse-specific protein sequences. By 'grafting' the variable (antibody-binding) regions on a human IgG tail, human-mouse chimeric antibodies were generated, in which the percentage of mouse protein was reduced to 25-30% (Fig. 2) [49]. Human-mouse chimeric antibodies may induce antibody responses (human antichimeric antibodies: HACA), and because these antibodies are directed against the murine antigen-binding domains of the antibody, they frequently interfere with ligand-binding, and hence reduce the biological activity. A further refinement was the 'humanization' of antibodies, in which only the hypervariable (mouse-derived) complemen-

Targeted therapies for inflammatory bowel disease tary-determining regions (CDRs) of the antibody are inserted in a human antibody backbone. Such antibodies contain about 5% of mouse protein, and although experience is still limited, the available data suggest that humanization indeed reduces the incidence of immunogenic responses [50]. Finally, by several techniques, cytokine-neutralizing antibodies can be generated that consist entirely of human protein sequences. One of the major obstacles in this development is the difficulty of generating highaffinity antibodies. In addition, for reasons that are incompletely understood, some completely human antibodies are still immunogenic when administered to humans. Nonetheless, high-affinity human antiTNF-a antibodies are now available and being tested in clinical trials [51]. Apart from complete antibodies, F(ab)2 fragments have been used, which have a significantly shorter half-life, which may be enhanced by coupling to polyethylene glycol (PEG) [52, 53]. All TNF-a neutralizing antibodies that are currently developed for clinical use are of the IgG class. Human IgG i, but not IgG4, antibodies are able to bind and activate complement, and it has been reported that TNF-a/antibody immune complex formation and subsequent complement activation may cause febrile responses. However, no such reactions have been observed in the large number of patients treated with TNF-a neutralizing antibodies of the IgGi class. A second class of TNF-a neutralizing proteins consists of recombinant soluble (PEG-linked) receptors, or dimericTNF type I or II receptors which are fused on an IgGi antibody tail [54]. Clinical studies have been performed with the latter two constructs. In sepsis patients the dimeric TNFRl/IgGi construct did not significantly reduce mortality, and the TNFRII/IgGi construct (etanercept) caused a dosedependent increase in mortality. No definitive explanation for this finding has been provided, but it should be noted that TNF-a-binding proteins and antibodies have the ability to retain TNF-a within the circulation, and may even shuttle compartmentalized TNF-a form tissue sites into the blood. As long as the bound TNF-a is effectively neutralized, this does not cause adverse effects. Although the TNFRII/IgGi construct has a high TNF-a binding affinity, compared to TNRI/IgGi constructs or TNF-a-neutralizing antibodies, it has a high 'off' rate, which implies that bound ligand (TNF-a) is relatively easily released. By such a mechanism TNFRII/IgGi constructs may act as a TNF 'sponge', and this may explain the deleterious effect of the use

Sander J.H. van Deventer of such constructs in sepsis patients. Etanercept has been administered to a large number of rheumatoid arthritis patients without causing adverse effects, but a small number of treated patients developed severe sepsis. Because all currently reported clinical trials in IBD have used anti-TNF-a antibodies, the further discussion is restricted to this class of TNF-a binding proteins. CDP571 CDP571 is a humanized high-affinity IgG^ antibody that has been used in clinical trials in rheumatoid arthritis patients, in UC and in CD. Most reported data concern short-term studies, and the antibody has been demonstrated to be safe and well tolerated. In a small trial in patients with active CD a single CDP571 dose (5 mg/kg) reduced the median Crohn Disease Activity Index (CDAI) from 263 to 167 and 29% of patients were in remission 2 weeks after the start of treatment [55]. At a larger dose (10 or 20 mg/ kg every 8-12 weeks) CDP517 induced a response in about 45% of patients with active CD (27%o in placebo-treated patients) that seemed to be maintained by repeated infusions [56]. Interestingly, patients treated with the lower 10 mg/kg dose with the longer interval (12 weeks) showed the best responses. A third trial investigated the potential steroid-sparing effects of CDP571. Patients with CD in remission (CDAI < 150), who were using corticosteroids and had previously failed an attempt to taper the steroid dose, were treated with CDP571 (an induction dose of 20 mg/kg, followed by 10 mg/kg at week 8) or placebo. After 40 weeks, in 44% of CDP571-treated patients, but in only 22% of placebo-treated patients, corticosteroids could be successfully discontinued [57]. Hence, CDP571 is effective in chronic steroid-dependent CD. Infliximab Infliximab is a first-generation human-mouse chimeric antibody, with a high affinity for human TNF-a. The antibody binds soluble as well as membrane-bound TNF-a and potently interfered with the biological activity of TNF-a in vitro. The rationale for the use of anti-TNF-a antibodies in chronic inflammatory disease came from pivotal studies in collagen-induced arthritis in mice and by using human synovial explants obtained from patients with rheumatoid arthritis. Subsequent clinical studies demonstrated that intravenous administration of infliximab (10 or 20 mg/kg) to patients with severe therapy-refractory rheumatoid arthritis

559 resulted in dramatic and rather prolonged (weeks) clinical responses. These responses were associated with a rapid reduction of C-reactive protein serum concentrations and of the erythrocyte sedimentation rate, and with a marked functional improvement. It has been subsequently reported that these clinical responses can be maintained by repeated infliximab infusions (every 4 or 8 weeks), which were generally well tolerated. Importantly, long-term treatment with infliximab and etanercept reduces radiological progression of joint destruction. Hence, in rheumatoid arthritis, TNF-a neutralizing proteins are not merely anti-inflammatory, but also modify the natural course of the disease. The (pre)clinical development of infliximab for sepsis, as well as the initial results in rheumatoid arthritis patients, revealed that the drug was well tolerated and no serious sideeffects were observed. The first clinical experience with infliximab in CD was obtained in a child with therapy-refractory Crohn's colitis associated with severe perianal disease and a remarkable clinical and endoscopic response to infliximab infusion was reported (two 10 mg/kg doses were infused within a 2-week period) [58]. Similar findings were reported in an open-label study including 10 patients who received a single dose of 10 or 20 mg/kg [59]. In this study rapid endoscopic improvement was observed in several patients, C-reactive protein concentrations decreased, as well as mucosal concentrations of secretory phospholipase A2. In addition, it was demonstrated that the increased intravascular generation of thrombin, which occurs in active CD, was rapidly normalized by infliximab treatment. Because the proinflammatory effects of thrombin are well known, this latter effect is considered to add to the anti-inflammatory effects of TNF-a neutralization. Two controlled clinical trials have reported the efficacy of infliximab in chronic active CD and in patients with fistulas. The first study included 108 patients with active CD who were randomised to receive a single infusion of infliximab (either 5, 10, or 20 mg/kg) or placebo [60]. All included patients had failed prior medical treatment, and most were treated with corticosteroids or immunomodulatory drugs at enrolment. Four weeks after the infusion a clinical response, defined as a reduction of the CDAI by more than 70 points, was observed in 65%o of aU infliximab-treated patients combined, as opposed to 17% of placebo-treated patients. Moreover, in about one-third of infliximab-treated patients a complete clinical remission was induced. Non-responding patients received an open-label infusion (10 mg/kg)

560 and all responding patients were re-randomized to receive 8-weekly infusions of infliximab (10 mg/kg) or placebo, for a total of four infusions. Analysis of the second part of this study revealed that the therapeutic response was gradually lost in placebotreated patients, but therapeutic responses and remissions were well maintained by repeated infliximab infusions [61]. Remarkably, clinical responses following administration of infliximab correlate well to endoscopic responses, and in some patients all mucosal lesions disappear [62]. Because healing of perianal fistulas was observed in infliximab-treated patients, a second controlled trial in patients with enterocutaneous or perianal fistulas was undertaken. Although previous uncontrolled trials had suggested that antibiotics, 6mercaptopurine (6-MP) and cyclosporine were useful in the treatment of perianal fistulas, this trial represented the first controlled clinical trial in CD patients with fistulas. Administration of infliximab at three time points (0, 2, and 6 weeks) resulted in the closure of all fistulas in 46% of treated patients, as compared to 13% of patients who received placebo infusions in addition to standard medical therapy [63]. The median time to closure was 14 days in infliximab-treated patients (more than 40 days in the placebo group) and the duration of the response was about 3 months. These two controlled trials clearly demonstrated the efficacy of infliximab in chronic active (therapy-refractory) CD and in fistulas secondary to CD, and were the basis of the approval of the drug for these indications both in North America and Europe. Subsequently, a dose-escalating clinical trial in pediatric patients suggested that the efficacy of treatment of children was similar to that in adults. Systemic complications of CD, such as spondyloarthritis and erythema nodosum, as well as 'metastatic' CD, have been reported to respond favorably to infliximab treatment, and CD in ileoanal reservoirs has also been successfully treated [64-66]. The experience in UC is very limited, but some patients may show therapeutic responses [67]. A large controlled clinical trial has been initiated. Side-effects of anti-TNF-oc antibodies Infliximab treatment is associated with several sideeff'ects. As expected, some patients (about 13%) develop HACA, which usually are low titer and do not necessarily preclude future treatment. Concurrent treatment with corticosteroids or immunosuppressive drugs reduces the incidence of HACA

Targeted therapies for inflammatory bowel disease induction (unpublished data). About 25% of patients who received infliximab infusions after having being treated in the initial clinical trials developed a clinical reaction characterized by fever and severe muscle and joint tenderness, which is now referred to as 'delayed hypersensitivity reaction'. These reactions are caused by a rapid induction of very high titers of HACA, even in patients who had not shown any HACA response either during the initial treatment or shortly before the repeat infusion. Delayed hypersensitivity reactions occur only after a long 'drug holiday', and have not been observed in patients who were repeatedly infused at 4-12-week intervals. Patients with delayed hypersensitivity reactions do not develop end-organ damage and respond well to treatment with corticosteroids or NSAID. Nonetheless, such patients should not receive subsequent infliximab infusions, because the high-titer HACA interfere with the TNF-oc-binding capacity of infliximab. Non-HACA-related infusion reactions include headache, (localized) urticaria, and wheezing, all occurring during or rapidly after the infusion. These reactions can be managed by reducing the infusion rate (the 5 mg/kg dose is usually administered within 2 h), and by administration of corticosteroids or antihistamines. Occasionally skin rashes occur in infliximab-treated patients and about 9% of all infused patients (transiently) develop anti-dsDNA antibodies, which is only rarely associated with lupus-like clinical features such as the typical butterfly rash, pleuritis, or arthritis. A comparable incidence of anti-dsDNA antibody formation has been reported following CDP571 treatment, and this may represent a class-specific side-eff'ect. Although lymphomas have occurred in infliximab-treated patients, most cases concerned patients who had an increased lymphoma risk because of the underlying disease (AIDS, rheumatoid arthritis). The background incidence of lymphomas in patients with CD is low, but two patients have been reported who were diagnosed with lymphomas after infliximab treatment. It should be noted that the incidence of lymphomas associated with immunosuppressive drugs is aff'ected by the underlying disease (the incidence is high in rheumatoid arthritis and transplant patients) and by the duration of follow-up. Because of the wide confidence limits and the short follow-up period it is at present not possible to reliably estimate the lymphoma risk in anti-TNF-oc antibody-treated patients. TNF-a is known to have a pivotal role in the immune response to infections, in particular those

Sander J.H. van Deventer from intracellular pathogens, and many investigators predicted that TNF-oc neutralizing therapies would result in a greatly increased infection rate. This prediction has not been substantiated, and it is now known that the infection rate in infliximabtreated patients with CD is only slightly increased. However, neutralization of TNF-oc can have serious consequences in patients with current intracellular infections, in particular tuberculosis, or in immunocompromised patients, and such conditions should be excluded before initiation of treatment (by chest X-ray and PPD skin test). After more than 140000 patients had been infused with infliximab, 70 cases of tuberculosis have been reported [68]. Many of these were extrapulmonary, and in most cases the cause was reactivation of latent disease. The effect of treatment using TNF-a neutralizing proteins on the safety and efficacy of vaccinations, in particular those with life-attenuated microorganisms, is not known, and the experience in pregnant women is very limited. Clinical use of TNF neutralizing proteins Although TNF-oc neutralizing therapies are an important step forward in the medical treatment of CD, many questions remain to be answered. What is the long-term effect of repeated infusions and what are the side-effects of such strategies? Does inffiximab treatment modify the natural course of CD and can surgery be prevented? Why are about one-third of CD patients refractory to infliximab therapy? Do standard immunosuppressive therapies, such as 6MP/azathioprine and methotrexate synergize with TNF-oc neutralizing treatments in CD? Obviously, it is important to know the answers, and several clinical trials addressing these questions have been initiated. At present, because of the high costs and the uncertainty about long-term side-effects, anti-TNFoc antibodies are second- or third-line therapies of IBD. Before considering such therapies it is mandatory to ensure that patients are appropriately managed, and in steroid-refractory or -dependent patients treatment with azathioprine or methotrexate should be considered first. It is important to exclude underlying fibrotic stenoses, inflammatory infiltrates, short bowel syndrome, and gastrointestinal infections. The approved infliximab dose in active CD is 5 mg/kg administered as a 2 h infusion. Data reported by large IBD centers indicate that the response rate in clinical practice is similar to the results of controlled clinical trials (about two-thirds

561 of patients showing a response, one-third a clinical remission). Patients who do not respond may be offered a second dose, but the majority of these patients are genuinely refractory to anti-TNF-a therapy or have complications such as bowel stenosis. Only relatively small clinical studies have reported the effects of repeated anti-TNF-a administrations, and at present the best strategy for repeated treatment of responding patients is uncertain. Because most patients will relapse 8-12 weeks after a single infliximab dose, I usually prescribe three doses (5 mg/kg) with an 8-week interval, and in order to reduce side-effects and prolong the duration of remission I treat all patients with either azathioprine (2.5 mg/kg per day orally) or methotrexate (25 mg/ week subcutaneously). If relapse occurs after the third infliximab infusion, patients can be retreated at an interval that is somewhat shorter than the observed relapse-free interval after the last dose. Infliximab is very useful in treating patients with perianal fistulas, but only in the context of appropriate (multidisciplinary) management. Before institution of treatment abscesses should be appropriately drained, setons should be inserted whenever appropriate, and removed after one or two infliximab doses. In conclusion, anti-TNF-a strategies are an important addition to the therapeutic armamentarium in CD and may induce important therapeutic responses in difficult-to-treat patients. Successful application of these therapies requires optimal baseline treatment including appropriately dosed immunomodulatory drugs in most patients. Short-term side-effects occur in a minority of patients, are usually not severe, and do not preclude further treatment in most patients. All patients considered for infliximab therapy should be screened for (latent) tuberculosis. The incidence and nature of long-term side-effects of this class of drugs remains uncertain, because of the relatively short follow-up of anti-TNF-a antibodytreated patients.

Interference with the IL-12/IL-18/ IFN-y axis IL-12 and IL-18 are cytokines that cause Thl differentiation of T lymphocytes following antigen-dependent stimulation (Fig. 3). Thl-differentiated T lymphocytes are characterized by the production of IL-2 and IFN-y. Increased mucosal concentrations of IFN-y occur in CD, and more recently increased

562 production of both IL-12 and IL-18 has been demonstrated to be a specific feature of CD (as opposed to UC). Although IFN-y has long been considered to be a proinflammatory cytokine, and in some animal models of IBD anti-IFN-y antibodies decrease disease activity, recent data from animal models have led to a more differentiated view. Mice that are IFN-y or IFN-y receptor-deficient are not protected against the induction of colitis by intracolonic administration of TNBS, and IFN-y may have anti-inflammatory effects by stimulating apoptosis of activated T lymphocytes. In a small series of patients with CD, administration of low doses of IFN-y did not exacerbate disease activity. Nonetheless, anti-human IFNy antibodies have been generated for clinical use, and studies in CD patients are planned. IL-12 is a heterodimeric cytokine that is composed of a p35 and a p40 chain. The heterodimeric p75 molecule is recognized by a specific receptor expressed by T lymphocytes and NK cells. In mice, homodimeric p40 is able to antagonize p75, and recently it has been demonstrated that p40 may form heterodimers with proteins other than p35. Hence, several cytokines that share the IL-12 p40 chain have now been identified, and these seem to have different biological activities. For example, experiments performed in p35- and p40-deficient mice demonstrated that the presence of p40 (in the absence of IL-12 p75) is protective in TNBS colitis. This is of importance, because most anti-IL-12 antibodies developed for clinical use bind p40. On the other hand, in wild-type mice anti-p40 antibodies were protective in TNBSinduced colitis, and the mechanism of action was related to the induction of apoptosis. Anti-human IL-12 antibodies have been generated, but no clinical results have been reported. More recently, IL-18 has been identified as a major IFN-y inducer. Within the intestinal mucosal of patients with CD, epithelial as well as lamina propria cells produce increased amounts of IL-18. Like IL-1, IL-18 is translated as a propeptide that needs to be cleaved by ICE in order to generate the mature bioactive protein. The IL-18 receptor with its associated proteins, as well as the associated signaling pathways, are very similar to the IL-1 receptor. IL-18 is antagonized by a naturally occurring soluble protein, IL-18 binding protein, that has a very high affinity. IL-18 antibodies, as well as recombinant IL18 binding protein, are currently being tested for efficacy in various animal models of chronic inflammation.

Targeted therapies for inflammatory bowel disease

Anti-inflammatory cytokines IL-10 IL-IO is a small homodimeric cytokine that antagonizes production of proinflammatory cytokines by T lymphocytes as well as monocytes, decreases expression of HLA class II molecules, and strongly reduces T lymphocyte proliferation. The importance of IL-10 for regulation of mucosal immune responses became apparent when mice made IL-10 deficient by homologous recombination developed cachexia and anemia as a consequence of chronic inflammatory disease of the small and large intestines. The mucosal inflammation in these mice responds to treatment with anti-TNFor anti-IFN-y antibodies, but remarkably, IL-10 administration is protective only when administered early in the course of the disease. Recombinant human IL-10 was found to be safe in human volunteers when administered intravenously or subcutaneously, and down-regulated the production of TNF-oc and other proinflammatory cytokines in ex-vivo stimulation experiments. In endotoxinchallenged volunteers, IL-10 pretreatment resulted in a decrease of the production of TNF-oc, IL-6 and IL-8, and activation of the coagulation system was prevented. A short-term phase II study in patients with active CD demonstrated that intravenously administered IL-10 was well tolerated, and suggested a beneficial clinical effect. Two large long-term clinical studies were performed in CD patients [69, 70]. One study enrolled mild to moderately active CD patients, who were not using corticosteroids or azathioprine, and daily subcutaneous doses of IL-10 or placebo were administered. The second study enrolled chronic active patients who had failed treatment with corticosteroids or immunosuppressive drugs. In both studies a moderate reduction of disease activity was observed at rhIL-10 doses between 5 and 10 |ig/kg. At higher doses patients started to experience side-eff*ects and therapeutic effects were lost. In a search for a possible mechanism, blood from patients being treated with IL-10 was stimulated ex vivo with endotoxin or phytohemagglutin (PHA). A dose-dependent reduction of endotoxin-induced TNF-a production was found, but after PHA stimulations in patients treated with IL-10 doses above 10 |ig/kg resulted in an increased production of TNF-a and IFN-y. These data indicate that IL-10 dose-dependently decreased the production of TNF-oc by monocytes, but at higher doses increased the production of TNF-oc and IFN-y by lymphocytes (most likely CD8'^ T lymphocytes

563

Sander J.H. van Deventer and NK cells). In retrospect, a similar increase in IFN-y production by IL-10 was also found in endotoxin-challenged volunteers (who received a single rhIL-10 dose of 25 |Lig/kg) [71]. Hence, when systemically administered, rhIL-10 at higher doses caused side effects that interfered with therapeutic efficacy. This has led to the development of methods that would more efficiently deliver IL-10 into the intestinal mucosa, for example by genetically modifying gut-homing T lymphocytes to express IL-10.

IL-11 Interleukin 11 (IL-11) is a member of the IL-6 family of cytokines. These cytokines activate cells through activation of heterodimeric membrane receptors, which consist of the transmembrane protein gpl30, and cytokine-specific oc-unit (the IL-11 receptor occhain in case of IL-11). IL-11 is produced by mesenchymal cells and has a wide spectrum of biological activities, including the stimulation of thrombopoiesis, and the protection of the gut epithelium from injury induced by irradiation and chemotherapeutic drugs. IL-11 also suppresses TNF-a production, probably through interference with NFKB signaling. In the HLA-B27 transgenic rat model, rhIL-11 reduced the severity of colitis, and reduced the mucosal production of TNF-oc, IL-1 (3 and IFN-y. In a multicenter controlled clinical trial, enrolling 76 patients with active CD, the safety and activity of several doses (5, 15, 40 |xg/kg, administered two or five times weekly) of rhIL-11 were investigated. Treatment was well tolerated, and an increase in thrombocyte count was observed in patients who received the higher doses or were most frequently treated. The highest clinical responses were observed in the patient group that received 16 jig/kg per week, in whom a therapeutic response (a decrease of the CDAI ^100 points) was observed in 5/12 patients (42%). Based on these preliminary results, further controlled clinical trials were started.

Interference with activation or recruitment of inflammatory leul(ocytes Inflammatory bowel disease is characterized by an influx of inflammatory cells into the intestinal mucosa. The mechanisms responsible for recruitment of neutrophils and monocytes, and the more

specific 'homing' of lymphocytes, have been characterized and provided well-defined therapeutic targets.

Anti-CD4 antibodies In view of the large body of evidence for a dominant role of CDA^ T lymphocytes in the initiation and control of mucosal immune responses, it is somewhat surprising that so few studies have aimed to alter T lymphocyte function using monoclonal antibodies. This may be a result of the unfavorable safety profiles of some of these interventions (i.e. OKT3 treatment). In an open-label dose-escalating study 12 patients with active CD were infused with seven daily doses of a depleting anti-CD4 antibody (cMT412). In patients who received daily doses of 10 and 30 mg (for a total dose of 100 and 210 mg, respectively), transient reductions of the circulating CD4 counts were induced, and a moderate therapeutic eff'ect was observed. In the 4 patients who received a 100 mg daily dose (a total dose of 700 mg) a long-lasting reduction (months) of the CD4"^ lymphocyte count was induced, and all patients had a clinical response. Although the antibody infusion was associated with mild acute toxicity, the long duration of depletion of circulating CD4"^ count is a severe drawback of this therapy. It should be noted that non-depleting antiCD4^ antibodies with a good safety profile have been clinically tested for other indications. Interest in antiCD4"^ antibodies may revive in view of the finding of therapeutic synergism of antibodies that target CD4 and TNF-a in collagen-induced arthritis in mice. In addition, other surface molecules expressed by T lymphocytes (CD2, CD40) are attractive therapeutic targets and are developed for use in inflammatory diseases.

ICAIVI1 antisense oligonucleotides Antisense oligonucleotides are short stretches of single-stranded DNA (18-25 bases) designed to target specific mRNA. Binding of the antisense oligonucleotide to the target mRNA results in the formation of a RNA/DNA hybrid that is rapidly degraded by RNAse H. Single-strand oligonucleotides are rapidly degraded within plasma, and either large amounts need to be infused, or the (phosphorothioate) backbone needs to be modified in order to resist breakdown. Alternatively, the oligonucleotides can be administered locally. A general side-effect of

564 oligonucleotides with a phosphorothioate backbone is prolongation of the activated partial thromboplastin time (PTT) through inhibition of the intrinsic tenase activity, and very high doses may also activate complement. ISIS 2302 is an antisense oligonucleotide that specifically hybridizes to human mRNA encoding the adhesion molecule ICAMl, which is important for the binding of neutrophils to cytokinestimulated endothelium. Because this neutrophilendothelial cell interaction is necessary for transmigration of neutrophils into the tissue, this intervention is expected to specifically inhibit neutrophil influx into the mucosal in IBD. Indeed, preclinical studies in mice have demonstrated that administration of a murine analog of ISIS 2302 (ISIS 3082) ameliorated dextran sulfate-induced colitis in mice. A small controlled cHnical study included 20 patients with active steroid-refractory CD who were treated with 13 infusions of ISIS 2302 (0.5, 1, or 2 mg/kg) or placebo in a period of 26 days. At the end of the study 47% of all ISIS 2302-treated patients, and 20% of placebo-treated patients, were in remission (CDAI < 1 5 0 ) . Unfortunately, a subsequent larger controlled study failed to confirm these promising data [72].

Anti-a4 antibodies The homing of intestinal T- and B lymphocytes is in large part a consequence of interaction of the 064(37 integrin and the mucosal addressin MadCAMl [7375]. Antibodies that block this interaction interfere with recruitment of potential proinflammatory mononuclear cells into the mucosa, and in animal models of intestinal inflammation this principle has been demonstrated to have anti-inflammatory effects [76]. Clinical studies using anti-0C4-antibodies in CD have been initiated.

Inhibitors of signal transduction pathways Inflammation is a result of altered expression of many genes, as a consequence of cell activation by various stimuli such as bacterial products, osmotic stress, oxygen radicals, radiation, or cytokines. These stimuli activated two major families of transcription activators, i.e. activator protein-1 (AP-1) and the N F - K B family, that regulate the fate of the cell (proliferation or apoptotic cell death) and the transcription of inflammatory cytokines (Fig. 4). In their turn the transcription activators are activated by

Targeted therapies for inflammatory bowel disease interacting networks of mitogen-activated protein kinases (MAPK). At least 12 mammalian MAPK members have been recognized, and it has become apparent that these kinases are activated in a threestep process that involves a MAP-kinase-kinasekinase (MKKK), a MAP kinase-kinase (MKK) and a MAP kinase (Fig. 4). Although the M K K K MKK-MAPK pathways extensively interact, specificity is conferred by the limited MAPK specificity of MKK and the presence of 'scaffold' proteins that anchor specific M K K K / M K K / M A P K proteins. Scaffolding results in the formation of multi-protein activation complexes that control signal transduction pathways. Development of specific protein kinase inhibitors is hindered by their redundancy and the presence of isoforms with different expression patterns and biological effects, but recently great progress has been made. Clinical development has been most rapid for MAPK p38 inhibitors, many of which are based on anti-inflammatory pyridinyl imidazole compounds. Several other small molecular inhibitors of the p38 MAP kinase have been developed, which act either by preventing phosphorylation of the kinase, or by blocking activation of kinases downstream of p38 through steric hindrance. The prototype MAPK p38 inhibitor is SB 203580, which inhibits LPS-induced TNF-oc production, and is effective in several animal models of acute and chronic inflammation. SB 203580 is not very specific for MAPK p38, but second- and thirdgeneration specific p38 inhibitors are currently in clinical development. This class of small molecular compound is expected to have anti-inflammatory effects resulting from inhibition of several proinflammatory cytokines, including TNF-a. Specific inhibitors of C-jun aminoterminal kinases (JNK), the major route for AP-1 activation, have not been reported, but are likely to be developed. These molecules are expected to have an anti-inflammatory profile that differs from MAPK inhibitors. Of the three JNK isomers, JNK2 is specifically involved in Thl differentiation, and JNK2 inhibitors will be attractive candidate drugs for treatment of CD, multiple sclerosis and rheumatoid arthritis. In conclusion, inhibition of specific protein kinases may be a potent anti-inflammatory strategy, and several (orally effective) lead compounds have been developed. However, the (long-term) safety of these compounds is unknown.

Sander J.H. van Deventer

stimuli

Growth factors

565

TNFa, IL-1, LPS, stress •

TNFc

membrane

MEKK1,2 M

MEKK1,2 •

MEKK3

MEK1,2

MEK4,7 •

MEK3,6

ERK1,2

JMK1,2,3 •

p38a,P,

cytoplasm

nucleus

\i/

^

^

liiiiiiii liiiiiiiiii;:

AP-1 (jun, fos)

I

NFKB

Figure 4. MAP kinases transmit signals from activated membrane receptors to the nucleus, causing increased transcription of multiple genes Involved in inflammation. Three major pathways have been identified, leading to activation of ERK1,2, JNK1,2,3, and p38a,p,y, respectively.

NF-KB

In resting cells dimeric N F - K B is bound to IKB, which prevents translocation to the nucleus. A multi-unit protein kinase, the IKB kinase (two isoforms, IKKl and IKK2 have been identified) that is activated by proinflammatory stimuli is able to phosphorylate IKB, thereby rendering IKB a target for a ubiquitin ligase complex. The subsequent ubiquination of IKB results in degradation within the 26S proteasome. The free N F - K B migrates into the nucleus and binds to a specific motif within the promoter of several genes involved in inflammation, including the TNFa gene (Fig. 5). By regulating the degradation of IkB the IKK complex critically controls activation of the

TNF-oc gene, and this has recently received substantial attention as a therapeutic target. In addition, low molecular weight N F - K B inhibitors have been developed, and these are able to reduce production of proinflammatory cytokines in animal models. Another interesting approach is interference with N F - K B activation by interfering with the degradation of IKB, either by preventing IKK activity or by reducing the production of N F - K B components such as the p65 protein. Curcumin, a food derivative, has been demonstrated to down-regulate N F - K B translocation through an effect that is located upstream of I K K activation, whereas interference with the production of N F - K B p65 protein by local adminis-

Targeted therapies for inflammatory bowel disease

566

Activation of Rel/NFi(B

'ne-^^f.^€ine

Phorbol ester

proteasome

(jerondakis

modified from: el al ( iirr Opin Immunol

cell cycle cell adhesion (anti) apoptosis inflammation 1998

Figure 5. Activation of several cytokine receptors, including the TNF-a p55 receptor and the IL-1 receptor causes activation of the NF-KB receptor, IKB is phosphorylated and degraded in the proteasome, thereby freeing the NF-KB, that is now able to translocate into the nucleus, where it activates many genes involved in cell adhesion, apoptosis and Inflammation.

tration of antisense oligonucleotides reduced the severity of TNBS-induced colitis in mice. The efficacy of this latter approach in IBD is currently being investigated in clinical trials. Finally, the antiinflammatory effects of IL-10 may in part be related to interference with N F - K B signaling.

Gene therapy The principle of gene therapy is simple: introduction of a functionally active gene into somatic cells in order to express a therapeutic protein. Early gene therapy projects were designed to correct 'inborn errors of metabolism' or to counteract abnormal proliferation of cancer cells. More recently, gene therapy techniques are also being applied to alter the course of chronic inflammatory diseases. Because IBD are under complex polygenic control, and the

pivotal genes have not been characterized, there is at present no reason to believe that these disease can be cured by gene therapy. However, as will be discussed, gene therapy can be used to alter the function of cells that are critically involved in the regulation of mucosal inflammation. One of the critical factors for success of gene therapy in chronic inflammatory diseases is the long-lasting and sufficient expression of the therapeutic gene in the target tissue, and many early gene therapy protocols failed simply because no therapeutic protein was expressed. When applied in chronic inflammatory diseases the intervention itself should not elicit inflammatory responses and longterm expression is necessary. Because the in-vivo efficiency of transfection using naked DNA is low, most gene therapy projects have been performed using viral vectors. These differ greatly in their efficacy and immunogenicity, and the duration of expression of the therapeutic gene.

Sander J.H. van Deventer

567

Adenovirus [77,78]

Adeno-associated virus

Most gene therapy trials have used adenovirusderived vectors for gene deHvery. Adenovirus is a double-stranded virus that causes common cold and gastroenteritis in humans. Of the more than 50 known serotypes, type 2 (Ad2) and type 5 (Ad5) have been used to construct vectors. The double-stranded 30-40 kbp adenovirus genome is flanked by two inverted terminal repeat sequences that contain the origins for DNA replication. The genome encodes for 'early' (El-4) and 'late' genes. Because viral gene therapy vectors should be replication-incompetent, and because the products of early genes are immunogenic, in most currently used vectors one or more of these genes has been deleted. Production of such defective viruses is achieved using 'packaging' cell lines that encode for the deleted genes. The adenovirus particle is an icosahedral virion and has no envelope. From each penton base a trimeric fiber protein protrudes that ends in a 'knob' structure, which binds to the receptors on target cells (a 'ball with spikes'). The most important cellular receptor for adenovirus is the Coxsackie/adenovirus receptor (CAR). Viral entry requires a second interaction of the capsid with cell surface aV integrins, leading to endocytosis of the particle. Once in the cell nucleus, adenovirus DNA remains extrachromosomal and is transiently (days to weeks) transcribed. The advantage of adenoviral vectors is the ability to accommodate large DNA cassettes, and large amounts of vector can be produced and purified. A major disadvantage is the relatively short duration of expression (days to weeks) of the transgene, and the immunogenicity of virtually all currently available vectors. Of particular importance for application in the gut is that epithelial cells express only low amounts of the CAR, and hence transfection efficiency is low even when locally applied. It is possible to increase the transfection efficiency by re-targeting adenoviral vectors through alterations of the knob structure, or by application of bispecific antibodies that bind both to the knob structure and to a cellular receptor. Despite these alterations it is doubtful whether adenoviral vectors will be useful in chronic inflammatory diseases, in particular because these vectors themselves are immunogenic and cause inflammatory responses.

Adeno-associated viruses (AAV) [79] are parvoviruses, and five human serotypes have been identified. These viruses are widely prevalent and, despite the fact that most adult humans have been infected with AAV, the virus is not known to cause a disease. AAV is a DNA virus that is packaged in a nonenveloped icosahedral capsid. The viral genome is a small (4.7 kb) linear DNA flanked by two terminal repeats. It is transcribed in three regions that produce seven transcripts. Because the primary cellular receptor for AAV is heparan sulfate, AAV is able to infect a wide range of (non-dividing) cells. Within the cefl the single-stranded DNA needs to be converted into a duplex, and subsequently is integrated at a specific locus on chromosome 19. The second strand formation is the rate-limiting step in AAV infection, and is greatly enhanced by the presence of adeno (helper) virus. This ability of adenovirus has been traced back to a product of the E4 gene, and based on this knowledge adenovirus-free systems for AAV production have been developed. Site-specific integration in chromosome 19 is a potential advantage of AAV-based viral vectors, but in most vectors the rep gene, that is necessary for site-specific integration, is deleted; consequently AAV-based vectors integrate non-specifically. The great advantage of AAV vectors is the efficient and long-term transduction of a wide range of cells, including the intestinal epithelium, and the absence of immunogenic responses. Long-term therapeutic expression (up to a year) of coagulation factor IX has been reported in factor IXdeficient mice and hemophilic dogs, and clinical trials in patients suffering from hemophilia B are ongoing [80]. Long-lasting expression of P-galactosidase within the small bowel of rats has been achieved using orally administered AAV vectors [81]. The gene was expressed by the epithelium, as well as by lamina propria cells, and treated rats became lactose-tolerant (normal adult rats do not express lactase). Because the turnover of the small bowel epithelium is rapid, the longevity of the P-galactosidase expression presumably indicates that epithelial stem cells were infected. As compared to adenovirus vectors, the maximal size of the therapeutic gene in AAV vectors is limited, and large-scale production and purification are more difficult. Despite these technical difficulties, AAV vectors are promising tools for (mucosa) targeted gene therapy in the intestinal tract.

Targeted therapies for inflammatory bowel disease

568

Retroviruses Retroviruses [82] are a large family of viruses that include Antiviruses and oncoviruses. They consist of a core nucleocapsid that is enclosed in a phospholipid envelope containing the specific glycoproteins that interact with cell surface receptors. The viral genome is present as two single-stranded RNA strands, that end in long terminal repeats (LTR) containing the promoter and enhancer elements for viral transcription. The wild-type genome of the murine leukemia viruses that are frequently used for vector construction contains three genes, gag, pol, and env. These encode reverse transcriptase, the nucleocapsid and envelope proteins as well as a packaging signal (v|/) that is necessary for viral assembly at the cellular membrane. The most frequently used retroviral vector is based on the murine Moloney leukemia virus (MMLV), which is nonpathogenic in humans. In this vector almost the entire viral genome can be replaced by a gene of interest. Non-infectious vectors are produced in 'packaging' cell lines that encode gag, pol, and env, MMLV-based vectors are sensitive to inactivation by complement, infect only dividing cells, and when genes are expressed under the control of viral promoters, these may be silenced by methylation of deletion of proviral DNA. Despite these disadvantages, using MMLV-based vectors, c^.x-v/v^;-transduced hematopoietic stem cells have been demonstrated to express therapeutic genes for many months. At present MMLV-based vectors are the most attractive tools for targeting T lymphocytes, and it has recently been demonstrated that human peripheral blood CD4'*' T lymphocytes acquire a T regulatory phenotype after transfection with an MMLV vector c o n t a i n i n g IL-10 [83]. Such approaches may be effective in a range of T lymphocyte-dependent chronic inflammatory diseases, including CD. Recent improvements in retroviral vector technology include cellular retargeting by replacing the retroviral envelope gene with the vesicular stomatitis virus G envelope protein ('pseudotyping') that is recognized by a ubiquitously expressed membrane phospholipid [84]. In addition, lentiviral vectors are now available that are able to integrate into the genome of non-dividing cells and can enter intestinal epithelial cells through the apical membrane. These characteristics render these vectors extremely interesting tools for intestinal gene therapy.

Conclusion New therapies for IBD have resulted from specific interference with mucosal proinflammatory cascades. Biotechnology, in particular the ability to rapidly generate monoclonal antibodies, has dramatically shortened the time to develop new interventions. In 1992 the first patient with CD was infused with infliximab, and today this therapy is approved and widely used in North America and Europe. Several other interventions, including the administration of 'regulatory' cytokines (IL-10) and antisense oligonucleotides (blocking ICAMl and NFKB p65) have also been tested in clinical trials. More recently, small molecules that block signal transduction pathways, and gene therapy approaches, have become available for clinical testing in IBD. Clearly, many of these new interventions will eventually fail on the route to clinical application. However, some of the drugs that make their way into clinical practice will become new standards of treatment, infliximab being a prominent example. These new therapies come with new and sometimes unexpected toxicities, and are expensive. It is conceivable that within the near future small molecules with therapeutic profiles similar to anti-TNF-oc antibodies will become available at lower costs. Using these new therapies, remission induction can now be achieved in most patients with active IBD. The new biological therapies have been less successful in maintaining remissions, and this remains an unmet therapeutic need.

References 1. 2. 3. 4. 5. 6. 7.

Elson CO, Cong Y, Brandwein S et al. Experimental models to study molecular mechanisms underlying intestinal inflammation. Ann NY Acad Sci 1998; 859: 85-95. Dinarello CA. Biologic basis for interleukin-1 in disease. Blood 1996; 87: 2095-147. Beutler BA. The role of tumor necrosis factor in health and disease. J Rheumatol 1999; 26: 16-21. Dinarello CA. Interleukin-1 beta, interleukin-18, and the interleukin-1 beta converting enzyme. Ann NY Acad Sci 1998; 856: 1-11. Fantuzzi G, Dinarello CA. Interleukin-18 and interleukin-1 beta: two cytokine substrates for ICE (caspase-1), J Clin Immunol 1999; 19: 1-11. Dinarello CA. Interleukin-1, interleukin-1 receptors and interleukin-1 receptor antagonist. Int Rev Immunol 1998; 16:457-99. Thomassen E, Bird TA, Renshaw BR, Kennedy MK, Sims JE. Binding of interleukin-18 to the interleukin-1 receptor homologous receptor IL-lRrpl leads to activation of signaling pathways similar to those used by interleukin-1. J Interferon Cytokine Res 1998; 18: 1077-88.

Sander J.H. van Deventer 8.

9.

10.

11.

12.

13.

14.

15.

16. 17. 18. 19. 20. 21.

22. 23. 24.

25. 26.

Brynskov J, Tvede N, Andersen CB, Vilien M. Increased concentrations of interleukin Ibeta, interleukin-2, and soluble interleukin-2 receptors in endoscopical mucosal biopsy specimens with active inflammatory bowel disease. Gut 1992; 33: 55-8. Mahida YR, Wu K, Jewell DP. Enhanced production of interleukin 1-beta by mononuclear cells isolated from mucosa with active ulcerative colitis or Crohn's disease. Gut 1989;30:835-8. Casini-Raggi V, Kam L, Chong YJ, Fiocchi C, Pizarro TT, Cominelli F. Mucosal imbalance of IL-1 and IL-1 receptor antagonist in inflammatory bowel disease. A novel mechanism of chronic intestinal inflammation. J Immunol 1995; 154:2434-40. Stokkers PC, van Aken BE, Basoski N, Reitsma PH, Tytgat GN, van Deventer SJ. Five genetic markers in the interleukin 1 family in relation to inflammatory bowel disease. Gut 1998; 43: 33-9. Andus T, Daig R, Vogl D et al. Imbalance of the interleukin 1 system in colonic mucosa - association with intestinal inflammation and interleukin 1 receptor agonist genotype 2. Gut 1997;41:651-7. Cominelli F, Nast CC, Duchini A, Lee M. Recombinant interleukin-1 receptor antagonist blocks the proinflammatory activity of endogenous interleukin-1 in rabbit immune colitis. Gastroenterology 1992; 103: 65-71. Wallach D, Varfolomeev EE, Malinin NL, Goltsev YV, Kovalenko AV, Boldin MP. Tumor necrosis factor receptor and Fas signaling mechanisms. Annu Rev Immunol 1999; 17: 331-67. Bielekova B, Lincoln A, McFarland H, Martin R. Therapeutic potential of phosphodiesterase-4 and -3 inhibitors in Thl-mediated autoimmune diseases. J Immunol 2000; 164: 1117-24. Corral LG, Kaplan G. Immunomodulation by thalidomide and thalidomide analogues. Ann Rheum Dis 1999; 58: 1107-13. Muller GW, Shire MG, Wong LM et al. Thalidomide analogs and PDE4 inhibition. Bioorg Med Chem Lett 1998; 8:2669-74. Reimund JM, Dumont S, Muller CD et al In vitro effects of oxpentifylline on inflammatory cytokine release in patients with inflammatory bowel disease. Gut 1997; 40: 475-80. Black RA, Ranch CT, Kozlosky CJ et al. A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells. Nature 1997; 385: 729-33. Killar L, White J, Black R, Peschon J. Adamalysins. A family of metzincins including TNF-alpha converting enzyme (TACE). Ann NY Acad Sci 1999; 878: 442-52. Moss ML, Jin SL, Milla ME et al. Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-alpha [Published erratum appears in Nature 1997; 386: 738]. Nature 1997; 385: 733-6. Moss ML, Jin SL, Becherer JD et al. Structural features and biochemical properties of TNF-alpha converting enzyme (TACE). J Neuroimmunol 1997; 72: 127-9. Tartaglia LA, Rothe M, Hu YF, Goeddel DV. Tumor necrosis factor's cytotoxic activity is signaled by the p55 TNF receptor. Cell 1993; 73: 213-16. Tartaglia LA, Goeddel DV, Reynolds C et al. Stimulation of human T-cell proliferation by specific activation of the 75kDa tumor necrosis factor receptor. J Immunol 1993; 151: 4637-41. Beg AA, Baltimore D. An essential role for NF-kappaB in preventing TNF-alpha-induced cell death [See comments]. Science 1996; 274: 782-4. Levi M, Ten Cate H, Bauer KA et al. Inhibition of endotoxin-induced activation of coagulation and fibrinolysis by

569

27.

28.

29. 30.

31.

32. 33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

pentoxifylline or by a monoclonal anti-tissue factor antibody in chimpanzees. J Chn Invest 1994; 93: 114-20. Bauditz J, Haemling J, Ortner M, Lochs H, Raedler A, Schreiber S. Treatment with tumour necrosis factor inhibitor oxpentifylline does not improve corticosteroid dependent chronic active Crohn's disease [See comments]. Gut 1997; 40: 470-4. MacKenzie SJ, Houslay MD. Action of rolipram on specific PDE4 cAMP phosphodiesterase isoforms and on the phosphorylation of cAMP-response-element-binding protein (CREB) and p38 mitogen-activated protein (MAP) kinase in U937 monocytic cells. Biochem J 2000; 347: 571-8. Zou LP, Deretzi G, Pehdou SH et al. Rolipram suppresses experimental autoimmune neuritis and prevents relapses in Lewis rats. Neuropharmacology 2000; 39: 324-33. Hartmann G, Bidlingmaier C, Siegmund B et al. Specific type IV phosphodiesterase inhibitor rolipram mitigates experimental coHtis in mice. J Pharmacol Exp Ther 2000; 292: 22-30. Singhal S, Mehta J, Desikan R et al. Antitumor activity of thalidomide in refractory multiple myeloma [See comments] [Published erratum appears in N Engl J Med 2000; 342: 364]. N Engl J Med 1999; 341: 1565-71. Postema PT, den Haan P, van Hagen PM, van Blankenstein M. Treatment of colitis in Behcet's disease with thalidomide. Eur J Gastroenterol Hepatol 1996; 8: 929-31. Bekker LG, Haslett P, Maartens G, Steyn L, Kaplan G Thalidomide-induced antigen-specific immune stimulation in patients with human immunodeficiency virus type 1 and tuberculosis. J Infect Dis 2000; 181: 954-65. Tramontana JM, Utaipat U, Molloy A et al. Thalidomide treatment reduces tumor necrosis factor alpha production and enhances weight gain in patients with pulmonary tuberculosis. Mol Med 1995; 1: 384-97. Sato EI, Assis LS, Lourenzi VP, Andrade LE. Long-term thalidomide use in refractory cutaneous lesions of systemic lupus erythematosus. Rev Assoc Med Brasil 1998; 44: 28993. Jacobson JM, Spritzler J, Fox L et al. Thalidomide for the treatment of esophageal aphthous ulcers in patients with human immunodeficiency virus infection. National Institute of Allergy and Infectious Disease AIDS Clinical Trials Group. Immunol Today 1999; 20: 538-40. Vasiliauskas EA, Kam LY, Abreu-Martin MT et al. An open-label pilot study of low-dose thalidomide in chronically active, steroid-dependent Crohn's disease. Gastroenterology 1999; 117:1278-87. Ehrenpreis ED, Kane SV, Cohen LB, Cohen RD, Hanauer SB. Thalidomide therapy for patients with refractory Crohn's disease: an open-label trial. Gastroenterology 1999; 117: 1271-7. Haslett PA, Klausner JD, Makonkawkeyoon S et al. Thahdomide stimulates T cell responses and interleukin 12 production in HIV-infected patients. AIDS Res Hum Retroviruses 1999; 15: 1169-79. Verbon A, Juffermans NP, Speelman P et al. A single oral dose of thalidomide enhances the capacity of lymphocytes to secrete gamma interferon in healthy humans. Surg Endosc 2000; 14: 721-5. Juffermans NP, Verbon A, Olszyna DP, van Deventer SJ, Speelman P, van Der Poll T. Thalidomide suppresses upregulation of human immunodeficiency virus coreceptors CXCR4 and CCR5 on CD4"' T cells in humans. J Infect Dis 2000; 181:2067-70. Corral LG, Haslett PA, Muller GW et al. Diff'erential cytokine modulation and T cell activation by two distinct classes of thalidomide analogues that are potent inhibitors of TNF-alpha. J Immunol 1999; 163: 380-6.

Targeted therapies for inflammatory bowel disease

570 43.

44.

45.

46.

47. 48. 49. 50. 51. 52.

53.

54. 55. 56.

57.

58. 59.

60.

Sykes AP, Bhogal R, Brampton C et al The effect of an inhibitor of matrix metalloproteinases on colonic inflammation in a trinitrobenzenesulphonic acid rat model of inflammatory bowel disease. Aliment Pharmacol Ther 1999; 13: 1535-42. Williams LM, Gibbons DL, Gearing A, Maini RN, Feldmann M, Brennan FM. Paradoxical effects of a synthetic metalloproteinase inhibitor that blocks both p55 and p75 TNF receptor shedding and TNF alpha processing in RA synovial membrane cell cultures. J Clin Invest 1996; 97: 2833-41. Dekkers PE, Lauw FN, ten Hove T et al. The effect of a metalloproteinase inhibitor (GI5402) on tumor necrosis factor-alpha (TNF-alpha) and TNF-alpha receptors during human endotoxemia. Blood 1999; 94: 2252-8. Dekkers PE, ten Hove T, Lauw FN et al. The metalloproteinase inhibitor GI5402 inhibits endotoxin-induced soluble CD27 and CD 16 release in healthy humans. Infect Immun 2000; 68: 3036 9. Tracey KJ, Abraham E. From mouse to man: or what have we learned about cytokinc-based anti-inflammatory therapies? Shock 1999; 11:224 5. Abraham E. Why immunomodulatory therapies have not worked in sepsis. Nursing 1999; 29: 59 63; quiz 64. Knight DM, Trinh H, Lc J et al. Construction and initial characterization of a mouse human chimeric anti-TNF antibody. Cytokine 1995; 7: 15 25. Holliger P, Bohlen H. Engineering antibodies for the clinic. Mol Immunol 1993; 30: 1443 53. Kcmpcni J. Preliminary results of early clinical trials with the fully human anti-TNFalpha monoclonal antibody D2E7. Cancer Metastasis Rev 1999; 18: 411 19. Bendelc AM, McComb J, Gould T et al. Combination benefit of PEGylated soluble tumor necrosis factor receptor type I (PEG sTNF-RI) and dexamethasonc or indomethacin in adjuvant arthritic rats. Clin Exp Rheumatol 1999; 17: 553 60. Edwards CK, 3rd. PEGylated recombinant human soluble tumour necrosis factor receptor type I (r-Hu-sTNF-RI): novel high affinity T N F receptor designed for chronic inflammatory diseases. J Rheumatol 2000; 27: 601 9. Epstein WV. Treatment of rheumatoid arthritis with a tumor necrosis factor receptor-Fc fusion protein [Letter; comment]. J Immunol 1998; 160: 4098-103. Stack WA, Mann SD, Roy AJ et al. Randomised controlled trial of CDP571 antibody to tumour necrosis factor-alpha in Crohn's disease [See comments]. Lancet 1997; 349: 521-4. Sandborn WJ, Feagan BG, Hanauer SB et al. An engineered human antibody to TNF (CDP571) for active Crohn's disease: a randomized double-blind placebo-controlled trial. Gastroenterology 2001; 120: 1330-8. Feagan BG, Sandborn WJ, Baker JP et al. A randomized, double-blind, placebo-controlled, multi-center trial of the engineered human antibody to TNF (CDP571) for steroid sparing and maintenance of remission in patients with steroid-dependent Crohn's disease. Gastroenterology 2000; 118: A655. Derkx B, Taminiau J, Radema S et al. Tumour-necrosisfactor antibody treatment in Crohn's disease [Letter] [See comments]. Lancet 1993; 342: 173-4. Van Dullemen HM, Van Deventer SJH, Hommes DW et al. Treatment of Crohn's disease with anti-tumor necrosis factor chimeric monoclonal antibody (cA2). Gastroenterology 1995; 109: 129-35. Targan SR, Hanauer SB, Vandeventer SJH et al. A shortterm study of chimeric monoclonal antibody Ca2 to tumor necrosis factor alpha for Crohns-disease. N Engl J Med 1997; 337: 1029-35.

61.

62.

63. 64.

65.

66. 67. 68. 69.

70.

71. 72.

73.

74. 75. 76. 77. 78. 79.

Rutgeerts P, D'Haens G, Targan S et al. Efficacy and safety of retreatment with anti-tumor necrosis factor antibody (infliximab) to maintain remission in Crohn's disease. Gastroenterology 1999; 117: 761-9. D'Haens G, Van Deventer S, Van Hogezand R et al. Endoscopic and histological healing with infliximab antitumor necrosis factor antibodies in Crohn's disease: a European multicenter trial. Gastroenterology 1999; 116: 1029-34. Present DH, Rutgeerts P, Targan S et al. Infliximab for the treatment of fistulas in patients with Crohn's disease. N Engl J Med 1999; 340: 1398-405. Ricart E, Panaccione R, Loftus EV, Tremaine WJ, Sandborn WJ. Successful management of Crohn's disease of the ileoanal pouch with infliximab. Gastroenterology 1999; 117:429-32. Brandt J, Haibel H, Comely D et al. Successful treatment of active ankylosing spondylitis with the anti-tumor necrosis factor alpha monoclonal antibody infliximab (In process citation). Arthritis Rheum 2000; 43: 1346 52. Van der Linde K, Meijssen MA, Van Bodegraven AA et al. Infliximab for treatment of perineal metastatic Crohn's disease. Gastroenterology 2000; 118: A588. Sands BE, Tremaine WJ, Sandborn WJ et al. Infliximab in the treatment of severe, steroid-refractory ulcerative colitis: a pilot study. Inflamm Bowel Dis 2001; 7: 83 8. Kcane J, Gcrshon S, Wise RP et al. Tuberculosis associated with inffiximab, a tumor necrosis factor alpha-neutralizing agent. N Engl J Med 2001; 345: 1098 104. Schrciber S, Fcdorak RN, Nielsen OH et al. Safety and efficacy of recombinant human interleukin 10 in chronic active Crohn's disease. Crohn's Disease IL-10 Cooperative Study Group. Gastroenterology 2000; 119: 1461 72. Fedorak RN, Gangl A, Elson CO et al. Recombinant human interleukin 10 in the treatment of patients with mild to moderately active Crohn's disease. Interleukin 10 Inflammatory Bowel Disease Cooperative Study Group. Gastroenterology 2000; 119:1473 82. Lauw FN, Pajkrt D, Hack CE, Kurimoto M, van Deventer SJ, van Der Poll T. Proinflammatory effects of IL-10 during human endotoxemia. J Infect Dis 2000; 182: 888-94. Schreiber S, Nikolaus S, Malchow H et al. Absence of efficacy of subcutaneous antisense ICAM-l treatment of chronic active Crohn's disease. Gastroenterology 2001; 120: 1339-46. Mebius RE, Streeter PR, Michie S, Butcher EC, Weissman IL. A developmental switch in lymphocyte homing receptor and endothelial vascular addressin expression regulates lymphocyte homing and permits CD4"^ CD3 cells to colonize lymph nodes. Proc Natl Acad Sci USA 1996; 93: 11019-24. Berg EL, McEvoy LM, Berlin C, Bargatze RF, Butcher EC. L-selectin-mediated lymphocyte rolling on MAdCAM-l [See comments]. Nature 1993; 366: 695-8. Berlin C, Berg EL, Briskin MJ et al. Alpha 4 beta 7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM-1. Cell 1993; 74: 185. Podolsky DK, Lobb R, King N et al. Attenuation of colitis in the cotton-top tamarin by anti-alpha 4 integrin monoclonal antibody. J Clin Invest 1993; 92: 372-80. Richards CD, Braciak T, Xing Z, Graham F, Gauldie J. Adenovirus vectors for cytokine gene expression. Ann NY Acad Sci 1995; 762: 282-92; discussion 292-3. Wilson JM. Adenoviruses as gene-dehvery vehicles. N Engl J Med 1996; 334: 1185-7. Flotte TR, Carter BJ. Adeno-associated virus vectors for gene therapy. Gene Ther 1995; 2: 357-62.

Sander J. H. van Deventer 80.

Chao H, Samulski R, Bellinger D, Monahan P, Nichols T, Walsh C. Persistent expression of canine factor IX in hemophiha B canines. GeneTher 1999; 6: 1695-704. 81. During MJ, Xu R, Young D, Kaplitt MG, Sherwin RS, Leone P. Peroral gene therapy of lactose intolerance using an adeno-associated virus vector [See comments]. Nat Med 1998; 4:1131-5. 82. Laine F, Blouin V, Ferry N, Evaluation of recombinant retrovirus and adenovirus for gene transfer to normal and pathologic intestine. Gastroenterol Clin Biol 1999; 23: 221-

571 83.

Te Velde AA, Van Montfrans C, Spits H, Hooijberg E, Van Deventer SJH. Therapeutic potential of genetically modified T-lymphocytes in Crohn's disease. Gastroenterology 2000; 118: A873. 84. Abe A, Chen ST, Miyanohara A, Friedmann T. In vitro cellfree conversion of noninfectious Moloney retrovirus particles to an infectious form by the addition of the vesicular stomatitis virus surrogate envelope G protein. J Virol 1998; 72:6356-61.

28 Role of antibiotics and probiotics in the management of inflammatory bowel disease PHILIPPE MARTEAU AND FERGUS SHANAHAN

Introduction An infectious cause or contribution to the primary pathogenesis of inflammatory bowel disease (IBD) has not been demonstrated, ahhough many studies have followed that track [1]. However, the deleterious role of some microorganisms from the endogenous flora is also now well estabhshed in the majority of models of colitis or enteritis [1-4]. Antibiotics have a weU-estabHshed role in the management of complications of IBD such as abscess and pouchitis. Whether antibiotics are of use as a primary therapy for either Crohn's disease (CD) or ulcerative colitis (UC) is less clear. An alternative way of influencing intestinal ecology is the use of probiotics or prebiotics. Probiotics are defined as living non-pathogenic microorganisms which, when ingested, exert a positive influence on host health or physiology [5, 6]. Prebiotics have been defined as non-digestible food ingredient that beneficially affects the host by selectively stimulating the growth and/or activity of one or a limited number of bacteria in the colon, that have the potential to improve host health [7]. As some animal studies have shown that probiotics or prebiotics may help in preventing or curing experimental colitis, clinical trials have begun in IBD. This chapter focuses on the use of metronidazole and ciprofloxacin in CD and pouchitis. In addition, existing data on the use of other antimicrobial agents and on the potential for probiotics in IBD is reviewed.

Background - rationale for tlierapeutic manipulation of enteric flora in IBD Although the etiology is unclear, considerable advances have been made in understanding the

pathogenesis of CD and UC. This involves an interaction between three contributory cofactors: host susceptibility, mucosal immunity and the resident enteric bacterial flora. Experimental models suggest that the inflammatory response in IBD is the result of abnormal immunologic reactivity to normal flora in a genetically susceptibile host, rather than an appropriate immune response to an abnormal or pathogenic flora [2-4]. Although a specific pathogenic infection has not been implicated that could account for most cases of IBD, heterogeneity may exist and enteric infections might also act as cofactors influencing mucosal immune development and/or have an impact on the timing or clinical onset and phenotype of IBD. The mucosal immune system senses and interprets the local microenvironment; under normal circumstances it exhibits a restrained response to commensal flora (tolerance) which is coupled with the capacity to react effectively to episodic challenge from pathogenic bacteria (immunity). This requires exquisitely precise regulation. In susceptible individuals, IBD may arise due to a breakdown in immunologic tolerance to the normal enteric bacteria [1, 2]. The pattern of the pathologic immunoinflammatory response is determined by the profile of mucosal cytokines generated which is, in part, genetically regulated and probably influenced by previous immunologic experience including types and timing of exposure to childhood mucosal infections. CD is associated with type 1 helper T cell (Thl) cytokines, such as interferon-y, tumor necrosis factor (TNF)-oc, and interleukin-(IL)-12, although there is some evidence that the profile of mucosal cytokines in early lesions may differ from that of chronic lesions [1, 8, 9]. Ulcerative colitis, in contrast, does not fit easily into the Thl/Th2 paradigm; it does not represent a Thl response and appears, at least in established disease, to be a modified Th2 response associated with cytokines such as IL-5 and IL-10 [8-10].

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 573-585. © 2003 Kluwer Academic Publishers. Printed in Great Britain

574 Most currently available drug therapies for either CD or UC suppress or modulate the host immune and inflammatory response at both systemic and mucosal levels [11]. Until recently, little attention has been directed toward modifying the enteric microenvironment. A striking example of the importance of controlling environmental influences on IBD is the adverse effect of smoking on the clinical course of CD [12]. In addition, the impact of environmental infections on the mucosal immunoinflammatory response is dramatically illustrated by the apparent capacity of helminthic infections to switch mucosal cytokine production in patients with CD from a Th 1 profile toward a Th2 [13]. Indeed, one of the suggested reasons for the increase in prevalence of CD within developed countries within recent decades is reduced exposure to helminthic and other parasitic infections [14]. For patients with established disease, therapeutic modification of the resident bacterial flora with antibiotics or probiotics has recently attracted considerable scientific and clinical research interest and off'ers another opportunity to influence the mucosal cytokine milieu. However, modification of the enteric flora is not a simple exercise - the flora within the human intestine are complex and dynamic with 400-500 bacterial species accounting for 1-2 kg of intestinal content and a collective metabolism tantamount to that of a virtual organ the neglected organ [15, 16]. While antibiotics may appear an obvious method for modifying the flora, they are too non-specific, pose a risk of overgrowth with pathogens such as Clostridium difficile, and the potential for development of antibiotic resistance. Probiotics offer the advantage of more measured and controlled

Antibiotics andprobiotics in inflammatory bowel disease manipulation of the flora in IBD and they have properties beyond a simple antimicrobial effect [4-6, 17-18]. In addition, probiotics offer the prospect of genetically modified organisms designed for delivery of important biological molecules including IL-10, vaccines and other biologically important molecules [18-20]. The rationale for therapeutic modification of the enteric flora in patients with established IBD is based on two main lines of evidence. First, is the beneficial conditioning influence of the commensal bacterial flora on intestinal structure and function under normal circumstances. Second is a body of evidence implicating the enteric flora in driving the mucosal inflammatory response in susceptible individuals (Table 1) [21-38]. Epithelial cells have been described as sensors of pathogenic infection, and cytokine signaling by the epithelium has been well described in the context of enteric pathogenic infections [39]. However, bidirectional dialog between the host and the normal flora is also evident [22, 40]. Indeed, evidence for influence of the enteric flora on mucosal integrity is well illustrated in germfree animals, where there is reduced epithelial turnover, vascularity, luminal enzyme activity, mucosal-associated lymphoid tissue and thickness of bowel wall. In addition, there are alterations in peristalsis and immune function [23]. Thus, the flora exchange regulatory signals not only with the epithelium but with the subepithelial cells within the intestine [4, 41, 42]. Evidence implicating the resident bacterial flora as an essential cofactor driving the inflammatory process in IBD is based on observational yet persuasive data from human studies and direct experimental results using animal

Table 1. Rationale for therapeutic manipulation of enteric flora in IBD Observational and experimental lines of evidence

References

Conditioning effects of resident flora on mucosal structure and function including epithelial turnover, blood flow, peristalsis, immune development and function

4,16,21-23

Lesions in IBD occur in areas of highest bacterial exposure Clinical response to diversion of fecal stream, relapse on restoration

24,25

Experimental induction of inflammation by exposure to fecal material

26,27

Immunologic reactivity to flora in patients with IBD (loss of tolerance)

28-33

Attenuation of inflammation in animal models of IBD when germfree

34,35

Experimental transfer of colitis with T cells reactive to bacterial antigens

36

Influence of antibiotics and probiotics in experimental animals and in pouchitis

37,38

*References are representative, not comprehensive.

Philippe Marteau and Fergus Shanahan models (Table 1). Perhaps the most persuasive argument is the transfer of enterocolitis using T cells that are reactive to enteric bacteria in an animal model of IBD [36]. In summary, although the lessons of Helicobacter pylori and peptic ulcer disease are sobering, and a search for a specific infectious etiology for IBD will continue [4, 8, 43], the bulk of evidence suggests that IBD is not a simple struggle between microbe and humans but rather a more complex interplay amongst genes, commensals and immune cells. There is a sound rationale for therapeutic alteration of the enteric flora and, as discussed below, the evidence for its efficacy as an adjunct to existing therapies is promising.

Metronidazole Metronidazole is a nitroimidazole compound that has a clinically useful antimicrobial spectrum against anaerobes, particularly gut anaerobes, as well as certain intestinal protozoan infections. Metronidazole is a prodrug which has no antimicrobial properties until it is activated by bacterial enzymes [44]. The size of the bacterial flora in the colon and the proportion of anaerobes in comparison with the small bowel may account for differences in efficacy of metronidazole in colonic and small bowel CD. The actual mode of action of metronidazole in CD is uncertain and may not be limited to its antimicrobial activity [4, 44] as metronidazole also has immunomodulatory effects and tissue-healing properties [46]. A role for metronidazole in CD was first suggested by Ursing and Kamme in 1975 [47]. Since then, its clinical usefulness in CD has been the subject of few well-designed studies but much debate. Metronidazole has some beneficial effect in ileocolonic CD. As discussed later, this has been shown in randomized controlled trials. Metronidazole has also been shown to be effective in perineal CD, although a randomized controlled study proving the efficacy in this setting has not been performed. Metronidazole is not effective in isolated small bowel CD but decreases the risk of recurrence of ileal lesions after ileocolonic resection for CD. Open studies and two randomized controlled trials have shown that it is not effective in UC when used alone or as an adjunct to corticosteroids [48-50]. The results of the randomized trials are summarized in Table 2.

575

Ileocolonic Crohn's disease Blichfeldt et al. [51] evaluated metronidazole as adjunctive therapy (patients were allowed to use prednisone or salazosulfapyridine) compared to placebo in a double-blind, crossover study. They found no overall benefit of metronidazole except in the subset of six patients with colitis only, in which there was considerable improvement. The Cooperative Crohn's Disease Study in Sweden [52] compared metronidazole, 800 mg/day, as primary therapy to sulfasalazine, 3 g/day. Although this was a doubleblind, crossover study, there was no placebo control. The authors found metronidazole and sulfasalazine to be equally effective therapy. Small bowel disease alone did not benefit, as only two subjects of 12 in this subset on metronidazole improved. Of note, in this trial, just over half the patients responded to either therapy and only 25% of metronidazole patients achieved quiescent disease. Ambrose et al. [53] compared metronidazole alone, metronidazole plus cotrimoxazole, cotrimoxazole alone and placebo in a randomized trial of 1 month involving 72 patients. No difference of response was observed between the four groups. In a 16-week randomized, placebo-controlled trial by Sutherland etal. [54], two different doses of metronidazole were compared with placebo in patients with active CD. By the end of the trial there was no difference in attaining remission in either of the metronidazole groups compared to placebo (27% for both metronidazole groups versus 25% for placebo). However, more individual patients on metronidazole showed some improvement in their Crohn's Disease Activity Index (CDAI) compared with patients using placebo. Prantera et al. [55] compared, in a 12-week randomized trial, the efficacy of a combination of metronidazole (250 mg four times daily) plus ciprofloxacin (500 mg twice daily) vs methylprednisolone (0.7-1 mg/kg followed by tapering) in 41 patients with active CD. Clinical remission was observed in 45.5%) of the patients receiving the antibiotics vs 63% in the steroid group (n.s.). There were more side-effects in the antibiotic group (27.3%) vs 10.6%). The authors concluded that, despite the high incidence of adverse events, this combination of antibiotics could be an alternative to steroids in the treatment of acute CD for the patients who tolerate the drugs. Rutgeerts et al. performed two randomized controlled trails demonstrating that metronidazole and ornidazole are significantly eff'ective in the prevention of postoperative recurrence of CD after ileal resection [56, 57]. In the first trial, 60 patients

Antibiotics andprobiotics in inflammatory bowel disease

576

Table 2. Randomized controlled trials with nitrolmldazole compounds in Crohn's disease Indication

No

Treatment and design

Dose of IVIZ and duration

Results (nitroimidazole-control)

Ref.

Active CD

21

IVIZ vs placebo crossover

1 g/day, 1 month

n.s.; MZ effective for colonic CD

51

Active CD

78

MZ vs salazopyrin crossover

0.8 g/day, 4 months

n.s.

52

Active CD

72

MZ vs cotrimoxazole vs both vs placebo

0.8 g/day, 4 weeks

n.s.

53

Active CD

105

MZ vs placebo

10 or 20 mg/kg per day, 16 weeks

Remission: 27% vs 25% n.s.; the mean decrease of CDAI was

54

greater in MZ treated patients. Greater effect of MZ in patients with colonic involvement Active CD

41

MZ -I- ciprofloxacin vs methylprednisolone

1 g/d of MZ and ciprofloxacin, 12 weeks

Remission 45.5% vs 63% (n.s.)

55

Prevention of postoperative recurrence

60

MZ vs placebo

20 mg/kg per day, 3 months

Severe endoscopic relapse; at 3 months: clinical relapse: 4% vs 25%; at 1 year (p = 0.02, n.s. after 1 year

56

Prevention of postoperative recurrence

71

Ornidazole vs placebo

1 g/day, 12 months

Severe endoscopic relapse; at 3 months: 41 %vs 74%*; at 12 months: 62% vs 94% (n.s.)

57

Pouchitis

13

MZ vs placebo, crossover

1.2 g/day, 2 weeks

Significative decrease in the number of bowel movements; endoscopy: n.s.

63

CD, Crohn's disease; CDAI, Crohn's Disease Activity Index; MZ, metronidazole. *p. Another searchable resource with MEDLINE links to journal articles about herbs is . Finally, a non-profit organization to promote research on herbal remedies maintains a website with links to recent studies at < http: / / www.herbs.org > .

Gregg W. Van Citters and Henry C. Lin

Physical examination The physical examination helps identify manifestations of nutritional deficiency. Volume depletion is marked by delayed capillary refill, decreased skin turgor, dry oral mucosa, tachycardia, and orthostatic hypotension. Protein-calorie malnutrition is associated with growth retardation, tissue wasting, loss of subcutaneous fat, edema, and dehydration. Symptoms of malnutrition include changes in hair or

593 skin color, diarrhea or steatorrhea, anorexia, low energy level, apathy, and abdominal distension. Many signs and symptoms of micronutrient deficiencies involve the skin, eyes, and digestive system (see Table 1). Rashes and dermatitis are associated with deficiencies of B vitamins, zinc, and essential fatty acids. Ocular dysfunctions are features of vitamin A or riboflavin deficiency. Maldigestion and with diarrhea may not only be a cause of malnutrition but

Table 1. External signs and symptoms of nutritional deficiencies Possible deficit(s) Signs HyperkeratinizatJon of the skin Extensive dermatitis Pigmentation on pressure points (e.g. elbows) Skin fissures or scales Seborrheic dermatitis, especially nasolabial Glossitis, stomatitis Cheilosis Hemorrhages, especially gingival Rachitic rosary bowed legs, knock-knees, pigeon chest, scoliosis Bilateral symmetric peripheral neuropathy polyneuropathy Loss of reflexes, loss of vibratory sense Vertical nystagmus Arrhythmias Cardiac enlargement, congestive heart failure Growth retardation, delayed sexual maturation Dehydration

Vitamin A Niacin Niacin Niacin (riboflavin, iron for mouth fissures) Riboflavin, zinc, essential fatty acids Niacin, riboflavin, iron Riboflavin Vitamin C Vitamin D Thiamine Thiamine, vitamin B12 Magnesium Potassium Thiamine Zinc (children) Sodium

Symptoms Xeroderma Skin rashes Impaired wound healing Bone pain (osteomalacia) Easy bone fractures Night-blindness Xerophthalmia Photophobia, burning or itching eyes Increased salivation, swollen tongue Maldigestion, gas Vomiting Diarrhea Irritability Anorexia Depression Poor memory inability to concentrate Confusion, hallucinations, delirium, paranoia Tremors, convulsions, paralysis, dysphagia Loss of taste Palpitations Peripheral edema Weight loss, aches, pains Impaired clotting Weakness Decreased libido, decreased fertility

Vitamin A Niacin, riboflavin, zinc, essential fatty acids Vitamin C Vitamin D Calcium Vitamin A Vitamin A Riboflavin Niacin Niacin Niacin, sodium Niacin, zinc, potassium Vitamin D, thiamine, pyridoxine Thiamine, zinc, sodium Thiamine, niacin, pyridoxine, biotin Thiamine Niacin, sodium Magnesium Zinc Thiamine Thiamine Vitamin C Vitamin K Pyridoxine, iron, magnesium, potassium, sodium Zinc

594 also, in time, be a complication of malnutrition since both digestive and absorptive capacity are reduced by protein-calorie malnutrition. Such a complication would result in a vicious cycle of impairment whereby the patient's symptoms and physical signs of malnutrition would progressively worsen. In addition, diarrhea may reflect deficiency or electrolyte imbalance that require specific intervention. A number of excellent reviews are available on the physical findings associated with nutritional deficiencies [168-172].

Laboratory studies Laboratory studies should be used to confirm suspicions of nutritional deficiencies and to identify subclinical deficiencies. Protein-energy malnutrition should be assessed with anthropometric measurements and serum albumin and prealbumin. Fat malabsorption can be quantified with 24-hour stool collection for fecal fat. Non-specific indicators of fat malabsorption include low serum concentrations of fat-soluble vitamins and P-carotene and low serum concentrations of essential fatty acids. Mean corpuscular hemoglobin and volume may be associated with low folate and vitamin B12 concentrations. In this case, low concentrations of the water-soluble vitamins should also be suspected. Screening should include iron, zinc, magnesium, and copper concentrations [173] in addition to electrolytes. Calcium homeostasis may be upset, particularly in the setting of fat malabsorption [173] and colonic resection [174]. Altered bone metabolism should be suspected for all IBD patients [175, 176], particularly those with a history of glucocorticoid therapy [177]. At the minimum, for the patient-at-risk, serum calcium, phosphate, alkaline phosphatase measurements, and bone densitometry should be obtained. Signs of metabolic bone disease should lead to a referral to an endocrinologist for aggressive management. The risk for abnormal bone metabolism and nephrolithiasis is elevated in IBD patients with hyperoxaluria and hypercalciuria [178]. Diarrhea should be given a specific diagnosis as well as quantified in volume. Factors contributing to diarrhea in the IBD patient include concurrent mucosal disease, small intestinal bacterial overgrowth, and rapid gastrointestinal transit. Proteinlosing enteropathy can be evaluated by direct measurement of protein content in whole-gut lavage [90] or more practically, by elevated oci-antitrypsin clearance [92]. Disaccharidase deficiencies can be

Nutrition in inflammatory bowel disease detected with carbohydrate challenge testing, measuring breath excretion of gaseous byproducts of carbohydrate fermentation [94, 95]. The lactulose breath hydrogen test [179-181] can be used to distinguish between small intestinal bacterial overgrowth and rapid gastrointestinal transit by carefully examining the timing and magnitude of the rise in breath hydrogen in response to oral lactulose and by comparing breath hydrogen profiles before and after antibiotic treatment. In an unoperated patient, rapid transit will show a premature rise of modest magnitude with a hydrogen profile demonstrating usually only a single peak, whereas small intestinal bacterial overgrowth typically manifests as a premature rise of great magnitude with a hydrogen profile demonstrating double peaks (small and large intestine). The picture is complicated in the IBD patient with ileocolonic resection, as rapid transit may coexist with bacterial overgrowth. In this setting it is advisable to treat the patient with a short course of poorly absorbed antibiotic [182] and re-evaluate after a time period suitable for recolonization by normal flora. If bacterial overgrowth was the primary cause of diarrhea there should be resolution of symptoms and a normalized breath hydrogen test. If rapid transit were to be an important contributor to diarrhea, then rapid transit will still show in the breath hydrogen test as an early and moderate rise in breath hydrogen. Although a potential drawback of the lactulose breath hydrogen test for quantifying rapid transit is that oral lactulose may contribute to a shorter absolute value for the transit time [181] by increasing the osmotic load, this eff'ect does not usually limit the usefulness of the test. Within each diagnostic unit it is the relative rather than the absolute value of the transit time that distinguishes a normal from an abnormal result. Glucose may be used instead of lactulose for the breath hydrogen test [183, 184]. However, since even a large load of glucose can be normally absorbed by the proximal small intestine, complete absorption of the test substance may reduce the sensitivity of the glucose breath hydrogen test for diagnosing small intestinal bacterial overgrowth involving only the distal onehalf of the small bowel.

Treating malnutrition in IBD Regardless of the nutritional approach, treatment should include a psychosocial component. The therapeutic strategy should be developed in coopera-

595

Gregg W. Van Citters and Henry C. Lin tion with the patient. Patients should be educated about their disease and encouraged to build their support systems, including family involvement and self-help groups [157]. A psychiatric referral should be considered if there is suspicion of confounding eating disorder.

Nutritional management in IBD Nutrition has at least two applications in the clinical management of IBD. As a primary therapy, nutrition can be used to induce remission as well as to sustain the patient. As an adjunct therapy, nutrition can correct deficiencies and modulate the immune response to maintain long-term remission.

Nutrition as primary therapy Bowel rest has often been suggested as a therapeutic approach to calming the inflammatory response during a flare of CD, particularly when clinical manifestations include complications such as obstructive or fistulating disease. Specifically, the role of bowel rest in inducing or maintaining a remission has been pursued but remains controversial [50, 147, 185-187]. Greenberg etal. [185] randomized 51 CD patients in acute flare who were unresponsive to medical management to receive total parenteral nutrition (TPN) only, enteral nutrition (EN) only, or a mixture of oral and parenteral nutrition. They found no difference between groups in remission induction or maintenance of remission at 1 year. Similarly, Wright and Adler [147] found no eflfect of nutrient delivery route on the short-term outcome of CD patients hospitalized for disease exacerbations. Pediatric CD patients showed similar rates of remission and relapse but had improved relapse and linear growth rates when nocturnal EN feedings were continued in conjunction with the resumption of normal oral intake [50]. When TPN plus bowel rest was coupled with aggressive medical management [186], remission was induced and maintained in most CD patients (UC patients overwhelmingly required surgical intervention). Within 1 week following major upper gastrointestinal surgery there was no clinical difference detected in patients receiving TPN plus bowel rest compared with patients receiving only EN [187]. TPN is typically the preferred nutritional support method for CD patients with active disease or complications. TPN may result in a more favorable outcome where bowel obstruction is a prominent

finding [145]. TPN may also improve recovery from bowel resections when used perioperatively [188]. Caution should be used in interpreting clinical signs when TPN is given in the context of a CD flare, as primary acute-phase indicators may be falsely improved with TPN [189]. Additionally, TPN plus bowel rest may contribute to mucosal atrophy [190], allowing bacterial overgrowth [191] and translocation [192] and increased release of macrophage tumor necrosis factor (TNF-oc) [193], a putative mediator of the autoimmune response in both UC [23] and CD [16]. Therefore, because TPN therapy is not necessarily associated with less costly outcomes [194], may not avoid surgical interventions, and is associated with significant morbidity and mortality [191, 195, 196], the anticipated benefits of TPN therapy must be carefully weighed against these risks. Enteral nutrition alone, whether elemental [197], polymeric [196,198], or whole protein [45], may be as effective as bowel rest plus TPN for inducing or maintaining remission in CD [147, 199, 200]. However, it remains controversial whether EN may be as effective as steroid therapy in inducing [201-206] or maintaining [46, 197, 202, 207, 208] remission. Meta-analyses of studies involving EN and/or steroid therapy [201, 204, 207] found that EN was no better than, and often inferior to, steroid therapy. Peptide-based diets may be as effective as prednisolone in improving CD activity scores after brief periods of treatment [202, 205], although steroid therapy is clearly superior in the long term. Larger and longer studies controlled for disease activity, location, and bowel resection location and extent are needed to better determine whether EN is effective in inducing or maintaining remission of IBD.

Nutrition as adjunct therapy The above studies suggest that nutritional support, regardless of route of delivery, can be eff^ective primary therapy for IBD for brief periods. Because malnutrition and its detrimental effects are so prevalent in IBD, supplemental nutrition may improve patient outcomes over extended periods. Although nutritional interventions are not conclusively superior to medical management of IBD, they may enhance the capability of drugs such as steroids or immunomodulators to induce or maintain remission [186, 195, 209-211]. TPN is most eff'ective as an adjunct therapy when absorptive capacity is compromised, as with active jejunoileal disease or short

596 bowel [55, 186], or in the setting of chronic intestinal obstruction [145]. EN may not result in better outcomes than TPN [187, 199], although protein stores may improve with EN [199]. Therefore, because cost and potential for complications are greater with TPN, EN should be used when conditions permit. As an adjunct treatment, EN may improve clinical course by maintaining intestinal bacteria homeostasis [191, 192] or by improving latent malnutrition. Malnutrition may be a particularly important factor in children, who exhibited improved linear growth with EN compared to steroids [46] and with nocturnal EN feedings supplementing daily oral intake [50]. Additional factors such as disease status [206], prior interventions, and absorptive capacity may determine the efficacy of enteral feedings in managing CD. A major potential problem with defined diets is that they may lack sufficient trace elements, vitamins, and other important nutrients [52, 212-218] that are normal constituents of a varied diet. Thus, the ultimate nutritional goal should be to normalize the patient's diet to oral intake. One strategy for introducing whole foods is the use of food challenges to personally tailor an exclusion diet [200]. However, success rates for this approach to maintaining remission [34, 200, 219] are limited. There are insufficient data currently available to support dietary restriction as a long-term therapeutic approach to IBD. Unnecessarily restricting food choices of the patient with well-controlled disease or disease in remission could even backfire and contribute to disordered eating and malnutrition [65].

Conventional nutrition prescriptions Dietary restrictions are frequently recommended to the IBD patient for symptom amelioration [34, 36, 220]. However, imposing limitations on dietary intake may induce iatrogenic malnutrition. Thus, it is critically important that recommendation of unnecessary and potentially harmful dietary restrictions be avoided. In many instances dietary supplementation, not restriction, is more appropriate. Protein-energy malnutrition can be addressed by providing sufficient calories to more than meet total energy expenditure needs. Total energy expenditure is estimated as the mathematical product of the basal metabolic rate and an activity factor; but correction of protein-energy malnutrition requires more than a maintenance-level intake; protein and energy must be ingested in greater quantities in order to correct macronutrient deficiency. Thus, we recommend a

Nutrition in inflammatory bowel disease diet that provides 1^ to 2 times the protein and energy necessary to meet total energy expenditure requirements. Many IBD patients, particularly those with CD, have at least marginal micronutrient deficiencies. Restoration of vitamin levels requires oral supplementation, or parenteral supplementation in severe malabsorption such as that seen in short bowel syndrome. Vitamin B12 may not be absorbed from the intestine in patients with active ileal disease or resection and should be administered by intramuscular injection. Fat-soluble vitamins should be taken along with a fat-containing meal to optimize absorption. Calcium supplements may be required at 1-2 g/ day, particularly if antiresorptive therapies such as bisphosphonates are being taken for loss of bone mineral density, if steatorrhea is present and the patient is at risk for oxaluria.

Alternative nutrition prescriptions: nutriceuticals Essential fatty acids Essential fatty acid deficiency (EFAD), low plasma concentrations of linoleic and/or linolenic acids, has long been considered a feature of CD [221]. Although metabolic pathways exist to produce them, the co-3 polyunsaturated fatty acids (PUFA) eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are now also considered essential [222, 223]. These fatty acids may modulate the inflammatory response by competing with other eicosanoid fatty acids for cyclooxygenase and lipoxygenase [224, 225] and by altering expression or release of cytokines [26]. CD patients without a history of intestinal resection may not have frank deficiency of linoleic and linolenic acids [58, 226] but (JL)-3 PUFA metabolism is clearly disturbed [226, 227]. DHA is lower in plasma from subjects with long-standing clinical remission of CD [227] but DHA is elevated in the setting of active disease or intestinal resection in both UC and CD patients compared to controls [41]. This suggests differential control of PUFA metabolism dependent on disease status, although longitudinal studies with strict dietary controls are required to determine the role of co-3 PUFA metabolism in modulating the inflammatory response. Fish-oil preparations vary in relative amounts of EPA and DHA but may be effective in treating CD [228, 229] and UC [28, 29, 230]. A double-blind, placebo-controlled study offish-oil supplementation

Gregg W. Van Citters and Henry C. Lin

597

demonstrated improved 1-year relapse rates in CD patients with quiescent disease [228]. A similar design demonstrated no difference in 1 -year relapse rates for UC patients given fish-oil supplements, although there was a modest corticosteroid-sparing effect [230]. Although objective indicators of UC activity may be improved by fish-oil supplements [29, 230, 231] there may [225] or may not [232] be subjective improvement. Modulation of eicosanoid formation [233] by fatty acids in fish oil could explain altered activity of cytotoxic cells [224, 231] or cytokine release [26]. However, fish-oil preparations may be contaminated if obtained from livers, may contribute to gastrointestinal disturbances, and may be related to increased risk for hemorrhage, so therapeutic application should be considered experimental and approached with caution.

tying. These conflicting results may be explained by possible selection of patients in the initial study that may have predisposed a patient to favorable outcome with the combination therapy [241], in particular greater bowel remnant length and ratio of bowel length to body weight. Thus, at present, the evidence to support the use of growth hormone and glutamine to encourage mucosal hyperplasia in patients with short bowels remains insufficient. Treatment with herbal remedies is fraught with difficulties. Herbal remedies are notoriously lacking in standards of purity and concentration and, for most remedies, the active ingredient has not been identified with any degree of certainty. Until active ingredients can be identified, and purity and concentration are standardized, herbal therapeutics should be considered highly risky and experimental.

Glutamine and growth hormone

Nutriceutical control of rapid transit

As a major donor of nitrogen in amino acid synthesis and a major fuel source for the mucosa [234, 235], glutamine has been studied as a possible means of improving the intestinal barrier or increasing absorptive area in the setting of small bowel disease or resection, including short bowel syndrome. However, CD patients with impaired intestinal permeability did not exhibit improved permeability, disease status, or nutritional status following 1 month of oral glutamine supplementation [236]. Byrne and colleagues first noted increased lean body mass and protein deposition in patients receiving growth hormone therapy plus standard nutritional therapy after gastrointestinal surgery [237]. Following these promising observations this group studied a combination therapy consisting of growth hormone and glutamine plus a high-carbohydrate diet for 3 weeks in patients with short bowel syndrome [238]. They observed significant improvements in absorption of fluid, electrolytes, and macronutrients in eight patients maintained on EN or TPN compared to two control patients [238]. However, their next study of 47 TPN-dependent short bowel syndrome patients found improvements only in protein absorption and stool output after 4 weeks of the combination therapy [239]. In contrast, Scolapio and colleagues performed a randomized, double-blind placebo-controlled crossover study of the combination therapy for 3 weeks in eight patients [240]. They failed to observe any improvements with combined growth hormone, glutamine, and high-carbohydrate diet beyond slight but significant increases in sodium and potassium absorption and delayed gastric emp-

We have recently developed a novel, nutrient-based therapy to manage diarrhea in IBD patients. Because this therapy is directed specifically at slowing gastrointestinal transit, an important beneficial side-effect of this therapy is increased contact time of the luminal contents with digestive enzymes and absorptive surfaces. In addition to improving diarrhea this therapy is likely to improve nutrition by increasing digestion and absorption of a meal. As described above, the end-products of fat digestion are the most potent nutrient triggers of the intestinal brake systems. To jump-start nutrienttriggered inhibitory feedback we administered orally the trigger for the brake in the form of a premeal drink consisting of a fatty acid mixed in a liquid nutritional supplement to 45 patients with chronic diarrhea including 11 IBD patients and to 7 controls [242]. In both groups, including the IBD patients, gastrointestinal transit time as measured by the lactulose breath hydrogen test was increased significantly by the fatty acid premeal drink. Although patients with the shortest small intestine remnants remain slightly malnourished and reliant on nutritional supplementation, most have enjoyed markedly improved health and quality of life with this nutrientbased treatment. We have also shown recently [243] that abnormally rapid transit contributes to poor bioavailability of an oral drug in IBD patients with chronic diarrhea, and that drug bioavailability can be readily and safely increased with the fatty acid-based therapy.

Nutrition in inflammatory bowel disease

598

The future of nutrition in IBD Many, if not most, patients with IBD suffer from some degree of malnutrition. As such, greater awareness of the problem of malnutrition and the application of more effective treatments are both critically important to achieving the clinician's goal of improved overall health of patients with IBD. A common cause of malnutrition in both active and quiescent IBD may be chronic diarrhea, as chronic diarrhea may be associated with accelerated intestinal transit, a mechanism for poor assimilation of both food and oral drugs. Accordingly, the future of the management of malnutrition is to deal with both the issues of what the patient needs as well as how to get it in.

References 1.

2.

3. 4. 5. 6. 7. 8.

9.

10. 11. 12. 13.

Anonymous. Dietary and other risk factors of ulcerative colitis. A case control study in Japan. Epidemiology Group of the Research Committee of Inflammatory Bowel Disease in Japan. J Clin Gastroenterol 1994; 19:166 71. Anonymous. A case control study of ulcerative colitis in relation to dietary and other factors in Japan. Epidemiology Group of the Research Committee of Inflammatory Bowel Disease in Japan. J Gastroenterol 1995; 3()(Suppl. 8):9 12. O'Morain C, Tobin A, Suzuki Y, O'Riordan T. Risk factors in inflammatory bowel disease. Scand J Gastroenterol (Suppl.) 1989; 170: 58 60. Persson PG, Ahlbom A, Hellers G. Diet and inflammatory bowel disease: a case control study. Epidemiology 1992; 3: 47 52. Probert CS, Bhakta P, Bhamra B, Jayanthi V, Mayberry JF. Diet of South Asians with inflammatory bowel disease. Arq Gastroenterol 1996; 33: 132-5. Reif S, Klein I, Lubin F, Farbstein M, Hallak A, Gilat T. Pre-illness dietary factors in inflammatory bowel disease. Gut 1997; 40: 754 60. Russel MG, Stockbrugger RW. Epidemiology of inflammatory bowel disease: an update. Scand J Gastroenterol 1996; 31:417-27. Russel MG, Engels LG, Muris JW et al. Modern life in the epidemiology of inflammatory bowel disease: a case-control study with special emphasis on nutritional factors. Eur J Gastroenterol Hepatol 1998; 10: 243-9. Shoda R, Matsueda K, Yamato S, Umeda N. Epidemiologic analysis of Crohn disease in Japan: increased dietary intake of n-6 polyunsaturated fatty acids and animal protein relates to the increased incidence of Crohn disease in Japan. Am J ChnNutr 1996;63:741-5. Tragnone A, Valpiani D, Miglio F et al. Dietary habits as risk factors for inflammatory bowel disease. Eur J Gastroenterol Hepatol 1995; 7: 47-51. Achord JL, Gunn CHJ, Jackson JF. Regional enteritis and HLA concordance in multiple siblings. Dig Dis Sci 1982; 27: 330-2. Roth MP, Petersen GM, McElree C, Feldman E, Rotter JL Geographic origins of Jewish patients with inflammatory bowel disease. Gastroenterology 1989; 97: 900-4. Orholm M, Iselius L, Sorensen TI, Munkholm P, Langholz E, Binder V. Investigation of inheritance of chronic inflam-

14.

15. 16.

17.

18. 19.

20.

21.

22.

23. 24.

25.

26.

27. 28. 29. 30.

31.

matory bowel diseases by complex segregation analysis. Br Med J 1993;306:20-4. Yang H, McElree C, Roth MP, Shanahan F, Targan SR, Rotter JI. Familial empirical risks for inflammatory bowel disease: differences between Jews and non-Jews. Gut 1993; 34:517-24. Gilberts EC, Greenstein AJ, Katsel P, Harpaz N, Greenstein RJ. Molecular evidence for two forms of Crohn disease. Proc Natl Acad Sci USA 1994; 91: 12721^. Plevy SE, Targan SR, Yang H, Fernandez D, Rotter JI, Toyoda H. Tumor necrosis factor microsatellites define a Crohn's disease-associated haplotype on chromosome 6. Gastroenterology 1996; 110: 1053-60. Ohmen JD, Yang HY, Yamamoto KK et al. Susceptibility locus for inflammatory bowel disease on chromosome 16 has a role in Crohn's disease, but not in ulcerative colitis. Hum Mol Genet 1996; 5: 1679-83. Reinshagen M, Loeliger C, Kuehnl P ct al. HLA class II gene frequencies in Crohn's disease: a population based analysis in Germany. Gut 1996; 38: 538 42. De La Concha EG, Fernandez-Arquero M, Santa-Cruz S ct al. Positive and negative associations of distinct HLA-DR2 subtypes with ulcerative colitis (UC). Clin Exp Immunol 1997; 108:392 5. McLeod RS, Stcinhart AH, Siminovitch KA ct al. Preliminary report on the Mount Sinai Hospital Inflammatory Bowel Disease Genetics Project. Dis Colon Rectum 1997; 40: 553 7. Mirza MM, Lee J, Teare D ct al. Evidence of linkage of the inflammatory bowel disease susceptibility locus on chromosome 16 (IBDl) to ulcerative colitis. J Med Genet 1998; 35: 218 21. Bouma G, Oudkerk PM, Crusius JB ct al. Evidence for genetic heterogeneity in inflammatory bowel disease (IBD); HLA genes in the predisposition to suflcr from ulcerative colitis (UC) and Crohn's disease (CD). Clin Exp Immunol 1997; 109: 175 9. Bouma G, Crusius JB, Garcia-Gonzalez MA ct al. Genetic markers in clinically well defined patients with ulcerative colitis (UC). Clin Exp Immunol 1999; 115: 294 300. Cavanaugh JA, Callen DF, Wilson SR ct al. Analysis of Australian Crohn's disease pedigrees refines the localization for susceptibility to inflammatory bowel disease on chromosome 16. Ann Hum Genet 1998; 62: 291 8. Papo M, Quer JC, Gutierrez C ct al. Genetic heterogeneity within ulcerative colitis determined by an interleukin-1 receptor antagonist gene polymorphism and antineutrophil cytoplasmic antibodies. Eur J Gastroenterol Hepatol 1999; 11:413-20. Endres S, Ghorbani R, Kelley VE ct al. The effect of dietary supplementation with n-3 polyunsaturated fatty acids on the synthesis of interleukin-1 and tumor necrosis factor by mononuclear cells. N Engl J Med 1989; 320: 265-71. Kim YL Can fish oil maintain Crohn's disease in remission? NutrRev 1996; 54:248-52. Salomon P, Kornbluth AA, Janowitz HD. Treatment of ulcerative colitis with fish oil n-3-omega-fatty acid: an open trial. J Clin Gastroenterol 1990; 12: 157-61. Stenson WF, Cort D, Rodgers J et al. Dietary supplementation with fish oil in ulcerative colitis. Ann Intern Med 1992; 116:609-614. Lorenz-Meyer H, Bauer P, Nicolay C et al. Omega-3 fatty acids and low carbohydrate diet for maintenance of remission in Crohn's disease. A randomized controfled multicenter trial. Study Group Members (German Crohn's Disease Study Group). Scand J Gastroenterol 1996; 31: 778-85. Burr ML, Merrett TG. Food intolerance: a community survey. Br J Nutr 1983; 49: 217-19.

Gregg W. Van Citters and Henry C. Lin 32.

33.

34.

35. 36.

37. 38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

Young E, Stoneham MD, Petruckevitch A, Barton J, Rona R. A population study of food intolerance. Lancet 1994; 343: 1127-30. Bischoff SC, Herrmann A, Manns MR Prevalence of adverse reactions to food in patients with gastrointestinal disease. Allergy 1996; 51: 811-18. Giaffer MH, Cann P, Holdsworth CD. Long-term effects of elemental and exclusion diets for Crohn's disease. Alimen Pharmacol Ther 1991; 5: 115-25. Pearson M, Teahon K, Levi AJ, Bjarnason L Food intolerance and Crohn's disease. Gut 1993; 34: 783-7. Ballegaard M, Bjergstrom A, Brondum S, Hylander E, Jensen L, Ladefoged K. Self-reported food intolerance in chronic inflammatory bowel disease. Scand J Gastroenterol 1997;32:569-71. Dieleman LA, Heizer WD. Nutritional issues in inflammatory bowel disease. Gastroenterol Clin N Am 1998; 27: 43551. Han PD, Burke A, Baldassano RN, Rombeau JL, Lichtenstein GR. Nutrition and inflammatory bowel disease. Gastroenterol Clin N Am 2000; 28: 423-43. Sturniolo GC, Mestriner C, Lecis PE et al. Altered plasma and mucosal concentrations of trace elements and antioxidants in active ulcerative colitis. Scand J Gastroenterol 1998; 33: 644-9. Geerling BJ, Badart-Smook A, Stockbrugger RW, Brummer RJ. Comprehensive nutritional status in patients with longstanding Crohn disease currently in remission. Am J Clin Nutr 1998; 67: 919-26. Esteve-Comas M, Nunez MC, Fernandez-Banares F et al. Abnormal plasma polyunsaturated fatty acid pattern in non-active inflammatory bowel disease. Gut 1993; 34: 1370-3. Dady IM, Thomas AG, Miller V, Kelsey AJ. Inflammatory bowel disease in infancy: an increasing problem? J Pediatr Gastroenterol Nutr 1996; 23: 569-76. Seidman E, LeLeiko N, Ament M et al. Nutral issues in pediatric inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1991; 12:424-38. Thomas AG, Taylor F, Miller V. Dietary intake and nutritional treatment in childhood Crohn's disease. J Pediatr Gastroenterol Nutr 1993; 17: 75-81. Ruuska T, Savilahti E, Maki M, Ormala T, Visakorpi JK. Exclusive whole protein enteral diet versus prednisolone in the treatment of acute Crohn's disease in children. J Pediatr Gastroenterol Nutr 1994; 19: 175-80. Papadopoulou A, Rawashdeh MO, Brown GA, McNeish AS, Booth IW. Remission following an elemental diet or prednisolone in Crohn's disease. Acta Paediatr 1995; 84: 7983. Polk DB, Hattner JA, Kerner JA, Jr. Improved growth and disease activity after intermittent administration of a defined formula diet in children with Crohn's disease. J Parent Ent Nutr 1992; 16:499-504. Beattie RM, Schiff"rin EJ, Donnet-Hughes A et al. Polymeric nutrition as the primary therapy in children with small bowel Crohn's disease. Alimen Pharmacol Ther 1994; 8: 609-15. Khoshoo V, Reifen R, Neuman MG, Griffiths A, Pencharz PB. Effect of low- and high-fat, peptide-based diets on body composition and disease activity in adolescents with active Crohn's disease. J Parent Ent Nutr 1996; 20: 401-5. Wilschanski M, Sherman P, Pencharz P, Davis L, Corey M, Griffiths A. Supplementary enteral nutrition maintains remission in paediatric Crohn's disease. Gut 1996; 38: 5438. Rossi TM, Lee PC, Young C, Tjota A. Small intestinal mucosa changes, including epithelial cell proliferative activ-

599

52.

53. 54. 55.

56.

57.

58.

59.

60. 61. 62.

63. 64.

65.

66.

67.

68.

69.

70.

71.

ity, of children receiving total parenteral nutrition (TPN). DigDisSci 1993; 38: 1608-13. Hirose S, Hirata M, Azuma N, Shirai Z, Mitudome A, Oda T. Carnitine depletion during total parenteral nutrition despite oral L-carnitine supplementation. Acta Paediatr Jpn 1997; 39: 194-200. Gagliardi E, Brathwaite CE. Nutral implications of significant small bowel resection. N Jersey Med 1995; 92: 155-7. Nightingale JM. The short-bowel syndrome. Eur J Gastroenterol Hepatol 1995; 7: 514-20. Nightingale JM, Lennard-Jones JE, Gertner DJ, Wood SR, Bartram CI. Colonic preservation reduces need for parenteral therapy, increases incidence of renal stones, but does not change high prevalence of gall stones in patients with a short bowel. Gut 1992; 33: 1493-7. Woolf GM, Miller C, Kurian R, Jeejeebhoy K N Nutral absorption in short bowel syndrome. Evaluation of fluid, calorie, and divalent cation requirements. Dig Dis Sci 1987; 32: 8-15. Engels LG, van den Hamer CJ, van Tongeren JH. Iron, zinc, and copper balance in short bowel patients on oral nutrition. Am J Clin Nutr 1984; 40: 1038-41. Edes TE, Walk BE, Thornton WHJ, Fritsche KL. Essential fatty acid sufficiency does not preclude fat-soluble-vitamin deficiency in short-bowel syndrome. Am J Clin Nutr 1991; 53: 499-502. Someya N, Yoshida T, Haraguchi Y, Tanaka K. A case of short bowel syndrome in which remarkable amelioration of mental deterioration was achieved by administration of trace elements. Gastroenterol Jpn 1985; 20: 465-9. Hiroi K, Goto Y, Ishikawa J, Kida K, Matsuda H. A case of beriberi accompanying short bowel. Acta Paediatr Jpn 1995; 37: 84-7. Gryboski JD. Eating disorders in inffammatory bowel disease. Am J Gastroenterol 1993; 88: 293-6. Holaday M, Smith KE, Robertson S, Dallas J. An atypical eating disorder with Crohn's disease in a fifteen-year-old male: a case study. Adolescence 1994; 29: 865-73. Mallett P, Murch S. Anorexia nervosa complicating inflammatory bowel disease. Arch Dis Childh 1990; 65: 298-300. Musch MW, Chang EB. Diarrhea in inflammatory bowel diseases. In: Targan SR, Shanahan F, eds. Inflammatory Bowel Disease: From Bench to Bedside. Baltimore, MD: Williams Wilkins, 1994: 239-54. Rigaud D, Angel LA, Cerf M et al. Mechanisms of decreased food intake during weight loss in adult Crohn's disease patients without obvious malabsorption. Am J Clin Nutr 1994;60:775-81. Mufler MJ, Schmidt LU, Korber J et al. Reduced metabolic efficiency in patients with Crohn's disease. Dig Dis Sci 1993; 38: 2001-9. Chan AT, Fleming CR, O'Fallon WM, Huizenga KA. Estimated versus measured basal energy requirements in patients with Crohn's disease. Gastroenterology 1986; 91: 75-8. Mingrone G, Greco AV, Benedetti G et al. Increased resting lipid oxidation in Crohn's disease. Dig Dis Sci 1996; 41: 726. Stokes MA, Hill GL. Total energy expenditure in patients with Crohn's disease: measurement by the combined body scan technique. J Parent Ent Nutr 1993; 17: 3-7. Kane SV, Sable K, Hanauer SB. The menstrual cycle and its effect on inflammatory bowel disease and irritable bowel syndrome: a prevalence study. Am J Gastroenterol 1998; 93: 1867-72. Annese V, Bassotti G, Napolitano G, Usai P, Andriulli A, Vantrappen G. Gastrointestinal motility disorders in patients with inactive Crohn's disease. Scand J Gastroenterol 1997;32: 1107-17.

Nutrition in inflammatory bowel disease

600 72. 73.

74.

75.

76. 77.

78. 79.

80. 81. 82. 83.

84. 85.

86.

87.

88.

89. 90.

Fallingborg J, Pedersen P, Jacobsen BA. Small intestinal transit time and intraluminal pH in ileocecal resected patients with Crohn's disease. Dig Dis Sci 1998; 43: 702-5. McHugh K, Castonguay TW, Collins SM, Weingarten HP. Characterization of suppression of food intake following acute colon inflammation in the rat. Am J Physiol 1993; 265: RlOOl-5. Toft-Nielsen MB, Madsbad S, Hoist JJ. Continuous subcutaneous infusion of glucagon-like peptide 1 lowers plasma glucose and reduces appetite in type 2 diabetic patients. Diabetes Care 1999; 22: 1137-43. Koch TR, Roddy DR, Go VL. Abnormalities of fasting serum concentrations of peptide YY in the idiopathic inflammatory bowel diseases. Am J Gastroenterol 1987; 82: 321-6. Wiklund B, Liljeqvist L, Rokaeus A. Neurotensin increases net fluid secretion and transit rate in the small intestine of man. Reg Pept 1984; 8: 33-9. Lavin JH, Wittert GA, Andrews J et al. Interaction of insulin, glucagon-like peptide 1, gastric inhibitory polypeptide, and appetite in response to intraduodenal carbohydrate. Am J Clin Nutr 1998; 68: 591 8. Woods SC, Gibbs J. The regulation of food intake by peptides. Ann NY Acad Sci 1989; 575: 236 43. Zimmerman J. Drug interactions in intestinal transport of folic acid and methotrexate. Further evidence for the heterogeneity of folate transport in the human small intestine. Biochcm Pharmacol 1992; 44: 1839 42. Reid IR. Glucocorticoid osteoporosis mechanisms and management. Eur J Endocrinol 1997; 137: 209 17. Haave NC, Innis SM. Effects of cholestyramine feeding on tissue lipase activities and plasma fatty acids in the pregnant rat. J Dcv Physiol 1989; 12: 11 14. Knodel LC, Talbert RL. Adverse effects of hypolipidacmic drugs. Med Toxicol 1987; 2: 10 32. Watkins DW, Khalafi R, Cassidy MM, Vahouny GV. Alterations in calcium, magnesium, iron, and zinc metabolism by dietary cholestyramine. Dig Dis Sci 1985; 30: 477 82. Hanauer SB, Stathopoulos G. Risk benefit assessment of drugs used in the treatment of inflammatory bowel disease. Drug Safety 1991; 6: 192-219. Singleton JW, Law DH, Kelley ML Jr, Mekhjian HS, Sturdevant RA. National Cooperative Crohn's Disease Study: adverse reactions to study drugs. Gastroenterology 1979; 77: 870-82. Pironi L, Cornia GL, Ursitti MA et al. Evaluation of oral administration of folic and folinic acid to prevent folate deficiency in patients with inflammatory bowel disease treated with salicylazosulfapyridine. Int J Clin Pharmacol Res 1988; 8: 143-8. Lyoumi S, Tamion F, Petit J et al. Induction and modulation of acute-phase response by protein malnutrition in rats: comparative effect of systemic and localized inflammation on interleukin-6 and acute-phase protein synthesis. J Nutr 1998; 128: 166-74. Cederholm T, Wretlind B, Hellstrom K et al. Enhanced generation of interleukins 1 beta and 6 may contribute to the cachexia of chronic disease. Am J Clin Nutr 1997; 65: 876-82. Beisel WR. Herman Award Lecture, 1995: infection-induced malnutrition-from cholera to cytokines. Am J Clin Nutr 1995;62:813-19. Acciuffi S, Ghosh S, Ferguson A. Strengths and limitations of the Crohn's disease activity index, revealed by an objective gut lavage test of gastrointestinal protein loss. Aliment Pharmacol Ther 1996; 10: 321-6.

91. 92.

93. 94. 95. 96. 97.

98. 99. 100. 101. 102.

103.

104. 105.

106. 107. 108. 109. 110. 111.

Baert D, Wulfrank D, Burvenich P, Lagae J. Lymph loss in the bowel and severe nutritional disturbances in Crohn's disease. J Clin Gastroenterol 1999; 29: 277-9. Karbach U, Ewe K, Bodenstein H. Alpha 1-antitrypsin, a reliable endogenous marker for intestinal protein loss and its application in patients with Crohn's disease. Gut 1983; 24:718-23. Arvanitakis C. Abnormalities of jejunal mucosal enzymes in ulcerative colitis and Crohn's disease. Digestion 1979; 19: 259-66. Bernstein CN, Ament M, Artinian L, Ridgeway J, Shanahan F. Milk tolerance in adults with ulcerative colitis. Am J Gastroenterol 1994; 89: 872-7. Marteau P, Messing B, Arrigoni E et al. Do patients with short-bowel syndrome need a lactose-free diet?. Nutr 1997; 13: 13-16. Mishkin S. Dairy sensitivity, lactose malabsorption, and elimination diets in inflammatory bowel disease. Am J Clin Nutr 1997; 65: 564-7. Ziambaras T, Rubin DC, Perlmutter DH. Regulation of sucrase-isomaltase gene expression in human intestinal epithelial cells by inflammatory cytokines. J Biol Chem 1996;271: 1237 42. Saltzman JR, Russell RM. Nutral consequences of intestinal bacterial overgrowth. ComprTher 1994; 20: 523 30. Hoverstad T, Bjorneklett A, Fausa O, Midtvedt T. Shortchain fatty acids in the small-bowel bacterial overgrowth syndrome. Scand J Gastroenterol 1985; 20: 492 9. Bjorneklett A, F'ausa O, Midtvedt T. Bacterial overgrowth in jejunal and ileal disease. Scand J Gastroenterol 1983; 18: 289 98. Thompson JS, Quiglcy EM, Adrian TE, Path FR. Role of the ileocecal junction in the motor response to intestinal resection. J Gastrointcst Surg 1998; 2: 174 85. Thompson JS, Quigley EM, Adrian TE. Factors affecting outcome following proximal and distal intestinal resection in the dog: an examination of the relative roles of mucosal adaptation, motility, luminal factors, and enteric peptides. Dig Dis Sci 1999; 44: 63 74. Riordan SM, Mclver CJ, Wakefield D, Bolin TD, Duncombe VM, Thomas MC. Small intestinal bacterial overgrowth in the symptomatic elderly. Am J Gastroenterol 1997;92:47-51. Hauge T, Persson J, Danielsson D. Mucosal bacterial growth in the upper gastrointestinal tract in alcoholics (heavy drinkers). Digestion 1997; 58: 591-5. Morris TH, Sorensen SH, Turkington J, Batt RM. Diarrhoea and increased intestinal permeability in laboratory beagles associated with proximal small intestinal bacterial overgrowth. Lab Anim 1994; 28: 313-19. MacMahon M, Lynch M, Mullins E et al. Small intestinal bacterial overgrowth - an incidental finding? J Am Geriatr Soc 1994; 42: 146-9. May GR, Sutherland LR, Meddings JB. Is small intestinal permeability really increased in relatives of patients with Crohn's disease?. Gastroenterology 1993; 104: 1627-32. Hollander D. Permeability in Crohn's disease: altered barrier functions in healthy relatives?. Gastroenterology 1993; 104: 1848-51. Riordan SM, Mclver CJ, Thomas DH, Duncombe VM, Bolin TD, Thomas MC. Luminal bacteria and small-intestinal permeability. Scand J Gastroenterol 1997; 32: 556-63. Yamazaki Y, Fukushima T, Sugita A, Takemura H, Tsuchiya S. The medical, nutritional and surgical treatment of fistulae in Crohn's disease. Jpn J Surg 1990; 20: 376-83. Dardai E, Pirityi S, Nagy L. Parenteral and enteral nutrition and the enterocutaneous fistula treatment. II. Factors influencing the outcome of treatment. Acta Chirurg Hung 1991; 32: 305-18.

Gregg W. Van Citters and Henry C. Lin 112. Hitosugi M, Kitamura O, Takatsu A. Sudden death of a patient with Crohn's disease. Nippon Hoigaku Zasshi (Japanese Journal of Legal Medicine) 1998; 52: 211-14. 113. Thompson JS. Management of the short bowel syndrome. Gastroenterology Clin N Am 1994; 23: 403-20. 114. Fromm H, Malavolti M. Bile acid-induced diarrhoea. Clin Gastroenterol 1986; 15: 567-82. 115. Hafkenscheid JC. Influence of bile acids on the (Na+-K+)activated- and Mg2+-activated ATPase of rat colon. Pflugers Archiv Eur J Physiol 1977; 369: 203-6. 116. Kirwan WO, Smith AN, Mitchell WD, Falconer JD, Eastwood MA. Bile acids and colonic motility in the rabbit and the human. Gut 1975; 16: 894-902. 117. Krag B, Krag E. Regional ileitis (Crohn's disease). II. Electrolyte and water movement in the ileum during perfusion with bile acids. Scand J Gastroenterol 1976; 11: 48790. 118. Kruis W, Haddad A, Phillips SF. Chenodeoxycholic and ursodeoxycholic acids alter motility and fluid transit in the canine ileum. Digestion 1986; 34: 185-95. 119. Hofmann AF, Poley JR. Cholestyramine treatment of diarrhea associated with ileal resection. N Engl J Med 1969;281:397-402. 120. Nyhlin H, Merrick MV, Eastwood MA. Bile acid malabsorption in Crohn's disease and indications for its assessment using SeHCAT. Gut 1994; 35: 90-3. 121. Bochenek W, Rodgers JBJ, BaHnt JA. Effects of changes in dietary lipids on intestinal fluid loss in the short-bowel syndrome. Ann Intern Med 1970; 72: 205-13. 122. Hofmann AF, Poley JR. Role of bile acid malabsorption in pathogenesis of diarrhea and steatorrhea in patients with ileal resection. I. Response to cholestyramine or replacement of dietary long chain triglyceride by medium chain triglyceride. Gastroenterology 1972; 62: 918-34. 123. Borgstrom B, Lundh G, Hofmann A. The site of absorption of conjugated bile salts in man. Gastroenterology 1963; 45: 229-38. 124. Barrett KE, Dharmsathaphorn K. Pharmacological aspects of therapy in inflammatory bowel diseases: antidiarrheal agents. J Clin Gastroenterol 1988; 10: 57-63. 125. Read N W Diarrhee motrice. Clin Gastroenterol 1986; 15: 657-86. 126. Read NW, Miles CA, Fisher D et al Transit of a meal through the stomach, small intestine, and colon in normal subjects and its role in the pathogenesis of diarrhea. Gastroenterology 1980; 79: 1276-82. 127. Binder HJ. Pathophysiology of acute diarrhea. Am J Med 1990; 88: 2-4S. 128. Chang EB, Sitrin MD, Black DD. Gastrointestinal, Hepatobiliary, and Nutritional Physiology. Philadelphia, PA: Lippincott-Raven Pubhshers, 1996. 129. Kremen AJ, Linner JH, Nelson CH. An experimental evaluation of the nutritional importance of proximal and distal smaU intestine. Ann Surg 1954; 140: 439-48. 130. Reynell P, Spray G Small intestinal function in the rat after massive resections. Gastroenterology 1956; 31: 361-8. 131. Read NW, McFarlane A, Kinsman RI et al. Effect of infusion of nutrient solutions into the ileum on gastrointestinal transit and plasma levels of neurotensin and enteroglucagon. Gastroenterology 1984; 86: 274-80. 132. Spiller RC, Trotman IF, Higgins BE et al. The ileal brakeinhibition of jejunal motility after ileal fat perfusion in man. Gut 1984; 25: 365-74. 133. Spiller RC, Trotman IF, Adrian TE et al. Further characterisation of the 'ileal brake' reflex in man - effect of ileal infusion of partial digests of fat, protein, and starch on jejunal motility and release of neurotensin, enteroglucagon, and peptide YY. Gut 1988; 29: 1042-51.

601 134. Jeejeebhoy KN. Short bowel syndrome. In: Kinney JM, Jeejeebhoy KN, Hill GL, Owen OE, eds. Nutrition and MetaboHsm in Patient Care. Philadelphia, PA: W.B. Saunders, 1992: 259-65. 135. Lin HC, Zhao XT, Reddy SN, Wang LJ, Gu Y-G. Inhibition of intestinal transit by fat depends on length of exposure to nutrient. Gastroenterology 1994; 106, A531 (Abstract). 136. Anonymous. Short bowel syndrome [Clinical Conference]. Gastroenterology 1979; 77: 572-9. 137. Lin HC, Zhao XT, Wang L. Jejunal brake: inhibition of intestinal transit by fat in the proximal small intestine. Dig DisSci 1996; 41: 326-9. 138. Wen J, Luque-de LE, Kost LJ, Sarr MG, Phillips SF. Duodenal motility in fasting dogs: humoral and neural pathways mediating the colonic brake. Am J Physiol 1998; 274: G192-5. 139. Lin HC, Zhao XT, Wang L. Intestinal transit is more potently inhibited by fat in the distal (ileal brake) than in the proximal (jejunal brake) gut. Dig Dis Sci 1997; 42: 1925. 140. Wen J, Phillips SF, Sarr MG, Kost LJ, Hoist JJ. PYY and GLP-1 contribute to feedback inhibition from the canine ileum and colon. Am J Physiol 1995; 269: G945-52. 141. Herbst F, Kamm MA, Nicholls RJ. Effects of loperamide on ileoanal pouch function. Br J Surg 1998; 85: 1428-32. 142. Fonkalsrud EW, Bustorff-Silva J. Reconstruction for chronic dysfunction of ileoanal pouches. Ann Surg 1999; 229:197-204. 143. BeUiveau P, Trudel J, Vasilevsky CA, Stein B, Gordon PH. Ileoanal anastomosis with reservoirs: complications and long-term resuhs. Can J Surg 1999; 42: 345-52. 144. Welsh FK, Farmery SM, MacLennan K et al. Gut barrier function in malnourished patients. Gut 1998; 42: 396-401. 145. Cravo M, Camilo ME, Correia JP. Nutral support in Crohn's disease: which route? Am J Gastroenterol 1991; 86:317-21. 146. Greenberg GR. Nutral support in inflammatory bowel disease: current status and future directions. Scand J Gastroenterol (Suppl.) 1992;192: 117-22. 147. Wright RA, Adler EC. Peripheral parenteral nutrition is no better than enteral nutrition in acute exacerbation of Crohn's disease: a prospective trial. J Clin Gastroenterol 1990; 12: 396-9. 148. Best WR, Becktel JM, Singleton JW, Kern F, Jr. Development of a Crohn's disease activity index. National Cooperative Crohn's Disease Study. Gastroenterology 1976; 70: 439-44. 149. Singleton JW. The Crohn's Disease Activity Index is a useful tool in clinical practice of gastroenterology. Inflam Bowel Dis 1999; 5: 138. 150. Simonis B, Heine M, Heene DL, Gladisch R. Evaluation and validation of a Crohn's disease inflammatory activity index reflecting pattern of endoscopic severity. Scand J Gastroenterol 1998; 33: 283-8. 151. Yoshida EM. The Crohn's Disease Activity Index, its derivatives and the Inflammatory Bowel Disease Questionnaire: a review of instruments to assess Crohn's disease. Can J Gastroenterol 1999; 13: 65-73. 152. Powefl-Tuck J, Bown RL, Lennard-Jones JE. A comparison of oral prednisolone given as single or multiple daily doses for active proctocolitis. Scand J Gastroenterol 1978; 13: 833-7. 153. Walmsley RS, Ayres RC, Pounder RE, Allan RN. A simple clinical colitis activity index. Gut 1998; 43: 29-32. 154. Kirschner BS. Differences in the management of inflammatory bowel disease in children and adolescents compared to adults. Neth J Med 1998; 53: S13-18. 155. Turnbull GK, VaUis TM. Quality of hfe in inflammatory bowel disease: the interaction of disease activity with

Nutrition in inflammatory bowel disease

602

156. 157. 158. 159.

160. 161. 162.

163. 164. 165.

166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176.

177.

psychosocial function. Am J Gastroenterol 1995; 90: 14504. Berin MC, Perdue MH. Effect of psychoneural factors on intestinal epithelial function. Can J Gastroenterol 1997; 11: 353-7. Talal AH, Drossman DA. Psychosocial factors in inflammatory bowel disease. Gastroenterol Clin N Am 1995; 24: 699-716. Guyatt G, Mitchell A, Irvine EJ et ai A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology 1989; 96: 804-10. Drossman DA, Leserman J, Li ZM, Mitchell CM, Zagami EA, Patrick DL. The rating form of IBD patient concerns: a new measure of health status. Psychosom Med 1991; 53: 701 12. Drossman DA, Li Z, Leserman J, Patrick DL. Ulcerative colitis and Crohn's disease health status scales for research and clinical practice. J Clin Gastroenterol 1992; 15: 104 12. Johnston KR, Govel LA, Andritz MH. Gastrointestinal effects of sorbitol as an additive in liquid medications. Am J Med 1994; 97: 185 91. Brcitcnbach RA, Simon J. Cases from the aerospace medicine resident teaching file. Case 59. A case of 'unbearable' grcmlincnteritis [Clinical Conference]. Aviat Space Environ Med 1994:65:432 3. Goldberg LD, Ditchek NT. Chewing gum diarrhea. Am J DigDis 1978:23:568. Lin HC, Sanders SL, Gu YG, Doty JE. Erythromycin accelerates solid emptying at the expense of gastric sieving. DigDisSci 1994; 39: 124 8. Eisenberg DM, Kessler RC, Foster C, Norlock FE, Calkins DR, Delbanco TL. Unconventional medicine in the United States. Prevalence, costs, and patterns of use. N Engl J Med 1993:328: 246 52. Borins M. The dangers of using herbs. What your patients need to know. Postgrad Med 1999; 104: 91 5. Foster S, Tyler VE. Tyler's Honest Herbal: A Sensible Guide to the Use of Herbs and Related Remedies. 4th edn. Binghamton, NY: Haworth Press, Inc., 1999. Wright RA, Heymsfield S. Nutral Assessment. Boston: Blackwcll Scientific Publications, 1984. Bates CJ. Diagnosis and detection of vitamin deficiencies. Br Med Bull 1999; 55: 643 57. Balint JP. Physical findings in nutritional deficiencies. Pediatr Clin N Am 1998; 45: 245 60. Touger-Decker R. Oral manifestations of nutrient deficiencies. Mount Sinai J Med 1998; 65: 355-61. Ehrenpreis ED, Popovich TL, Cravero R. A practical guide to the recognition of nutritional deficiencies. Compr Ther 1997;23:218-22. Ovesen L, Chu R, Howard L. The influence of dietary fat on jejunostomy output in patients with severe short bowel syndrome. Am J Clin Nutr 1983; 38: 270-7. Hylander E, Ladefoged K, Jarnum S. Calcium absorption after intestinal resection. The importance of a preserved colon. Scand J Gastroenterol 1990; 25: 705-10. Pigot F, Roux C, Chaussade S et al. Low bone mineral density in patients with inflammatory bowel disease. Dig DisSci 1992; 37: 1396-403. Silvennoinen J, Lamberg-Allardt C, Karkkainen M, Niemela S, Lehtola J. Dietary calcium intake and its relation to bone mineral density in patients with inflammatory bowel disease. J Intern Med 1996; 240: 285-92. Robinson RJ, al-Azzawi F, Iqbal SJ, Kryswcki T, Almond L, Abrams K et al. Osteoporosis and determinants of bone density in patients with Crohn's disease. Dig Dis Sci 1998; 43: 2500-6.

178. Kuczera M, Wiecek A, Kokot F. Markers of bone turnover in patients with nephrolithiasis. Int Urol Nephrol 1997; 29: 523-8. 179. Staniforth DH, Rose D. Statistical analysis of the lactulose/ breath hydrogen test in the measurement of orocaecal transit: its variability and predictive value in assessing drug action. Gut 1989; 30: 171-5. 180. Riordan SM, Mclver CJ, Walker BM, Duncombe VM, Bolin TD, Thomas MC. The lactulose breath hydrogen test and small intestinal bacterial overgrowth. Am J Gastroenterol 1996;91:1795-803. 181. Miller MA, Parkman HP, Urbain JL et al. Comparison of scintigraphy and lactulose breath hydrogen test for assessment of orocecal transit: lactulose accelerates small bowel transit. Dig Dis Sci 1997; 42: 10-18. 182. Attar A, Flourie B, Rambaud JC et al. Antibiotic efficacy in small intestinal bacterial overgrowth-related chronic diarrhea: a crossover, randomized trial. Gastroenterology 1999; 117:794 7. 183. Sellin JH, Hart R. Glucose malabsorption associated with rapid intestinal transit. Am J Gastroenterol 1992; 87: 584 9. 184. de BD, Chaussain M, Badoual J, Raymond J, Dupont C. Small-bowel bacterial overgrowth in children with chronic diarrhea, abdominal pain, or both. J Pediatr 1996; 128: 203 7. 185. Grcenberg GR, Fleming CR, Jeejccbhoy KN, Rosenberg IH, Sales D, Trcmainc WJ. Controlled trial of bowel rest and nutritional support in the management of Crohn's disease. Gut 1988:29: 1309 15. 186. Sitzmann JV, Converse RLJ, Baylcss TM. Favorable response to parenteral nutrition and medical therapy in Crohn's colitis. A report of 38 patients comparing severe Crohn's and ulcerative colitis. Gastroenterology 1990; 99: 1647 52. 187. Reynolds JV, Kanwar S, Welsh FK et al. 1997 Harry M. Vars Research Award. Docs the route of feeding modify gut barrier function and clinical outcome in patients after major upper gastrointestinal surgery? J Parent Ent Nutr 1997; 21: 196 201. 188. Steffcs C, Fromm D. Is preoperative parenteral nutrition necessary for patients with predominantly ileal Crohn's disease? Arch Surg 1992; 127: 1210 12. 189. Carlson GL, Gray P, Barber D, Shaffer JL, Mughal M, Irving MH. Total parenteral nutrition modifies the acute phase response to Crohn's disease. J Roy Coll Surg Edinb 1994:39:360 4. 190. Mukau L, Talamini MA, Sitzmann JV. Elemental diets may accelerate recovery from total parenteral nutrition-induced gut atrophy. J Parent Ent Nutr 1994; 18: 75-8. 191. Kaufman SS, Loseke CA, Lupo JV et al. Influence of bacterial overgrowth and intestinal inflammation on duration of parenteral nutrition in children with short bowel syndrome. J Pediatr 1997; 131: 356-61. 192. Weber TR. Enteral feeding increases sepsis in infants with short bowel syndrome. J Pediatr Surg 1995; 30: 1086-8. 193. Pappo I, Bercovier H, Berry E, Gallilly R, Feigin E, Freund HR. Antitumor necrosis factor antibodies reduce hepatic steatosis during total parenteral nutrition and bowel rest in the rat. J Parent Ent Nutr 1995; 19: 80-2. 194. Eisenberg JM, Glick HA, Buzby GP, Kinosian B, Williford WO. Does perioperative total parenteral nutrition reduce medical care costs?. J Parent Ent Nutr 1993; 17: 201-9. 195. Galandiuk S, O'Neifl M, McDonald P, Fazio VW, Steiger E. A century of home parenteral nutrition for Crohn's disease. Am J Surg 1990; 159: 540-4. 196. Gonzalez-Huix F, de Leon R, Fernandez-Banares F et al. Polymeric enteral diets as primary treatment of active Crohn's disease: a prospective steroid controlled trial. Gut 1993;34: 778-82.

Gregg W. Van Citters and Henry C. Lin 197. Teahon K, Bjarnason I, Pearson M, Levi AJ. Ten years' experience with an elemental diet in the management of Crohn's disease. Gut 1990; 31: 1133-7. 198. Giaffer MH, North G, Holdsworth CD. Controlled trial of polymeric versus elemental diet in treatment of active Crohn's disease. Lancet 1990; 335: 816-19. 199. Gonzalez-Huix F, Fernandez-Banares F, Esteve-Comas M et al. Enteral versus parenteral nutrition as adjunct therapy in acute ulcerative colitis. Am J Gastroenterol 1993; 88: 227-32. 200. Jones VA. Comparison of total parenteral nutrition and elemental diet in induction of remission of Crohn's disease. Long-term maintenance of remission by personalized food exclusion diets. Dig Dis Sci 1987; 32: 100S-7S. 201. Fernandez-Banares F, Cabre E, Esteve-Comas M, Gassull MA. How effective is enteral nutrition in inducing clinical remission in active Crohn's disease? A meta-analysis of the randomized clinical trials. J Parent Ent Nutr 1995; 19: 35664. 202. Gorard DA, Hunt JB, Payne-James JJ et al. Initial response and subsequent course of Crohn's disease treated with elemental diet or prednisolone. Gut 1993; 34: 1198-202. 203. Lindor KD, Fleming CR, Burnes JU, Nelson JK, Ilstrup DM. A randomized prospective trial comparing a defined formula diet, corticosteroids, and a defined formula diet plus corticosteroids in active Crohn's disease. Mayo Clin Proc1992; 67: 328-33. 204. Messori A, Trallori G, D'Albasio G, Milla M, Vannozzi G, Pacini F. Defined-formula diets versus steroids in the treatment of active Crohn's disease: a meta-analysis. Scand J Gastroenterol 1996; 31: 267-72. 205. Okada M, Yao T, Yamamoto T, Takenaka K, Imamura K, Maeda K et al. Controlled trial comparing an elemental diet with prednisolone in the treatment of active Crohn's disease. Hepato-Gastroenterology 1990; 37: 72-80. 206. Zoli G, Care M, Parazza M et al. A randomized controlled study comparing elemental diet and steroid treatment in Crohn's disease. Alimen Pharmacol Ther 1997; 11: 735^0. 207. Griffiths AM, Ohlsson A, Sherman PM, Sutherland LR. Meta-analysis of enteral nutrition as a primary treatment of active Crohn's disease. Gastroenterology 1995; 108: 105667. 208. O'Brien CJ, Giaffer MH, Cann PA, Holdsworth CD. Elemental diet in steroid-dependent and steroid-refractory Crohn's disease. Am J Gastroenterol 1991; 86: 1614-18. 209. Abad-Lacruz A, Gonzalez-Huix F, Esteve M et al. Liver function tests abnormalities in patients with inflammatory bowel disease receiving artificial nutrition: a prospective randomized study of total enteral nutrition vs total parenteral nutrition. J Parent Ent Nutr 1990; 14: 618-21. 210. Cosgrove M, Jenkins HR. Experience of percutaneous endoscopic gastrostomy in children with Crohn's disease. Arch Dis Childh 1997; 76: 141-3. 211. di Costanzo J, Cano N, Martin J et al. Treatment of external gastrointestinal fistulas by a combination of total parenteral nutrition and somatostatin. J Parent Ent Nutr 1987; 11: 465-70. 212. Fernandez-Banares F, Mingorance MD, Esteve M, et al. Serum zinc, copper, and selenium levels in inflammatory bowel disease: effect of total enteral nutrition on trace element status. Am J Gastroenterol 1990; 85: 1584-9. 213. Goode HF, Robertson DA, Kelleher J, Walker BE. Effect of fasting, self-selected and isocaloric glucose and fat meals and intravenous feeding on plasma zinc concentrations. Ann Clin Biochem 1991; 28(Pt 5): 442-5. 214. Myung SJ, Yang SK, Jung HY et al. Zinc deficiency manifested by dermatitis and visual dysfunction in a patient with Crohn's disease. J Gastroenterol 1998; 33: 876-9.

603 215. Rannem T, Ladefoged K, Hylander E, Hegnhoj J, Staun M. Selenium depletion in patients with gastrointestinal diseases: are there any predictive factors? Scand J Gastroenterol 1998;33:1057-61. 216. Tominaga M, lida M, Aoyagi K, Kohrogi N, Matsui T, Fujishima M. Red cell folate concentrations in patients with Crohn's disease on parenteral nutrition. Postgrad Med J 1989;65:818-20. 217. Touloukian RJ, Gertner JM. Vitamin D deficiency rickets as a late complication of the short gut syndrome during infancy. J Pediatr Surg 1981; 16: 230-5. 218. Zak J, Burns D, Lingenfelser T, Steyn E, Marks IN. Dry beriberi: unusual complication of prolonged parenteral nutrition. J Parent Ent Nutr 1991; 15: 200-1. 219. Riordan AM, Hunter JO, Cowan RE et al. Treatment of active Crohn's disease by exclusion diet: East Anglian multicentre controlled trial. Lancet 1993; 342: 1131-4. 220. Lykins TC, Stockwell J. Comprehensive modified diet simplifies nutrition management of adults with short-bowel syndrome. J Am Diet Assoc 1998; 98: 309-15. 221. Holman RT. Control of polyunsaturated acids in tissue lipids. J Am Coll Nutr 1986; 5: 183-211. 222. Gerster H. Can adults adequately convert alpha-linolenic acid (18: 3n-3) to eicosapentaenoic acid (20: 5n-3) and docosahexaenoic acid (22: 6n-3)?. Int J Vit Nutr Res 1998; 68: 159-73. 223. Spector AA. Essentiality of fatty acids. Lipids 1999; 34(Suppl): S l - 3 . 224. Fisher M, Upchurch KS, Levine PH et al. Effects of dietary fish oil supplementation on polymorphonuclear leukocyte inflammatory potential. Inflammation 1986; 10: 387-92. 225. McCall TB, O'Leary D, Bloomfield J, O'Morain CA. Therapeutic potential of fish oil in the treatment of ulcerative colitis. Aliment Pharmacol Ther 1989; 3: 415-24. 226. Kuroki F, lida M, Matsumoto T, Aoyagi K, Kanamoto K, Fujishima M. Serum n3 polyunsaturated fatty acids are depleted in Crohn's disease. Dig Dis Sci 1997; 42: 1137-41. 227. Geerling BJ, Houwelingen AC, Badart-Smook A, Stockbrugger RW, Brummer, RJ. Fat intake and fatty acid profile in plasma phospholipids and adipose tissue in patients with Crohn's disease, compared with controls. Am J Gastroenterol 1999; 94: 410-17. 228. Belluzzi A, Brignola C, Campieri M, Pera A, Boschi S, Miglioli M. Effect of an enteric-coated fish-oil preparation on relapses in Crohn's disease. N Engl J Med 1996; 334: 1557-60. 229. Belluzzi A, Campieri M, Brignola C, Gionchetti P, Miglioh M, Barbara L. Polyunsaturated fatty acid pattern and fish oil treatment in inflammatory bowel disease [Letter]. Gut 1993;34: 1289-90. 230. Hawthorne AB, Daneshmend TK, Hawkey CJ et al. Treatment of ulcerative colitis with fish oil supplementation: a prospective 12 month randomised controlled trial. Gut 1992; 33: 922-8. 231. Almallah YZ, Richardson S, O'Hanrahan T et al. Distal procto-colitis, natural cytotoxicity, and essential fatty acids. Am J Gastroenterol 1998; 93: 804-9. 232. Greenfield SM, Green AT, Teare JP et al. A randomized controlled study of evening primrose oil and fish oil in ulcerative colitis. Aliment Pharmacol Ther 1993; 7: 159-66. 233. Lee TH, Hoover RL, Williams JD et al. Effect of dietary enrichment with eicosapentaenoic and docosahexaenoic acids on in vitro neutrophil and monocyte leukotriene generation and neutrophil function. N Engl J Med 1985; 312:1217-24. 234. Matthews DE, Marano MA, Campbell RG. Splanchnic bed utihzation of glutamine and glutamic acid in humans. Am J Physiol 1993; 264: E848-54.

604 235. Reeds PJ, Burrin DG, Jahoor F et al. Enteral glutamate is almost completely metabolized in first pass by the gastrointestinal tract of infant pigs [Published errata appear in Am J Physiol 1996; 271: section E following table of contents]. Am J Physiol 1996; 270: E413-18. 236. Den Hond E, Hiele M, Peeters M, Ghoos Y, Rutgeerts P. Effect of long-term oral glutamine supplements on small intestinal permeability in patients with Crohn's disease. J Parent Ent Nutr 1999;23:7-11. 237. Byrne TA, Morrissey TB, Gatzen C et al. Anabolic therapy with growth hormone accelerates protein gain in surgical patients requiring nutritional rehabilitation. Ann Surg 1993; 218:400 16. 238. Byrne TA, Morrissey TB, Nattakom TV, Ziegler TR, Wilmore DW. Growth hormone, glutamine, and a modified diet enhance nutrient absorption in patients with severe short bowel syndrome. J Parent Ent Nutr 1995; 19: 296 302.

Nutrition in inflammatory bowel disease 239. Byrne TA, Persinger RL, Young LS, Ziegler TR, Wilmore, DW. A new treatment for patients with short-bowel syndrome. Growth hormone, glutamine, and a modified diet. Ann Surg 1995;222:243-54. 240. Scolapio JS, Camilleri M, Fleming CR et al. Effect of growth hormone, glutamine, and diet on adaptation in short-bowel syndrome: a randomized, controlled study. Gastroenterology 1997; 113: 1074^81. 241. Wilmore DW, Lacey JM, Soultanakis RP, Bosch RL, Byrne, TA. Factors predicting a successful outcome after pharmacologic bowel compensation. Ann Surg 1997; 226: 288-92. 242. Lin HC, Van Citters GW, Heimer F, Bonorris G. Slowing of gastrointestinal transit by oleic acid: a preliminary report of a novel, nutrient-based treatment in humans. Dig Dis Sci 2001; 46: 223-9. 243. Van Citters GW, Lin HC. Oleic acid improves drug absorption in patients with inflammatory bowel disease. Gastroenterology 1998; 114: A1105 (Abstract).

30 Medical management of ulcerative colitis WILLIAM J. SANDBORN

Introduction The optimal medical treatment of patients with ulcerative colitis (UC) requires that the treating physician obtain a history and perform any necessary diagnostic procedures, and then prescribe an appropriate medication regimen based on a knowledge of clinical pharmacology and the evidence from controlled clinical trials. This chapter will review the results from clinical trials that assessed the efficacy of medications used to treat UC and then provide an integrated therapeutic approach to the medical treatment of specific disease settings for patients with UC.

Pretreatment evaluation of the patient Prior to initiating or altering the medical treatment regimen of patients with UC, the physician should evaluate the patient [1]. This evaluation begins with a medical history to determine the age of onset, the duration of disease, the extent of disease, the disease course over time, prior and current medication use (including duration and dose), and the symptoms currently being experienced by the patient. At the time of diagnosis the patients should undergo colonoscopy with mucosal biopsy and small bowel X-ray to provide a baseline characterization of the UC and to exclude Crohn's disease (CD). Infectious and medication-associated causes of colitis should be excluded. In patients with an estabhshed diagnosis of UC it is useful to repeat these tests when patients relapse and fail to respond to empiric therapy with 5aminosalicylates and/or corticosteroids, prior to instituting immune modifier therapy or referring the patient for surgery. Adherence to this methodical approach allows the treating physician to make observations that would lead to a change in therapy such as: a change in the proximal extent of colitis; endoscopic findings of severe colitis; features that are more compatible with a diagnosis of CD; infectious

Pancolitis

Figure 1. Distinguishing the various states of ulcerative GOlitis - proctitis, proctosigmoiditis, left-sided colitis, and pancolitis - depends on both the degree of mucosal inflammation and the extent of colonic mucosal involvement. (Reprinted with permission from: Miner PB, Peppercorn MA, Targan SR. A rational approach to 5-aminosalicylic acid therapy in ulcerative colitis. Hosp Pract 1993; 2a(Suppl 3): 3-24.)

colitis or medication-associated colitis; and patients with UC in endoscopic remission who may be experiencing symptoms of concomitant irritable bowel syndrome. Classification of the patient according to the anatomic extent of involvement is shown in Fig. 1 [2]. Determining the extent of involvement is important because many 5-aminosalicylate-based medications and corticosteroid preparations are delivered topically, and do not distribute uniformly throughout the colon at a high concentration. The expected sites of drug delivery for various topically delivered 5-aminosalicylate formulations are shown in Table 1 [3]. Suppositories can be expected to release medication only in the rectum (approximately the last 10 cm of the colon) [4, 5]. Enemas will reach the ascending colon/splenic flexure in approximately 80-90% of patients (Fig. 2) [6-13].

Stephan R. Targan, Fergus Shanahan and Lor en C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 605-629. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Medical management of ulcerative colitis

606 Table 1.5-Aminosalicylate preparations Generic name

Proprietary name

Formulation

Sites of delivery

Daily dose Active Maintenance

Indication

Mesalamine

Rowasa Salofalk

Enema suspension

Distal to splenic flexure

4g

1-4 g

Active distal UC, remission maintenance distal UC

Mesalamine

Rowasa Canasa

Suppository

Rectum

1-1.5 g

0.5-1 g

Active proctitis, remission maintenance distal UC

Mesalamine

Asacol

Eudragit-s-coated tablets (release at pH ^ 7 . 0 )

Terminal ileum, colon

1.6-4.8 g

0.8-4.8 g

Active UC, remission maintenance UC

Mesalamine

Salofalk Mesasal Claversal

Mesalamine in a sodium /glycerine buffer coated with Eudragit-I (release at pH ^ 6 . 0 )

Distal jejenum, proximal ileum

1.5-4 g

0.75-4 g

Active UC, remission maintenance UC

Mesalamine

Pentasa

Ethylcellulose-coated microgranules (time and pH-dependent release)

Duodenum, jejenum, ileum, colon

2-4 g

1.5-4

Active UC, remission maintenance UC

Olsalazine

Dipentum

5-ASA dimer linked by azo-bond

Colon

2-3 g

ig

Active UC, remission maintenance UC

Sulfasalazine

Azulfadine

5-ASA linked to sulfapyridine by azo-bond

Colon

2-4 g

2-4 g

Active UC, remission maintenance UC

Balsalazide

Colazide Colazal

5-ASA linked to inert carrier by azo-bond

Colon

2-6.75 g

2-6.75 g

Active UC, maintenance UC

Reprinted after modification with permission from: Loftus EV, Sandborn WJ. Drug therapy for inflammatory bowel disease. Contemp Intern Med 1995;7:21-34.

The Truelove and Witts classification can be used to determine whether patients have mild to moderately active or severely active UC (Table 2) [14]. This assessment of disease severity is important in determining whether or not to hospitalize the patient and whether steroid therapy is mandatory. Other disease activity indexes such as the Sutherland Index or the Mayo Index are more useful for distinguishing patients with remission, mildly active disease, and moderately active disease for the purposes of assessing efficacy of sulfasalazine and other 5-aminosalicylate-based medications (Table 3) [15, 16].

Goals of treatment The primary goals of medical therapy are to induce and then maintain significant clinical improvement or remission, resulting in a reduction or resolution of

the signs and symptoms of active UC. Secondary goals, which often occur in parallel with clinical changes, are induction of endoscopic improvement and remission. The efficacy of various medical therapies in achieving these endpoints in patients with UC is reviewed in the following sections.

5-Aminosalicylate-based medications Sulfasalazine, oral mesalamine (Pentasa, Asacol, Salofalk, Mesasal, Claversal) rectal mesalamine (Rowasa, Canasa, Salofalk, Pentasa), olsalazine, and balsalazide are all drugs that deliver 5-aminosalicylate to the colon (Table 1) [3]. The clinical pharmacology of these medications is reviewed in detail elsewhere in this book.

William J. Sandborn

607

Figure 2. The individual colonic spread of radiolabeled low-viscosity 100 ml budesonide enemas in five patients with distal ulcerative colitis 15 min after administration. The filled areas of the colon represent areas where radioactivity was found. Modified with permission from: Nyman-Pantelidis M, Nillson A, Wagner GW, Borga 0. Pharmacokinetics and retrograde colonic spread of budesonide enemas in patients with distal ulcerative colitis. Aliment Pharmacol Ther 1994; 8: 617-22.

Table 2. Truelove and Witts criteria for evaluating the severity of ulcerative colitis* Variable

Mild disease

Severe disease

Fulminant disease

6

>10

Intermittent

Frequent

Continuous

Temperature (°C)

Normal

>37.5

>37.5

Pulse (beats/min)

Normal

>90

>90

Hemoglobin

Normal

< 75% of normal value

Transfusion required

^30

>30

>30

Air, edematous wall, thumbprinting

Dilatation

Abdominal tenderness

Abdominal distention and tenderness

Stools (number/day) Blood in stool

Erythrocyte sedimentation rate (mm/h) Colonic features on X-ray Clinical signs

*Moderate disease includes features of both mild and severe disease. Reprinted with permission from: Truelove SO, Witts LT. Cortisone in ulcerative colitis: final report on a therapeutic trial. Br Med J 1955; 2:1041-8.

Medical management of ulcerative colitis

608 Table 3. Mayo scoring system for assessment of ulcerative colitis activity

Stool frequency* 0 = Normal no. stools for this for this patient 1 = 1 - 2 stools more than normal 2 = 3-4 stools more than normal 3 = 5 or more stools more than normal * Each patient served as his or her own control to establish the degree of abnormality of the stool frequency Rectal bleeding** 0 = No blood seen 1 = Streaks of blood with stool less than half the time 2 = Obvious blood with stool most of the time 3 = Blood alone passed ** The daily bleeding score represented the most severe day of bleeding Findings of flexible proctosigmoidoscopy 0 = Normal or inactive disease 1 = Mild disease (erythema, decreased vascular pattern, mild friability) 2 = Moderate disease (marked erythema, absent vascular pattern, friability erosions) 3 = Severe disease (spontaneous bleeding, ulceration Physician's global assessment*** 0 = Normal 1 = Mild disease 2 = Moderate disease 3 = Severe disease *** The physician's global assessment acknowledged the three other criteria, the patient's daily record of abdominal discomfort and general sense of well-being, and other observations, such as physical findings and the patient's performance status A total Mayo ulcerative colitis activity score of 0-2 points indicates remission/minimally active disease; a score of 3-5 points indicates mildly active disease; a score of 6-10 points indicates moderately active disease; and a score of 11-12 may indicate moderate or severe disease, depending on the patient's Truelove and Witt's score. Reprinted with permission from: Schroeder KW, Tremaine WJ, llstrup DM. Coated oral 5aminosalicylic acid therapy for mildly to moderately active ulcerative colitis. N Engl J Med 1987; 317:1625-9.

Sulfasalazine In 1942 Svartz reported on both the therapeutic results and the toxic effects of a novel sulfanilamide preparation, sulfasalazine, in patients with UC [17]. Sulfasalazine is composed of 5-aminosalicylate linked to sulfapyridine by a diazo bond. Placebocontrolled trials demonstrated that sulfasalazine administered orally at doses of 2-6 g/day or as a 3 g enema was effective in inducing remission in patients with mildly to moderately active UC and ulcerative proctitis [18-21]. Studies comparing sulfasalazine with a combination of low-dose oral and rectal steroids for active UC concluded that steroid therapy acted more rapidly, and perhaps was more effective, than sulfasalazine [22, 23]. Additional placebo-controlled trials demonstrated that sulfasalazine at doses of 2-4 g/day was effective in maintaining remission in patients with UC, and that the 4 g dose was more effective whereas the 2 g dose was better tolerated [24-26]. One placebo-controlled trial failed to demonstrate a maintenance benefit [27]. Com-

parative studies of sulfasalazine versus a high-fiber diet or mast cell stabilizers such as cromoglycate also showed maintenance benefits for sulfasalazine [2831]. Approximately 10-20% of orally administered sulfasalazine is absorbed systemically with the remainder passing unaltered to the colon [32]. Sulfasalazine undergoes metabolism in the colon by bacterial azoreductase enzymes to 5-aminosalicylate and sulfapyridine [33, 34]. The active moiety of sulfasalazine was determined to be the poorly absorbed molecule 5-aminosalicylate and not the well-absorbed molecule sulfapyridine [32, 35-38]. Adverse events occurring in patients with inflammatory bowel disease (IBD) treated with sulfasalazine include: headache, epigastric pain, nausea and vomiting, cyanosis, skin rash, fever, hepatitis, autoimmune hemolysis, transient reticulosis, aplastic anemia, leukopenia, agranulocytosis, folate deficiency, pancreatitis, systemic lupus erythematosus, sulfonamide-induced toxic epidermal necrolysis, Stevens-Johnson syndrome, pulmonary dysfunction, and male infertility [39, 40]. For the

William J. Sandborn most part the side-effects from sulfasalazine can be attributed to the systemic absorption of sulfapyridine, and they occur more commonly in patients who are genetically predisposed to 'slow' acetylation of sulfapyridine to A^-acetylsulfapyridine in the liver [39]. Headache, nausea and vomiting, and epigastric pain often appear to be related to the sulfasalazine dose, and it is frequently possible to desensitize patients by discontinuing sulfasalazine for 1-2 weeks, and then restarting at 0.125-0.25 g/day and increasing by 0.125 g/week up to a maintenance dose of 2 g/day [40]. Sulfasalazine therapy may also lead to a paradoxical worsening of diarrhea in patients with UC [41].

609 suppositories was effective in maintaining remission in patients with left-sided UC, distal UC, and ulcerative proctitis [65-69]. There is little evidence of a mesalamine dose response in maintenance therapy, but more frequent dosing intervals (twice-daily suppositories or daily enemas) are associated with lower relapse rates than less frequent dosing intervals of once-daily suppositories or enemas every second or third night [66, 68, 70]. Comparative studies suggest that rectal mesalamine 4 g is superior to oral mesalamine 1.5 or 2 g and that rectal mesalamine 4 g in combination with oral mesalamine 1.6 g is superior to monotherapy with oral mesalamine 1.6 g for maintaining remission [71-73]. It is unclear whether this represents a dose response or an advantage for rectal delivery of mesalamine.

Rectal mesalamine (5-aminosalicylate) After it was demonstrated that mesalamine (5-aminosalicylate; 5-ASA) was the active moiety of sulfasalazine (discussed above), oral and rectal drug delivery systems were devised to avoid absorption of mesalamine in the proximal small intestine, instead targeting the colon as the site of drug release. Placebo-controlled trials demonstrated that mesalamine administered rectally as a suspension enema at doses of 1-4 g/day or as a suppository at doses of 0.5-1.5 g/day was effective in inducing remission in patients with mildly to moderately active left-sided UC, ulcerative proctosigmoiditis, and ulcerative proctitis [4, 15, 42-44]. There does not appear to be a dose response across this range of mesalamine doses [42, 45]. Studies comparing novel foam or suppository formulations of mesalamine to standard mesalamine enemas or suppositories have shown equivalent results [46-48]. Relatively small studies comparing rectally administered 5-ASA with rectal steroids demonstrated similar efficacy rates [49-59]. However, a meta-analysis suggested that rectally administered mesalamine is superior to rectal steroids for inducing remission [45]. Other small comparative studies have suggested that rectally administered mesalamine is equivalent to oral sulfasalazine, 4-aminosalicylate, and bismuth but not sucralfate [22, 23, 60-63]. A comparison of oral mesalamine 2.4 g/day, rectal mesalamine 4 g/day, and a combination of the two therapies demonstrated a benefit for rectal mesalamine and combination therapy [64]. Whether this result represents a mesalamine dose response or an advantage for rectal delivery of mesalamine is unclear. Additional placebo-controlled trials d e m o n s t r a t e d that mesalamine administered rectally as 1 or 4 g enemas or 0.5 or 1 g

Oral mesalamine (5-ASA) Placebo-controlled trials demonstrated that oral mesalamine administered at doses of 1.6-4.8 g/day (Asacol 1.6-4.8 g/day, Pentasa 2-4 g/day) was effective in inducing remission in patients with mildly to moderately active UC [16, 74-76]. Meta-analyses indicate that a dose response for oral mesalamine probably exists [77, 78]. Comparative studies have demonstrated that oral mesalamine at doses of 0.84.0 g/day has comparable efficacy to sulfasalazine 2.0-4.0 g/day, but that side-effects occur more frequently in sulfasalazine-treated patients [79-81]. Additional placebo-controlled trials demonstrated that oral mesalamine at doses of 0.8-4.0 g/day was effective in maintaining remission in patients with UC [82, 83]. Comparative studies of oral mesalamine versus sulfasalazine demonstrated comparable efficacy for maintaining remission in UC, again with fewer side-effects for oral mesalamine [84-88]. More recent comparative studies for maintaining remission in patients with UC have suggested that oral mesalamine may be equivalent to dietary fiber and probiotic bacteria [88-91].

Olsalazine Olsalazine is a dimer, composed of two 5-ASA molecules linked by a diazo bond. Placebo-controlled trials demonstrated that olsalazine administered orally at doses of 0.75-3 g/day or rectally as a 1 g enema was not consistently effective in inducing remission in patients with mildly to moderately active UC, due to a higher than expected dropout

610 rate in olsalazine-treated patients for worsened diarrhea [92-97]. The worsened diarrhea is a result of ileal secretion [98]. In contrast to the placebocontrolled trials, studies that compared olsalazine with sulfasalazine for active UC for the most part concluded that the two agents had comparable efficacy [99-102]. Additional placebo controlled trials demonstrated that olsalazine at doses of 1-2 g/day was effective in maintaining remission in patients with UC [103, 104]. Comparative studies of olsalazine against oral mesalamine and sulfasalazine demonstrated similar efficacy for maintenance of remission in patients with UC [105-110].

Balsalazide Balsalazide is composed of 5-ASA Hnked to an inert carrier molecule by a diazo bond. A comparative study of balsalazide 6.75 g/day with oral mesalamine (Asacol) 2.4 g/day in patients with active UC demonstrated similar efficacy [112]. Maintenance studies in patients with UC demonstrated that balsalazide 4 g/ day was more effective than 2 g/day, balsalazide 3 g/ day had similar efficacy to balsalazide 6 g/day and to oral mesalamine (Asacol) 1.2 g/day, and balsalazide 2 g/day had similar efficacy to sulfasalazine 2 g/day in maintaining remission [111, 113-115].

Toxicity of mesalamine, olsalazine, and balsalazide Adverse events attributable to 5-ASA occur infrequently in patients with IBD treated with mesalamine, olsalazine, and balsalazide. Rare but serious events include pulmonary toxicity, pericarditis, hepatitis, and pancreatitis [116-119]. Interstitial nephritis has been reported in patients treated with mesalamine but whether the mesalamine causes the renal lesion is unclear [120-123]. Several studies have demonstrated that renal tubular proteinuria may be related to the disease activity of the IBD [124, 125]. Hanauer et al. reported on the safety of Asacol in 2940 patients with UC at doses up to 7.2 g/day for up to 5.2 years and concluded that there were no clinically significant dose or duration effects on renal function [126]. A minority of patients will experience worsening diarrhea and abdominal pain due to a hypersensitivity reaction to 5-ASA, and treatment with olsalazine will lead to an ileal secretory diarrhea in some patients, as discussed above [127].

Medical management of ulcerative colitis

Corticosteroid-based medications Cortisone is produced by the adrenal cortex, and prednisone must be activated in the liver to hydrocortisone and prednisolone, respectively. Prednisolone and methylprednisolone have the same glucocorticoid and anti-inflammatory activity as hydrocortisone but less mineralocorticoid activity. Rectal administration of hydrocortisone, prednisolone, methylprednisolone, and betamethasone; oral administration of cortisone, prednisone, and prednisolone; and intravenous administration of prednisolone, methylprednisolone, and corticotropin are all methods of delivering corticosteroids for a systemic effect (Table 4). In contrast to systemically administered corticosteroids, rectal administration of beclomethasone, tixicortol, budesonide, and prednisolone metasulfobenzoate; and oral administration of fluticasone and controlled colonic release budesonide are all methods of delivering corticosteroids directly to the colon for a nonsystemic effect (Table 4). Topical administration of beclomethasone, fluticasone, tixocortol, or budesonide to the colon results in a predominantly nonsystemic effect because these newer corticosteroids have high affinities for the glucocorticoid receptors and undergo extensive first-pass hepatic metabolism. Topical administration of prednisolone metasulfobenzoate and tixocortol to the colon results in a predominantly non-systemic effect because they are poorly absorbed.

Oral corticosteroids (systemic effect) In 1954 and 1955 Truelove and colleagues reported the preliminary and final results of a placebo-controlled trial which demonstrated that a tapering dose of cortisone beginning at 100 mg/day was effective in inducing remission in patients with mildly to severely active UC [14, 128]. Studies comparing sulfasalazine with a combination of low-dose oral and rectal steroids for active UC concluded that steroid therapy acted more rapidly, and perhaps was more effective, than sulfasalazine [22, 23]. A dose-ranging study demonstrated that prednisone 40-60 mg/day was more effective than 20 mg/day, and that 60 mg/day was no more effective than 40 mg/day but resulted in a greater frequency of side-effects [129]. A subsequent study demonstrated that a single daily dose of prednisone 40 mg was equally effective as prednisone 10 mg four times daily [130]. Placebo-controlled trials of cortisone 25 mg twice daily and prednisone

William J. Sandborn

611

Table 4. Rectal corticosteroid preparations

Generic name

Proprietary name

Formulation

Sites of delivery

Site and mechanism of action

Hydrocortisone acetate

Anusol-HC 25 mg

Suppository 25 mg

Rectum

Systemic

50-100

Hydrocortisone

Cortenema

Enema 100mg/60ml

Distal to splenic flexure

Systemic

100

Hydrocortisone acetate

Cortifoam

Foam 80 mg/ 900 mg foam

Distal rectum

Systemic

80-160

Hydrocortisone acetate

Colifoam*

Foam 125 ml/5 ml foam

Distal to splenic flexure

Systemic

125-250

Active distal UC

Hydrocortisone acetate

Proctocort

Suppository 30 mg

Rectum

Systemic

60-120

Active proctitis

Hydrocortisone acetate + pramoxine hydrochloride (local anesthetic)

Proctofoam HC

Topical aerosol 1% hydrocortisone (approx. 7 mg) 1 % pramoxine (approx. 7 mg)

Anus, distal rectum

Systemic

7-28

Active procitis

Prednisolone phosphate

Predsol enema*

Enema 20mg/100ml

Distal to splenic flexure

Systemic

20

Active distal UC

Betamethasone valerate

Betnesor

Enema 5mg/100ml

Distal to splenic flexure

Systemic

5

Active distal UC

Prednisolone metasulphobenzoate

Predenema*

Enema 20mg/100ml

Distal to splenic flexure

Non-systemic: poorly absorbed

20

Active distal UC

Prednisolone metasulphobenzoate

Predfoam*

Foam 20 mg/20 ml foam

Distal to splenic flexure

Non-systemic: poorly absorbed

20

Active distal UC

Tixocortol pivalate

Rectovalone*

Enema 250mg/100ml

Distal to splenic flexure

Non-systemic: poorly absorbed and first pass metabolism

250

Active distal UC

Budesonide

Entocort enema*

Enema 2mg/100ml

Distal to splenic flexure

Non-systemic: first pass metabolism

2

Active distal UC

*lndicates not available in the United States.

Daily dose (mg)

Indication Active proctitis

Active distal UC

Active proctitis in the distal rectum

612 at doses of 15 mg/day or 40 mg every other day failed to demonstrate a maintenance benefit for oral corticosteroids [131-133].

Intravenous corticosteroids and corticotropin (ACTH) Patients with severe UC and those refractory to oral coticosteroids are hospitaHzed and treated with intravenous corticosteroids. The rationale for this practice is altered corticosteroid absorption and metabolism in patients with UC. Oral administration of a 40 mg dose of prednisolone resulted in a lower peak and a slower rate of decrease in the plasma concentration of prednisolone in patients with severe UC compared with the time versus concentration curve observed in healthy volunteers; although total prednisolone absorption was similar [134]. In contrast, intravenous administration of prednisolone to patients with UC resulted in serum concentrations similar to volunteers [135]. Continuous infusion of prednisolone resulted in greater mean serum concentrations over time compared to bolus intravenous dosing; and both intravenous dosing strategies resulted in greater mean serum concentrations than oral dosing [135]. Uncontrolled studies have reported that approximately 60% of patients hospitalized for severe UC will respond to intravenous corticosteroid therapy [136-138]. Dosing strategies have included prednisolone 60 mg/day in four divided doses [136], betamethasone 3 mg twice daily [138], and hydrocortisone 300-400 mg/day [139141]. Methylprednisolone 40-60 mg/day is preferred by many clinicians because it has minimal mineralocorticoid effect. There was no apparent advantage in increasing the dose of methylprednisolone to 1000 mg/day [142]. No placebo-controlled trials of intravenous corticosteroid therapy for severe UC have been performed. A comparative study of intramuscular corticotropin (adrenal corticotropin hormone, ACTH) 80 U/ day and cortisone 200 mg/day demonstrated a similar overall benefit in patients with active UC and in the subgroup of patients with a first attack, with a possible advantage for corticotropin in patients in patients with a relapse of established colitis [131]. Subsequent studies in patients with severe UC showed that overall the response to corticotropin 80-120 U/day is similar to hydrocortisone 300-400 mg/day, with trends toward better response to hydrocortisone in patients recently treated with cortico-

Medical management of ulcerative colitis steroids and better response to corticotropin in patients not recently treated with corticosteroids [139-141].

Rectal corticosteroids (systemic effect) Hydrocortisone, prednisolone, methylprednisolone, and betamethasone administered directly to the rectum as enemas or suppositories are well absorbed (similar to oral dose) [143, 144] and can result in suppression of the adrenal axis [49, 145-148]. Placebo-controlled trials have demonstrated efficacy for rectal administration of hydrocortisone 100 mg and prednisolone 5 mg in patients with active ulcerative proctitis or proctosigmoiditis [149-151]. Relatively small studies comparing rectally administered 5ASA with rectal hydrocortisone 100-178 mg/day or prednisolone 20-30 mg/day, demonstrated similar efficacy rates in patients with active ulcerative proctitis or proctosigmoiditis [49-57, 59, 152, 153]. However, a meta-analysis suggested that rectally administered mesalamine is superior to rectal steroids for inducing remission [45]. Other small comparative studies have suggested that rectally administered hydrocortisone 100 mg, prednisolone 20-30 mg, or betamethasone 5 mg is equivalent to oral prednisone, oral sulfasalazine, and rectal 4-aminosalicylate, sodium cromoglycate, hydrocortisone foam 100 mg, and Ridogrel in patients with active ulcerative proctitis or proctosigmoiditis [23, 154-160]. Rectal hydrocortisone 100 mg every 2 nights each week for 6 months did not demonstrate a maintenance benefit compared with placebo [149].

Rectal corticosteroids (non-systemic effect) Placebo-controlled trials demonstrated that budesonide administered rectally in a suspension enema at doses of 2-8 mg/day was effective in inducing remission in patients with mildly to moderately active leftsided UC, ulcerative proctosigmoiditis, and ulcerative proctitis; a 0.5 mg/day dose was not effective [48, 161-164]. Relatively small studies comparing rectally administered budesonide 2.0-2.5 mg/day with other rectal steroids (methylprednisolone 20 mg, prednisolone 25-31 mg, hydrocortisone 100-125 mg) demonstrated similar efficacy rates [147, 148, 163, 165-167]. Similarly, relatively small studies comparing rectally administered budesonide 2 mg/ day with rectal 5-ASA 1-4 g/day demonstrated similar efficacy rates [58, 168]. Other small compara-

William J. Sandborn tive studies have demonstrated that rectally administered non-systemic corticosteroids (prednisolone metasulfobenzoate 20 mg, beclomethasone 0.5-3 mg, tixocortol 250 mg) are equivalent to oral prednisolone, oral 5-ASA, rectal betamethasone 5 mg, rectal hydrocortisone 100 mg, rectal prednisolone 30 mg, rectal 5-ASA, and Ridogrel but not sucralfate [59, 157, 169-174]. Rectal budesonide 2 mg 2 nights each week for 6 months did not demonstrate a maintenance benefit compared with placebo [164].

Oral corticosteroids (non-systemic effect) A placebo-controlled trial demonstrated that oral fluticasone 20 mg/day was not ejffective in inducing remission in patients with mildly to moderately active distal UC [175]. A study comparing oral fluticasone 20 mg/day with prednisolone 40 mg/day tapered to 10-20 mg/day in patients with active UC showed a greater benefit and more rapid response in the prednisolone group [176]. A comparative study of controlled colonic release budesonide 10 mg and prednisolone 40 mg/day and tapered to 0 mg demonstrated similar response rates in the two groups, with fewer side-effects in the budesonide group [177].

Toxicity of corticosteroids Corticosteroid toxicity occurred frequently in patients with active CD treated with prednisone at an initial dose of 60 mg/day tapered over 17 weeks. Toxicities observed included a moon face in 47%, acne in 30%, infection in 27%, ecchymoses in 17%, hypertension in 15%, hirsutism in 7%o, petechial bleeding in 6%, and striae in 6% [178]. A similar short-term toxicity profile can be expected in patients with UC. Prolonged corticosteroid therapy at low to intermediate doses (doses frequently utilized in patients with steroid-dependent UC) is associated with the potential for multiple serious side-effects [179]. Hypertension occurs in up to 20% of patients [180]. New-onset diabetes mellitus requiring initiation of hypoglycemic therapy occurs at a frequency 2.23 times greater than in the general population [181]. Infection occurs at a frequency of 13-20%o [182]. Osteonecrosis occurs at a frequency of approximately 5% [183]. The frequency of steroid-associated osteoporosis may be as high as 50% [184]. Neurologic side-effects occur often and can include myopathy at a frequency of 7%o and psychosis at a

613 frequency of 3-5%) [185]. Ophthalmologic sideeffects also occur often and can include cataracts at a frequency of 22%) (dose-dependent) and glaucoma (frequency unclear, response genetically determined) [186, 187]. These frequencies of side-effects from prolonged exposure to corticosteroids were generally confirmed in a study of patients with UC who had undergone colectomy for medically refractory disease [188].

Immune modifer medications The antimetabolites 6-mercaptopurine (Purinethol), its pro-drug azathioprine (Imuran), and methotrexate; the calcineurin inhibitors cyclosporine (Sandimmune, Neoral) and tacrolimus (FK506, Prograf); and the T-cell inhibitor mycophenolate mofetil (Cellcept) are all medications with immune modifier activity. The clinical pharmacology of these medications is reviewed in detail elsewhere in this book.

Azathioprine and 6-mercaptopurine The first uncontrolled reports of the treatment of UC with 6-mercaptopurine (6-MP) and azathioprine (AZA) date back to the early 1960s [189-191]. A placebo-controlled trial of AZA 2.5 mg/kg per day in combination with a tapering dose of corticosteroids in 80 patients with active UC showed no benefit at 1 month and a trend toward a benefit at 1 year that was not significant [192, 193]. A comparative study of AZA 2.5 mg/kg per day versus sulfasalazine 65 mg/kg per day in patients with active UC showed similar efficacy for the two agents [194]. Two small placebo-controlled trials suggested that AZA at 1.5 mg/kg per day and 2.0-2.5 mg/kg per day was steroid-sparing in patients with steroid-dependent UC [195, 196]. Finally, a placebo-controlled withdrawal trial in patients maintained with AZA 100 mg/day demonstrated a maintenance of remission benefit [197]. These studies have been supplemented by five recently published open series of AZA and 6MP therapy for U C [198-202]. Overall, these controlled and uncontrolled studies demonstrate that 6-MP and AZA are effective for steroid-sparing and maintenance of remission, and probably for inducing remission in patients with chronically active and treatment-refractory UC. Adverse events occurring in patients with IBD treated with 6-MP and AZA include: pancreatitis

Medical management of ulcerative colitis

614 (3%); fever, rash, arthralgias, malaise, nausea, diarrhea, leukopenia (2-5%); thrombocytopenia, infection, and hepatitis [203-205]. It appears that there is not an increased risk of malignancy when AZA/6-MP is used as a monotherapy in IBD, and in three large series only one lymphoma occurred in 1308 patients [203, 206, 207].

Methotrexate Two uncontrolled reports of the treatment of UC with methotrexate date occurred in the late 1980s and early 1990s [208, 209]. These studies suggested that intramuscular methotrexate 25 mg/week might be beneficial whereas oral methotrexate 15 mg/week appeared less promising. In patients with CD these uncontrolled observations regarding dose response and route of administration have been substantiated, with placebo-controlled trials demonstrating efficacy for intramuscular methotrexate 15-25 mg/week but not for oral methotrexate 12.5-15 mg/week [210-213]. In patients with UC a single placebocontrolled trial of oral methotrexate 12.5 mg/week did not demonstrate efficacy for either inducing or maintaining remission [214]. Whether intramuscular or subcutaneous methotrexate at doses of 15-25 mg/ week would be effective for patients with UC is unknown, although the uncontrolled study discussed above is encouraging [208]. Based on the currently available evidence, patients with UC should not be treated with methotrexate.

Cyclosporine The use of oral cyclosporine for UC was first reported in 1984 [215]. Subsequently multiple uncontrolled studies have reported a beneficial effect of cyclosporine administered at relatively high doses (5-15 mg/kg per day orally or 2-7 mg/kg per day intravenously) in patients with severe UC unresponsive to intravenous corticosteroids [216-221]. A placebo-controlled trial demonstrated that the addition of intravenous cyclosporine administered at a dose of 4 mg/kg per day as a continuous infusion to intravenous corticosteroids was effective for inducing remission in patients with severe steroid-refractory UC (82% versus 0% response) [222]. However, only 45% of the patients treated with cyclosporine had avoided colectomy after 6 months of follow-up [223], an observation confirmed by other uncontrolled studies [217-219, 224]. Maintenance therapy

with AZA or 6-MP in patients with UC who require treatment with cyclosporine appears to reduce the rate of relapse and colectomy [217, 225, 226]. Two more recent controlled trials have demonstrated that monotherapy with intravenous cyclosporine 4 mg/ kg per day has efficacy comparable to intravenous corticosteroids or intravenous corticosteroids combined with intravenous cyclosporine in patients with active refractory UC [227, 228]. Overall, these studies demonstrate that intravenous cyclosporine is effective in inducing remission in patients with severe, steroid-refractory UC, but that the benefit is of limited duration unless AZA or 6-MP is initiated for maintenance of remission. Uncontrolled pilot studies of low-dose cyclosporine enemas suggested a potential beneficial effect in patients with active leftsided UC [229-232], but a placebo-controlled trial was negative [233]. Adverse events occurring in patients with IBD treated with cyclosporine include: hypertension, headaches, paresthesias, seizures, gingival hyperplasia, hypertrichosis, anaphylaxis, opportunistic infection, and renal insufficiency [217, 221, 234, 235]. There is an increased risk of opportunistic infection in patients with IBD treated with intravenous cyclosporine combined with corticosteroids and AZA or 6-MP; Pneumocystis carinii pneumonia, invasive apergillosis, lung abscess, mycotic aneurysm, and overwhelming sepsis have all been reported, with death rates ranging from 1% to 2% in larger series [217, 234-237]. There does not appear to be an increase in perioperative morbidity or mortality in patients with UC who receive intravenous cyclosporine and then require colectomy within a short period of time [238]. Another study reported that 20% of 99 patients with IBD treated with intravenous cyclosporine had a decrease in estimated renal function greater than 30% [239]. Results from a previous study suggest that patients with autoimmune diseases treated with intravenous cyclosporine have a significant likelihood of having histologic evidence of irreversible nephrotoxicity on renal biopsy (which to date has not been performed in patients with IBD) [240].

Tacrolimus Small uncontrolled case series have suggested that tacrolimus may be beneficial in patients with refractory UC [241-243]. To date no controlled trials of tacrolimus have been performed in patients with UC. Given the similarities in mechanism of action to

William J. Sandborn cyclosporine, it is reasonable to believe that tacrolimus would have similar efficacy and toxicity as cyclosporine in this patient group.

615 frequently and include contact dermatitis, nausea, vomiting, headaches, sleep disturbance, diaphoresis, tremor, and lightheadedness [253-255].

Mycophenolate mofetil

Heparin

A small controlled trial reported that mycophenolate mofetil 15 mg/kg per day had similar efficacy as azathioprine 2.5 mg/kg per day in patients with chronically active CD [244]. In contrast, a small controlled trial of mycophenolate mofetil 20 mg/kg per day versus AZA 2.0 mg/kg per day in patients with chronically active UC demonstrated a superior outcome with azathioprine [245]. Similarly, an uncontrolled study did not report a benefit with mycophenolate in patients with UC [246]. Based on the currently available evidence, patients with UC should not be treated with mycophenolate mofetil.

The inflammatory bowel diseases have been associated with a hypercoagulable state, and heparin has been proposed as a potential therapy in UC both for its anticoagulant and anti-inflammatory effects [260]. Four uncontrolled trials have suggested that both intravenous and subcutaneous heparin may be of benefit in patients with active UC [261-264]. A placebo-controlled trial of subcutaneous porcine heparin 10 000 units two or three times daily in patients with active UC reported a response rate of 42% for the heparin group and 19% for the placebo group [265]. Paradoxically, heparin appears to improve active UC rather than to exacerbate bleeding from the friable and ulcerated colonic mucosa.

Miscellaneous agents 4-Aminosalicylate (4-ASA)

Anti-tumor necrosis factor oc antibody

4-Aminosalicylate (para-aminosalicylate, PAS) is an isomer of 5-ASA, the active component of sulfasalazine. Controlled trials with oral or rectal enema formulations of 4-ASA have shown superiority to placebo and equivalence to corticosteroids and 5ASA [61, 154, 247-252].

Placebo-controlled trials have demonstrated that the mouse/human chimeric monoclonal antibody to tumor necrosis factor (TNF)-oc, infliximab, and the humanized antibody to TNF, CDP571, are effective for the treatment of CD [266-269]. Two small studies conducted in patients with active UC have suggested a potential beneficial effect [270, 271]. Use of antiTNF-oc antibodies for the treatment of UC in clinical practice is premature at the present time.

Nicotine Epidemiologic studies have demonstrated that nonsmokers are at increased risk to develop UC. There are multiple potential mechanistic explanations for this beneficial effect [253]. Two placebo-controlled trials of transdermal nicotine patches reported that the highest tolerated dose of nicotine (up to 25 mg/24 h and 22 mg/24 h, respectively) was effective in patients with active UC [254, 255]. Similarly, a controlled trial of transdermal nicotine at the highest tolerated dose (up to 25 mg/16 h) showed equivalence to prednisolone for active UC [256]. A controlled trial using a lower dose of transdermal nicotine (15 mg/16 h) for maintenance of remission was negative [257]. Two uncontrolled studies of nicotine enemas for active distal UC reported a beneficial clinical effect and a reduction in both blood concentrations of nicotine (by first-pass hepatic metabolism of nicotine) and side-effects [258, 259]. Side-effects from transdermal nicotine occur

Stiort-chain fatty acids Luminal «-butyrate and other short-chain fatty acids (SCFA) produced by bacterial carbohydrate fermentation are the major luminal source of energy for colonocytes, and colonocytes from patients with UC may have reduced capacity to oxidize SCFA [272]. Two small and poorly designed controlled trials suggested that SCFA enemas may be beneficial in active distal UC [273, 274]. Subsequently, four placebo-controlled trials with larger sample sizes have been conducted [275-278]. Two of two studies comparing butyrate enemas to placebo [275, 276], and two of three studies comparing SCFA enemas to placebo [276-278] failed to demonstrate efficacy in patients with active left-sided UC.

Medical management of ulcerative colitis

616

Fish oil Fish oil (eicosapentanoic acid and docosahexenoic acid) inhibits 5-lipoxygenase and other enzymes involved in arachidonate metabolism. A small controlled trial suggested that fish oil may be of benefit in patients with active UC [279]. However, three large controlled trials with adequate sample sizes did not demonstrate clinical efficacy of fish oil compared with placebo in patients with active UC [280, 281] or for maintenance of remission [282]. These negative clinical results were observed in spite of the finding that fish oil did result in a significant decrease in colonic luminal concentrations of leukotriene B4 [280].

eff'ective than 5-ASA 1.6 g/day and similar to placebo for maintaining remission in patients with UC [295].

Anti-piateiet-activating factor agents Platelet-activating factor is elevated in the stool and mucosa of patients with UC. Randomized, doubleblind placebo-controlled trials have not demonstrated that antibodies directed against platelet-activating factor are effective in the treatment of moderately or severely active UC [296, 297].

Bismuth Antibiotics Given the attractive but unsubstantiated hypothesis that UC may be a specific infectious colitis, studies evaluating treatment with antibiotics were logical. A controlled trial of oral tobramycin in patients with mild to moderately active UC demonstrated efficacy but there was no maintenance benefit [283, 284]. Only one of two placebo-controlled trials of oral ciprofloxacin in patients with active UC demonstrated efficacy [285, 286]. Controlled trials of oral vancomycin [287], intravenous metronidazole [288], and intravenous tobramycin and metronidazole [289] in patients with severely active UC demonstrated no clinical benefit compared with placebo. Based on the nearly uniform negative outcome of these studies, antibiotics have no role in the treatment of UC except in the case of a documented or suspected co-existing infection with a specific infectious organism.

Bismuth compounds have both antidiarrheal and antibacterial properties. Uncontrolled studies of bismuth subsalicylate enemas [298] and tripotassium dicitrato bismuth enemas [299] suggested benefit in patients with mild to moderately active distal UC. These observations were confirmed by a small controlled trial comparing bismuth citrate enemas to 5ASA enemas that demonstrated equivalence in patients with mild to moderately active distal UC [62]. Systemic absorption of bismuth after rectal administration was minimal.

Interferon Because of the probable central role of the immune system in the pathogenesis of UC, it seems reasonable to evaluate the potential efficacy of the interferons (alpha, beta, and gamma) as a therapy. Interferon-alpha has been used in a pilot study to treat patients with mild to moderately active UC with reported success [300].

Ziieuton and ottier ieul(otriene inhibitors Initial studies suggested that compounds which act to inhibit 5-lipoxygenase might be of benefit in patients with UC [290, 291]. However, larger controlled trials with adequate sample sizes were negative [292-295]. The Merck compound MK-591 resulted in a significant decrease in colonic luminal concentrations of leukotriene B4 but did not demonstrate clinical efficacy when compared with placebo in patients with mild to moderately active UC [292]. There was a trend toward clinical benefit for ziieuton compared with placebo in patients with mild to moderately active UC, but the results did not reach statistical significance [293, 294]. Ziieuton was less

Interieul(in-10 Interleukin-lO is a cytokine produced by T-helper type 2 cells, B cells, monocytes, and macrophages that has multiple immune modifier effects. A small controlled trial of recombinant human interleukin10 (rhuIL-10) (Schering Plough Research Institute, Kenilworth, NJ) in patients with mild to moderately active CD demonstrated a modest beneficial effect [301]. A placebo-controlled trial of interleukin-10 in 94 patients with mild to moderately active UC was negative [302].

William J. Sandborn

617

Lidocaine/ropivacaine Topical anesthetic agents including both lidocaine and ropivacaine may have anti-inflammatory effects. Uncontrolled studies of 2% lidocaine enemas [303, 304] and ropivacaine gel enemas [305] have been reported to be of benefit in patients with mild to moderately active distal UC. A placebo-controlled trial in this same patient group has been completed, but to date the results have not been made available. Systemic absorption of these agents after rectal administration is minimal.

Treatment indications and algoritiim and specific treatment approaclies The indications for therapy in patients with UC are summarized in Table 5 and a suggested treatment algorithm is proposed in Fig. 3. The specific approaches to the medical treatment of patients with proctitis/distal UC, extensive UC, refractory UC, and severe UC are each reviewed separately below.

Proctitis/distal ulcerative colitis

Probiotic bacteria Probiotic bacteria may be of benefit in patients with UC and pouchitis. Two controlled studies have reported that E. coli Nissle is therapeutically equivalent to 5-ASA for maintaining remission in patients with UC [90, 91]. These trials were not statistically powered to demonstrated equivalence. A placebocontrolled trial is needed to determine if probiotic therapy is a useful treatment modality for UC.

The treatment of patients with mildly to moderately active proctitis/distal UC may include rectal therapy, oral therapy, or a combination of both. Patient preferences with regard to rectal therapy for induction of remission must be considered. Rectal mesalamine may be more efficacious than orally administered mesalamine [64)]or rectally administered corticosteroids [45]. Thus, for patients who will accept rectal therapy, mesalamine suppositories (500 mg b.i.d.) for proctitis and mesalamine enemas (1 g or 4 g nightly) for distal colitis are the treatment of choice. For patients who prefer oral therapy, oral mesalamine 2.0-4.8 g/day, sulfasalazine 2-4 g/day, and balsalazide 6.75 g/day are all equivalent first-line treatments [77, 78, 112]. Olsalazine should be

Table 5. Ulcerative colitis: indications for treatment Drug Sulfasalazine Rectal mesalamine Oral mesalamine Olsalazine Balsalazide Rectal corticosteroids Oral corticosteroids Intravenous corticosteroids Azathioprine/6-mercaptopurine Cyclosporine Transdermal nicotine

Mildly to moderately active Distal Extensive

Yes Yes Yes No' Yes Yes Yes No No No No

Yes No Yes No' Yes No Yes No No No No

Refractory Yes' Yes' Yes'

No' Yes' Yes' Yes' Yes^

Yes No Yes

Severely active No''

No^ No^ No' No^ Yes'

No Yes No Yes No

Remission maintenance Distal Extensive

Yes Yes Yes Yes Yes No No No Yes No No

Yes No Yes Yes Yes No No No Yes No No

typically continued as a carryover of treatment for mildly to moderately active disease when additional agents are added. h"ypically discontinued because of the possibility of intolerence to sulfasalazine, mesalamine, or balsalazide. 'Diarrhea occurs frequently at higher doses in patients with active ulcerative colitis. 'Adjunctive therapy to intravenous corticosteroids. ^Some patients who fail oral corticosteroids will respond to hospitalization with intravenous administration of corticosteroids. Reprinted after modification with permission from: Loftus EV, Sandborn WJ. Drug therapy for inflammatory bowel disease. Contemp Intern Med 1995;7:21-34.

Medical management of ulcerative colitis

618

Severe colitis

Mild-to-moderate distal colitis

IV steroids (hydrocortisone, 300 mg/d or the equivalent) plus topical hydrocortisone

(proctitis, left-sided colitis) 5-ASA enemas or suppositories; active doses .

Mild-to-moderate extensive colitis Sulfasalazine or oral 5-ASA active doses

Maintenance ^_ doses

Topical hydrocortisone (hs or bid)

-^

^

No response in 7-10 d, surgery or IV cyclosporine at referral center

Oral corticosteroids (for example, prednisone, 40-60 - ^ mg/d tapered over 4-12 wk) Maintenance with sulfasalazine or oral 5-ASA Azathioprine, 1.5-2.5 mg/kg/d, or 6-mercaptopurine, 1.5 mg/kg/d Maintenance with Azathioprine or 6-mercaptopurine

•f response -

no response

Consider maintenance with sulfasalazine or rectal/oral 5-ASA

Surgery

Figure 3. Suggested treatment guidelines for ulcerative colitis are offered in this algorithm. Reprinted after modification with permission from: Loftus EV, Sandborn WJ. Drug therapy for inflammatory bowel disease. Contemp Intern Med 1995; 7: 21-34.

avoided due to a high frequency of diarrhea in patients with active UC. Sulfasalazine therapy is associated with more side effects but is less expensive than oral mesalamine or balsalazide. In one study the combination of mesalamine enemas 4 g/day and oral mesalamine 2.4 g/day was more effective than either agent alone [64] and another study demonstrated that combination therapy results in greater colonic tissue concentrations of 5-ASA [306]. Based on these data, many clinicians use rectal and oral mesalamine in combination to induce remission and then continue the oral mesalamine for maintenance of remission. In patients who fail to respond to oral or rectal mesalamine, sulfasalazine, or balsalazide, rectal corticosteroids can be tried. When all of these treatment approaches are ineffective, then oral prednisone is added to the other

oral and rectal therapies. The preferred initial prednisone dosing regimen is 40 mg/day administered as a single dose. The optimal tapering strategy has not been determined, but experienced clinicians will typically treat the patient with prednisone 40 mg/day for 2-4 weeks, then taper by 5 mg/week to a daily dose of 20 mg/day, then slow the taper to 2.5 mg/week until prednisone is discontinued. Once the patients have achieved clinical remission, a long-term maintenance of remission strategy must be undertaken to avoid relapse. Again, patient preferences with regard to rectal therapy must be considered. Rectal mesalamine may be more efficacious than orally administered sulfasalazine or mesalamine for maintaining remission [71, 72]. Again, for patients who will accept rectal therapy, mesalamine suppositories (500 mg once or twice daily) for

William J. Sandborn proctitis and mesalamine enemas (4 g nightly, every other night or every third night) for distal colitis are the treatment of choice. For patients who prefer oral therapy, oral mesalamine 1.6-4.8 g/day, sulfasalazine 2-4 g/day, olsalazine 1.0 g/day, and balsalazide 3.0-6.75 g/day are all equivalent first-line maintenance treatments [77, 78, 113]. Sulfasalazine therapy is associated with more side-effects but is less expensive than oral mesalamine, olsalazine, or balsalazide. In one study the combination of mesalamine enemas 4 g/day and oral mesalamine 1.2 g/day was more effective than oral mesalamine alone [73]; however, most patients find maintenance with both oral and rectal mesalamine unacceptable. There is not agreement among expert clinicians as to whether patients should taper rectal mesalamine to the least frequent effective dose interval, or oral mesalamine, sulfasalazine, olsalazine, or balsalazide to the lowest effective dose; or instead continue maintenance therapy with the same dose interval or dose required to induce remission. The former strategy is less expensive and may improve patient compliance by reducing the amount and frequency of medication administered, whereas the latter strategy may result in effective maintenance of remission in a larger proportion of patients. Rectal corticosteroids are not effective for maintaining remission and should not be used for that indication. Clinical trials have not demonstrated that oral corticosteroids at low to moderate doses are effective for maintaining remission. Nevertheless, some patients who respond to higher doses of prednisone will relapse with steroid tapering and can be maintained nearly asymptomatic by increasing the prednisone dose back to 1525 mg/day These patients are classified as 'steroiddependent'. Because of the toxicity associated with long-term corticosteroid therapy, this is not an acceptable form of maintenance therapy and such patients should be treated for refractory disease as described below.

619 of diarrhea in patients with active UC. There appears to be a dose response for oral medications that deliver 5-ASA [77]. There is not agreement among expert clinicians as to whether the preferred strategy is to begin with the lowest dose proven to be effective for active disease, increasing the dose in those patients who fail to respond; or rather to begin with the maximally tolerated dose and then titrate the dose downward when the patient comes into clinical remission. When treatment with one of these agents at an optimal dose has failed, then oral prednisone is added to the oral mesalamine, sulfasalazine, or balsalazide. The preferred initial prednisone dosing regimen is 40 mg/day administered as a single dose, with tapering as described above for proctitis/distal UC. Once the patients have achieved clinical remission, a long-term maintenance of remission strategy must be undertaken to avoid relapse. Oral therapy with a drug that delivers 5-ASA to the colon is the treatment of choice. Oral mesalamine 1.6-4.8 g/day, sulfasalazine 2-4 g/day, olsalazine 1.0 g/day, and balsalazide 3.0-6.75 g/day are all equivalent first-line maintenance treatments [77, 78, 113]. Sulfasalazine therapy is associated with more side-effects but is less expensive than oral mesalamine, olsalazine, or balsalazide. As described above for patients with proctitis/distal UC, there is not agreement among expert clinicians as to whether patients with extensive UC should taper oral mesalamine, sulfasalazine, olsalazine, or balsalazide to the lowest effective dose; or instead continue maintenance therapy with the same dose required to induce remission. Clinical trials have demonstrated that oral corticosteroids at low to moderate doses are not effective for maintaining remission. As discussed above for proctitis/distal UC, treatment with prednisone 15-25 mg/day in patients with extensive UC who are 'steroid-dependent' is not acceptable because of the toxicity associated with long-term corticosteroid therapy. Such patients should be treated for refractory disease as described below.

Extensive ulcerative colitis Oral administration of medications that deliver 5ASA to the colon is the treatment of choice in patients with mildly to moderately active extensive UC. Oral mesalamine 2.0-4.8 g/day, sulfasalazine 2 4 g/day, and balsalazide 6.75 g/day are all equivalent first-line treatments [77, 78, 112]. Sulfasalazine therapy is associated with more side effects but is less expensive than oral mesalamine or balsalazide. Olsalazine should be avoided due to a high frequency

Refractory ulcerative colitis Patients with mildly to moderately active UC who fail to respond to oral prednisone at a dose of 40-60 mg/day in combination with mesalamine, sulfasalazine, or balsalazide can be considered to have refractory UC. One potential approach to treatment is hospitalization for intravenous administration of corticosteroids. The rationale for this treatment

620 approach is a clinical trial that demonstrated greater mean serum prednisolone concentrations with intravenous compared to oral dosing [135]. The mainstay of treatment in patients who have failed combination therapy with maximal doses of oral and rectal mesalamine and oral corticosteroids is AZA or 6-MP. The prodrug AZA is approximately 50% 6-MP by molecular weight, requiring a conversion factor of 2 to convert a dose of AZA to a therapeutically equivalent dose of 6-MP. The doses of AZA and 6-MP shown to be effective for UC and CD in controlled trials range from 100 mg/day to 3.0 mg/kg per day. For patients with normal AZA and 6MP metabolism (based on normal thiopurine methyltransferase (TPMT) activity), a starting AZA dose of 2.0-2.5 mg/kg per day or 6-MP dose of 1.01.5 mg/kg per day is recommended (see Chapter 25 on Clinical pharmacology for further details). Patients with decreased TPMT activity should have their starting AZA or 6-MP dose reduced by 50% to 1.0-1.25 mg/kg per day and 0.5-0.75 mg/kg per day, respectively. Patients with absent TPMT activity should not be treated with AZA or 6-MP. AZA and 6-MP are slow-acting antimetabolite drugs, requiring at least 1-2 months and perhaps 3-4 months to reach the full clinical effect. Thus, concomitant therapy with corticosteroids should not be tapered below a dose of 15-20 mg/day for 2-3 months in patients who are beginning AZA or 6-MP. Concomitant therapy with mesalamine, sulfasalazine, olsalazine, or balsalazide is continued in most cases. Measurement of a total leukocyte count every 1-2 months as long as patients are receiving AZA or 6MP is mandatory to monitor for leukopenia. Indications for treatment with AZA or 6-MP in patients with UC include: induction of remission in steroidrefractory disease; steroid sparing in steroid-dependent disease; and maintenance of remission in patients who have failed maintenance therapy with high-dose mesalamine, sulfasalazine, olsalazine, or balsalazide. Transdermal nicotine may be of benefit as an alternative to AZA or 6-MP in patients with active UC refractory to mesalamine, sulfasalazine, olsalazine, balsalazide, and corticosteroids. Because nonsmokers frequently experience side-effects from nicotine when initally beginning therapy, a dose escalation strategy in which patients are treated with 7 mg/day of transdermal nicotine for 1 week, then 14 mg/day for 1 week, then 21 mg/day is recommended. Clinical trials have demonstrated efficacy for transdermal nicotine over 4-6 weeks of treatment. The

Medical management of ulcerative colitis role of maintenance therapy with transdermal nicotine beyond 6 weeks is unknown.

Severe ulcerative colitis Severe UC is defined using the Truelove and Witts criteria' (Table 1) [14]. Toxic (fulminant) colitis is defined as the sudden extension of mucosal inflammation through all layers of the colonic wall to the serosa and presents clinically as focal visceral tenderness to deep palpation. Megacolon is defined as dilation of the colon (5-6 cm or more) demonstrated by X-ray and presents clinically as abdominal distension, decreased or absent bowel sounds, and in some cases decreased stool frequency. Approximately 10% of all patients with UC will develop a severe flare at some point in their disease course, whereas only 12% progress to toxic (fulminant) colitis and/or megacolon. The mortality for severe UC is 2%, but remains approximately 30% for toxic (fulminant) colitis. In patients with severe or toxic colitis, hospitalization is mandatory. The treatment regimen outlined by Truelove and Jewell, consisting of intravenous fluids, electrolyte supplements, bowel rest, transfusion if indicated, intravenous antibiotics, intravenous corticosteroids, and rectal corticosteroids, remains in use today [136], although controlled trials have demonstrated that intravenous antibiotics are not of benefit. Sixty percent of patients treated with this regimen will be symptom-free by the end of 5 days, 15% will have significant improvement, and 25% will not improve and should be treated with cyclosporine or surgery. Most patients hospitalized with severe UC should continue to receive a normal diet. Two randomized controlled trials have demonstrated that bowel rest does not affect the outcome of severe UC in patients treated with intravenous prednisone [307, 308]. Patients with toxic colitis or megacolon should be made nil per os because of the potential for eminent surgical intervention. Peripheral or central intravenous nutrition should be instituted if there is evidence of malnutrition. The goal of intravenous nutrition is to replace nutritional deficits rather than for any primary therapeutic benefit. Factors which have been implicated in the development of toxic megacolon should be avoided including barium enema, narcotic antidiarrheals (codeine, tincture of opium, loperamide, and diphenoxylate), anticholinergic agents, antidepressants, and electrolyte imbalance. Patients should be mon-

William J. Sandborn itored frequently. Abdominal X-ray may be indicated daily in patients with severe colitis and twice daily in patients with megacolon. Frequent physical examination by both an experienced gastroenterologist and surgeon is also of great importance, as is frequent monitoring of the complete blood count, electrolytes, and nutritional parameters. Mesalamine, sulfasalazine, olsalazine, and balsalazide should in general be temporarily discontinued in patients hospitalized with severe or toxic UC because of the possibility of a drug-induced exacerbation of colitis which can be indistinguishable from a flare of colitis. Controlled clinical trials have demonstrated that antibiotics have no role in the treatment of severe UC unless a specific infection is suspected. Nevertheless, many authorities continue to advocate broad-spectrum antibiotic therapy with either a combination of metronidazole, an aminoglycoside, and a broad-spectrum penicillin or with a third-generation cephalosporine. Intravenous corticosteroid therapy should be initiated with hydrocortisone 300-400 mg/day or methylprednisolone 40-60 mg/day. There are no controlled data to determine the frequency of administration but a pharmacokinetic study showed that corticosteroid blood levels are better maintained within the presumptive therapeutic range when administered as a continuous infusion rather than as intermittent bolus therapy. For patients who have not been recently treated with steroids there is some evidence to suggest that they may respond better to intravenous ACTH administered at a dose of 40 units every 8 h than to conventional corticosteroids. Patients with severe UC who fail to respond to 710 days of intravenous corticosteroid therapy may be considered for 'rescue' therapy with intravenous cyclosporine at a dose of 4 mg/kg per day administered as a continuous infusion over 24 h. This dose should be adjusted to maintain a whole blood cyclosporine A concentration (high-performance liquid chromatography or monoclonal radioimmunoassay) of approximately 300-350 ng/ml. Patients with known infection, hypocholesterolemia (risk of seizure), or significant renal insufficiency should not be treated with cyclosporine. Significant abdominal pain is probably a relative indication since this may represent transmural inflammation and potentially early perforation. Patients who fail to respond within 7-10 days should undergo colectomy. Patients who respond to intravenous cyclosporine should be discharged on oral cyclosporine at a dose approximately twice the total daily dose that they received

621 intravenously. This may be administered as standard oral cyclosporine (Sandimmune) or as the microemulsion oral formulation of cyclosporine (Neoral). Again, the oral cyclosporine dose should be adjusted to maintain a whole blood concentration of cyclosporine A in the range of approximately 300-350 ng/ ml. Oral cyclosporine should be overlapped for 3-4 months with the slow-acting immune modifier agents AZA or 6-MP, which are then continued long-term for maintenance of remission. Corticosteroids can be tapered over 2-3 months. Prophylaxis for Pneumocystis carrini pneumonia with trimethoprim/sulfomethoxazole is recommended while patients are receiving triple-drug immunosuppression.

Conclusions Initial treatment of mild to moderately active UC may be sulfasalazine, oral or rectal mesalamine, balsalazide, corticosteroid enemas, or oral corticosteroids. Patients with persistent mild to moderate symptoms of active UC in spite of these therapies (treatment-refractory) may require AZA/6-MP, and in some cases intravenous corticosteroid or transdermal nicotine. Patients with severely active UC should be treated with intravenous conventional corticosteroids, and intravenous cyclosporine may be considered in those who do not respond. Remission should be maintained with sulfasalazine, oral or rectal mesalamine, olsalazine, or balsalazide, and in some cases with AZA/6-MP.

References 1. 2. 3. 4.

5. 6. 7.

Tremaine WJ, Sandborn WJ. Practice guidelines for inflammatory bowel disease: an instrument for assessment. Mayo Clin Proc 1999; 74: 495-501. Miner PB Jr, Peppercorn MA, Targan SR. A rational approach to 5-aminosalicylic acid therapy in ulcerative colitis. Hosp Pract (Off"Ed) 1993; 28(Suppl. )3: 1-24. Loftus EV Jr, Sandborn WJ. Drug therapy for inflammatory bowel disease. Contemp Intern Med 1995; 7: 21-34. Williams CN, Haber G, Aquino JA. Double-blind, placebocontrolled evaluation of 5-ASA suppositories in active distal proctitis and measurement of extent of spread using 99mTclabeled 5-ASA suppositories. Dig Dis Sci 1987; 32: 71-5S. Jay M, Beihn RM, Digenis GA, Deland FH, Caldwell L, Mlodozeniec AR. Disposition of radiolabelled suppositories in humans. J Pharm Pharmacol 1985; 37: 266-8. Chapman NJ, Brown ML, Phillips SF et al. Distribution of mesalamine enemas in patients with active distal ulcerative colitis. Mayo Clin Proc 1992; 67: 245-8. Tiel-van Buul MM, Mulder CJ, van Royen EA, Wiltink EH, Tytgat GN. Retrograde spread of mesalazine (5-aminosalicylic acid)-containing enema in patients with ulcerative colitis. Clin Pharmacokinet 1991; 20: 247-51.

Medical management of ulcerative colitis

622 8. 9.

10. 11.

12. 13.

14. 15.

16.

17.

18. 19. 20. 21.

22. 23.

24. 25. 26. 27.

Jay M, Digenis GA, Foster TS, Antonow DR. Retrograde spreading of hydrocortisone enema in inflammatory bowel disease. Dig Dis Sci 1986; 31: 139-44. Campieri M, Lanfranchi GA, Brignola C et al. Retrograde spread of 5-aminosalicylic acid enemas in patients with active ulcerative colitis. Dis Colon Rectum 1986; 29: 10810. Kruis W, Bull U, Eisenburg J, Paumgartner G. Retrograde colonic spread of sulphasalazine enemas. Scand J Gastroenterol 1982; 17:933-8. Nyman-Pantelidis M, Nilsson A, Wagner ZG, Borga O. Pharmacokinetics and retrograde colonic spread of budesonide enemas in patients with distal ulcerative colitis. Aliment Pharmacol Ther 1994; 8: 617-22. Swarbrick ET, Loose H, Lennard-Jones JE. Enema volume as an important factor in successful topical corticosteroid treatment of colitis. Proc R Soc Med 1974; 67: 753-4. van Bodegraven AA, Boer RO, Lourens J, Tuynman HA, Sindram JW. Distribution of mesalazine enemas in active and quiescent ulcerative colitis. Aliment Pharmacol Ther 1996; 10: 327 32. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Final report on a therapeutic trial. Br Med J 1955; 2: 1041 8. Sutherland LR, Martin F, Greer S ct al. 5-Aminosalicylic acid enema in the treatment of distal ulcerative colitis, proctosigmoiditis, and proctitis. Gastroenterology 1987; 92: 1894 8. Schroedcr KW, Trcmaine WJ, Ilstrup DM. Coated oral 5aminosalicylic acid therapy for mildly to moderately active ulcerative colitis. A randomized study. N Engl J Med 1987; 317: 1625 9. Svartz N. Salazyopyrin, a new sulfanilamide preparation: A. Therapeutic results in rheumatic polyarthritis. B. Therapeutic results in ulcerative colitis. C. Toxic manifestations in treatment with sulfanilamide preparation. Acta Med Scand 1942;110: 557 90. Baron JH, Conncll AM, Lennard-Jones JE, Jones FA. Sulphasalazine and salicylazosulphadimidine in ulcerative colitis. Lancet 1962; 1: 1094 6. Dick AP, Grayson MJ, Carpenter RG, Petrie A. Controlled trial of sulphasalazine in the treatment of ulcerative colitis. Gut 1964; 5: 437 42. Palmer KR, Goepel JR, Holdsworth CD. Sulphasalazine retention enemas in ulcerative colitis: a double-blind trial. Br Med J (Clin Res Ed) 1981; 282: 1571-3. Fruhmorgen P, Demling L. On the efficacy of ready-madeup commercially available salicylazosulphapyridine enemas in the treatment of proctitis, proctosigmoiditis and ulcerative colitis involving rectum, sigmoid and descending colon. Hepatogastroenterology 1980; 27: 473-6. Truelove SC, Watkinson G, Draper G. Comparison of corticosteroid and sulphasalazine in the treatment of ulcerative colitis. Br Med J 1962; 2: 1708-11. Lennard-Jones JE, Longmore AJ, Newell AC et al. An assessment of prednisone, salazopyrine and topical hydrocortisone hemisuccinate used as outpatient treatment for ulcerative colitis. Gut 1960; 1: 217-22. Misiewicz JJ, Lennard-Jones JE, Connell AM, Parson JH, Avery Jones F. Controlled trial of sulphasalazine in maintenance therapy for ulcerative colitis. Lancet 1965; 1: 185-8. Dissanayake AS, Truelove SC. A controlled therapeutic trial of long-term maintenance treatment of ulcerative colitis with sulphazalazine (Salazopyrin). Gut 1973; 14: 923-6. Azad Khan AK, Howes DT, Piris J, Truelove SC. Optimum dose of sulphasalazine for maintenance treatment in ulcerative colitis. Gut 1980; 21: 232-40. Riis P, Anthonisen P, Wullf HR, Folkenborg O, Bonnevie O, Binder V. The prophylactic effect of salazosulphapyridine in ulcerative colitis during long-term treatment. A double-

28. 29. 30.

31. 32.

33. 34. 35.

36. 37.

38.

39.

40. 41. 42. 43.

44. 45. 46.

blind trial on patients asymptomatic for one year. Scand J Gastroenterol 1973; 8: 71-4. Davies PS, Rhodes J. Maintenance of remission in ulcerative colitis with sulphasalazine or a high-fibre diet: a clinical trial. Br Med J 1978; 1: 1524-5. Dronfield MW, Langman MJ. Comparative trial of sulphasalazine and oral sodium cromoglycate in the maintenance of remission in ulcerative colitis. Gut 1978; 19: 1136-9. Willoughby CP, Heyworth MF, Piris J, Truelove SC. Comparison of disodium cromoglycate and sulphasalazine as maintenance therapy for ulcerative colitis. Lancet 1979; 1: 119-22. Davies PS, Rhodes J, Counsell B, Evans BK. Maintenance of remission in ulcerative colitis. Effect of an orally absorbed mast cell stabilizer. Am J Gastroenterol 1980; 74: 150-3. Das KM, Eastwood MA, McMAnus JPA, Sircus W. The metabolism of salicylazosulphapyridine in ulcerative colitis IL The relationship between metabolites and the progress of the disease studied in out-patients. Gut 1973; 14: 637-41. Peppercorn MA, Goldman P. The role of intestinal bacteria in the metabolism of salicylazosulfapyridine. J Pharmacol Exp Ther 1972; 181: 555 62. Azad Khan AK, Guthrie G, Johnston HH, Truelove SC, Williamson DH. Tissue and bacterial splitting of sulphasalazine. Clin Sci 1983; 64: 349 54. Das KM, Eastwood MA, McManus JP, Sircus W. The metabolism of salicylazosulphapyridine in ulcerative colitis. I. The relationship between metabolites and the response to treatment in inpatients. Gut 1973; 14: 631 41. Azad Khan AK, Piris J, Truelove SC. An experiment to determine the active therapeutic moiety of sulphasalazine. Lancet 1977; 2: 892 5. van Hccs PA, Bakkcr JH, van Tongercn JH. Effect of sulphapyridine, 5-aminosalicylic acid, and placebo in patients with idiopathic proctitis: a study to determine the active therapeutic moiety of sulphasalazine. Gut 1980; 21: 632 5. Klotz U, Maier K, Fischer C, Heinkel K. Therapeutic efficacy of sulfasalazine and its metabolites in patients with ulcerative colitis and Crohn's disease. N Engl J Med 1980; 303: 1499 502. Das KM, Eastwood MA, McManus JP, Sircus W. Adverse reactions during salicylazosulfapyridine therapy and the relation with drug metabolism and acetylator phenotype. N Engl J Med 1973; 289: 491 5. TafTet SL, Das KM. Sulfasalazine. Adverse effects and desensitization. Dig Dis Sci 1983; 28: 833-42. Shanahan F, Targan S. Sulfasalazine and salicylate-induced exacerbation of ulcerative colitis [Letter]. N Engl J Med 1987; 317: 455. Campieri M, Gionchetti P, Belluzzi A et al. Optimum dosage of 5-aminosalicylic acid as rectal enemas in patients with active ulcerative colitis. Gut 1991; 32: 929-31. Hanauer SB. Dose-ranging study of mesalamine (PENTASA) enemas in the treatment of acute ulcerative proctosigmoiditis: results of a multicentered placebo- controlled trial. The U.S. PENTASA Enema Study Group. Inflam Bowel Dis 1998;4:79-83. Sutherland LR, Martin F. 5-Aminosalicylic acid enemas in treatment of distal ulcerative colitis and proctitis in Canada. Dig Dis Sci 1987; 32: 64-6S. Marshall JK, Irvine EJ. Rectal corticosteroids versus alternative treatments in ulcerative colitis: a meta-analysis. Gut 1997;40:775-81. Marteau P, Florent C. Comparative, open, randomized trial of the efficacy and tolerance of slow-release 5-ASA suppositories once daily versus conventional 5-ASA suppositories twice daily in the treatment of active cryptogenic proctitis:

William J. Sandborn

47.

48. 49.

50. 51.

52.

53.

54.

55.

56. 57.

58.

59.

60.

61.

62. 63.

French Pentasa Study Group. Am J Gastroenterol 2000; 95: 166-70. Campieri M, Paoluzi P, D'Albasio G, Brunetti G, Pera A, Barbara L. Better quality of therapy with 5-ASA colonic foam in active ulcerative colitis. A multicenter comparative trial with 5-ASA enema. Dig Dis Sci 1993; 38: 1843-50. Gionchetti P, Rizzello F, Venturi A et al. Comparison of mesalazine suppositories in proctitis and distal proctosigmoiditis. Aliment Pharmacol Ther 1997; 11: 1053-7. Danish 5-ASA Group. Topical 5-aminosahcylic acid versus prednisolone in ulcerative proctosigmoiditis. A randomized, double-bhnd multicenter trial. Dig Dis Sci 1987; 32: 598602. Campieri M, Lanfranchi GA, Bazzocchi G G et al. Treatment of ulcerative colitis with high-dose 5-aminosahcylic acid enemas. Lancet 1981; 2: 270-1. Porro GB, Ardizzone S, Petrillo M, Fasoli A, Molteni P, Imbesi V. Low Pentasa dosage versus hydrocortisone in the topical treatment of active ulcerative colitis: a randomized, double-bhnd study Am J Gastroenterol 1995; 90: 736-9. Mulder CJ, Tytgat GN, Wiltink EH, Houthoff HJ. Comparison of 5-aminosahcylic acid (3 g) and prednisolone phosphate sodium enemas (30 mg) in the treatment of distal ulcerative colitis. A prospective, randomized, double-blind trial. Scand J Gastroenterol 1988; 23: 1005-8. Farup PG, Hovde O, Halvorsen FA, Raknerud N, Brodin U. Mesalazine suppositories versus hydrocortisone foam in patients with distal ulcerative colitis. A comparison of the efficacy and practicality of two topical treatment regimens. Scand J Gastroenterol 1995; 30: 164-70. Friedman LS, Richter JM, Kirkham SE, DeMonaco HJ, May RJ. 5-Aminosalicylic acid enemas in refractory distal ulcerative colitis: a randomized, controlled trial. Am J Gastroenterol 1986; 81: 412-18. Lee FI, Jewell DP, Mani V et al. A randomised trial comparing mesalazine and prednisolone foam enemas in patients with acute distal ulcerative colitis. Gut 1996; 38: 229-33. Campieri M, Gionchetti P, Belluzzi A et al. Efficacy of 5aminosalicylic acid enemas versus hydrocortisone enemas in ulcerative colitis. Dig Dis Sci 1987; 32: 67-70S. Lucidarme D, Marteau P, Foucault M, Vautrin B, Filoche B. Efficacy and tolerance of mesalazine suppositories vs. hydrocortisone foam in proctitis. Aliment Pharmacol Ther 1997; 11:335-40. Lemann M, Galian A, Rutgeerts P et al. Comparison of budesonide and 5-aminosalicylic acid enemas in active distal ulcerative colitis. Aliment Pharmacol Ther 1995; 9: 557-62. Cobden I, al Mardini H, Zaitoun A, Record CO. Is topical therapy necessary in acute distal colitis? Double-blind comparison of high-dose oral mesalazine versus steroid enemas in the treatment of active distal ulcerative colitis. Aliment Pharmacol Ther 1991; 5: 513-22. Kam L, Cohen H, Dooley C, Rubin P, Orchard J. A comparison of mesalamine suspension enema and oral sulfasalazine for treatment of active distal ulcerative colitis in adults. Am J Gastroenterol 1996; 91: 1338-42. Campieri M, Lanfranchi GA, Bertoni F et al. A doubleblind clinical trial to compare the effects of 4-aminosahcylic acid to 5-aminosalicylic acid in topical treatment of ulcerative colitis. Digestion 1984; 29: 204-8. Pullan RD, Ganesh S, Mani V et al. Comparison of bismuth citrate and 5-aminosalicylic acid enemas in distal ulcerative coHtis: a controlled trial. Gut 1993; 34: 676-9. Campieri M, Gionchetti P, Belluzzi A et al. Sucralfate, 5aminosalicylic acid and placebo enemas in the treatment of distal ulcerative colitis. Eur J Gastroenterol Hepatol 1991; 3:41-4.

623 64.

65.

66.

67.

68.

69.

70. 71.

72.

73.

74. 75.

76.

77. 78. 79.

80. 81.

Safdi M, DeMicco M, Sninsky C et al. A double-blind comparison of oral versus rectal mesalamine versus combination therapy in the treatment of distal ulcerative colitis. Am J Gastroenterol 1997; 92: 1867-71. Biddle WL, Greenberger NJ, Swan JT, McPhee MS, Miner PB Jr. 5-Aminosalicylic acid enemas: effective agent in maintaining remission in left-sided ulcerative colitis [Published erratum appears in Gastroenterology 1989; 96: 1630]. Gastroenterology 1988; 94: 1075-9. Miner P, Daly R, Nester T and the Rowasa Study Group. The effect of varying the dose intervals of mesalamine enemas for the prevention of relapse in distal ulcerative colitis. Gastroenterology 1994; 106: A736. DArienzo A, Panarese A, DArmiento FP et al. 5-Aminosahcyhc acid suppositories in the maintenance of remission in idiopathic proctitis or proctosigmoiditis: a double-bhnd placebo-controlled clinical trial. Am J Gastroenterol 1990; 85: 1079-82. DAlbasio G, Paoluzi P, Campieri M et al. Maintenance treatment of ulcerative proctitis with mesalazine suppositories: a double-blind placebo-controlled trial. Italian IBD Study Group. Am J Gastroenterol 1998; 93: 799-803. Marteau P, Grand J, Foucault M, Rambaud JC. Use of mesalazine slow release suppositories 1 g three times per week to maintain remission of ulcerative proctitis: a randomised double blind placebo controlled multicentre study. Gut 1998; 42: 195-9. Sutherland L, Martin F. 5-Aminosalicylic acid enemas in the maintenance of distal ulcerative colitis and proctitis. Can J Gastroenterol 1987; 1: 3-6. DAlbasio G, Trallori G, Ghetti A et al. Intermittent therapy with high-dose 5-aminosalicylic acid enemas for maintaining remission in ulcerative proctosigmoiditis. Dis Colon Rectum 1990; 33: 394^7. Mantzaris GJ, Hatzis A, Petraki K, Spihadi C, Triantaphyllou G. Intermittent therapy with high-dose 5-aminosalicylic acid enemas maintains remission in ulcerative proctitis and proctosigmoiditis. Dis Colon Rectum 1994; 37: 58-62. DAlbasio G, Pacini F, Camarri E et al. Combined therapy with 5-aminosalicylic acid tablets and enemas for maintaining remission in ulcerative colitis: a randomized doubleblind study. Am J Gastroenterol 1997; 92: 1143-7. Sninsky CA, Cort DH, Shanahan F et al. Oral mesalamine (Asacol) for mildly to moderately active ulcerative colitis. A multicenter study. Ann Intern Med 1991; 115: 350-5. Sutherland LR, Robinson M, Onstad G et al. A doubleblind, placebo controUed, multicenter study of the efficacy and safety of 5-aminosalicylic acid tablets in the treatment of ulcerative colitis. Can J Gastroenterol 1990; 4: 463-7. Hanauer S, Schwartz J, Robinson M et al. Mesalamine capsules for treatment of active ulcerative colitis: results of a controlled trial. Pentasa Study Group. Am J Gastroenterol 1993; 88: 1188-97. Sutherland LR, May GR, Shaffer EA. Sulfasalazine revisited: a meta-analysis of 5-aminosalicylic acid in the treatment of ulcerative colitis. Ann Intern Med 1993; 118: 540-9. Sutherland LR, Roth DE, Beck PL. Alternatives to sulfasalazine: a meta-analysis of 5-ASA in the treatment of ulcerative colitis. Infiam Bowel Dis 1997; 3: 65-78. Munakata A, Yoshida Y, Muto T et al. Double-blind comparative study of sulfasalazine and controUed-release mesalazine tablets in the treatment of active ulcerative colitis. J Gastroenterol 1995; 30(Suppl. 8): 108-11. Rachmilewitz D. Coated mesalazine (5-aminosahcyhc acid) versus sulphasalazine in the treatment of active ulcerative colitis: a randomised trial. Br Med J 1989; 298: 82-6. Riley SA, Mani V, Goodman MJ, Herd ME, Dutt S, Turnberg LA. Comparison of delayed release 5 aminosahcyhc acid (mesalazine) and sulphasalazine in the treatment of

Medical management of ulcerative colitis

624

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93. 94.

95.

96.

97.

mild to moderate ulcerative colitis relapse. Gut 1988; 29: 669-74. Mesalamine Study Group. An oral preparation of mesalamine as long-term maintenance therapy for ulcerative colitis. A randomized, placebo-controlled trial. Ann Intern Med 1996; 124:204-11. Miner P, Hanauer S, Robinson M, Schwartz J, Arora S. Safety and efficacy of controlled-release mesalamine for maintenance of remission in ulcerative colitis. Pentasa UC Maintenance Study Group. Dig Dis Sci 1995; 40: 296-304. Ardizzone S, Petrillo M, Molteni P, Desideri S, Bianchi PG. Coated oral 5-aminosalicylic acid (Claversal) is equivalent to sulfasalazine for remission maintenance in ulcerative colitis. A double-blind study. J Clin Gastroenterol 1995; 21: 287 9. Dew MJ, Hughes P, Harries AD, Williams G, Evans BK, Rhodes J. Maintenance of remission in ulcerative colitis with oral preparation of 5-aminosalicylic acid. Br Med J (CHnResEd) 1982; 285: 1012. Mulder CJ, Tytgat GN, Weterman IT ct al. Double-blind comparison of slow-release 5-aminosalicylate and sulfasalazine in remission maintenance in ulcerative colitis. Gastroenterology 1988; 95: 1449 53. Riley SA, Mani V, Goodman MJ, Herd ME, Dutt S, Turnberg LA. Comparison of delayed-relcasc 5-aminosalicylic acid (mcsalazine) and sulfasalazine as maintenance treatment for patients with ulcerative colitis. Gastroenterology 1988; 94: 1383 9. Rutgecrts P. Comparative cllicacy of coated, oral 5-aminosalicylic acid (Claversal) and sulphasalazine for maintaining remission of ulcerative colitis. International Study Group. Aliment Pharmacol Ther 1989; 3: 183 91. Fernandez-Banares F, Hinojosa J, Sanchez-Lombrana JL ct al. Randomized clinical trial of Plantui^o ovata seeds (dietary fiber) as compared with mesalamine in maintaining remission in ulcerative colitis. Spanish Group for the Study of Crohn's Disease and Ulcerative Colitis (GETECCU). Am J Gastroenterol 1999; 94: 427 33. Rembacken BJ, Snelling AM, Hawkey PM, Chalmers DM, Axon AT. Non-pathogenic Escherichia coli versus mcsalazine for the treatment of ulcerative colitis: a randomised trial. Lancet 1999; 354: 635 9. Kruis W, Schutz E, Fric P, Fixa B, Judmaier G, Stolte M. Double-blind comparison of an oral Escherichia coli preparation and mesalazine in maintaining remission of ulcerative colitis. Aliment Pharmacol Ther 1997; 11: 853-8. Meyers S, Sachar DB, Present DH, Janowitz HD. Olsalazine sodium in the treatment of ulcerative colitis among patients intolerant of sulfasalazine. A prospective, randomized, placebocontroUed, double-blind, dose-ranging clinical trial. Gastroenterology 1987; 93: 1255-62. Zinberg J, Molinas S, Das KM. Double-blind placebocontrolled study of olsalazine in the treatment of ulcerative colitis. Am J Gastroenterol 1990; 85: 562-6. Hanauer SB, Barish C, Pambianco D et al. A multi-center, double-blind, placebo-controlled, dose-ranging trial of olsalazine for mild-moderately active ulcerative colitis. Gastroenterology 1996; 110: A921. Feurle GE, Theuer D, Velasco S et al. Olsalazine versus placebo in the treatment of mild to moderate ulcerative colitis: a randomised double blind trial. Gut 1989; 30: 1354-61. Hetzel DJ, Shearman DJ, Labrooy J et al. Olsalazine in the treatment of active ulcerative colitis: a placebo controlled clinical trial and assessment of drug disposition. Scand J Gastroenterol 1988; 148(Suppl.): 61-9. Selby WS, Barr GD, Ireland A, Mason CH, Jewell DP. Olsalazine in active ulcerative colitis. Br Med J (Clin Res Ed) 1985; 291: 1373-5.

98.

99.

100. 101.

102.

103.

104. 105. 106. 107.

108.

109.

110. 111.

112.

113.

114.

115.

Pamukcu R, Hanauer SB, Chang EB. Effect of disodium azodisalicylate on electrolyte transport in rabbit ileum and colon in vitro. Comparison with sulfasalazine and 5-aminosalicylic acid. Gastroenterology 1988; 95: 975-81. WiUoughby CP, Cowan RE, Gould SR, Machell RJ, Stewart JB. Double-blind comparison of olsalazine and sulphasalazine in active ulcerative colitis. Scand J Gastroenterol 1988; 148(Suppl.):40-4. Ewe K, Eckardt V, Kanzler G. Treatment of ulcerative colitis with olsalazine and sulphasalazine: efficacy and sideeffects. Scand J Gastroenterol 1988; 148(Suppl.): 70-5. Rao SS, Dundas SA, Holdsworth CD, Cann PA, Palmer KR, Corbett CL. Olsalazine or sulphasalazine in first attacks of ulcerative colitis? A double blind study. Gut 1989;30:675-9. Ferry GD, Kirschner BS, Grand RJ et al. Olsalazine versus sulfasalazine in mild to moderate childhood ulcerative colitis: results of the Pediatric Gastroenterology Collaborative Research Group Clinical Trial. J Pediatr Gastroenterol Nutr 1993; 17: 32 8. Sandberg-Gertzcn H, Jarnerot G, Kraaz W. Azodisal sodium in the treatment of ulcerative colitis. A study of tolerance and relapse-prevention properties. Gastroenterology 1986;90: 1024 30. Wright JP, O'Keefe EA, Cuming L, Jaskiewicz K. Olsalazine in maintenance of clinical remission in patients with ulcerative colitis. Dig Dis Sci 1993; 38: 1837 42. Courtney MG, Nunes DP, Bergin CF ct al. Randomised comparison of olsalazine and mesalazine in prevention of relapses in ulcerative colitis. Lancet 1992; 339: 1279 81. Nilsson A, Danielsson A, Lotlierg R ct al. Olsalazine versus sulphasalazine for relapse prevention in ulcerative colitis: a multicentcr study. Am J Gastroenterol 1995; 90: 381 7. Kiilerich S, Ladcfoged K, Rannem T, Ranlov PJ. Prophylactic effects of olsalazine v sulphasalazine during 12 months maintenance treatment of ulcerative colitis. Danish Olsalazine Study Group. Gut 1992; 33: 252 5. Kruis W, Judmaier G, Kayasseh L ct al. Double-blind dosefinding study of olsalazine versus sulphasalazine as maintenance therapy for ulcerative colitis. Eur J Gastroenterol Hepatol 1995:7:391 6. Rijk MC, van Lier HJ, van Tongeren JH. Relapse-preventing effect and safety of sulfasalazine and olsalazine in patients with ulcerative colitis in remission: a prospective, double- blind, randomized multicenter study. Ulcerative Colitis Multicenter Study Group. Am J Gastroenterol 1992;87:438-42. Ireland A, Mason CH, Jewell DP. Controlled trial comparing olsalazine and sulphasalazine for the maintenance treatment of ulcerative colitis. Gut 1988; 29: 835-7. Green JR, Swan CH, Rowlinson A et al. Short report: comparison of two doses of balsalazide in maintaining ulcerative colitis in remission over 12 months. Aliment Pharmacol Ther 1992; 6: 647-52. Green JR, Lobo AJ, Holdsworth CD et al. Balsalazide is more effective and better tolerated than mesalamine in the treatment of acute ulcerative colitis. Abacus Investigator Group. Gastroenterology 1998; 114: 15-22. Green JR, Gibson JA, Kerr GD et al. Maintenance of remission of ulcerative colitis: a comparison between balsalazide 3 g daily and mesalazine 1.2 g daily over 12 months. ABACUS Investigator group. Aliment Pharmacol Ther 1998;12:1207-16. Giaffer MH, Holdsworth CD, Lennard-Jones JE et al. Improved maintenance of remission in ulcerative colitis by balsalazide 4 g/day compared with 2 g/day. Aliment Pharmacol Ther 1992; 6: 479-85. Mclntyre PB, Rodrigues CA, Lennard-Jones JE et al. Balsalazide in the maintenance treatment of patients with

William J. Sandborn

116. 117.

118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129.

130.

131. 132. 133.

134. 135. 136. 137.

ulcerative colitis, a double-blind comparison with sulphasalazine. Aliment Pharmacol Ther 1988; 2: 237-43. Bitton A, Peppercorn MA, Hanrahan JP, Upton MP. Mesalamine-induced lung toxicity. Am J Gastroenterol 1996; 91: 1039-40. laquinto G, Sorrentini I, Petillo FE, Berardesca G. Pleuropericarditis in a patient with ulcerative colitis in longstanding 5-aminosalicylic acid therapy. Ital J Gastroenterol 1994; 26: 145-7. Deltenre P, Berson A, Marcellin P, Degott C, Biour M, Pessayre D. Mesalazine (5-aminosalicylic acid) induced chronic hepatitis. Gut 1999; 44: 886-8. Fernandez J, Sala M, Panes J, Feu F, Navarro S, Teres J. Acute pancreatitis after long-term 5-aminosalicylic acid therapy. Am J Gastroenterol 1997; 92: 2302-3. Brouillard M, Gheerbrant JD, Gheysens Y et al [Chronic interstitial nephritis and mesalazine: 3 new cases?]. Gastroenterol Clin Biol 1998; 22: 724-6. Calvino J, Romero R, Pintos E et al. Mesalazine-associated tubulo-interstitial nephritis in inflammatory bowel disease. Clin Nephrol 1998; 49: 265-7. Colombel JF, Brabant G, Gubler MC et al. Renal insufficiency in infant: side-effect of prenatal exposure to mesalazine? [Letter; see comments]. Lancet 1994; 344: 620-1. Corrigan G, Stevens PE. Review article: interstitial nephritis associated with the use of mesalazine in inflammatory bowel disease. Aliment Pharmacol Ther 2000; 14: 1-6. Riley SA, Lloyd DR, Mani V. Tests of renal function in patients with quiescent colitis: effects of drug treatment. Gut 1992;33: 1348-52. Schreiber S, Hamling J, Zehnter E et al. Renal tubular dysfunction in patients with inflammatory bowel disease treated with aminosahcylate. Gut 1997; 40: 761-6. Hanauer SB, Verst-Brasch C, Regalli G. Renal safety of long-term mesalamine therapy in inflammatory bowel disease (IBD). Gastroenterology 1997; 112: A991. Sturgeon JB, Bhatia P, Hermens D, Miner PB, Jr. Exacerbation of chronic ulcerative colitis with mesalamine. Gastroenterology 1995; 108: 1889-93. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Preliminary report on a therapeutic trial. Br Med J 1954; 2: 375-8. Baron JH, Connell AM, Kanaghinis TG, Lennard-Jones JE. Out-patient treatment of ulcerative colitis. Comparison between three doses of oral prednisone. Br Med J 1962; 2: 441-3. Powell-Tuck J, Bown RL, Lennard-Jones JE. A comparison of oral prednisolone given as single or multiple daily doses for active proctocolitis. Scand J Gastroenterol 1978; 13: 833-7. Truelove SC, Witts LJ. Cortisone and corticotropin in ulcerative colitis. Br Med J 1959; 1: 387-94. Lennard-Jones JE, Misiewicz JJ, Connell AM, Baron JH, Avery Jones F. Prednisone as maintenance treatment for ulcerative colitis in remission. Lancet 1965; 1: 188-9. PoweU-Tuck J, Bown RL, Chambers TJ, Lennard-Jones JE. A controlled trial of alternate day prednisolone as a maintenance treatment for ulcerative colitis in remission. Digestion 1981; 22: 263-70. Elliott PR, Powell-Tuck J, Gillespie PE J et al. Prednisolone absorption in acute colitis. Gut 1980; 21: 49-51. Berghouse LM, Elliott PR, Lennard-Jones JE, EngHsh J, Marks V. Plasma prednisolone levels during intravenous therapy in acute colitis. Gut 1982; 23: 980-3. Truelove SC, Jewell DP. Intensive intravenous regimen for severe attacks of ulcerative colitis. Lancet 1974; 1: 1067-70. Truelove SC, Willoughby CP, Lee EG, Kettlewell MG. Further experience in the treatment of severe attacks of ulcerative colitis. Lancet 1978; 2: 1086-8.

625 138. Jarnerot G, Rolny P, Sandberg-Gertzen H. Intensive intravenous treatment of ulcerative colitis. Gastroenterology 1985;89:1005-13. 139. Powell-Tuck J, BuckeU NA, Lennard-Jones JE. A controfled comparison of corticotropin and hydrocortisone in the treatment of severe proctocolitis. Scand J Gastroenterol 1977; 12: 971-5. 140. Kaplan HP, Portnoy B, Binder HJ, Amatruda T, Spiro H. A controlled evaluation of intravenous adrenocorticotropic hormone and hydrocortisone in the treatment of acute colitis. Gastroenterology 1975; 69: 91-5. 141. Meyers S, Sachar DB, Goldberg JD, Janowitz HD. Corticotropin versus hydrocortisone in the intravenous treatment of ulcerative colitis. A prospective, randomized, double-blind clinical trial. Gastroenterology 1983; 85: 351-7. 142. Ireland A, Rosenberg W, Jewell DP. High dose methylprednisolone in the treatment of active ulcerative colitis. Gastroenterology 1988; 29: A1466. 143. Powell-Tuck J, Lennard-Jones JE, May CS, Wilson CG, Paterson J W Plasma prednisolone levels after administration of prednisolone-21- phosphate as a retention enema in colitis. Br Med J 1976; 1: 193-5. 144. Lee DA, Taylor GM, James VH, Walker G. Plasma prednisolone levels and adrenocortical responsiveness after administration of prednisolone-21-phosphate as a retention enema. Gut 1979; 20: 349-55. 145. Farmer RG, Schumacher OP. Treatment of ulcerative colitis with hydrocrotisone enemas: relationship of hydrocortisone absorption, adrenal suppression, and clinical response. Dis Colon Rectum 1970; 13: 355-61. 146. Luman W, Gray RS, Pendek R, Palmer KR. Prednisolone metasulphobenzoate foam retention enemas suppress the hypothalamo-pituitary-adrenal axis. Aliment Pharmacol Ther 1994; 8: 255-8. 147. Bianchi Porro G, Prantera C, Campieri M et al. Comparative trial of methylprednisolone and budesonide enemas in the treatment of adctive distal colitis. Eur J Gastroenterol Hepatol 1994; 6: 125-30. 148. Danielsson A, Hellers G, Lyrenas R et al. A controlled randomized trial of budesonide versus prednisolone retention enemas in active distal ulcerative colitis. Scand J Gastroenterol 1987; 22: 987-92. 149. Truelove SC. Treatment of ulcerative colitis with local hydrocortisone hemisuccinate sodium. A report on a controlled therapeutic trial. Br Med J 1958; 1072-7. 150. Lennard-Jones JE, Baron JH, Connell AM, Avery Jones F. A double blind controlled trial of prednisolone-21-phosphate suppositories in the treatment of idiopathic proctitis. Gut 1962; 3: 207-10. 151. Watkinson G Treatment of ulcerative colitis with topical hydrocortisone hemisuccinate sodium: a controlled trial employing restricted sequential analysis. Br Med J 1958; 2: 1077-82. 152. Porro GB, Ardizzone S, Petrillo M, Fasoli A, Molteni P, Imbesi V. Low Pentasa dosage versus hydrocortisone in the topical treatment of active ulcerative colitis: a randomized, double-Wind study. Am J Gastroenterol 1995; 90: 736-9. 153. Campieri M, Lanfranchi GA, Bazzocchi G et al. Treatment of ulcerative colitis with high-dose 5-aminosalicylic acid enemas. Lancet 1981; 2: 270-1. 154. O'DonneU LJ, Arvind AS, Hoang P et al. Double blind, controlled trial of 4-aminosalicylic acid and prednisolone enemas in distal ulcerative colitis. Gut 1992; 33: 947-9. 155. Sharma MP, Duphare HV, Dasarathy S. A prospective randomized double blind trial comparing prednisolone and 4-aminosalicylic acid enemas in acute distal ulcerative cohtis. J Gastroenterol Hepatol 1992; 7: 173-7.

626 156. Grace RH, Gent AE, Hellier MD. Comparative trial of sodium cromoglycate enemas with prednisolone enemas in the treatment of ulcerative colitis. Gut 1987; 28: 88-92. 157. Halpern Z, Sold O, Baratz M, Konikoff F, Halak A, Gilat T. A controlled trial of beclomethasone versus betamethasone enemas in distal ulcerative colitis. J Clin Gastroenterol 1991;13:38-41. 158. Lennard-Jones JE. Betamethasone 17-valerate and prednisolone 21-phosphate retention enemas in proctocolitis. Br Med J 1971;3:84-6. 159. Ruddell WS, Dickinson RJ, Dixon MP, Axon AT. Treatment of distal ulcerative colitis (proctosigmoiditis) in relapse: comparison of hydrocortisone enemas and rectal hydrocortisone foam. Gut 1980; 21: 885-9. 160. Van Outryve M, Huble F, Van Eeghem P, De Vos M. Comaprison of ridogrel versus prednisolone, both administered rectally, for the treatment of active ulcerative colitis. Gastroenterology 1996; 110: A1035. 161. Danielsson A, Loll^erg R, Persson T ct al. A steroid enema, budesonide, lacking systemic effects for the treatment of distal ulcerative colitis or proctitis. Scand J Gastroenterol 1992; 27: 9 12. 162. Hanaucr SB, Robinson M, Pruitt R ct al. Budesonide enema for the treatment of active, distal ulcerative colitis and proctitis: a dose-ranging study. U.S. Budesonide Enema Study Group. Gastroenterology 1998; 115: 525 32. 163. Baylcss T, Sninsky C, US Budesonide Enema Study Group. Budesonide enema is an effective alternative to hydrocortisone enema in active distal ulcerative colitis. Gastroenterology 1995; 108: A778. 164. Lindgren S, Suhr O, Persson T, Panlzar N. Treatment of active distal ulcerative colitis (UC) and maintenance of remission with Entocort enema: a randomised controlled dosage study. Gut 1997; 41: A223. 165. Danish Budesonide Study Group. Budesonide enema in distal ulcerative colitis. A randomized dose response trial with prednisolone enema as positive control. Scand J Gastroenterol 1991; 26: 1225 30. 166. Lol>)erg R, Ostergaard TO, Langholz E et al. Budesonide versus prednisolone retention enemas in active distal ulcerative colitis [Published erratum appears in Aliment Pharmacol Ther 1995; 9: 213]. Aliment Pharmacol Ther 1994; 8: 623 9. 167. Tarpila S, Turunen U, Seppala K ct al. Budesonide enema in active haemorrhagic proctitis - a controlled trial against hydrocortisone foam enema. Aliment Pharmacol Ther 1994; 8: 591 5. 168. Lamers C, Meijer J, Engels L ct al. Comparative study of the topically acting glucocorticoid budesonide and 5-aminosalicylic acid enema therapy of proctitis and proctosigmoiditis. Gastroenterology 1991; 100: A223. 169. Hamilton I, Pinder IF, Dickinson RJ, Ruddell WS, Dixon MF, Axon AT. A comparison of prednisolone enemas with low-dose oral prednisolone in the treatment of acute distal ulcerative colitis. Dis Colon Rectum 1984; 27: 701-2. 170. Hanauer SB, Kirsner JB, Barrett WE. The treatment of leftsided ulcerative colitis with tixicortol pivalate. Gastroenterology 1986; A1449. 171. Mulder CJ, Endert E, van der HH ct al. Comparison of beclomethasone dipropionate (2 and 3 mg) and prednisolone sodium phosphate enemas (30 mg) in the treatment of ulcerative proctitis. An adrenocortical approach. Neth J Med 1989; 35: 18-24. 172. Mulder CJ, Fockens P, Van der Heide H, Tytgat GNJ. A controlled randomized trial of beclomethasone dipropionate (3 mg) versus 5-ASA (1 g) versus the combination of both (3 mg/1 g) as retention enemas in active distal ulcerative colitis. Gastroenterology 1994; 106: A739.

Medical management of ulcerative colitis 173. Riley SA, Gupta I, Mani V. A comparison of sucralfate and prednisolone enemas in the treatment of active distal ulcerative colitis. Scand J Gastroenterol 1989; 24: 1014-18. 174. van der Heide H, van den Brandt-Gradel V, Tytgat GN et al. Comparison of beclomethasone dipropionate and prednisolone 21-phosphate enemas in the treatment of ulcerative proctitis. J Clin Gastroenterol 1988; 10: 169-72. 175. Angus P, Snook JA, Reid M, Jewell DP. Oral fluticasone propionate in active distal ulcerative colitis. Gut 1992; 33: 711-14. 176. Hawthorne AB, Record CO, Holdsworth CD ct al. Double blind trial of oral fluticasone propionate v prednisolone in the treatment of active ulcerative colitis. Gut 1993; 34: 1258. 177. Lofberg R, Danielsson A, Suhr O ct al. Oral budesonide versus prednisolone in patients with active extensive and left-sided ulcerative colitis. Gastroenterology 1996; 110: 1713-18. 178. Singleton JW, Law DH, Kelley ML Jr, Mekhjian HS, Sturdevant RA. National Cooperative Crohn's Disease Study: adverse reactions to study drugs. Gastroenterology 1979; 77: 870 82. 179. Talar-Williams C, Sneller MC. Complications of corticosteroid therapy. Eur Arch Dotorhinolaryngol 1994; 251: 131 6. 180. Whitworth JA. Mechanisms of glucocorlicoid-induced hypertension [Clinical conference]. Kidney Int 1987; 31: 1213 24. 181. Gurwitz JH, Bohn RL, Glynn RJ, Monane M, Mogun H, Avorn J. Glucocorticoids and the risk for initiation of hypoglycemic therapy. Arch Intern Med 1994; 154: 97 101. 182. Stuck AE, Minder CE, Frey FJ. Risk of infectious complications in patients taking glucocorticosteroids. Rev Infect Dis 1989; 11:954 63. 183. Mankin HJ. Nontraumatic necrosis of bone (osteonecrosis). N Engl J Med 1992; 326: 1473 9. 184. Lukert BP, Raisz LG. Glucocorticoid-induced osteoporosis. Rheum Dis Chn N Am 1994; 20: 629 50. 185. Lacomis D, Samuels MA. Adverse neurologic effects of glucocorticosteroids. J Gen Intern Med 1991; 6: 367 77. 186. Urban RC Jr, Dreyer EB. Corticosteroid-induced glaucoma. Int Ophthalmol Clin 1993; 33: 135 9. 187. Urban RC Jr, Cotlier E. Corticosteroid-induced cataracts. Surv Ophthalmol 1986; 31: 102 10. 188. Kusunoki M, Moeslein G, Shoji Y ct al. Steroid complications in patients with ulcerative colitis. Dis Colon Rectum 1992;35:1003-9. 189. Bean RHD. The treatment of chronic ulcerative colitis with 6-mercaptopurine. Med J Austral 1962; 2: 592-3. 190. McKay IR, Wall AJ, Goldstein G. Response to azathioprine in ulcerative colitis. Report of 7 cases. Am J Dig Dis 1966; 11:536-45. 191. Bowen GE, Irons GV, Rhodes JB, Kirsner JB. Early experiences with azathioprine in ulcerative colitis. A note of caution. J Am Med Assoc 1966; 195: 166-70. 192. Jewell DP, Truelove SC. Azathioprine in ulcerative colitis: an interim report on a controlled therapeutic trial. Br Med J 1972; 1: 709-12. 193. Jewell DP, Truelove SC. Azathioprine in ulcerative cohtis: final report on controlled therapeutic trial. Br Med J 1974; 4: 627-30. 194. Caprilh R, Carratu R, Babbini M. Double-blind comparison of the effectiveness of azathioprine and sulfasalazine in idiopathic proctocolitis. Preliminary report. Am J Dig Dis 1975; 20: 115-20. 195. Kirk AP, Lennard-Jones JE. Controlled trial of azathioprine in chronic ulcerative colitis. Br Med J (Clin Res Ed) 1982; 284: 1291-2.

William J. Sandborn 196. Rosenberg JL, Wall A J, Levin B, Binder HJ, Kirsner JB. A controlled trial of azathioprine in the management of chronic ulcerative colitis. Gastroenterology 1975; 69: 96-9. 197. Hawthorne AB, Logan RF, Hawkey CJ et al. Randomised controlled trial of azathioprine withdrawal in ulcerative colitis. Br Med J 1992; 305: 20-2. 198. Adler DJ, Korelitz BL The therapeutic efficacy of 6-mercaptopurine in refractory ulcerative colitis. Am J Gastroenterol 1990;85:717-22. 199. Steinhart AH, Baker JP, Brzezinski A, Prokipchuk EJ. Azathioprine therapy in chronic ulcerative colitis. J Clin Gastroenterol 1990; 12: 271-5. 200. Lobo AJ, Foster FN, Burke DA, Johnston D, Axon AT. The role of azathioprine in the management of ulcerative colitis. Dis Colon Rectum 1990; 33: 374-7. 201. George J, Present DH, Pou R, Bodian C, Rubin PH. The long-term outcome of ulcerative colitis treated with 6mercaptopurine. Am J Gastroenterol 1996; 91: 1711-14. 202. Ardizzone S, Molteni P, Imbesi V, Bollani S, Bianchi PG, Molteni F. Azathioprine in steroid-resistant and steroiddependent ulcerative colitis [Published erratum appears in J Clin Gastroenterol 1998; 26: 231]. J Clin Gastroenterol 1997; 25: 330-3. 203. Present DH, Meltzer SJ, Krumholz MP, Wolke A, Korelitz BL 6-Mercaptopurine in the management of inflammatory bowel disease: short- and long-term toxicity. Ann Intern Med 1989; 111:641-9. 204. Council WR, Kamm MA, Ritchie JK, Lennard-Jones JE. Bone marrow toxicity caused by azathioprine in inflammatory bowel disease: 27 years of experience. Gut 1993; 34: 1081-5. 205. Kirschner BS. Safety of azathioprine and 6-mercaptopurine in pediatric patients with inflammatory bowel disease. Gastroenterology 1998; 115: 813-21. 206. Council WR, Kamm MA, Dickson M, Balkwill AM, Ritchie JK, Lennard-Jones JE. Long-term neoplasia risk after azathioprine treatment in inflammatory bowel disease. Lancet 1994; 343: 1249-52. 207. Bouhnik Y, Lemann M, Mary JY et al. Long-term follow-up of patients with Crohn's disease treated with azathioprine or 6-mercaptopurine. Lancet 1996; 347: 215-19. 208. Kozarek RA, Patterson DJ, Gelfand MD, Botoman VA, Ball TJ, Wilske KR. Methotrexate induces clinical and histologic remission in patients with refractory inflammatory bowel disease. Ann Intern Med 1989; 110: 353-6. 209. Baron TH, Truss CD, Elson CO. Low-dose oral methotrexate in refractory inflammatory bowel disease. Dig Dis Sci 1993; 38: 1851-6. 210. Feagan BG, Rochon J, Fedorak RN et al. Methotrexate for the treatment of Crohn's disease. North American Crohn's Study Group Investigators. N Engl J Med 1995; 332: 292-7. 211. Feagan BG, Fedorak RN, Irvine EJ et al. A comparison of methotrexate with placebo for the maintenance of remission in Crohn's disease. North American Crohn's Study Group Investigators. N Engl J Med 2000; 342: 1627-32. 212. Oren R, Moshkowitz M, Odes S et al. Methotrexate in chronic active Crohn's disease: a double-blind, randomized, Israeli multicenter trial. Am J Gastroenterol 1997; 92: 2203-9. 213. Arora S, Katkov W, Cooley J et al. Methotrexate in Crohn's disease: results of a randomized, double-blind, placebocontroUed trial. Hepatogastroenterology 1999; 46: 1724-9. 214. Oren R, Arber N, Odes S et al. Methotrexate in chronic active ulcerative colitis: a double-blind, randomized, Israeli multicenter trial. Gastroenterology 1996; 110: 1416-21. 215. Gupta S, Keshavarzian A, Hodgson HJ. Cyclosporin in ulcerative colitis [Letter]. Lancet 1984; 2: 1277-8.

621 216. Lichtiger S, Present DH. Preliminary report: cyclosporin in treatment of severe active ulcerative colitis. Lancet 1990; 336: 16-19. 217. Cohen RD, Stein R, Hanauer SB. Intravenous cyclosporin in ulcerative colitis: a five-year experience. Am J Gastroenterol 1999; 94:1587-92. 218. Hermida-Rodriguez C, Cantero PJ, Garcia-Valriberas R, Pajares Garcia JM, Mate-Jimenez J. High-dose intravenous cyclosporine in steroid refractory attacks of inflammatory bowel disease. Hepatogastroenterology 1999; 46: 2265-8. 219. Hyde GM, Thillainayagam AV, Jewell DP. Intravenous cyclosporin as rescue therapy in severe ulcerative colitis: time for a reappraisal? Eur J Gastroenterol Hepatol 1998; 10:411-13. 220. Santos J, Baudet S, Casellas F, Guarner L, Vilaseca J, Malagelada JR. Efficacy of intravenous cyclosporine for steroid refractory attacks of ulcerative colitis. J Clin Gastroenterol 1995; 20: 285-9. 221. Sandborn WJ. A critical review of cyclosporine therapy in inflammatory bowel disease. Inflam Bowel Dis 1995; 1: 4 8 63. 222. Lichtiger S, Present DH, Kornbluth A et al. Cyclosporine in severe ulcerative colitis refractory to steroid therapy. N Engl J Med 1994; 330: 1841-5. 223. Kornbluth A, Lichtiger S, Present D, Hanauer S. Long-term results of oral cyclosporin in patients with severe ulcerative coUtis: a double-blind, randomized, multi-center trial. Gastroenterology 1994; 106: A714. 224. Stack WA, Long RG, Hawkey CJ. Short- and long-term outcome of patients treated with cyclosporin for severe acute ulcerative colitis. Aliment Pharmacol Ther 1998; 12: 973-8. 225. Ramakrishna J, Langhans N, Calenda K, Grand RJ, Verhave M. Combined use of cyclosporine and azathioprine or 6-mercaptopurine in pediatric inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1996; 22: 296-302. 226. Fernandez-Banares F, Bertran X, Esteve-Comas M et al. Azathioprine is useful in maintaining long-term remission induced by intravenous cyclosporine in steroid-refractory severe ulcerative colitis. Am J Gastroenterol 1996; 91: 24989. 227. Svanoni F, Bonassi U, Bagnolo F, Caporuscio S. Effectiveness of cyclosporine A (CsA) in the treatment of active refractory ulcerative colitis (UC). Gastroenterology 1998; 114: A1096. 228. D'Haens G, Lemmens L, Hiele M et al. Intravenous cyclosporine (CyA) monotherapy versus intravenous methylprednisolone (MP) in severe ulcerative colitis: a randomized, double blind controlled trial. Gastroenterology 1998; 114: A963. 229. Sandborn WJ, Tremaine WJ, Schroeder KW, Steiner BL, Batts KP, Lawson GM. Cyclosporine enemas for treatmentresistant, mildly to moderately active, left-sided ulcerative colitis. Am J Gastroenterol 1993; 88: 640-5. 230. Ranzi T, Campanini MC, Velio P, Quarto dP, Bianchi P. Treatment of chronic proctosigmoiditis with cyclosporin enemas [Letter; comment]. Lancet 1989; 2: 97. 231. Brynskov J, Freund L, Thomsen OO, Andersen CB, Rasmussen SN, Binder V. Treatment of refractory ulcerative colitis with cyclosporin enemas [Letter; see comments]. Lancet 1989; 1: 721-2. 232. Winter TA, Dalton HR, Merrett MN, Campbell A, Jewell DP. Cyclosporin A retention enemas in refractory distal ulcerative colitis and 'pouchitis'. Scand J Gastroenterol 1993; 28: 701-4. 233. Sandborn WJ, Tremaine WJ, Schroeder KW et al. A placebo-controlled trial of cyclosporine enemas for mildly to moderately active left-sided ulcerative colitis. Gastroenterology 1994; 106: 1429-35.

628 234. Stein R, Cohen R, Hanauer S. Complications during cyclosporine therapy for inflammatory bowel disease. Gastroenterology 1997; 112, A1096 235. Sternthal, M, George J, Kornbluth A, Present D. Toxicity associated with the use of cyclosporin in patients with inflammatory bowel disease (IBD). Gastroenterology 1996; 110, A1019. 236. Scott AM, Myers GA, Harms BA. Pneumocystis carinii pneumonia postrestorative proctocolectomy for ulcerative colitis: a role for perioperative prophylaxis in the cyclosporine era? Report of a case and review of the literature. Dis Colon Rectum 1997; 40: 973-6. 237. Quan VA, Saunders BP, Hicks BH, Sladen GE. Cyclosporin treatment for ulcerative colitis complicated by fatal Pneumocystis carinii pneumonia. Br Med J 1997; 314: 363-4. 238. Flcshner PR, Michelassi F, Rubin M, Hanauer SB, Plevy SE, Targan SR. Morbidity of subtotal colectomy in patients with severe ulcerative colitis unresponsive to cyclosporin. Dis Colon Rectum 1995; 38: 1241 5. 239. Mahadaven U, Kornbluth AA, Goldstein E, Gcrogc J, Lichtiger S, Present DH. Is cyclosporine (CS) induced nephrotoxicity permanent or progressive in patients with inflammatory bowel disease (IBD)? Gastroenterology 1997; 112: A1030. 240. Fculren G, Mihatsch MJ. Risk factors for cyclosporincinduced nephropathy in patients with autoimmune diseases. International Kidney Biopsy Registry of Cyclosporine in Autoimmune Diseases. N Engl J Med 1992; 326: 1654 60. 241. Bousvaros A, Wang A, Leichtncr AM. Tacrolimus (FK-506) treatment of fulminant colitis in a child. J Pedialr Gastroenterol Nulr 1996; 23: 329 33. 242. Bousvaros A, Kirschner B, Wcrlin J ct al. Oral tacrolimus treatment of severe colitis in children. Gastroenterology 1997; 112: A941. 243. Fellcrmann K, Ludwig D, Stahl M, David-Walek T, Stange EF. Steroid-unresponsive acute attacks of inflammatory bowel disease: immunomodulation by tacrolimus (FK506). Am J Gastroenterol 1998; 93: 1860 6. 244. Neurath MF, Wanitschke R, Peters M, Krummcnaucr F, Meyer zum Buschenfelde KH, Schlaak JF. Randomised trial of mycophenolatc mofetil versus azathioprine for treatment of chronic active Crohn's disease. Gut 1999; 44: 625 8. 245. Orth T, Peters M, Schlaak JF et al. Mycophenolatc mofetil versus azathioprine in patients with chronic active ulcerative colitis: a 12-month pilot study. Am J Gastroenterol 2000; 95: 1201-7. 246. Fellermann K, Steffen M, Stein J et al. Mycophenolatc mofetil: lack of efficacy in chronic active inflammatory bowel disease. Aliment Pharmacol Ther 2000; 14: 171-6. 247. Gandolfo J, Farthing M, Powers G et al. 4-Aminosalicylic acid retention enemas in treatment of distal colitis. Dig Dis Sci 1987; 32: 7 0 0 ^ . 248. Ginsberg AL, Beck LS, Mcintosh TM, Nochomovitz LE. Treatment of left-sided ulcerative colitis with 4-aminosalicylic acid enemas. A double-blind, placebo-controlled trial. Ann Intern Med 1988; 108: 195-9. 249. Ginsberg AL, Davis ND, Nochomovitz LE. Placebo-controlled trial of ulcerative colitis with oral 4-aminosalicylic acid. Gastroenterology 1992; 102: 448-52. 250. Marteau P, Halphen M. [Comparative randomized open study of the efficacy and tolerance of enemas with 2 gr of 4amino-salicylic acid (4-ASA) and 1 gr of 5-amino- salicylic acid (5-ASA) in distal forms of hemorrhagic rectocolitis]. Gastroenterol Clin Biol 1995; 19: 31-5. 251. Selby WS, Bennett MK, Jewell DP. Topical treatment of distal ulcerative colitis with 4-amino-salicyHc acid enemas. Digestion 1984;29:231-4.

Medical management of ulcerative colitis 252. Beeken W, Howard D, Bigelow J et al. Controlled trial of 4ASA in ulcerative colitis. Dig Dis Sci 1997; 42: 354-8. 253. Sandborn WJ. Nicotine therapy for ulcerative colitis: a review of rationale, mechanisms, pharmacology, and clinical results. Am J Gastroenterol 1999; 94: 1161-71. 254. Pullan RD, Rhodes J, Ganesh S et al. Transdermal nicotine for active ulcerative colitis. N Engl J Med 1994; 330: 81115. 255. Sandborn WJ, Tremaine WJ, Offbrd KP et al. Transdermal nicotine for mildly to moderately active ulcerative colitis. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1997; 126:364-71. 256. Thomas GA, Rhodes J, Ragunath K et al. Transdermal nicotine compared with oral prednisolone therapy for active ulcerative colitis. Eur J Gastroenterol Hepatol 1996; 8: 769 76. 257. Thomas GA, Rhodes J, Mani V et al. Transdermal nicotine as maintenance therapy for ulcerative colitis. N Engl J Med 1995:332:988 92. 258. Green JT, Thomas GA, Rhodes J et al. Nicotine enemas for active ulcerative colitis a pilot study. Aliment Pharmacol Ther 1997; 11:859 63. 259. Sandborn WJ, Tremaine WJ, Leighton J A et al. Nicotine tartrate liquid enemas for mildly to moderately active leftsided ulcerative colitis unresponsive to first-line therapy: a pilot study. Aliment Pharmacol Ther 1997; 11: 663 71. 260. Korzcnik JR. IBD: a vascular disorder? The case for heparin therapy. Inflam Bowel Dis 1997; 3: 87 94. 261. Gafl'ney PR, O'Lcary JJ, Doyle CT et al. Response to heparin in patients with ulcerative colitis [Letter]. Lancet 1991; 337: 238 9. 262. Gafi'ney PR, Doyle CT, Gaff'ney A, Hogan J, Hayes DP, Annis P. Paradoxical response to heparin in 10 patients with ulcerative colitis. Am J Gastroenterol 1995; 90: 220 3. 263. Folwaczny C, Fricke H, Endres S, Hartmann G, Jochum M, Loeschke K. Anti-inflammatory properties of unfractioned heparin in patients with highly active ulcerative colitis: a pilot study [Letter]. Am J Gastroenterol 1997; 92: 911 12. 264. Evans RC, Wong VS, Morris Al, Rhodes JM. Treatment of corticosteroid-resistant ulcerative colitis with heparin a report of 16 cases. Aliment Pharmacol Ther 1997; 11: 1037 40. 265. Korzenik JR. Heparin: an emerging counterintuitive therapy for inflammatory bowel disease. Curr Treat Options Gastroenterol 2000; 3: 95 8. 266. Targan SR, Hanauer SB, Van Deventer SJ et al. A shortterm study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn's disease. Crohn's Disease cA2 Study Group. N Engl J Med 1997; 337: 1029-35. 267. Present DH, Rutgeerts P, Targan S et al. Infliximab for the treatment of fistulas in patients with Crohn's disease. N Engl J Med 1999; 340: 1398^05. 268. Rutgeerts P, D'Haens G, Targan S et al. Efficacy and safety of retreatment with anti-tumor necrosis factor antibody (infliximab) to maintain remission in Crohn's disease. Gastroenterology 1999; 117: 761-9. 269. Stack WA, Mann SD, Roy A J et al. Randomised controlled trial of CDP571 antibody to tumour necrosis factor-alpha in Crohn's disease. Lancet 1997; 349: 521-4. 270. Evans RC, Clarke L, Heath P, Stephens S, Morris Al, Rhodes JM. Treatment of ulcerative colitis with an engineered human anti-TNFalpha antibody CDP571. Aliment Pharmacol Ther 1997; 11: 1031-5. 271. Sands BE, Podolsky DK, Tremaine WJ et al. Chimeric monoclonal anti-tumor necrosis factor antibody (cA2) in the treatment of severe, steroid-refractory ulcerative colitis (UC). Gastroenterology 1996; 110: A1008. 272. Roediger WE. The colonic epithelium in ulcerative colitis: an energy-deficiency disease? Lancet 1980; 2: 712-15.

629

William J. Sandborn 273. Scheppach W, Sommer H, Kirchner T et al. Effect of butyrate enemas on the colonic mucosa in distal ulcerative colitis. Gastroenterology 1992; 103: 51-6. 274. Senagore AJ, MacKeigan JM, Scheider M, Ebrom JS. Short-chain fatty acid enemas: a cost-effective alternative in the treatment of nonspecific proctosigmoiditis. Dis Colon Rectum 1992; 35: 923-7. 275. Steinhart AH, Hiruki T, Brzezinski A, Baker JP. Treatment of left-sided ulcerative colitis with butyrate enemas: a controlled trial. Aliment Pharmacol Ther 1996; 10: 729-36. 276. Scheppach W. Treatment of distal ulcerative colitis with short-chain fatty acid enemas. A placebo-controlled trial. German-Austrian SCFA Study Group. Dig Dis Sci 1996; 41:2254-9. 277. Breuer RI, Soergel KH, Lashner BA et al. Short chain fatty acid rectal irrigation for left-sided ulcerative colitis: a randomised, placebo controlled trial. Gut 1997; 40: 485-91. 278. Vernia P, Marcheggiano A, Caprilli R et al. Short-chain fatty acid topical treatment in distal ulcerative colitis. Aliment Pharmacol Ther 1995; 9: 309-13. 279. Asian A, Triadafilopoulos G. Fish oil fatty acid supplementation in active ulcerative colitis: a double-bhnd, placebocontrolled, crossover study. Am J Gastroenterol 1992; 87: 432-7. 280. Stenson WF, Cort D, Rodgers J et al. Dietary supplementation with fish oil in ulcerative colitis. Ann Intern Med 1992; 116:609-14. 281. Hawthorne AB, Daneshmend TK, Hawkey CJ et al. Treatment of ulcerative colitis with fish oil supplementation: a prospective 12 month randomised controlled trial. Gut 1992; 33: 922-8. 282. Loeschke K, Ueberschaer B, Pietsch A et al. n-3 fatty acids only delay early relapse of ulcerative colitis in remission. Dig Dis Sci 1996; 41: 2087-94. 283. Burke DA, Axon AT, Clayden SA, Dixon MF, Johnston D, Lacey RW. The efficacy of tobramycin in the treatment of ulcerative colitis. Aliment Pharmacol Ther 1990; 4: 123-9. 284. Lobo AJ, Burke DA, Sobala GM, Axon AT. Oral tobramycin in ulcerative colitis: effect on maintenance of remission. AHment Pharmacol Ther 1993; 7: 155-8. 285. Mantzaris GJ, Archavlis E, Christoforidis P et al. A prospective randomized controlled trial of oral ciprofloxacin in acute ulcerative colitis. Am J Gastroenterol 1997; 92: 454-6. 286. Turunen UM, Farkkila MA, Hakala K et al. Long-term treatment of ulcerative colitis with ciprofloxacin: a prospective, double-blind, placebo-controlled study. Gastroenterology 1998; 115: 1072-8. 287. Dickinson RJ, O'Connor HJ, Pinder I, Hamilton I, Johnston D, Axon AT. Double blind controlled trial of oral vancomycin as adjunctive treatment in acute exacerbations of idiopathic colitis. Gut 1985; 26: 1380-4. 288. Chapman RW, Selby WS, Jewell DR Controlled trial of intravenous metronidazole as an adjunct to corticosteroids in severe ulcerative colitis. Gut 1986; 27: 1210-12. 289. Mantzaris GJ, Hatzis A, Kontogiannis P, Triadaphyllou G. Intravenous tobramycin and metronidazole as an adjunct to corticosteroids in acute, severe ulcerative colitis. Am J Gastroenterol 1994; 89: 43-6. 290. Laursen LS, Naesdal J, Bukhave K, Lauritsen K, RaskMadsen J. Selective 5-lipoxygenase inhibition in ulcerative colitis. Lancet 1990; 335: 683-5. 291. Hillingso J, Kjeldsen J, Laursen LS et al. Blockade of leukotriene production by a single oral dose of MK-0591 in active ulcerative coHtis. Clin Pharmacol Ther 1995; 57: 33541. 292. Roberts WG, Simon TJ, Berlin RG et al. Leukotrienes in ulcerative colitis: results of a multicenter trial of a leuko-

293.

294.

295.

296.

297.

298. 299.

300. 301.

302.

303. 304. 305.

306.

307. 308.

triene biosynthesis inhibitor, MK-591. Gastroenterology 1997;112:725-32. Laursen LS, Lauritsen K, Bukhave K et al. Selective 5lipoxygenase inhibition by zileuton in the treatment of relapsing ulcerative colitis. A randomised double blind palcebo controlled multicenter trial. Eur J Gastroenterol Hepatol 1994; 6: 209-15. Peppercorn M, Das K, Elson C et al. Zileuton a 5lipoxygenase inhibitor in the treatment of active ulcerative colitis - a double blind, placebo-controlled trial. Gastroenterology 1994; 106: A751. Hawkey CJ, Dube LM, Rountree LV, Linnen PJ, Lancaster JF. A trial of zileuton versus mesalazine or placebo in the maintenance of remission of ulcerative colitis. European Zileuton Study Group for Ulcerative Colitis, Gastroenterology 1997; 112:718-24. Stack WA, Jenkins D, Vivet P, Hawkey CJ. Lack of effectiveness of the platelet-activating factor antagonist SR27417A in patients with active ulcerative colitis: a randomized controlled trial. Platelet Activating Factor Antagonist Study Group in Ulcerative Colitis. Gastroenterology 1998; 115: 1340-5. Smithson JE, Grool TA, van Deventer SJH, Jewell DP. Controlled clinical trial of lexipafant in addition to corticiosteroids in severe ulcerative colitis. Gastroenterology 1997; 112: A1094. Ryder SD, Walker RJ, Jones H, Rhodes JM. Rectal bismuth subsalicylate as therapy for ulcerative colitis. Aliment Pharmacol Ther 1990; 4: 333-8. Srivastava ED, Swift GL, Wilkinson S, Williams GT, Evans BK, Rhodes J. Tripotassium dicitrato bismuthate enemas in the treatment of ulcerative proctitis. Aliment Pharmacol Ther 1990;4:577-81. Sumer N, Palabiyikoglu M. Induction of remission by interferon-alpha in patients with chronic active ulcerative cohtis. Eur J Gastroenterol Hepatol 1995; 7: 597-602. Van Deventer SJ, Elson CO, Fedorak RN. Multiple doses of intravenous interleukin 10 in steroid-refractory Crohn's disease. Crohn's Disease Study Group. Gastroenterology 1997; 113:383-9. Schreiber S, Fedorak RN, Wild G et al. Safery and tolerence of rHuIL-10 treatment in patients with mild/moderate active ulcerative colitis. Gastroenterology 1998; 114: A1080-1. Bjorck S, Dahlstrom A, Ahlman H. Topical treatment of ulcerative proctitis with lidocaine. Scand J Gastroenterol 1989; 24: 1061-72. Bjorck S, Dahlstrom A, Johansson L, Ahlman H. Treatment of the mucosa with local anaesthetics in ulcerative colitis. Agents Actions 1992; (Spec No.): C60-72. Arlander E, Ost A, Stahlberg D, Lofberg R. Ropivacaine gel in active distal ulcerative colitis and proctitis - a pharmacokinetic and exploratory clinical study. Aliment Pharmacol Ther 1996; 10: 73-81. Frieri G, Pimpo MT, Palumbo GC et al. Rectal and colonic mesalazine concentration in ulcerative colitis: oral vs. oral plus topical treatment. Aliment Pharmacol Ther 1999; 13: 1413-17. Mclntyre PB, Powell-Tuck J, Wood SR etal. Controlled trial of bowel rest in the treatment of severe acute colitis. Gut 1986; 27: 481-5, Dickinson RJ, Ashton MG, Axon AT, Smith RC, Yeung CK, Hill GL. Controlled trial of intravenous hyperalimentation and total bowel rest as an adjunct to the routine therapy of acute colitis. Gastroenterology 1980; 79: 1199204.

31 Surgical management of ulcerative colitis IAN LINDSEYAND NEIL J. MCC MORTENSEN

Introduction The past 25 years have seen great changes in the management of ulcerative colitis (UC) with the introduction of a new procedure, ileal pouch-anal anastomosis (IPAA), and a gradual development in terms of refinement of its technique. These include older refinements such as methods of anastomosis, pouch design, and the plane of rectal dissection, as well as more recent ones, including laparoscopicassisted IPAA, omission or otherwise of a loop ileostomy and IPAA for selected patients with Crohn's disease (CD). This chapter reviews the history of surgery for UC, and examines indications and goals of surgery, the various surgical options, and focuses detailed attention on aspects of IPAA, now considered the gold standard for surgery.

History and evolution of surgery for ulcerative colitis In the early 1900s major surgery was hazardous, and severe UC was treated by irrigating the colon with fluids introduced via appendicostomy [1], then subsequently by cecostomy [2]. As abdominal surgery became safer, more invasive excisional procedures evolved to allow surgeons to gain better long-term control of the disease in severe cases. In 1944 Strauss and Strauss established proctocolectomy and endileostomy, although there was major morbidity related to the ileostomy [3]. These difficulties were largely alleviated by the development of the everted ileostomy in 1952 by Brooke [4]. The field of stomal therapy was pioneered in the late 1950s at the Cleveland Chnic, leading to progress in the development of better stomal appliances and improved patient education [5]. However, the ileostomy remained 'incontinent', and this led to the development by Kock in the late 1960s of the 'continent' ileostomy with an intra-

abdominal reservoir [6]. Although addition of a nipple valve further improved its continence [7], the high complication and reoperation rate and the emergence of IPAA led to its decline except in a few dedicated units. Ileorectal anastomosis (IRA) was described by Devine [8] and championed by Aylett [9] in the UK in the 1960s as a means of avoiding an ileostomy. However, a high incidence of inflammation in the rectal stump, with need for subsequent rectal excision, added to the development of dysplasia and cancer in the rectum, has largely led to its abandonment in UC [10]. IPAA has naturally evolved as a procedure to eliminate mucosal disease in UC while maintaining intestinal continuity and preserving anal continence [11]. It was born of a combination of two major surgical concepts, each of which had a long history of development. Ileoanal anastomosis was attempted as early as 1900 but functional outcome was very poor [12]. Nissen [13] revisited the procedure in 1933, as did Ravitch and Sabiston [14] in 1947 with similar results. Soave [15] in 1964 popularized colonic pullthrough after distal rectal mucosectomy for Hirschdprung's disease in children. In 1973 Safaie-Shirafi and Soper [16] performed distal rectal mucosal stripping with straight ileoanal anastomosis for familial adenomatous polyposis, and in 1977 Martin et al. [17] performed similar procedures in patients with UC. However, straight ileoanal anastomosis was plagued by high stool frequency, urgency and incontinence. The concept of improving the stool frequency of an ileoanal anastomosis by construction of an ileal reservoir originated well prior to the latest of these developments. Valiente and Bacon (1955) [18] described a triple-limbed ileal pouch in dogs combined with an ileoanal anastomosis. Karlan etal [19] (1959) constructed a double-barrel pouch in dogs with good eff'ect. Parks and Nicholls (1978) [11] performed the first ileal reservoir in combination

Stephan R. Targan, Fergus Shanahan and Lor en C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 631-641. © 2003 Kluwer Academic Publishers. Printed in Great Britain

632 with ileoanal anastomosis in humans with acceptable results. Various technical modifications of this procedure have occurred over the past 20 years but essentially the procedure remains the same.

Indications for surgery in ulcerative colitis About 20-25% of patients with UC require surgery, usually in the first 10 years of the disease [20]. It is estimated that about 10% of patients submit to surgery within the first year of diagnosis, and about 2% per year do so thereafter [21]. Patients requiring surgery clearly have disease activity towards the severe end of the spectrum. In the past, surgery was offered only as a last resort when other options were exhausted and patients were malnourished, debilitated and often septic. As a consequence, results of surgery were often poor, reinforcing the gastroenterologists' notion of trying to avoid surgery at all costs. Nowadays, however, there is closer collaboration between surgeons and gastroenterologists, surgery is offered earlier in the course of the disease and surgical results in expert hands are generally very good. Indications for surgery in UC are considered absolute or relative, both in the emergency and elective setting (Table 1). Generally patients fall into one of three groups: acute severe attack, failed medical therapy or development of dysplasia or carcinoma. Most patients with UC these days undergo elective surgery, and we see fewer emergency presentations than in the past as a result of better medical management. The risk of requirement for surgery correlates well with the disease extent. With pancolitis, about 35-40% of patients will require surgery; left-sided colitis 5% and for distal colitis, this falls to about 2% [22]. When presenting acutely with severe disease or with failed medical therapy, the patient is not in ideal condition for major reconstructive surgery, and surgery is usually staged. However, the development of dysplasia or carcinoma presents the surgeon with an opportunity to perform cold elective surgery in good c o n d i t i o n s , possibly with single-stage reconstruction.

Surgical management of ulcerative colitis Table 1. Indications for surgery in ulcerative colitis Absolute (development of complications) Free perforation Hemorrhage Severe acute attack ('toxic colitis') unresponsive to medical therapy over 2 - 5 days ± toxic megacolon Stricture Cancer or severe dysplasia Relative (effect of disease or medical treatment on patient) Failed medical therapy Inability to wean steroids or immunosuppressives Side-effects of medications Chronic intractable symptoms Social factors: impact on life, work, geographic considerations

Absolute indications Fortunately free perforation is uncommon, occurring in about 3% of cases, since mortality is as high as 30%. The risk of perforation is related to the severity of the acute attack and the extent of the inflammatory process in the colon. As such it is almost always seen in the setting of a severe acute attack, particularly in pancolitis, where the risk of perforation is higher still. Perforation is usually seen in association with toxic megacolon, although perforation can occur without megacolon. Toxic colitis and megacolon should both be seen as part of the spectrum of severe acute disease. The severe acute attack is initially managed medically and in 50% of cases [23] (up to 80% with the use of cyclosporin) [24], settles accordingly, although the chances of these patients retaining their colon in the long term are less good [24]. In circumstances where the clinical state of the patient does not improve over the next 2-5 days, then surgery is required. These patients are managed by collaboration between gastroenterologist and colorectal surgeon. Hemorrhage severe enough to require emergency surgery is very rare. Usually massive bleeding occurs in the setting of a severe acute attack [25]. On most occasions bleeding stops and is managed by volume replacement and transfusion, and only occasionally is surgery required. Most surgeons would perform subtotal colectomy for bleeding, leaving the rectum in place, either as a mucous fistula or buried mucous fistula. This avoids the risks of difficult emergency pelvic surgery and keeps future reconstructive options open. In a small proportion of cases persis-

Ian Lindsey and Neil JMcC Mortensen tent bleeding from the retained rectum necessitates a low Hartmann's resection of the rectum. Although colonic strictures may be benign in longstanding UC, the risk of malignancy is high, unlike in CD in which colonic stricturing is common due to the full thickness disease. In one series of 70 strictures, 24% were malignant [26]. It may be difficult to distinguish the benign from the malignant stricture radiologically and endoscopically; therefore we regard stricture as an absolute indication for surgery. Development of a carcinoma or high-grade dysplasia is a risk in patients with long-standing UC, particularly those with extensive disease and those with early-age onset of disease. The risk is minimal in the first 8-10 years (2-3%) then increases cumulatively at a rate of about 1-2% per year, so that by 20 years the risk is about 10-20%o [27]. A colectomy has been recommended in the presence of high-grade dysplasia, dysplasia associated with a mass or persistent low-grade dysplasia on multiple examinations [28].

Relative indications These days failed medical therapy constitutes the largest group of patients requiring surgery. The patient may not respond to, or be able to be weaned off, steroids, or steroid-sparing drugs. Long-term side-effects of these potent drugs, such as osteoporosis or growth retardation in children, may be a cause for concern for both patient and physician. In about 5-10% of patients the disease pattern is one of chronic refractory symptoms r a t h e r t h a n a relapsing/remitting course, and this may eventuate in surgery. The particular social, economic or geographic circumstances of the patient may lead to a preferential choice of surgery. When surgical indications are relative, decisionmaking is not clear-cut. In general the philosophy of the gastroenterologist, the wishes and circumstances of the patient as well as their symptoms and sideeffects, will blend to produce a particular decision at a particular time in the course of the patient's illness. One cannot be dogmatic about this decision-making process; often several relative indications are involved, and patient wishes influence this process to a major degree.

633

Surgical goals and options: emergency setting Subtotal colectomy, ileostomy and buried mucous fistula For many reasons this operation is ideally suited to the emergency setting [29]. It is a usually straightforward abdominal operation that eliminates the majority of the disease and results in a marked and prompt improvement in the patient's condition. It can be performed with relatively little morbidity. An intra-abdominal anastomosis with its ensuing risks is avoided. Rectal dissection, with its risks of rectal perforation, pelvic bleeding and sepsis, bladder and sexual nerve damage, is left for the elective setting [30]. Most importantly future options, including reconstructive surgery, are left open to be considered at a later date, when the patient is off* medication and in better nutritional condition. Staged reconstruction reduces the chances of failing to distinguish between UC and CD [31], and the patient gains valuable experience in managing an ileostomy. The disadvantages of this approach include the generation of adhesions, which require lysis at subsequent IPAA, and the possibility of persistent symptomatic disease in the retained rectal stump, but these are easily outweighed by the advantages. The operation is occasionally difficult if there is severe extensive disease with or without toxic megacolon, when the bowel can be very friable, there may be perforations sealed by omentum and adjacent parietal peritoneum, and the risk of fecal spillage is high.

Technical points The appropriate preoperative preparation includes adequate resuscitation, avoidance of mechanical bowel preparation, siting of a stoma, and full counseling regarding surgery. A midline incision avoids compromising stoma sites. Care is taken during colonic mobilization in case of friability or sealed perforations, and quarantining of the abdomen with packs, particularly at the splenic ffexure, a typical site of surgical perforation, will reduce the risk of gross fecal contamination. Omentum or parietal peritoneum plastered to the colon should be resected 'en-bloc'. Several technical maneuvers facilitate future pouch reconstruction. The anastomosis between superior mesenteric and ileocolic arteries is preserved, by dividing the mesocolon close to the right

634 colon. No terminal ileum is resected in division of the ileum and colon and construction of the end ileostomy. The inferior mesenteric artery is preserved and individual sigmoid vessels are taken. The distal sigmoid is divided to allow easy reach to the anterior abdominal wall, the staple line is inverted with sutures, and the end of the stump brought out just into the subcutaneous tissue extra-fascially in the lowest part of the abdominal wound, securing the serosa to the fascia. A blow-out of the stump, a risk especially if the rectum is diseased, results in wound infection rather than pelvic sepsis. If the stump is badly diseased and will not hold sutures it can be brought out 5-7 cm beyond the abdominal wall and covered with moist gauze, and at 7-10 days amputated flush to mature a mucous fistula, with sufficient adherence to the abdominal wall. No breach is made in the pelvic planes, allowing dissection in virgin territory at the second stage.

Diverting loop ileostomy and 'blowhole' colostomy TurnbuU et ai [32] described this operation in the setting of multiple sealed perforations and distended, friable bowel, the setting of severe acute colitis, on the basis that colectomy was hazardous because of the risk of gross fecal spillage. These days, better medical management and earlier surgical referral have made this option almost redundant, but on rare occasions it may be useful, and is included for historical reasons. Technical points The lie of the transverse colon is gaged with supine abdominal X-ray and a radiopaque marker at the umbilicus, the abdomen is entered via a small lower midline incision and a loop ileostomy is fashioned. A new vertical left paramedian incision is made over the dilated transverse colon and, before the colon is opened, the omentum and seromuscular layer of the colon is sutured carefully to the parietal peritoneum. The bowel is opened and a skin-level colostomy is matured. The colon is removed electively 6 months later.

Other emergency procedures Emergency proctocolectomy and end-ileostomy This was once advocated as the operation of choice in the setting of acute colitis [33] but subtotal

Surgical management of ulcerative colitis colectomy is now preferred. Although controlled data are unlikely to become available, the risks of bleeding, damage to autonomic nerves and pelvic sepsis are likely to be higher than elective rectal dissection [30], and the option of future pouch surgery is eliminated. It may occasionally be required for massive bleeding or for limited severe acute proctosigmoiditis, where subtotal colectomy alone is unlikely to deal with the acute surgical problem. Emergency colectomy and ileorectal anastomosis (IRA) This option is rarely advised. Emergency IPAA and loop ileostomy Since this is major challenging reconstructive surgery with significant risk of complications even in the elective setting, this operation should not be undertaken in the acute setting.

Surgical goals and options: elective setting There are several surgical options for UC, and each has its particular advantages and disadvantages. Generally these relate to avoidance or otherwise of the following: a permanent ileostomy; a rectal excision with its risks of nerve damage and bleeding; surveillance of the rectal stump and risk of rectal cancer; anorectal function; a perineal dissection with its risks of non-healed perineal wound; staged multiple operations and their impact on work and life; complex surgery with its risks of complications.

Total abdominal colectomy and IRA IRA has gone out of fashion since the emergence of IPAA, having been popularized by Aylett in the 1960s and 1970s as the only surgical alternative to permanent ileostomy [9]. The main reason for declining interest in IRA is the cumulative requirement for rectal excision (primarily for poor function but also because of dysplasia and cancer in the rectum). This requirement ranges from 11% to 57% [34] (mean about 25%), and is proportional to length of followup. Poor function may result from persistent proctitis or rectal stricture, although the overall functional results are similar to IPAA, and most patients' function is slightly superior to those published for IPAA. The risk of cancer in the retained rectum is the

Ian Lindsey and Neil JMcC Mortensen greatest fear of the surgeon undertaking IRA, this risk increasing with time and estimated to be about 5% at 15-20 years [34]. Patients require dihgent follow-up for sigmoidoscopic surveillance of the rectum, yet may develop rectal cancer despite this. IRA does have its attractions, because it is simple, has low mortality and morbidity, both ileostomy and the complications of a rectal dissection are avoided, its functional results are similar to IPAA and the option of future restorative surgery is available [34]. If IRA is to be performed, several preconditions must be met. The rectum must be relatively spared and be distensible, the anal sphincters must be adequate to deal with liquid stool and IRA should not be contemplated when there is dysplasia or cancer in the colon or rectum. Perhaps IRA finds its niche in two circumstances assuming the above preconditions are met. First, in the adolescent [35] or young adult, who wishes to avoid a stoma or a pelvic dissection while growing up or having a family. Psychological as well as sexual function and fertility difficulties may be avoided, the procedure acting as a 'bridge' to later IPAA. Secondly, a diagnosis of indeterminate colitis may influence the surgeon to perform IRA on the basis that, if the patient has UC, IPAA is still feasible, yet if CD develops, IRA is an excellent restorative option if rectum is supple and (relatively) spared [36]. The advantages are avoidance of a stoma, biopsy of the rectum without confounding effects of diversion on rectal histology, and keeping future options open. The disadvantages are the risk of anastomotic leak, and lack of information on the natural history of the diverted rectal disease.

Subtotal colectomy and permanent endileostomy Most patients after urgent subtotal colectomy and ileostomy are keen to undergo IPAA in order to have their ileostomy closed. However, it should be recognized that, occasionally, the now-healthy asymptomatic patient does not wish to submit to further major surgery. Because of the long-term risks of cancer in the retained rectum [34] and difficulty in surveillance, the patient should be actively discouraged from this course of action. The notion of future proctectomy (with or without reconstruction) should be raised with the patient at the first outpatient review.

635

Proctocolectomy and permanent end-ileostomy (PC) Prior to the advent of IPAA this proctocolectomy and ileostomy (PC) was considered the gold standard in UC and its attractions are several. Unless the patient is very ill PC can be performed as a single operation. It eliminates all colonic and rectal mucosal disease and therefore UC symptoms and cancer risk. In the hands of a well-trained colorectal surgeon this is a relatively straightforward operation. For these reasons it is the alternative of choice to IPAA, and even in specialized centers a substantial proportion of patients will undergo this operation. In Oxford just under a quarter of our patients elect to undergo PC. The overriding disadvantage, of course, is the presence of a permanent ileostomy, which most patients, especially younger ones, want to avoid at all costs. However, a subset of patients will decline IPAA because of its complexities and risks, and this may relate to their age, social or work circumstances. In some patients IPAA is contraindicated because of age, poor sphincter function or advanced colon or rectal cancer. It is difficult to compare the benefits and drawbacks of PC and IPAA. The procedures are offered to different population groups with different expectations, which introduces selection bias. Additionally, the yardsticks by which we measure outcomes are different and thus difficult to directly compare, although some have tried to do so using quality-oflife instruments. Satisfaction with an ileostomy can be high, particularly for patients over 60. However, we also know that IPAA patients experience less difficulty performing daily activities, less restriction in sport and sexual activity than those with permanent ileostomy [37] and therefore overall may have a greater quality of hfe.

Technical points Colonic mobilization is conducted preserving the omentum. Low ligation of the inferior mesenteric vessels is performed. Rectal dissection is carried out carefully to avoid damaging the nerves governing sexual and bladder function. Either close rectal or mesorectal dissection may be employed, since close rectal dissection has not shown an advantage over mesorectal dissection in protecting the pelvic nerves. An intersphincteric perineal dissection is performed. Care is taken in fashioning the permanent endileostomy to minimize stoma-related complications.

636 The pre-marked site is used, and the stapled-off terminal ileum is brought out through the rectus abdominus muscle via a small fascial aperture. A good length of ileum is brought out to enable eversion with a decent 2-3 cm spout and the mesenteric edge is sutured to the adjacent anterior abdominal wall to prevent volvulus.

Continent ileostomy (Kocic pouch) Continent ileostomy was devised and developed by Nils Kock [6] in 1967 to improve the quality of life of ileostomates in whom, despite the great improvements achieved with everted ileostomy, physical and psychological difficulties were still common. Kock conceived the continent ileostomy as a reservoir fashioned from the terminal ileum allowing voluntary bowel control. An antiperistaltic pouch resulted in partial continence, but modification [7] with a 'nipple valve', a 10 cm segment of terminal ileum intussuscepted upon itself, resulted in total continence: the bowel contents remained in the pouch until self-catheterization was performed. However, the procedure did not gain wide acceptance, partly due to the emergence of I PA A in the late 1970s, but also because of the high complication rate, the greatest problems being nipple valve-related. Slippage was common (30%) despite modifications such as scarification, stapling and meshing, while fistularization, necrosis and prolapse were also encountered. Other non-nipple valve related problems include pouchitis and outflow obstruction. Quality-of-life studies have shown superior results with continent over conventional ileostomy, particularly in the area of sexual function and sporting activity [38] and satisfaction is high in those who retain well-functioning Kock pouches. In dedicated tertiary colorectal centers, the procedure may be reasonably considered in limited circumstances, such as when I PA A has failed, or is contraindicated because of age, poor sphincter function, as an alternative to conventional ileostomy.

IPAA Patient selection and counseling It must be remembered that, despite the attraction of IPAA, bowel function does not return to normal, and care must be taken to explain the alternatives and disadvantages to any potential candidate. There is a wide range of considerations to be taken into

Surgical management of ulcerative colitis account, including future employment, fertility, pregnancy and sexual function, age, compliance with future follow-up, in addition to purely technical considerations such as body build, co-morbiditiy, bowel and sphincter function. The patient should be well prepared in advance of surgery, having met and had discussion with a colorectal surgeon, a stomal therapist and a patient with an established pouch. Written information is very useful. Patients undergoing urgent surgery for acute severe colitis are rarely psychologically prepared for restorative surgery. Patients should be aware that IPAA is a complex operation, often conducted over three stages over 6-9 months, with a complication rate of 25-40%, a reoperation rate of 15-25%; a failure rate of 5-10% and variable pouch function. Anal sphincter function It is essential to establish that the sphincters are normal prior to surgery since, even with the bestfunctioning pouches, the stool is liquid or at best semisolid. A history of anal surgery, obstetric trauma or incontinence, as well as scars or weakness of clinical resting or squeeze pressures, should be noted. Anal ultrasound and anorectal physiology are indicated in these circumstances and some would go as far as performing these routinely preoperatively in case of postoperative incontinence disturbance for baseline or medicolegal reasons [39]. Abnormal sphincter function is unusual in the younger patient group, but the sphincters may deteriorate with age, beginning in the 50s, and is a particular problem in women aged over 60. Several studies have shown a decrease in resting and squeeze pressures with increasing age [40, 41], and continence disturbance after IPAA is known to be more common over 50 [42], though this may not necessarily translate to disturbance in quality of life. Extremes of age Patients as old as their early 80s and as young as 3 have undergone pouch surgery. In the past pouch surgery was offered with reluctance in some centers over the age of 55, particularly in women, but arbitrary age cut-offs are not recommended as long as elderly patients are counseled about these increased risks. Co-morbidity risks, as well as the implications of deteriorating sphincter function (above), need to be taken into account, but age/?^r se is not grounds for absolute contraindication.

Ian Lindsey and Neil JMcC Mortensen There are relatively few published pediatric studies on pouch surgery [43]. They indicate that pouch surgery is feasible with probably less pelvic sepsis, strictures and pouchitis than a comparable adult group. Ileostomy and pelvic dissection with the risks of psychological maladaptation, infertility and sexual dysfunction are major drawbacks, and colectomy with ileorectal anastomosis as a bridge to pouch surgery in those with relative rectal sparing is an option. Fertility, pregnancy and sexual function In females with or considering an IPAA, thought should be given to the risk of damage to the anal sphincters during vaginal delivery. In the short term, 7-35% [44, 45] of first vaginal deliveries are associated with an anal sphincter injury on anal ultrasound, and a rate of up to 13% [44] fecal incontinence, rising with second vaginal delivery. Persistent pudendal neuropathy occurs with prolonged second-stage or forceps delivery in a small group of patients [46]. Small series of vaginal delivery in pouch patients without significant disturbances of continence have been reported [47]; however, we believe strong consideration should be given to elective Cesarian section. There is also recent evidence that pouch surgery leads to reduced fertility and an increased rate of in-vitro fertilization [48] presumably due to the formation of pelvic adhesions. The incidence of male sexual dysfunction is said to be low (1-3%) [49]. There is recent evidence that when a sensitive, dedicated, patient-based sexual function questionnaire [50] is used, significant rates of partial impotence rates are picked up, which may be as high as 5-10% in the 20-50-year-old age group [51]. Consideration should be given to storage of a semen sample preoperatively. These issues should be covered during counseling of the patient of reproductive age prior to pouch surgery, and in female pouch patients prior to pregnancy. Crohn's disease and indeterminate colitis A diagnosis of CD is generally considered a contraindication to IPAA. The risk of recurrent CD in the pouch with loss of distensibihty and function, the risk of abscess formation and fistulization, and the risk of subsequent pouch excision and short bowel syndrome have been documented from experience with the both the Kock pouch [52] and IPAA [53-55]. A French group have recently suggested that IPAA may be justified in carefully selected patients with CD, those without anoperineal or small bowel dis-

637 ease, on the basis that the results of surgery do not seem to be any worse than those for other restorative operations in CD [56]. However, patients in this series may well have had indeterminate colitis [57], and it is generally agreed that the surgeon should ideally have a clear diagnosis of UC before undertaking IPAA. The preoperative diagnosis is not clear-cut in a small proportion of patients (2-8%), who are labeled as 'indeterminate colitis' [53]. Mostly these cases behave like UC and the risk of later CD is small (611%) [58]. Should IPAA be offered to these patients? Rates of operative morbidity and pouchitis are similar to IPAA in UC. Failure of the pouch (diversion or excision) is about twice as likely in one series, relating to the manifestation of CD in the anus or small bowel in a minority of these patients [58], but patients who retained their pouches had similar function to those in UC. One series suggested an increased risk of perineal complications in indeterminate colitis pouches [59], but this may reflect differences in defining indeterminate colitis histologically. What happens when IPAA is performed for indeterminate colitis, which later turns out to be CD? According to Cleveland Clinic data patients fall into two distinct groups [53]. Those who demonstrated distinct preoperative clinical features suspicious of CD did poorly, with poor function and a very high rate of pouch excision (up to 45%o); however, those who did not, managed to keep their pouches and had similar function to UC patients. In other words they behave much like patients with indeterminate colitis. Thus patients with indeterminate colitis may be offered IPAA fairly confidently on the basis that their outcome will be similar to UC patients. CD is still a possibility (6-11%) and they should be counseled accordingly about this small risk for increased sepsis, pouch failure, and pouch excision with permanent ileostomy. Even if the diagnosis is ultimately CD the outcome is still usually good if IPAA is diligently avoided in indeterminate colitis with any clinical suspicions of CD. Extraintestinal manifestations These are not a contraindication, but may not improve with surgery and seem to be a risk for pouchitis. Pyoderma gangrenosum, erythema nodosum, polyarteritis nodosum, arthritis and uveitis are often helped by surgery. However, the courses of primary sclerosing cholangitis (PSC) and ankylosing spondylitis are not influenced by proctocolectomy. In

638 PSC it is important to properly manage the UC because PSC patients are candidates for liver transplantation. Brooke ileostomy should be avoided in those not suitable for transplantation since 50% of these patients will bleed from peristomal varices [60]. In those candidates for transplantation, IPAA is controversial due to the high risk of chronic pouchitis despite immuno-suppressive therapy [61].

Technical issues and controversies Staging of surgery: one, two or three stages? IPAA is a complex reconstructive operation with a low pelvic anastomosis. The risk of leak is significant and pelvic sepsis is a major cause of poor function and pouch failure. Surgeons have been reluctant to perform this operation without the protection of a covering loop ileostomy to mitigate the consequences of pelvic sepsis. However, closure of ileostomy has its complications and increases overall hospital stay. Omission of ileostomy has been shown to be reasonable in selected low-risk cases when compared against use of a loop ileostomy in higherrisk patients [62]. Of high-risk factors considered by surgeons including high-dose steroids ( > 20 mg), hypoalbuminemia and anemia, only steroids have consistently shown to be a reliable factor. However, since it is not legitimate to compare high- and low-risk groups, because of the inherent biases, it remains open as to whether the low-risk patient is best served by protective loop ileostomy or not. A randomized trial after exclusion of these high-risk patients would answer this question, but not surprisingly the one small trial addressing this issue was inconclusive [63]. Single-stage surgery is seen most often in large North American tertiary colorectal units and has obvious advantages for the patient when it is safe to perform. The major drawback is the inability to clearly distinguish UC from CD, particularly when preoperative information is equivocal or when proctocolectomy is performed for severe disease, and there is a small risk of inadvertent pouch construction for CD [31]. Laparoscopic-assisted or open surgery for IPAA? The role of laparoscopic-assisted IPAA is being debated. There are doubts about the true advantages of minimal access in terms of speed of recovery. However, being a benign disease, minimal access surgery for inflammatory bowel disease (IBD) in principle has a lot to recommend it.

Surgical management of ulcerative colitis Laparoscopically assisted IPAA is certainly feasible, and some impressive results have been published [64]. Laparoscopic-assisted subtotal colectomy has been reported for acute severe colitis with results comparable to matched open controls [65]. The issues are those of surgical morbidity and mortality, potentially less surgical stress, with faster recovery and less immune suppression, as well as fewer adhesions and better cosmesis. The data comparing open to laparoscopically assisted IPAA are inadequate. Reported series in the literature are either retrospective reviews detailing general experience, morbidity and mortality results, as well as duration of ileus and length of stay, or nonrandomized comparisons with matched open controls. Results are similar to those reported in the open literature. Data from randomized controlled trials are awaited. It would be useful to know if immune suppression is less in minimal-access surgery. IBD patients are often immune-suppressed because of their underlying illness, poor nutrition and medical therapies. Difficulties exist in knowing which parameter best directly reflects the surgical stress response. Differences in serum interleukin-6 (IL-6) after open and laparoscopically assisted colonic surgery for IBD have consistently shown trends to increased IL-6 levels after open surgery but none with statistical significance [66-69]. This may be because of the heterogeneity of the data as there is a high variability of IL-6 response, particularly in open cases [68, 69]. Recently, granulocyte elastase (GE) has been studied as an alternative marker, in comparison to IL-6 and C-reactive protein (CRP), after open and laparoscopically assisted surgery for IBD [69]. It was found to be significantly elevated (p < 0.05) on days 1, 2 and 3 in open surgery compared to laparoscopically assisted patients. This marker shows promise, but clearly further research is required. Other studies have demonstrated a poor correlation between serum CRP and the extent of tissue damage [66-68]. Fewer adhesions appear to form after laparoscopic surgery when compared to open surgery. There is both anecdotal and scientific evidence to support this from observations at laparoscopy and from animal studies [70-75]. There are no data on the incidence of adhesions after laparoscopic surgery for IBD, but any reduction would clearly be an advantage, particularly during IPAA, when adhesions formed at subtotal colectomy and ileostomy can be problematic.

Ian Lindsey and Neil JMcC Mortensen Preservation of body image in this often-young group of patients is an advantage of the laparoscopic approach. A comparative non-randomized study of 22 patients undergoing laparoscopic or open resection for IBD showed that cosmetic score was significantly higher in the laparoscopic group (p < 0.01), and that body image correlated strongly with cosmesis and quality of life [76]. Close rectal or mesoreetal dissection? A particular concern to male patients contemplating IPAA is the risk of impotence secondary to pelvic nerve damage. If surgery is to be acceptable to patients then the risk of nerve damage and impotence must be kept to a minimum. Close rectal or perimuscular dissection is a surgical technique first described by Dowling and Lee at Oxford in 1972 [77]. By convention the mesoreetal plane is used for rectal cancer surgery, and by some surgeons for IBD. Close rectal dissection is conducted in the non-anatomical perimuscular plane in an effort to avoid damage to the pelvic autonomic nerves, which lie immediately outside the mesoreetal plane. It is, however, a more challenging dissection. According to recent data there appears to be no statistically significant difference in the rates of complete or partial impotence between those undergoing mesoreetal and close rectal dissection [51]. This is useful information since many surgeons operate in the mesoreetal plane out of convenience and familiarity, and it is important for such surgeons to be clear that they are not disadvantaging their patients by using this plane. Mucosectomy and handsewn anastomosis or doublestapling? The intense debate that this issue generated during the 1980s has largely subsided. The debate was centered on the question of whether it is safe to leave a 'columnar cuff' of mucosa in the upper half of the anal canal, as is done after double-stapling, or whether complete extirpation of all columnar epithelium down to the dentate line is necessary by mucosectomy. On one side of the argument the 'musosectomists' argue that, to eliminate or reduce the risk of columnar cuff dysplasia and cancer in the longer term, as well as that of recurrent colitis in the cuff ('cuffitis'), mucosectomy is necessary [78]. Double-stapling leaves potential disease-bearing columnar mucosa in the anal canal and sometimes inadvertently leaves a short rectal stump.

639 Antagonists to this view, however, favor the technical ease and speed of double-stapling with the elimination of the perineal phase of surgery. They cite the superior nocturnal continence subsequent to avoidance of the manual disturbance of the anal sphincters [79] and preservation of the important anal transition zone with its sensory discriminatory function [80]. They also point to the hitherto-unrecognized variabihty in transitional zone anatomy [81] as well as data showing that mucosectomy is often incomplete and can leave residual islands of columnar mucosa behind [82]. A paucity of good randomized data has hindered clarification of this issue [83]. However, it is fair to say that, after 15 years of follow-up of the doublestapled operation, the incidence of columnar cuff dysplasia and cancer is small [84], and most of the few cases of cancer [85] have been in patients undergoing mucosectomy. We routinely practice doublestapling but recognize the technical requirement to be able to undertake mucosectomy in certain circumstances, as well as the need to closely follow up double-stapled patients in the longer term. Pouch design: J, S or W configuration? There are several options in pouch design for IPAA, including the J, S and W pouches, and in essence these are double, treble and quadruple 15-20 cm single limbs of ileum turned back on themselves a multiple of times to create a reservoir of variable volume. Although, early on, pouches were composed of many different configurations, the J pouch is now the most widely used, because its ease of construction is a major advantage. W pouches probably carry a small advantage over J pouches in terms of pouch frequency, but this was during the phase of routine mucosectomy, when tension on J pouches using mucosectomy was greater than currently [86]. This advantage has been largely offset by the more common use of double-stapling, and the difference in stool frequency between different pouch configurations is regarded as negligible.

References 1. 2. 3.

Brown JY. Value of complete physiological rest of large bowel in treatment of certain ulcerations and obstetrical lesions of this organ. Surg Gynecol Obstet 1913; 16: 610-16. Brooke BN. Ulcerative Colitis and its Surgical Treatment. Edinburgh: E&S Livingstone, 1954. Strauss AA, Strauss SF. Surgical treatment of ulcerative colitis. Surg Clin N Am 1944; 24: 211-24.

Surgical management of ulcerative colitis

640 4. 5. 6.

7. 8. 9. 10.

11. 12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

25. 26. 27.

Brooke BN. The management of ileostomy. Lancet 1952; 102-4. Turnbull RB. Management of an ileostomy. Am J Surg 1953;86:617-24. Kock NG. Intra-abdominal 'reservoir' in patients with permanent ileostomy; preliminary observations on a procedure resulting in fecal 'continence' in five ileostomy patients. Arch Surg 1969;99:223-31. Faren A, Kock NG, Sundin T. The long-term functional results after ileocystoplasty. Scand J Urol Nephrol 1973;7: 127-30. Devine H, Devine J. Subtotal colectomy and colectomy in ulcerative colitis. Br Med J 1957; 1: 489-92. Aylett SO. Three hundred cases of diffuse ulcerative colitis treated by total colectomy and ileorectal anastomosis. Br Med J 1966; 5494: 1001-5. Pare R, Levy E, Frileux P, Loygue J. Current results: ileorectal anastomosis after total abdominal colectomy for ulcerative colitis. In: RR Dozois, ed. Alternatives to Conventional Ileostomy. Chicago: Year Book Medical Publishers, 1985:81 99. Parks AG, Nicholls RJ. Proctocolectomy without ileostomy for ulcerative colitis. Br Med J 1978; 2: 85 8. Hochenegg J. Meine operationcrfolge bei rectum carcinoma. Wein Klin Wochcnschr 1900; 13: 394 404. Nissen R. Meeting of the Berlin Surgical Society, Zentralb. Chirurgie 1933; 15: 888. Ravitch MM, Sabiston DC. Anal ileostomy with preservation of the sphincter: a proposed operation in patients requiring total colectomy for benign lesions. Surg Gynecol Obstet 1947; 84: 1095 9. Soave F. A new surgical technique for treatment of Hirschsprung's disease. Surgery 1964; 56: 1007 14. Safaie-Shirafi, Soper RT. Endorectal pull through procedure in the surgical treatment of familial polyposis coli. J Pedialr Surg. 1973; 8: 711 16. Martin LW, LeCoultre C, Shubert WK. Total colectomy and mucosal proctectomy with preservation of continence in ulcerative colitis. Ann Surg 1977; 186: 477 80. Valiente MA, Bacon HE. Construction of pouch using 'pantaloon' technique for pull-through following total colectomy. Am J Surg 1955; 90: 6621 43. Karlan M, McPherson RC, Watman RN. An experimental evaluation of fecal continence sphincter and reservoir in the dog. Surg Gynecol Obstet 1959; 108: 469-75. Lee EC, Truelove S. Proctocolectomy for ulcerative colitis. World J Surg 1980; 4: 195-201. Leijonmarck CE, Persson PG, Hellers G. Factors affecting colectomy rate in ulcerative colitis: an epidemiological study. Gut 1990; 31: 329-33. Langholtz E, Munkholm P, Davidsen M et al. Course of ulcerative colitis: analysis of changes in disease activity over years. Gastroenterology 1994; 107: 3-11. Jarnerot G, Rolny P, Sandberg-Gertzen H. Intensive intravenous treatment of ulcerative colitis. Gastroenterology 1985;89: 1005-13. Hyde GM, Thillainayagam AV, Jewell DP. Intravenous cyclosporine as rescue therapy in severe ulcerative colitis: time for a reappraisal? Eur J Gastroenterol Hepatol 1998; 10:411-3. Farmer RG, Easley KA, Rankin GB. Clinical patterns, natural history, and progression of ulcerative colitis. Dig DisSci 1993; 38: 1137-46. Gumaste V, Sachar DB, Greenstein AJ. Benign and malignant colorectal strictures in ulcerative coHtis. Gut 1992; 33: 938-41. Lennard-Jones JE, Morson BC, Ritchie JK, Williams CD. Cancer surveillance and ulcerative colitis: experience over 15 years. Lancet 1983; 16: 149-52.

28.

29.

30. 31. 32. 33. 34.

35. 36.

37. 38.

39.

40. 41. 42. 43. 44. 45. 46.

47. 48.

Lennard-Jones JE. Cancer risk in ulcerative colitis and Crohn's disease - strategies to avoid cancer death. In: RN Allen et al. editors. Inflammatory Bowel Diseases. New York: Churchill Livingston, 1997: 675-82. Keighley MR, Acute fulminating colitis and emergency colectomy. In: Keighley MR, Williams NS (eds). Surgery of the Anus, Rectum and Colon. London: WB Saunders, 1993: 1379-97. Hosking SW, Kane SP, Cour-Palais IJ. Reducing the surgical mortality of acute colitis. A district hospital experience. J Royal Coll Surg England 1985; 30: 255-7. Lucarotti ME, Freeman BJ, Warren BF, Durdey P. Synchronous proctocolecotmy and ileoanal pouch formation and the risk of Crohn's disease. Br J Surg 1995; 82: 755-6. Turnbull RB Jr, Hawk WA, Weakley FL. Surgical treatment of toxic megacolon: ileostomy and colostomy to prepare patients for colectomy. Am J Surg 1971; 122: 325 31. Binder SC, Miller HH, Deterling RA Jr. Emergency and urgent operations for ulcerative colitis. Arch Surg 1975; 110: 284 9. Hawley PR. Surgical treatment of ulcerative colitis subtotal colectomy and ileorectal anastomosis. In: RN Allen et al. eds. Inflammatory Bowel Diseases. New York: Churchill Livingston, 1997: 741 5. Jones PF, Orr G. Colectomy and ileorectal anastomosis In: RN Allen et al. eds. Inflammatory Bowel Diseases. New York: Churchill Livingston, 1983: 268 73. Lindsey I, Mortensen NJMcC. Indeterminate colitis: surgical approaches. In: TM Bayless and SB Hanauer, eds. Advanced Therapy of Inflammatory Bowel Disease. Hamilton, Ontario: BC Decker, 2001: 241 4. Pezim ME, Nicholls RJ. Quality of life after restorative proctocolectomy with pelvic ileal reservoir. Br J Surg 1985; 72: 31 3. Kohler LW, Pemberton JH, Zinmeister AR, Kelly KA. Quality of life after proctocolectomy. A comparison of Brooke ileostomy, Kock pouch and ileal pouch anal anastomosis. Gastroenterology 1991; 101: 679 84. McHugh SM, Diamant NE. Effect of age, gender and parity on anal canal pressures. Contribution of impaired anal sphincter function to faecal incontinence. Dig Dis Sci 1987; 32: 726-36. Matheson DM, Keighley MR. Manomctric evaluation of rectal prolapse and faecal incontinence. Gut 1981; 22: 1269. Dozois RR, Ileal 'J' pouch-anal anastomosis. Br J Surg 1985; 72(Suppl.): 580-2. Telander RL, Spencer M, Perrault J, Telander D, Zinsmeister AR. Long term follow up of the ileoanal anastomosis in children and young adults. Surgery 1990; 108: 717-25. Sultan AH, Kamm MA, Hudson CN, Thomas JM, Bartram CI. Anal-sphincter disruption during vaginal delivery. N Engl J Med 1993; 329: 1905-11. Varma A, Gunn J, Gardiner A, Lindow SW, Duthie GS. Obstetric anal sphincter injury. Prospective evaluation of incidence. Dis Colon Rectum 1999; 42: 1537-43. Kamm MA. Obstetric damage and faecal incontinence. Lancet 1994; 344: 730-3. Juhasz ES, Fozard B, Dozois RR, Ilstrup DM, Nelson H. Ileal pouch-anal anastomosis function following childbirth. An extended evaluation. Dis Colon Rectum. 1995; 38: 15965. Olsen K 0 , Joelsson M, Laurberg S, Oresland T. Fertility after ileal pouch-anal anatomosis in women with ulcerative colitis. Br J Surg 1999; 86: 493-5. Lindsey I, Guy RJ, Warren BF, Mortensen NJMcC. The anatomy of Denonvilliers' fascia and the pelvic nerves, impotence, and implications for the colorectal surgeon. Br J Surg 2000; 87: 1288-99.

Ian Lindsey and Neil JMcC Mortensen 49.

50.

51. 52. 53. 54. 55. 56. 57.

58.

59. 60.

61.

62. 63. 64. 65. 66.

67. 68.

69.

Rosen RC, Riley A, Wagner G, Osterloh IH, Kirkpatrick J, Mishra A. The International Index of Erectile Function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology 1997; 49: 822-30. Lindsey I, George BD, Kettlewell MGW, Mortensen NJMcC. Impotence after mesorectal and close rectal dissection for inflammatory bowel disease. Dis Colon Rectum 2001;44:831-5. Kock NG, Myrvold HE. Progress report on the continent ileostomy. World J Surg 1980; 4: 143-8. Hyman NH, Fazio VW, Tuckson WB, Lavery IC. Consequences of ileal pouch-anal anastomosis for Crohn's colitis. Dis Colon Rectum 1991; 34: 653-7. Sagar PM, Dozois RR, Wolff" BG. Long-term results of ileal pouch-anal anastomosis in patients with Crohn's disease. Dis Colon Rectum 1996; 39: 893-8. Deutsch AA, McLeod RS, Cullen J, Cohen Z. Results of the pelvic-pouch procedure in patients with Crohn's disease. Dis Colon Rectum 1991; 34: 475-7. Panis Y, Poupard B, Nemeth J, Lavergne A, Hautefeuille P, Valleur P. Ileal pouch/anal anastomosis for Crohn's disease. Lancet 1996; 347: 854-7. Phillips RKS. Ileal pouch-anal anastomosis for Crohn's disease. Gut 1998; 43: 303-4. Mclntyre PB, Pemberton JH, Wolff" BG, Dozois RR, Beart RW Indeterminate colitis: Long-term outcome in patients after ileal pouch-anal anastomosis. Dis Colon Rectum 1995; 38: 51-5. Koltun WA, Schoetz DJ, Roberts PL, Murray JJ, Coller JA, Veidenheimer MC. Indeterminate colitis predisposes to perineal complications after ileal pouch-anal surgery. Dis Colon Rectum 1991; 34: 857-60. Wiesner RH, LaRusso NF, Dozois RR, Beaver SJ. Peristomal varices after proctocolectomy in patients with primary sclerosing cholangitis. Gastroenterology 1986; 90: 316-22. Penna C, Dozois RR, Tremaine W et al. Pouchitis after ileal pouch-anal anastomosis for ulcerative colitis occurs with increased frequency in patients with associated primary sclerosing cholangitis. Gut 1996; 38: 234-9. Tjandra JJ, Fazio VW, Milsom JW, Lavery IC, Oakley JR, Fabre JM. Omission of temporary diversion in restorative proctocolectomy - is it safe? Dis Colon Rectum 1993; 36: 1007-14. Grobler SP, Hosie KB, Keighley MR. Randomised trial of loop ileostomy in restorative proctocolectomy. Br J Surg 1992;79: 903-6. Brown SR, Eu KW, Seow-Choen F. Consecutive series of laparoscopic-assisted vs minilaparotomy restorative proctocolectomies. Dis Colon Rectum 2001; 44: 397-400. Marcello PW, Milson JW, Wong SK, Brady K, Goormastic M, Fazio VW. Laparoscopic total colectomy for acute colitis. Dis Colon Rectum 2001; 44: 1441-5. Stage JG, Schultze S, Moller P et al. Prospective randomised study of laparoscopic versus open colonic resection for adenocarcinoma. Br J Surg 1997; 84: 391-6. Fukushima R, Kawamura YJ, Saito H. Interleukin-6 and stress hormone responses after uncomplicated gasless laparoscopic-assisted and open sigmoid colectomy. Dis Colon Rectum 1996; 39: S29-34. Hewitt PM, Ip SM, Kwok SP. Laparoscopic-assisted vs open surgery for colorectal cancer: comparative study of immune effects. Dis Colon Rectum 1998; 41: 901-9. Hildebrandt U, Kessler K, Pistorius G et al. Granulocyte elastase and systemic response after laparoscopic-assisted and open resections in Crohn's disease. Dis Colon Rectum 1999; 42: 1480-6. Krahenbiihl L, Schafer M, Kuzinkovas V, Renzulli P, Baer HU, Buchler MW. Experimental study of adhesion forma-

641

70. 71.

72.

73. 74.

75.

76. 77. 78.

79. 80.

81. 82.

83. 84. 85. 86.

tion in open and laparoscopic fundoplication. Br J Surg 1998; 85: 826-30. Garrard CL, Clements RH, Nanney L, Davidson JM, Richards WO. Adhesion formation is reduced after laparoscopic surgery. Surg Endosc 1999; 13: 10-13. Schippers E, Tittel A, Ottinger A, Schumpelick V. Laparoscopy versus laparotomy: comparison of adhesion formation after bowel resection in a canine model. Dig Surg 1998; 15: 145-7. Chen MD, Teigen GA, Reynolds HT, Johnson PR, Fowler JM. Laparoscopy versus laparotomy: an evaluation of adhesion formation after pelvic and paraaortic lymphadenectomy in a porcine model. Am J Obstet Gynecol 1998; 178: 499-503. Levrant SG, Bieber EJ, Barnes RB. Anterior abdominal wall adhesions after laparotomy or laparoscopy. J Am Assoc Gynecol Laparosc 1997; 4: 353-6. Jorgensen JO, Lalak NJ, Hunt DR. Is laparoscopy associated with a lower rate of postoperative adhesions than laparotomy? A comparative study in the rabbit. Aust NZ J Surg 1995; 65: 3 4 2 ^ . Dunker MS, Bemelman WA, Slors JF, van Duijvendijk P, Gouma DJ. Functional outcome, quality of life, body image and cosmesis in patients after laparoscopic-assisted and conventional restorative proctocolectomy: a comparative study. Dis Colon Rectum 2001; 44: 1800-7. Lee ECG, Dowhng BL. Perimuscular excision of the rectum for Crohn's disease and ulcerative colitis. A conservative technique. Br J Surg 1972; 59: 29-32. Tsunoda A, Talbot IC, Nicholls RJ. Incidence of dysplasia in the anorectal mucosa in patients having restorative proctocolectomy. Br J Surg 1990; 77: 506-9. Johnston D, Holdsworth PJ, Nasmyth DG et al. Preservation of the entire anal canal in conservative proctocolectomy for ulcerative colitis: a pilot study comparing end-to-end ileo-anal anastomosis without mucosal resection with mucosal proctectomy and endo-anal anastomosis. Br J Surg 1987; 74: 940-4. Miller R, Bartolo DC, Roe AM, Mortensen NJMcC. Anorectal sampling: a comparison of normal and incontinent patients. Br J Surg 1988; 75: 44-6. Thompson-Fawcett MW, Warren BF, Mortensen NJMcC. A new look at the anal transitional zone with reference to restorative proctocolectomy and the columnar cuff. Br J Surg 1998;85: 1517-21. O'Connell PR, Pemberton JH, Weiland LH et al. Does rectal mucosa regenerate after ileoanal anastomosis? Dis Colon Rectum 1987; 30: 1-5. Choen S, Tsunoda A, Nicholls RJ. Prospective randomised trial comparing anal function after hand-sewn ileoanal anastomosis with mucosectomy versus stapled ileoanal anastomosis without mucosectomy in restorative proctocolectomy. Br J Surg 1991; 78: 430-4. Thompson-Fawcett MW, Mortensen NJMcM. Anal transitional zone and columnar cuff" in restorative proctocolectomy. Br J Surg 1996; 83: 1047-55. Stern H, Walfisch S, Mullen B, McLeod R, Cohen Z. Cancer in an ileoanal reservoir: a new late complication? Gut 1990; 31:473-5. Tuckson WB, Fazio VW. Functional comparison between double and triple ileal loop pouches. Dis Colon Rectum 1991; 34: 17-21. Keighley MR, Yoshioka K, Kmiot W. Prospective randomised trial to compare the stapled double lumen pouch and the sutured quadruple pouch for restorative proctocolectomy. Br J Surg 1988; 75: 1009-11.

32 Pouchitis: clinical cliaracteristics and management UMA MAHADEVAN AND WILLIAM J. SANDBORN

Introduction The term 'pouchitis' was coined by Kock et al. [1] in 1977 to describe an inflammatory condition of the continent ileal reservoir in patients who have undergone a proctocolectomy. This procedure was first described in 1969 [2]. In 1980 Parks et al. and Utsunomiya et al [3, 4] independently described a procedure consisting of a proctocolectomy with ileal reservoir and anal anastomosis. With minor modifications the ileal pouch-anal anastomosis (IPAA) is currently the surgical option of choice in patients with familial adenomatous polyposis (FAP) and ulcerative colitis (UC) with either dysplasia or disease refractory to medical therapy. The functional outcome after IPAA is acceptable. The median number of bowel movements in 24 h is six. Soiling at night occurs in approximately onethird of patients and 24% need to wear a pad occasionally [5]. Despite these numbers, the healthrelated quality of life after IPAA has consistently been comparable to normal populations and better than in those with active UC or inflammatory bowel disease (IBD) in general [5-7]. However, Tiainen et al. found that the health-related quality of life decreased with poor functional status, increased number of bowel movements and chronic pouchitis [5]. Pouchitis is the most common long-term complication of IPAA in UC [8]. Its definition and diagnostic criteria have evolved over time and its etiology is unknown. This chapter will review the current knowledge base for this frequent and sometimes debilitating form of IBD.

Diagnosis The diagnosis of pouchitis should be made on a combination of clinical, endoscopic and histologic grounds. Patients will usually present with diarrhea, watery stools, fecal urgency, abdominal cramps and

occasional rectal bleeding. An endoscopy with biopsies of the pouch is done to confirm the diagnosis. The neoterminal ileum above the pouch should be normal; inflammation and ulceration here raises the suspicion of possible Crohn's disease (CD). Inflammation of the pouch mucosa with granularity, edema, mucosal hemorrhages, contact bleeding and superficial ulcers is present in pouchitis with varying degrees of severity [9]. Inflammation is either uniform throughout the pouch or more severe in the distal pouch [10]. In 1987 the Mayo Clinic defined pouchitis as a clinical syndrome of watery, frequent, at times bloody stool accompanied by urgency, incontinence, abdominal cramps, malaise and fever. The symptoms must be present for at least 2 days and should be relieved within 48 h with metronidazole therapy [8]. While this definition is sensitive for pouchitis, it is not specific and may include other clinical entities such as CD, pouch ischemia and anastomotic stricture. The St Marks Hospital developed a diagnostic criteria for pouchitis based on a triad of diarrhea (6 or more stools/day), endoscopic findings (four or more of edema, granularity, friability, loss of vascular pattern, mucosal hemorrhage or ulceration), and a minimum grade of 4 on a 6-point histopathologic index (polymorphonuclear leukocyte infiltration and percent ulceration per low-power field) [11]. This definition was very specific, but would exclude less severe cases of pouchitis. In 1994 Sandborn et al. developed the Pouchitis Disease Activity Index (PDAI) incorporating the Mayo Clinic definition and the pouchitis triad and histopathologic index defined by the St Marks group [12]. The PDAI attempts to provide a standardized definition of pouchitis based on clinical, endoscopic and histologic markers (Table 1), with pouchitis defined as a score of seven or more points. Diarrhea was defined based on an increase from the baseline number of bowel movements. Other clinical manifes-

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 643-658. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Pouchitis: clinical characteristics and management

644 Table I.The Pouchitis Disease Activity Index Clinical criteria

Score

Stool frequency Usual postoperative stool frequency 1-2 stools/day > postoperative usual 3 or more stools/day > postoperative usual

0 1 2

Rectal bleeding None or rare Present daily

0 1

Fecal urgency/abdominal cramps None Occasional Usual

0 1 2

Fever (temperature > 100 F) Absent Present

0 1

Endoscopic criteria Edema Granularity Friability Loss of vascular pattern Mucus exudates Ulceration

1 1 1 1 1 1

Acute histologic criteria Polymorph Infiltration Mild Moderate + crypt abscess Severe + crypt abscess Ulceration per low-power field (average) 50%

1 2 3 1 2 3

tion for dehydration, frequent incontinence). Duration is defined as acute ( < 4 weeks) or chronic ( > 4 weeks). The pattern of pouchitis is described as infrequent (one or two acute episodes), relapsing (three or more acute episodes), or continuous. Finally, the response to medical therapy is labeled as treatment-responsive or treatment-refractory with the medications for either case specified. If pouchitis is refractory to medical therapy, or has atypical components, further diagnostic tests should be done to exclude alternate diagnoses such as pouch dysfunction. Infectious etiologies should be ruled out by stool sampling and pouch biopsy. There are three cases in the literature of chronic treatment-refractory pouchitis eventually diagnosed as cytomegalovirus infection of the pouch. Therapeutic response to ganciclovir was achieved [14, 15]. Pouchogram X-ray (luminal contrast study) will demonstrate ileoanal anastomotic separations, pouch fistulas and anastomotic strictures. If CD is suspected, a small bowel follow-through X-ray will rule out disease above the pouch. A computerized axial tomograph (CT) of the pelvis will detect peripouch abscesses or inflammatory phlegmons. Endoluminal transpouch ultrasonography has also been used in pouch dysfunction with reported higher rates of fistula and abscess detection than both CT scan and pouchography [16]. Anorectal manometry can be used to rule out pelvic floor dysfunction. Finally, scintigraphic pelvic pouch emptying scans can be used to evaluate patients with inefficient or inadequate pouch evacuation [17].

Pouchitis is defined as a total PDAI score of seven or more points. Adapted with permission form: Sandborn WJ, Tremaine WJ, Batts KP, Pemberton JH, Phillips SF. Pouchitis following ileal pouch-anal anastomosis: a pouchitis disease activity index. Mayo Clin Proc 1994;69:409-15.

tations such as rectal bleeding, abdominal cramps and urgency were also included. The specificity and sensitivity of diagnosis is increased by defining the disease as a continuum from mild to severe pouchitis with symptoms individualized to the norms of each patient. A pouchitis classification scheme was also proposed by Sandborn in 1996 [13]. The degree of activity of pouchitis can be delineated as either remission (no active pouchitis), mild to moderate activity (increased stool frequency, urgency, infrequent incontinence), or severely active (hospitaliza-

Pathophysiology The ileal pouch undergoes adaptive changes once it is exposed to the fecal stream. Functionally, it changes from a primarily absorptive organ to an organ of storage. The histopathologic changes that follow likely reflect this transition. However, the efficiency of pouch emptying or fecal stasis does not appear to be a factor in the degree of mucosal change [18]. The mucosal permeability in the pouch also changes after ileostomy closure. The initially high permeability of the isolated pouch decreases with reanastomosis. This suggests an adaptive mucosal response to reanastomosis and exposure to the fecal stream. The permeability increases again in the presence of pouchitis. Whether this is a cause or effect of inflammation from the invasion of fecal bacteria is unknown [19].

Uma Mahadevan and William J. Sandborn Inflammatory and morphologic changes occur in the pouch mucosa early after ileostomy closure and subsequently do not show a significant difference from 6 weeks to 6 months after reanastomosis [20]. A normal pouch, in both FAP and UC patients, will undergo a minor decrease in the length of the vilH and an increase in crypt depth. With acute pouchitis there is a slight increase in villous length and pronounced crypt hyperplasia. Chronic pouchitis, however, will demonstrate severe villous atrophy and crypt hyperplasia [21]. Ileal pouches acquire certain colonic characteristics including goblet cells, villous atrophy and crypt hyperplasia; however, complete colonic metaplasia does not seem to occur [20, 22]. A study of 25 pouch patients found that the nine with the most severe villous changes produce colorectal-type sulfomucin. However, all patients had small bowel sucrase-isomaltase activity regardless of the severity of inflammation [23]. The inflammation tends to be macroscopically and histologically most severe in the distal versus the more proximal zone of the pouch [10]. Sheperd et al. showed that there was greater inflammation in the posterior wall where the mucosa is in contact with the fecal stream than in the anterior wall, which may be protected from prolonged contact by gas in the lumen. Some colonic phenotypic expression was found in 96% of pouches, without complete colonic metaplasia [24].

Etiology The etiology of pouchitis is unknown. Speculation has centered on the possibility of fecal stasis and/or bacterial overgrowth, nutritional deficiencies, ischemic complications of surgery, a novel third form of IBD, a recurrence of UC in the pouch, or a missed diagnosis of CD. The significantly higher occurrence of pouchitis in patients with UC versus FAP suggests that the mechanism is not related to surgical changes common to both diseases - i.e. ischemia and fecal stasis. However, the efficacy of allopurinol and antibiotics in treating pouchitis suggests that these mechanisms may play a role. The UC host may be genetically more susceptible to having an inflammatory response to these insults in the pouch, much as they are theorized to be susceptible to such insults in their now-resected colon. Lichtman etal. described a model of pouchitis [25]. in both Lewis and Sprague-Dawley rats that underwent ileal pouch rectal anastomosis. This model demonstrates many of the possible etiologic theories

645 of pouchitis. Increased anaerobic bacterial flora in the pouch versus normal ileum was noted and inflammation was present within 4 weeks of surgery. Postsurgical treatment with both metronidazole and allopurinol, a xanthine oxidase inhibitor thought to decrease post-ischemic damage mediated by free radicals, led to histologic improvement in the pouches. A more severe pouchitis developed in the Lewis rats versus the Sprague-Dawley rats, suggesting a host genetic susceptibility to pouchitis.

Autoimmune mechanisms There is evidence to suggest that pouchitis is a recurrence of UC or a novel, third form of IBD. When the ileal reservoir undergoes adaptive changes in its histology and physiology resembhng colonic mucosa, it may become more susceptible to the autoimmune mechanisms that underlie UC. Mucosal cellular infiltrates in ileal reservoirs demonstrate intraepithelial lymphocyte densities similar to counts in the normal colon and significantly less than expected in normal ileum. An increase in RFD9'^ macrophages is known to occur in IBD, but not in acute infections. The proportions of RFD9"^ and 3G8^ macrophages were found to be significantly higher in pouchitis reservoirs versus normal reservoirs [26]. Increased levels of interleukins IL-ip, IL-6 and IL-8 are found in patients with pouchitis versus those with normal ileal mucosa or ileal reservoirs without pouchitis [27]. These cytokine levels were also found to be significantly higher in active pouchitis and chronic active UC compared to healthy controls [28]. In addition, activation of cytokine transcription factors STAT-lot and STAT-ip in pouchitis resembles patterns of activation in UC [29]. These levels decrease after treatment of the pouchitis [30]. Mucosal biopsy specimens from patients with pouchitis demonstrate an increase in activated CD4'^ T ceUs and interferon-gamma (IFN-y) production. The authors suggested that this might lead to the mucosal destruction and crypt hyperplasia seen in pouchitis patients [31]. Leukotriene B4, an eicosanoid involved in mucosal inflammation, has increased concentrations in UC compared to normal controls. Concentrations have also been found to be increased in patients with pouchitis versus those without [32, 33]. Though these in-vitro studies may only be detecting markers of inflammation, they suggest that autoimmune mechanisms similar to UC are at play in the inflamed ileal reservoir.

646

Genetics: pANCA and extraintestinal manifestations The association of the serologic marker antineutrophil cytoplasmic antibody-perinuclear staining pattern (pANCA) with pouchitis is controversial. The prevalence of pANCA in UC patients is 60% [34]. Whether this number is decreased after proctocolectomy [35, 36] or unchanged [37-40] is a matter of debate. The literature is also divided as to whether there is a correlation between pANCA and the development of pouchitis. Four studies have found that the prevalence of pANCA is higher than expected in I PA A patients with pouchitis (89-100%) and lower than expected in patients without pouchitis (18-74%) [34, 38, 39, 41]. However, six more recent studies have shown that there is no correlation between pANCA and the occurrence of pouchitis [36, 40, 42-45]. Whether the failure of these later studies to demonstrate an association is based on the definitions of pouchitis used, the ANCA assay methodology, disease heterogenicity or a true absence of association remains to be determined. However, pANCA levels are increased in patients with primary sclerosing cholangitis (PSC) [46]. PSC, in turn, is a risk factor for pouchitis [35, 47, 48]. Patients with PSC who undergo IPAA have a 63% chance of developing pouchitis versus only 32% for those without PSC. The cumulative risk of developing pouchitis in PSC patients is also higher at 1, 2, 5 and 10 years than in UC patients without pouchitis (22%, 43%, 61% and 79% respectively versus 15.5%, 22.5%, 36% and 45.5%) [48]. Extraintestinal manifestations (EIM) are also increased in patients with pouchitis [42, 49-51]. The increased incidence of pouchitis in patients with PSC and other EIM suggests that there may be a particular genotype of UC that has a stronger predisposition to develop pouchitis. pANCA may or may not be a serological marker for that genotype. These correlations also support the theory that pouchitis may be either a recurrence of UC in the pouch or a third, new form of IBD.

IVIissed diagnosis of Cretin's disease Patients with pouchitis may have findings that overlap with CD. Often, ulcerations and fistulas occur at the site of anastomotic suture lines and have features consistent with CD [9, 52]. However, histologic and endoscopic studies of refractory pouchitis have shown no evidence of CD [9, 53]. Multiple investiga-

Pouchitis: clinical characteristics and management tors have suggested that a diagnosis of CD should never be made based on findings in the pouch. This diagnosis requires disease in the neoterminal ileum above the pouch, the presence of fistulas, reexamination of the original proctocolectomy specimen or disease in the remaining gut consistent with CD.

Fecal stasis and bacterial overgrowth After colectomy the ileal reservoir develops a microbial flora count that is between what is expected in the normal terminal ileum and the colon [54]. This finding, as well as the response of the inflamed pouch mucosa to antibiotics (see below) has led to the theory that pouchitis is the result of inflammation caused by fecal stasis and bacterial overgrowth in the new ileal reservoir. Studies of the ileal pouch have varied. Some authors found that patients with pouchitis did indeed have greater counts of anaerobic and/or aerobic bacteria, when compared to controls without pouchitis [55-59]. Others have found an increase in the bacterial count in all ileal pouches when compared to ileostomy controls and the normal colon, but without a significant difference between those with and without pouchitis [60-64]. This suggests that, if a relationship exists between bacterial overgrowth and pouchitis, it is not a direct one.

Fecal bile acids Alterations in bile acid metabolism occur after IPAA. Natori et al. found increased levels of cholic acid and chenodeoxycholic acid and decreased levels of the secondary bile acids deoxycholic acid and lithocholic acid in ileal pouches versus healthy controls. However, these secondary bile acids were greater in the ileal pouch compared to ileostomy patients, suggesting that a degree of normalization occurs in pouch patients secondary to fecal stasis and bacterial overgrowth in the pouch [65]. Bile acids were found to be cytotoxic to epithelial cells [66], suggesting they have a role in pouchitis and colonic metaplasia. However, later studies by the same author showed that the ileal dialysate was not cytotoxic to the epithelial monolayers. Epithelial resistance and histologic structure was actually preserved, suggesting the presence of a protective factor in the dialysate [67]. Further studies also disputed the role of bile acids in pouchitis by demonstrating no difference in the total bile acid concentration or daily fecal bile acid output in pouchitis patients versus those with a healthy ileal reservoir [52, 64, 68].

Uma Mahadevan and William J. Sandborn

647

Short-chain fatty acids (SCFA)

Epidemiology

Short-chain fatty acids (acetate, propionate, butyrate), are products of anaerobic bacterial fermentation of carbohydrates. They are the major source of energy for the colonic mucosa [69] and are involved in crypt cell production and the absorption of water and electrolytes [70]. While patients with IPAA have been found to have the same mean total concentrations of SCFA as normal controls [71] some authors have found decreased levels of SCFA in both active ulcerative colitis [72, 73] and active pouchitis [7477]. However, Sandborn et al. found no difference in the total fecal SCFA concentrations between patients with and without pouchitis, no correlation between fecal bacterial concentrations and total fecal SCFA, and no correlation between fecal concentrations of bacteria or SCFA and chronic inflammation or villous atrophy [64].

The cumulative risk of having one or more episode of pouchitis varies from 15% to 53% in patients with UC. This variation in range likely results from the different diagnostic criteria used for pouchitis over the years [48, 85-91]. The rate is much lower in patients with FAP, ranging from 3% tol4%o [85, 87, 92]. The rates of refractory pouchitis range from 4.5% to 5.5%, with severe intractable pouchitis leading to excision of the pouch in 0.3%o to 1.3%) of patients. Fig. 1 shows the clinical outcome of 100 UC patients at the Mayo Clinic who underwent IPAA [13].

incidence

Predictive factors ischemia Construction of the ileal pouch may require division of mesenteric vessels to avoid tension on the anastomosis, resulting in some degree of ischemia in the pouch [78]. This ischemia may not be a single postoperative injury, but chronic transient damage aggravated by distension of the pouch. Mild endoscopic and histologic changes consistent with ischemia have been found in the ileal reservoirs of IPAA patients [52, 79]. Patients with pouchitis were also found to have lower rates of mucosal blood flow on endoscopic Doppler readings in the pouch than healthy controls [80]. The villi of the small intestine are reported to contain the highest concentration of xanthine dehydrogenase of any tissue [81]. It has been hypothesized that transient ischemia in the pouch leads to conversion of xanthine dehydrogenase to xanthine oxidase. During subsequent reperfusion, xanthine oxidase catalyzes the formation of oxygen free radicals, possibly leading to cellular injury and pouchitis [82-84]. However, the same surgical techniques are used in both UC and FAP patients, the latter of whom have far fewer episodes of pouchitis. It is unlikely that ischemia is itself a direct cause of pouchitis, but it may play a role in the susceptible host.

Penna et al. found a cumulative risk of pouchitis in UC patients to be 15.5%, 22.5%, 36% and 45.5% at 1, 2, 5 and 10 years after IPAA, respectively. This risk was much higher in patients who had PSC, whose risk at 1, 2, 5 and 10 years was 22%, 43%, 61% and 79% respectively [48]. Stahlberg et al. found similar results with a cumulative risk of 51%o at 4 years. All six patients (100%) with PSC developed pouchitis, and extracolonic manifestations as a whole were a predictive factor for pouchitis [89]. Multiple other studies lend support to the belief that there is an increased incidence of pouchitis in patients with EIMs [42, 49-51]. Lohmuller et al. studied 734 patients who underwent IPAA. Patients who had preoperative EIM had a 39% incidence of pouchitis versus an incidence of 26%o for those who did not. Those who developed EIM postoperatively had a 53%o incidence of pouchitis versus 25%) in those who did not. As in UC smoking has been found to be protective against the development of pouchitis. Merrett et al. found an incidence of pouchitis of 33% in ex-smokers, 25% in non-smokers and 6% in current smokers [93]. These findings have been supported by others [49, 89]. The role of the extent of preoperative UC on the development of pouchitis is more controversial. Samarasekera et al. found no correlation between distal colitis or more extensive disease and the incidence of pouchitis in 177 patients studied [94]. However, Schmidt et al. found that colonic extension had a statistically significant association with the development of pouchitis after IPAA with a linear

Pouchitis: clinical characteristics and management

648

UC patients with IPAA 100 .

—

'

—

No pouchitis 68

.

Pouchitis 32

I Recurrent pouchitis 21 Chronic pouchitis 5

I Maintenance suppressive therapy 3

i

1-2 episodes of acute pouchitis 11

Frequent relapses of acute pouchitis 16

Pouch excision 2

Figure 1.The clinical outcome of 100 patients at the Mayo Clinic who underwent ileal pouch-anal anastomosis for ulcerative colitis. Republished with permission from: Sandborn WJ. Pouchitis: definition, risk factors, frequency, natural history, classification, and public health perspective. In McLeod RS, Martin F, Sutherland LR et al. (eds.). Trends in Inflammatory Bowel Disease 1996. Lancaster, UK: Kluwer Academic Publishers, 1997: 51-63.

association between the gross extent of colonic disease and pouch inflammation. However, the severity of the disease was not predictive [95]. Interestingly, inflammation in the terminal ileum with eosinophils and villous blunting was also predictive with a linear association with the degree of pouch inflammation [95]. The role of eosinophils is further supported by the finding of a threefold elevation in the preoperative colonic mucosa of patients who subsequently developed pouchitis versus those who did not [96]. Pouchitis does not appear to have a predilection for age, sex or race, though one small study did note a decreased incidence in African-Americans when compared to Caucasians [49]. The presence of PSC or other EIM, however, does appear to be predictive for the development of pouchitis. Likewise, the risk of developing pouchitis in smokers is decreased. The role played by the extent of preoperative disease and the histopathology of the colon and terminal ileum needs to be further assessed.

Dysplasia The risk of dysplasia and adenocarcinoma in the pelvic pouch has been mostly a theoretical concern, but one which may become increasingly real in the near future as pouches surpass the age of 20. While cancer in the rectal cuff' following IPAA was described in 1990 [97], an adenocarcinoma of the actual ileal pouch has yet to be described. This may just be a matter of time. In 1990 Heppell et al. demonstrated a rodent model of cancer in the Sprague-Dawley rat ileal reservoir using a chemical carcinogen, 1,2-dimethylhydrazine (DMH), and a tumor promoter, sodium butyrate [98]. Two of 20 rats with a total colectomy and ileal reconstruction developed ileal pouch carcinoma. This was similar to the rate of ileal cancer in the rats without surgery or with ileorectal anastomosis. However, there was a high death rate in the rats with an ileal reservoir, precluding randomization and possibly biasing the results. Lichtman et al. [25]

Uma Mahadevan and William J. Sandborn found lower rates of pouchitis in Sprague-Dawley rats versus Lewis rats undergoing ileal pouch formation. If pouchitis and chronic inflammation predispose to dysplasia, Lewis rats with higher pouchitis rates would theoretically demonstrate higher rates of adenocarcinoma in the ileal pouch. There have been at least 17 cases of adenocarcinoma arising in the permanent ileostomy of patients with UC. The case described by Reissman et al notes diffuse colonic metaplasia in the ileostomy around the adenocarcinoma with sulfomucin production [99]. In 1997 the first case of an adenocarcinoma arising in a continent ileostomy was described in a patient with UC. The pouch mucosa showed chronic inflammation with villous atrophy and mild to moderate dysplasia [100]. It was not clear if this patient suflered from recurrent pouchitis. Also in 1997 a case of large cell lymphoma arising in the pouch of a patient who had had UC was described. This patient suffered from chronic refractory pouchitis, which may in retrospect have been due to the invasive lymphoma, undetected until surgical resection of the pouch for pouch dysfunction [101]. The first reported case of dysplasia in a patient with an IPAA was in 1991 [102]. In 1995 the same group reported the results of 87 patients with IPAA for UC, followed for a mean of 6.3 years. Three types of mucosal adaptation were noted in the reservoir. Type A (51% of patients) was characterized by normal mucosa or a mild villous atrophy and no or mild inflammation. Type B (40% of patients) demonstrated transient atrophy with temporary moderate or severe villous atrophy followed by normalization. Finally, type C (9%) demonstrated constant atrophy with permanent total or subtotal villous atrophy accompanied by severe pouchitis. It was in this last group that low-grade dysplasia was found in three of eight patients. This group also had the greatest detection of sulfomucin-producing goblet cells in the pouch [103]. A prospective follow-up of seven patients with type C mucosa and 14 with type A patterns was done. Dysplasia was noted in five of seven pouches (71%) - four low-grade dysplasia and one high-grade dysplasia. There was no correlation with dysplasia in the colectomy specimen but there was an association with an early onset of UC. The authors felt that patients identified as having a type C response 4 years after ileostomy closure should have at least annual pouchoscopy with surveillance for dysplasia [104]. Other authors have not found dysplasia on surveillance of the pouch [105-108]. However, similar rates

649 of type A, B and C mucosa were found in adults [107] and children [108] with an ileoanal pouch for UC. Setti Carraro et al. confirmed the finding that only patients with type C mucosa developed chronic pouchitis. They also noted that the categorization of response type could be made at 6 months after ileostomy closure [107]. In 1991 Lofberg etal. [102] reported a case of a 36year-old Sri Lankan man with intractable UC who underwent a colectomy, mucosal proctectomy and IPAA with an S-type pelvic pouch. No dysplasia was noted in the colectomy specimen. The patient suffered from chronic pouchitis and was on longterm metronidazole therapy. Four years after pouch creation he was noted to have low-grade dysplasia on biopsy and DNA aneuploidy by flow cytometry. The patient then underwent periodic surveillance pouchoscopy with biopsy and in 1996 high-grade dysplasia was detected [104]. In 1997 the patient experienced abdominal pain. Further investigation led to the diagnosis of a primary cholangiocarcinoma, with likely underlying subclinical PSC. The patient expired 3 months later [109]. This case illustrates much of the epidemiology of pouchitis and dysplasia discussed above. The patient had an increased risk of developing pouchitis because of his underlying PSC. The presence of both chronic pouchitis and PSC (by extrapolation from the dysplasia data in patients with UC and PSC [110115]) increased his risk of dysplasia. Patients with chronic and severe inflammation have a greater degree of colonic metaplastic changes in their pouch. This may expose them to the same factors - autoimmune, bacterial, genetic - that lead to their initial UC. This partially colon-like mucosa is now susceptible to develop an IBD type illness with its inherent risk of dysplasia. No screening program is currently advocated for patients with IPAA after colectomy for UC. Further studies are needed to delineate which patients need screening, when, where in the pouch and how often. It is reasonable, based on the available data, to do random mucosal sampling in the reservoir of all patients with a history of UC and a pouch 4 years after closure of the ileostomy. Those found to have type C mucosal changes and/or chronic pouchitis should undergo annual surveillance pouchoscopy, much as is done in UC.

Pouchitis: clinical characteristics and management

650

Treatment The treatment of pouchitis is predominantly empiric given the few controlled trials available. In a review of medical therapy in pouchitis, Sandborn et al. identified only four randomized controlled trials for the treatment of pouchitis [116]. Antibiotics are the mainstay of treatment, lending credence to theories of bacterial overgrowth and fecal stasis having a primary role in the pathogenesis of pouchitis. Other treatments are based on therapy used in UC as well as pathophysiologic changes noted in the pouch. Table 2 lists the treatment options available [117].

Antibiotics Metronidazole is the first-line therapy for pouchitis. Support for metronidazole comes from a W-of-T

randomized trial [118] and a randomized controlled crossover trial which demonstrated a 73% response (defined as a decrease in stool frequency) in 13 patients with chronic pouchitis. The placebo response was 9% [119]. Hurst et al. found that 41 of 52 patients (79%) with acute pouchitis responded to a 7-day course of metronidazole at 250 mg orally three times a day with complete relief [92]. Two small series found metronidazole to have a response rate of 100% when given as a topical solution instilled at 75150 mg daily [120] or 40-160 mg daily [ 121 ]. The 20% of patients who fail to respond to oral metronidazole, or who suffer from side-effects such as dysgeusia, dyspepsia, nausea and peripheral neuropathy, can be tried on alternate antibiotics. Hurst et al. found that 11 of 52 patients did not respond to metronidazole. These patients were then given ciprofloxacin at 500 mg twice a day, with eight

Table 2. Treatment options for pouchitis Class 1.

Antibiotics

2.

Probiotics

3.

Example A. B. C. D. E. F. A.

Metronidazole Ciprofloxacin Amoxicillin/clavulanic acid Erythromycin Tetracycline Rifaximin Lactobacilli, Bifidobacteria, Streptococcus salivarius subsp. ttiermoptiilus

5-Aminosalicylates

A. B. C.

Mesalamine enemas Sulfasalazine Oral mesalamine

4.

Corticosteroids

A. B. C.

Conventional corticosteroid enemas Budesonide enemas Oral corticosteroids

5.

Nutritional agents

A. B.

Short-chain fatty acid enemas or suppositories Glutamine suppositories

6.

Immune modifier agents

A.

Cyclosporine enemas

B.

Azathioprine

7.

Oxygen free radical inhibitor A.

Allopurinol

8.

Smoking/nicotine A. B.

Smoking ?Transdermal nicotine

A. B.

Bismuth subsalicylate Bismuth carbomer enemas

A. B.

Ileal pouch exclusion Ileal pouch excision

9.

Antidiarrheal/antimicrobial

10. Surgical options

Modified with permission from: Sandborn WJ. Pouchitis following ileal pouch-anal anastomosis: definition, pathogenesis, and treatment. Gastroenterology 1994; 107:1856-60.

Uma Mahadevan and William J. Sandborn

651

(73%) responding. This gave an overall antibiotic response rate of 96% [86]. Gionchetti et al used rifaximin 1 g twice a day in combination with ciprofloxacin 500 mg twice a day for 15 days in 18 patients with chronic treatment-resistant pouchitis [122]. Sixof 18 (33%) had complete remission defined as a PDAI of 0. Ten of 18 (55.6%) had clinical improvement with a decrease of three points on their PDAI, for a total response rate of 88.8%). Other antibiotics used with some success include amoxicillin/clavulanic acid, erythromycin and tetracycline [123]. An initial episode of pouchitis should be treated with metronidazole 250 mg three times a day for 710 days. Response should be seen within 2-3 days. Responders who have repeat episodes and are able to tolerate the medication should be re-treated with the

same regimen. Some patients with chronic pouchitis will require anywhere from 250 mg every third day to 250 mg three times daily to maintain their response. Others may develop resistance and require a rotating schedule of three or more antibiotics. If antibiotics fail, other therapeutic options should be considered (Fig. 2) [117].

5-Aminosalicylates (5-ASA) Anecdotal reports suggest a benefit from oral and topical mesalamine [9, 52, 124]. Miglioli et al. describe three patients with pouchitis after IPAA for UC. They were given 5-ASA as a suppository or enema at 1.2-4 g daily. After 20-30 days a clinical and endoscopic improvement was noted with partial histological recovery [125].

Symptoms of pouchitis

1 Endoscopy with biopsy • Pouchitis \ -*• Metronidazole

I

Response

\ Distant recurrence

—hPrompt recurrence

— \

I .

Distant recurrence

[\/laintenance metronidazole

Anti-inflammatory drugs ° Immunosuppressive drugs ^

—I

Nutritional replacement^

No response
= 0.02 respectively). Closure of all fistulas was observed in 55% of patients on the 5 mg/kg dose and 38% of patients on the 10 mg/kg dose, and 13% of placebo-treated patients (p = 0.001 and p = 0.04 respectively). Infliximab was more beneficial in non-users of AZA or 6-MP (odds ratio of response: 1.8 and 8.3 for users and non-users respectively). This finding is not surprising since immunomodulators have shown efficacy in the treatment of perianal disease. Compared with the placebo-treated patients the infliximab-treated patients had significant decreases in the score on the Perianal Disease Activity Index at weeks 2 and 8. In the responders the median time to response with infliximab was 14 days, compared with 42 days for patients responding to placebo. The median duration of response was 12 weeks. Five patients (all on infliximab) suffered serious adverse reactions; four of these patients were assigned to the 10 mg/kg dose and, of these, three had infectious complications (pneumonia, furunculosis and anal abscess). The study again confirmed that 5 mg/kg was also an optimal dose for treatment of fistulizing CD, although the optimal number and timing of infusions will need to be addressed in future studies [105]. The most frequently reported events in infliximabtreated patients have been headache, nausea and upper respiratory infections [106]. Acute infusion reactions consisting of fever, chills, pruritus, urticaria and shortness of breath or dyspnea occurred in 16%) of infliximab-treated patients. Most infusion reactions were mild, and all patients recovered with treatment (e.g. diphenhydramine) and/or discontinuation of the infusion. Human antichimeric antibodies (HACA) developed in 13% of patients and correlated with the occurrence of infusion reactions. Concurrent immunomodulatory therapy with AZA,

665 6-MP or MTX protected against both infusion reactions and the development of HACA. It is uncertain what role HACA have regarding efficacy, duration of response or risk of infusion reactions. Many patients in clinical trials were re-treated despite the presence of HACA. Other immunologic responses include antibodies to double-stranded DNA that developed in 9% of patients, but only a few patients have developed lupus-like syndromes that appear to behave similar to other reports of drug-induced lupus [106]. The formation of anti-DNA antibodies may be a class eff'ect of related preparations of monoclonal antibodies as they have also been seen with etanercept [107]. A second type of delayed reactions occurred in patients who had received an initial infliximab infusion, followed by a second infusion 6-24 months later. In contrast to the acute infusion reactions these patients had features of a serum sickness-like illness that occurred 3-12 days after the second infusion. These 'delayed hypersensitivity reactions' presented with severe arthralgias, myalgias, fever, rash, and leukocytosis. All patients recovered, although several required hospitalization and treatment with high-dose corticosteroids. Of note, these patients had no evidence of HACA prior to the inciting infusion but rapidly developed high HACA titers coinciding with the delayed hypersensitivity reaction. Thus far there are no reports of these patients receiving subsequent infliximab infusions. Of 771 patients who have received infliximab in clinical trials (including trials in rheumatoid arthritis), six patients have developed lymphoproUferative disorders, including a 60-year-old patient with CD for 30 years who had also been receiving AZA and who developed a B-cell non-Hodgkin's lymphoma. A 29-year-old patient with CD for 4 years developed Hodgkin's disease 2 weeks after receiving an infliximab infusion at a dose of 5 mg/kg. In this case the rapid emergence of lymphoma makes a causative role for infliximab unlikely. Such reports linking infliximab to lymphoma must be examined carefully, as patients with CD are at increased risk for lymphoma as a result of their underlying disease and immunosuppressive therapies [108]. Overall, the available data provide no firm evidence linking infliximab treatment to infections, malignancy or autoimmune disorders [106]. The ultimate role for infliximab in CD has yet to be defined. The drug has been marketed in the US for only 2 years, and FDA approval has been limited to single infusions of 5 mg/kg for active disease not

666 responding to 'conventional' therapy or as a series of three infusions for fistuUzing disease. Nevertheless, most patients who respond to infliximab lose their clinical response by 8-12 weeks and do benefit from repeated infusions. In contrast, infliximab is approved for three initial doses at 0, 2 and 6 weeks followed by maintenance therapy every 8 weeks. The rheumatoid arthritis patients are treated concurrently with MTX and have a lower incidence of HACA formation and infusion reactions. Thus, it is likely that repeated early infusions (and possibly concurrent immune modulation) allow the development of tolerance whereas, as has been noted in the patients with delayed hypersensitivity reactions, single infusions may be tolerizing and increase the risk of delayed hypersensitivity when subsequent infusions are delayed beyond 8-12 weeks. We currently reserve infliximab therapy for patients who have failed or cannot tolerate therapy with AZA, 6-MP or MTX, and administer repeat infusions of infliximab at approximately 8-week intervals, or as needed according to the duration of the clinical response. Additional indications include patients who are intolerant of steroid therapy, or who have significant symptoms despite steroids, who cannot wait for the long-term response to conventional immune modulators. We have observed that most patients who respond require ongoing therapy with infliximab despite concurrent AZA, 6-MP or MTX. The response to infliximab is probably decreased in patients with intestinal strictures, especially in the presence of proximal bowel dilation [109]. For patients with fistulizing CD we utilize infliximab after a stepwise approach, starting with antibiotics and surgical drainage of purulence and possible placement of setons to prevent re-accumulation of fluctuance. We use infliximab along with AZA or 6-MP maintenance therapy, but also recognize that some patients may achieve long-term results after a single infliximab infusion. Patients are retreated with infliximab on an as-needed basis to control recurrent fistula drainage. CDP571 CDP571 is a humanized, high-affinity, anti-TNF-a monoclonal IgG4 antibody that contains the complementarity-determining region of a murine antiTNF-oc antibody linked to a human Y4 heavy chain and K light chain [110]. CDP571 contains 95% human and 5% murine sequences, and thus has the theoretical advantage of less immnunogenicity than infliximab. CDP571 has anti-TNF-a neutralizing

Medical therapy for Crohn's disease activity but does not fix complement or mediate antibody-dependent cytotoxicity [107]. In a double-blind, placebo-controlled study, 31 patients with mild-moderately active CD (CDAI 150-400) were randomly assigned to receive a single infusion of CDP571 (n = 2\; 5 mg/kg) or placebo (« = 10) and were followed for 8 weeks. The primary endpoint was the change in CDAI after 2 weeks. In the CDP571-treated group, median CDAI fell from 263 io \61 (p = 0.0003), whereas in placebo-treated patients the change was not significant (253 and 247, before and after treatment respectively). Six patients in the CDP571-treated group, but no patients in the placebo group, achieved remission (DCAI < 150) at 2 weeks (n.s.) [111]. In the CDP571-treated patients there were significant improvements in the HarveyBradshaw score, erythrocyte sedimentation rate and ocpglycoprotein at 2 weeks, but these improvements were not sustained at follow-up at 4, 6 and 8 weeks. The decline in clinical and laboratory parameters starting at 4 weeks paralleled the progressive fall in CDP571 plasma concentrations. In a larger, dose-ranging trial, 169 patients with moderate-severely active CD were randomized to placebo or four different regimens of CDP571 (10 or 20 mg/kg at week 0; and 10 mg/kg at weeks 8 and 16, or 10 mg/kg at week 12), and were followed for 24 weeks. At week 2 clinical improvement (defined as a CDAI decrease of > 7 0 points from baseline) occurred in 15/56 (27%) of the placebo group, 29/ 54 (54%) of the 10 mg/kg group {p = 0.005), and 21 / 57 (37%) of the 20 mg/kg group. Of interest, the group that received CDP571 at 12-week intervals experienced a significantly longer time before withdrawal from the trial, whereas there were no differences between CDP571 from placebo for patients receiving drug at shorter intervals. Two of 13 (15%) placebo-treated patients with draining fistulas had closure of ^ 5 0 % of fistulas, compared with 12/24 (50%) of CDP571-treated patients (p = 0.074). Among CDP571-treated patients, anti-CDP571 and anti-double-stranded DNA antibodies developed in 9% and 7% of patients respectively [112]. No patients developed HACA. A second randomized, double-blind, placebo-controlled trial examined the steroid-sparing properties of CDP571 [113]. Seventy-one patients with steroiddependent CD received CDP571 (20 mg/kg at week 0, and 10 mg/kg at week 8) or placebo. At week 40, 17/39 (43.6%) of CDP571-treated patients were in remission and off* steroids, compared with 7/32 (21.9%) of placebo-treated patients (p = 0.049). In

Stephen B. Hanauer and Themistocles Dassopoulos the CDP571-treated group, anti-CDP571 and antidouble-stranded DNA antibodies developed in 2.6% and 6.7% of patients respectively. Etanercept Etanercept is a genetically engineered fusion protein that contains two identical chains of the recombinant extracellular human p75 TNFR monomer fused to the Fc domain of human IgGi [114]. Etanercept binds soluble TNF-a and blocks its binding to TNFa receptors [114, 115]. In vitro, etanercept binds to transmembrane TNF-oc, but does not result in antibody-dependent or complement-dependent cytotoxicity [107]. In a pilot study, 10 patients with active CD (CDAI220-450) were treated with etanercept 25 mg subcutaneously twice a week for 12 weeks [116]. Clinical results demonstrated about 50%o improvement that was not accompanied by mucosal heahng. A larger study, yet to be reported, did not substantiate differences between etanercept and placebo in patients with chronic active CD (Sandborn, W, personal communication). The success of anti-TNF-a therapy clearly points to TNF-oc as a pivotal mediator in the pathogenesis of CD, yet many questions remain as to the precise mechanisms of anti-TNF-oc strategies in CD. An understanding of the relative roles of soluble and transmembrane TNF-oc, and the identification of the cellular source(s) of TNF-oc in CD, may allow the design of better-targeted anti-TNF-oc approaches. Comparative studies of non-responders, responders, and long-term responders to infliximab will illuminate the role of TNF and other mediators in the entire disease spectrum of CD. The suggestion of lower efficacy with higher doses of infliximab suggests that 'too much' anti-TNF-oc therapy may be detrimental to therapeutic eflicacy. Of equal importance, excessive amounts of anti-TNF-oc therapies may interfere with the homeostatic role of TNF-oc at the systemic level, and may predispose to infection and abnormal immune surveillance of neoplasia. Recently, the FDA has required warnings for both infliximab and etanercept regarding risks of demyelinating neurologic sequelae, as well as risks of miliary tuberculosis in high-risk populations. As with other targeted therapies in CD (topical mesalamine and corticosteroids), targeting anti-TNF therapy to the mucosal level could potentially improve efficacy and minimize toxicity.

667 Thalidomide Thalidomide, a drug originally released as a sedative and antiemetic, has multiple immunomodulatory and anti-inflammatory properties, has been reapproved by the FDA for treatment of lepromatous leprosy and has been used to treat aphthous stomatitis [117, 118]. Thalidomide suppresses production of TNF-oc [119, 120] and IL-12 [121], but also has other pharmacologic actions, including anti-angiogenesis, that are potentially relevant to IBD [122]. In an openlabel trial [123], 16 male and six female patients with steroid-refractory luminal {n = 9) and refractory fistulizing (n = 13) CD received thalidomide 200 mg/day (n= 18) or 300 mg/day (n = 4) for 12 weeks. The dose was reduced if patients developed excessive sedation. For lumenal disease, clinical response was defined as a reduction in the CDAI score of > 150 points, and remission was defined by a CDAI score < 150. For fistulizing disease, response was defined as a clinical improvement of ^ 1 + in three parameters of the goal interval score (GIS) [47], and remission as a score of ^ 2 + in all three parameters of the GIS. All 14 patients completing 12 weeks of treatment had a chnical response, and nine were in remission (three luminal and six fistula patients). Two patients with fistulizing disease resistant to infliximab responded to thalidomide. Median daily prednisone dose decreased from 26 mg initially, to 7.5 mg at week 12 (p0.22 |im) must be responsible for the inflammatory response [6,7]. Experimental models of inflammatory bowel disorders in immunologically altered rodents (transgenic, knockout, spontaneous) require the presence of normal luminal bacteria, especially Bacteroides spp. In HLA-B27 transgenic rats Bacteroides spp. preferentially induce colitis [8], and in humans B.fragilis is increased in CD ileal resections [9]. Besides Bacteroides spp. the human neoterminal ileum is also heavily colonized by Escherichia colidind Clostridium spp. (i.e. acoloniclike flora) after ileocolonic resection [10]. Two clinical studies indicate that the reduction of the load of anaerobic bacteria during intake of nitroimidazole antibiotics (metronidazole and ornidazole) led to less severe endoscopic recurrence [11, 12]. All these studies suggest a potential role of bacteria in inducing CD recurrence. In addition, other experiments focused on dietary components as a factor in luminal-content-induced recurrence. For instance ubiquitous titanium dioxide and aluminosalicylate particles used as food and pharmaceutical additives induce persistent intestinal injury [13]. Further studies, possibly using the infusion model, may elucidate the main responsible components in the fecal stream. It has also been suggested that increased intestinal permeability, leading to enhanced penetration of antigens, may be a basic feature of CD [14]. This hypothesis was not, however, confirmed by investigators in France. They were unable to find a correlation between enhanced permeability 6 weeks postoperatively and endoscopic recurrence again 6 weeks later [15].

699

Early immunologic changes A major factor in the development of intestinal inflammation is the delicate balance of cytokines. In the inflamed mucosa of full-blown CD, a typical Thlpattern of cytokines has been observed, with abundant presence of tumor necrosis factor alpha (TNFa), interleukin-1 (IL-1) and interferon gamma (IFNY). This pattern is different from the one observed in early ileal Crohn's lesions, where low levels of IFN and high levels of IL-4, a typical immunoregulatory Thelper 2 cytokine, were demonstrated [16]. IL-4 attenuates the barrier function of the intestinal epithelium and may induce an enhanced penetration of noxious agents. In addition, incubation of monocytes/macrophages with IL-4 stimulates IL-12 and TNF production by these cells and may explain a switch from aTh2 toward a more typical Thl response [17]. The same group of investigators also demonstrated that ileal CD recurrence was associated with an enhanced IL-5 and immunoglobulin E (IgE) production accompanied by eosinophilic infiltration, a finding which suggested the involvement of an allergic mechanism [18].

Predilection for the neoterminal ileum It remains unclear why postoperative recurrence of CD preferably develops in the ileum immediately proximal to the ileocolonic anastomosis. Scanning electron-microscopic studies have shown that a triad consisting of mucosal architectural alterations, epithelial bridge formation and goblet cell hyperplasia can be found in the majority of biopsies taken from resection specimens, both in the affected area and in the unaffected proximal ileal section margin. Macroscopic recurrence of CD would merely represent a transition from (electron-)microscopic towards endoscopically visible disease [19]. Reflux of colonic content after removal of the ileocecal valve may play a key role in this phenomenon. In addition, stasis and bacterial overgrowth may be involved. In the absence of a classic ileocolonic anastomosis, e.g. with primary ileostomies, the risk of severe recurrence is indeed much lower [20]. An alternative explanation has been sought in the inflammatory changes often encountered in the enteric nervous system. Even in the absence of microscopic inflammation in the ileal resection margins, subtle inflammation surrounding neural structures is common and may 'guide' the inflammation to the more proximal ileal segments [21]. Our group showed

700 that the presence and severity of neuritis in the ileal section margin were predictive of recurrent CD 3 months postoperatively [22]. Postoperative recurrence is not exclusively located in the neoterminal ileum. Recurrent inflammation in the colon after an ileal or ileocecal resection is rare, but may occur. We reported on a series of 18 patients in whom the rectum was the major site of symptomatic CD recurrence, often without accompanying ileal recurrence [23].

Incidence, natural history and characteristics of recurrence For the assessment of recurrence rates an actuarial analysis has to be used. This implies the calculation of the number of patients with recurrent disease over the number of patients at risk in each year of follow-up allowing recurrence in previous years. Systematic endoscopic study data are available only for the neoterminal ileum and ileocolonic anastomosis after ileal or ileocolonic resections. Recurrent lesions can be visualized as early as a few weeks to months after resection in the neoterminal ileum proximal to the anastomosis with a postanastomotic colon mostly free of macroscopic disease, although microscopic colonic lesions can also be identified [1]. Severe endoscopic recurrence is seen in about 30% of patients 3 months after surgery and 50-94% of patients at 12 months, as shown in the placebo arms of the postoperative recurrence prevention trials [24]. Symptoms recur later after curative resection averaging about 10% of the patients per follow-up year if one looks at figures of older follow-up series. In the same postoperative prevention drug studies the symptomatic recurrence rates in the placebo groups are higher and amount to 23-26% at 1 year and about 40% at 2 years [24]. Recurrence of tissue lesions precedes the appearance of recurrent symptoms. The severity and extent of tissue recurrence predicts the time to clinical relapse. Patients without lesions or presenting only a few aphthous ulcers at ileocolonoscopy at 1 year are not at risk for early symptomatic relapse, but more than half of the patients presenting with diff'use aphthous or ulcerative ileitis will have symptomatic relapse within 1 -3 years after operation. Patients with ulcers confined to the immediate preanastomotic region are probably prone to develop fibrostenosis of the anastomosis [25].

Postoperative prevention of recurrence of CD There is convincing evidence that in patients with postoperative recurrence the evolution of CD mimics the natural evolution of CD at its onset. The evolution goes from preulcerative inflammation over aphthoid ulcers to more extensive ulceration and nodularity, to result in complications including stenosis and fissures which can be complicated by abscesses and fistulas. The clinical presentation of recurrent Crohn's ileitis is often strikingly similar to the preoperative presentation. As early as 1971 de Dombal et al. suggested that there might be several subtypes of CD: an indolent' one, which tends to recur slowly, and an 'aggressive' one, recurring soon after surgical resection [26]. This hypothesis was further elaborated in a retrospective analysis at the Mount Sinai Hospitals in New York, in which patients were classified into a subgroup with 'perforating' CD (including fistulas and abscesses), and a non-perforating or stenosing subgroup. Second operations for perforating indications were performed more often among cases whose surgical indication had been perforation initially. Reoperation for perforating disease was required earlier for non-perforating CD [27]. We have not been able to reproduce this finding in a group of North American patients in whom recurrence was defined radiologically and not based on reoperation. In this same study the length of ileal recurrence measured on small bowel X-rays before resection and at the time of symptomatic recurrence was comparable before surgery and at the time of clinical recurrence. In seven patients who had sequential small bowel studies without intervening surgery, the length of measured inflammation remained unchanged, a finding which demonstrates that the extent of disease rarely changes once it is established [28].

Risl( factors for symptomatic recurrence after resection for Crohn's disease (Table 3) Several studies have looked for the potential risk factors for CD recurrence. We will discuss the main risk factors found and reproduced in several studies. The only independent factors generally reproduced in multivariate analyses are location of the disease prior to surgery, pattern of the disease and smoking status. Many other factors such as gender, duration of disease, first or subsequent resection, type of anastomosis and extent of the resection (see below), contraceptive pill use, age at onset of disease, presence of granulomas, blood tranfusions, or a

Filip Baert, Geert D'Haens and Paul Rutgeerts Table 3. Risk factors for Crohn's disease recurrence Risk factors for recurrence Active smoking (especially in women) Location of disease ileocolitis > ileitis > colitis Ileostomy^ lleocolonic anastomosis No risk factors for recurrence Type of anastomosis (hand-sewn versus stapled; end-to-end versus end-to-side or side-to-side) Resection free margins First versus second or third resection

family history have not been found to be risk factors in multivariate analyses in several studies [29, 30].

Location of diseased segment The anatomic site of bowel resection is without doubt an important determinant of recurrence. The highest rates are found after resection for ileocolitis or ileitis involving an ileocolonic anastomosis. Lower rates are found after colonic resection with colo-colic anastomosis. Surprisingly CD also recurs in the ileum after right colonic resection with ileocolonic anastomosis even when the ileum was not diseased prior to surgery. The rate of symptomatic recurrence proximal to an ileostomy is the lowest. It is uncertain whether the rate is really lower or whether the evolution of recurrent lesions is slower. Reoperation rates after resection with ileocolonic anastomosis range from 25% to 60% at 5 years and 42% to 91% at 15 years. After colo-colic anastomosis the rates range between 8.5% and 42%) at 5 years and 22%o to 40% at 15 years. Reoperation for end-ileostomies still amounts to 25% at 5 years and 45%) at 10 years. The rate of recurrence of CD in the neoterminal ileum after ileorectal anastomosis seems comparable to that after more proximal ileocolonic anastomosis but the disease more often spreads distally with anorectal complications as the main manifestation [20,31].

Ttie betiavior of Crohn's disease CD has been divided into three categories according to its behavior: an inflammatory type, a fibrostenotic type (forming strictures) and a perforating type (forming fistulas and abscesses). Perforating disease

701 was found to be a good predictor of recurrence as described by studies from de Dombal and coauthors [26] in Leeds and Greenstein et al. [27] in New York. De Dombal et al. as early as 1971 described a biphasic symptomatic recurrence graph, but offered no clear explanation. The Mount Sinai investigators distinguished perforating indications for resection and non-perforating indications. Perforating indications included acute free perforation, subacute perforation with abscess formation, and chronic perforation with intestinal fistula formation. Non-perforating indications comprised intestinal obstruction, medical intractability, hemorrhage, and toxic dilation without perforation. Operations for perforating indications in their retrospective analyses were followed by repeated resection much earlier than operations for non-perforating indications. Time to first reoperation was 4.7 years in the perforating group and 8.8 years in the non-perforating group. Time to second reoperation averaged 2.3 years for perforating versus 5.2 years for non-perforating indications. Second operations in the perforating indication group were undertaken for perforating indications again in 64%o of the patients with ileitis and 77% of the patients with ileocolitis. A third resection in patients with perforating indications for second resection was carried out for perforating complications in 81% overall. Another study confirms perforating disease as a significant predictor for clinical and surgical recurrence. N o t all investigators have confirmed this dichotomy of CD behavior. It should be stated that many of these studies are based on retrospective series dating back to the 1970s or earlier. Today, with the broader use of immunosuppressive and the new biological agents, it is probably fair to look at clinical recurrence rather than surgical recurrence. We studied the correlation between preoperative disease and recurrent ileitis in terms of length and all three types of disease behavior. We found that the majority of patients with initial fibrostenotic disease, and all patients with inflammatory disease, had recurrence with the same disease pattern. Perforating disease presented again with symptoms with any of the three features. There was a striking correlation between the pre- and postsurgical extent of ileal disease (r = 0.70;;? < 0.0001). We could not confirm perforating disease as being a more aggressive type in terms of onset of clinical recurrence [28].

702

Smoking status Generally speaking smoking has a deleterious effect on CD. The most important risk factor for clinical recurrence and reoperation generally reproduced in all studies is smoking. In an Italian series by M. Cottone in 1994 the 6-year clinical recurrence-free rate after surgery was 60% (95 CI 43-72%) for nonsmokers, 41% (95% CI 11-70%) for ex-smokers and 27% (95% CI 17-37%) for smokers [32]. The need for repeated surgery 5 years after surgery averages 20% in non-smokers versus 36% in smokers, as shown by Sutherland etai in 1990. At 10 years these rates amount to 41% and 70% respectively [33]. The risk is particularly high in female smokers with small bowel disease (odd ratio 9.2). Another study confirms the high risk with a higher cumulative recurrence risk in heavy smokers ( > 15 cigarettes/day). The risk of excisional surgery associated with smoking seems to be increased only in patients not given immunosuppressive drugs, as shown by the French Getaid Group [34]. It seems that the effects of smoking can be antagonized by long-term immunosuppression.

Prevention General measures and surgical factors Only one strategy has resulted in a long-term diseasefree perianastomotic region thus far. Diversion of intestinal contents through a proximal ileostomy protects the neoterminal ileum and ileocolonic anastomosis from recurrence [4]. All other strategies can help to reduce the incidence of CD recurrence but will never prevent it completely. As mentioned above, active smokers have a substantially higher risk of having recurrence after a radical resection. There is approximately a twofold increased risk of recurrence, and the risk is dependent on the number of cigarettes smoked. Ex-smokers have a similar risk to that of non-smokers, and giving up smoking soon after surgery is associated with a lower probability of recurrence [35, 36]. There is no role for the continuation of steroid use in general postoperatively. Steroid therapy should be completely tapered and stopped prior to or immediately after surgery. Several surgical factors have been studied. Surgeons have looked at various strategies to decrease the risk of postsurgical recurrence. Contrary to what was originally believed, the pathological features of the section margins and of the resected bowel loops do

Postoperative prevention of recurrence of CD not affect the incidence of recurrence. Neither the presence of granulomas in the resection specimen, nor the presence or absence of 'disease-free' resection margins (i.e. with or without inflammatory activity) have been shown to influence recurrence rates or severity, although the presence of neural inflammation may be a predictor of early recurrence [22,37,38]. A recent clinical trial in which patients were randomized to 2 or 10 cm margins, performed at the Cleveland Clinic, detected no difference in recurrence rates [39]. The type of anastomosis (end-to-end versus endto-side ileocolonic anastomosis) also has also no effect on the rate of recurrence [31]. In some centers an end-to-end ileocolonic anastomosis is preferred, because anastomotic strictures can more easily be treated endoscopically with this construction. Given these findings, resections should be as limited as possible and surgically safe. The recurrence rates of CD are much lower after surgical resection with ileostomy than after resection with ileocolonic anastomosis [40-42], unless the fecal stream is diverted from the anastomosis [21]. In a group of 182 patients with an end-ileostomy for CD 64% of those with ileocolitis as initial presentation had recurrence in the neoterminal ileum after 20 years, versus 15% of those with initial colitis (p < 0.001) [43]. The number of interventions seems to be of little importance with regard to recurrence. In a Swedish study the cumulative recurrence rates at 10 years were 65% after the second intervention versus 60% after the third one. This supports the hypothesis that disease behavior remains unchanged throughout the patient's history. The same study reported that an ileorectal anastomosis may carry an increased risk for postoperative recurrence, up to 70% at 10 years [44].

Medical intervention It seems obvious that if one can prevent the recurrence of new lesions in the remaining bowel symptomatic relapse will also be prevented. Eventually this would lead to cure of the disease. Prophylactic therapy after resection for CD should therefore aim primarily at keeping the remaining bowel completely free of the disease. A secondary endpoint could be the prevention of symptomatic recurrence only. We are at present far from reaching the primary goal. Standard anti-inflammatory approaches with glucocorticosteroids or 5-acetylsalicyHc (5-ASA) formulations are unable to prevent tissue recurrence. Studies with 5-ASA formulations show that in the short term the most severe endoscopic lesions can be

703

Filip Baert, Geert D'Haens and Paul Rutgeerts prevented, but the evolution seems only to be somewhat delayed. The problem with 5-ASA treatment may be the delivery of the highest concentrations to the site of the ileocolonic anastomosis. A major drawback of prophylactic studies available at present is the lack of good patient selection at inclusion. We present a summary of the current available evidence (see also Table 4) Sulfasalazine and 5-ASA Prophylactic therapy with sulfasalazine has been shown not to be beneficial in all but one study. By far the largest number of recurrence prevention trials has been performed with mesalamine (5-ASA). This medication is considered safe and well tolerated, but is rather expensive when used chronically in appropriate doses. Postoperative prevention studies with 5-ASA have produced varying results, possibly because of different pharmacological preparations, different doses and time point at the start of the postoperative treatment (from less than 10 days to 8 weeks following surgery). In addition, endpoints of the studies differed in that they used endoscopic, clinical and/or radiologic endpoints as primary criteria. Finally patient selection in the diferent studies may be an important bias. An Italian study by Caprilli and colleagues demonstrated a preventive effect of 2.4 g/day of Asacol, with a significant reduction in the 2-year endoscopic and

clinical recurrence rate. The authors observed a 39% prevention of all recurrences and 55% prevention of severe recurrences. This study, however, was not Winded, and therefore may have been subject to bias. Moreover recurrence rates in the group without treatment were rather low [45]. In another Italian study, by Brignola etal, using 3 g of Pentasa per day, severe endoscopic recurrence was seen in 56%o of patients compared to 24% in the treatment group (/?< 0.008) [46]. A large Canadian multicenter trial with an appropriate placebo group confirmed these results to a certain extent: the overall 3-year symptomatic recurrence was 27% on 1.5 g mesalamine twice daily (Salofalk) started within 8 weeks after curative resection as compared to 47% on placebo. The most important limitation of this study was the time point at which therapy was initiated. Eight weeks after surgery some patients undoubtedly already had microscopic (and endoscopic) signs of Crohn's recurrence. Based on the placebo results in this study one could argue that 53% of the patients received unnecessary treatment for 3 years [47]. Mesalamine in a slow-release formulation (Pentasa) was also compared with placebo more recently in the German-Austrian double-blind ECCDS IV study. Using 4 g Pentasa/day and examining at clinical parameters (Crohn's Disease Activity Index, CDAI) to determine recurrence and all types of resections for CD were included. In this study, after

Table 4. Summary of postoperative recurrence trials: active compound, endpoints and key results study

Product

Start

Endpoints

Result

Caprilli ef a/[45] (/7= 110)

Asacol 2.4 g/day x 2 years

< 15 days

Endo 6,12, 24 months

24 months: 52% vs 85% (p

1 B

00

A -

.2 -

0 200

250

Time (months)

Figure 1. Kaplan-Meier curves of survival free of joint complications in Crohn's disease patients who had never undergone surgery and patients after ileocecal resection (ICR).

an actively inflamed, and therefore abnormally permeable, gut. Whether this interaction is proximal or distal to the ileocecal valve is unclear, and further work is required. In contrast, the axial compHcations, such as AS, seem more likely to be mediated by typical or atypical interactions between HLA class I molecules and the immune system, although intestinal inflammation may make the conditions more favorable for the development of axial inflammation. This would be consistent with the observation that the HLA-B27 association in IBD is weaker than in idiopathic AS, suggesting that the presence of intestinal inflammation may allow the development of joint disease even in patients without the usual genetic predisposition. The pathogenetic mechanisms in type 2 peripheral arthritis are less clear, and although there is a strong HLA class I association this may be secondary to linkage disequilibrium with other genes in the region such as the non-classical class I gene MICA. All the hypotheses described above remain largely speculative, and there are others, including the suggestion that arthritis and other extraintestinal manifestations are an autoimmune reaction to self

antigens such as isoforms of tropomyosin found in the gut, eye and joint [34]. However, these do not account for the HLA associations found, and the evidence has not yet been widely replicated.

Clinical features Diagnosis Axial disease Symptomatic disease Low back pain is a common complaint in both IBD and the general population. In IBD it may represent inflammatory arthritis of the sacroiliac joints (sacroiliitis) or progressive AS. It has been suggested that the course of AS associated with IBD is less severe than idiopathic AS, but this is not universally accepted. In a patient presenting with back pain it is therefore important to distinguish between mechanical low back pain and the inflammatory back pain associated with sacroiliitis and AS. The clinical features of inflammatory low back pain are an insidious onset over months, morning stiffness and exacerbation of pain by rest and pain

752

Articular and ocular complications of inflammatory bowel disease

radiating into the buttocks (rather than central back pain). It tends to occur in patients under 40 years of age. Mechanical back pain may be of sudden onset, is often central and is better after rest, occurring generally in older patients. Plain radiographs of the sacroiliac joints are the conventional means of diagnosis, but X-ray changes occur only after several months and magnetic resonance imaging scanning (with or without gadolinium enhancement) is more sensitive and avoids the necessity of a radiation dose. If radiologic evidence of sacroiliitis is found then a further assessment should be made to detect evidence of progressive axial disease. This should include the modified Schober test of lumbar flexion, lateral lumbar flexion and chest expansion. The most useful blood test in diagnosis is H LABI? status, although in IBD a negative HLA-B27 status does not preclude the diagnosis of AS, and patients with low back pain or decreased spinal mobility in the presence of sacroiliitis should be treated as having AS . A symptomatic disease Up to 20% of IBD patients may have asymptomatic sacroiliitis detectable by plain radiology, and it may be diagnosed on the basis of routine abdominal Xrays. In many cases there is no history of inflammatory back pain even on direct questioning. If sacroilitis is diagnosed a clinical assessment of spinal mobility should be undertaken along with HLA-B27 status. If there is evidence of decreased spinal mobility patients should be treated as having early AS.

Peripheral arthritis (PeA) As mentioned above, there are two distinct forms of PeA, and in addition arthralgia may occur. This is often in conjunction with reducing doses of corticosteroids or the commencement of i m m u n o suppressants such as azathioprine. These normally settle with time and rarely require specific treatment. Both forms of arthritis are rheumatoid factornegative and are not generally erosive or deforming. If there is evidence of erosive disease then further investigation by a rheumatologist is required. Diagnosis of IBD PeA is largely based upon clinical grounds, but other joint disease should be excluded. For type 1 arthritis the differential diagnosis includes gout and pseudogout, septic arthritis and arthritis following genitourinary infections such as

Chlamydia and gonorrhea. These causes should be actively sought, particularly where there is no apparent relation to activity of the bowel disease, and serum urate and calcium should be checked routinely. Joint aspiration and examination should be performed in cases where there is any doubt. For type 2 arthritis a rheumatoid factor and autoantibody screen should be performed to exclude seropositive rheumatic disease. Arthritis associated with IBD is very rarely erosive or deforming and is not associated with other manifestations of rheumatoid arthritis such as subcutaneous nodules. X-rays of the most severely aff'ected joints should therefore be performed if symptoms are persistent, to exclude erosive disease.

Management of IBD arthritis General points The management of arthritis in IBD depends upon the nature and duration of the arthritis, as discussed below. The aims of treatment are to alleviate symptoms and preserve joint function for the future, particularly in axial disease. Because of the relatively small number of patients involved no controlled trials of treatment have been conducted and most treatment modalities rely on general principles. Physical treatments (rest, range of movement exercises and physiotherapy) are important components of management and should not be ignored. These may be enhanced by simple measures such as splinting affected joints and the use of assistive devices such as a walking stick. Analgesia is a key element in management, and is a potentially difficult area, as many analgesics have undesirable effects on the gut. A stepped approach to the prescription of analgesics should be used, starting with simple analgesia such as paracetamol, progressing to stronger analgesics as required. Analgesics containing opioids may cause problematic constipation, particularly in patients with resistant distal colitis. Non-steroidal anti-inflammatory drugs (NSAIDs) should be avoided if at all possible. They may cause enterocolitis in their own right and may trigger or exacerbate relapse of pre-existing IBD. This may be associated with significant lower gastrointestinal bleeding. They should not be used at all in patients with active disease, and in these situations other therapies should be used. The recent advent of cyclo-oxygenase2 (COX-2) specific NSAIDs has

Timothy R. Orchard and Derek P. Jewell

753

raised the prospect of fewer unwanted gastrointestinal problems, particularly in the stomach. However, experiments in animals have demonstrated that the enterocolitis of NSAIDs is not mediated by the cyclo-oxygenase pathway, and that the role of the COX isoforms in inflammation may be quite complex. It is therefore likely that COX-2 specific NSAIDs may not improve the tolerability of these drugs in IBD patients, and they should still be avoided.

The most important part in the successful management of patients with AS and IBD is good communication between patient, gastroenterologist and rheumatologist to maximize the effectiveness of the available therapeutic options. Isolated sacroiliitis should be treated symptomatically with simple analgesia, and regular exercise should be encouraged. However, further treatment should not be required in the absence of decreased spinal mobility or severe pain.

Axial disease

Periplieral joint disease

Patients with AS should be managed in conjunction with a rheumatologist. Physical therapies are of particular importance and all patients should take regular exercise to maintain the mobility of the spine. This may include swimming and spinal exercises with regular physiotherapy if necessary. These forms of exercise should also be recommended to patients with sacroiliitis associated with any form of low back pain or decreased spinal mobility. For patients with severely reduced spinal mobility or severe low back pain injection of steroid into the sacroiliac joints may help, particularly in association with a period of intensive inpatient physiotherapy. The period of relief from sacroiliac injection alone may be brief.

Type 1 arthritis This is usually self-limiting, so treatment is largely symptomatic. Resting the joint is important, and use of a walking stick or splint to take pressure off'the joint may lead to a significant improvement. Range of movement exercises should be performed to minimize any periarticular muscle atrophy and to prevent contractures. In severe cases formal physiotherapy may be required to improve function. Analgesia should be with simple analgesia, and, as type 1 arthritis is usually associated with active bowel disease, NSAIDs should not be used. A good, but relatively seldom used, therapy is intra-articular injection of steroid. This may provide very effective symptom relief and may remove the requirement for other treatments. If oral steroids are used to treat the active bowel disease then these will normally treat the arthritis effectively. If not, an empirical change of 5-ASA drug to sulfasalazine may give good symptom relief. If sulfasalazine fails then low doses of oral steroid specifically for the joint disease may be effective, but this should not be prolonged for more than a few weeks. Long-term treatment is not usually required, although maintenance with sulfasalazine as the 5ASA of choice may be appropriate, particularly in patients at risk of recurrent disease such as those who are HLA-DR103-positive. In the minority of patients with persistent problems the treatment options are those of type 2 arthritis (see below).

Drug treatments Simple analgesics should be used if possible. NSAIDs are the drugs of choice in idiopathic AS, and if there is active spinal disease in the absence of active IBD then it is reasonable to use NSAIDs. However, they should be stopped if the IBD flares up. Sulfasalazine can be used for both the IBD and joint symptoms, but it is most effective in patients with associated peripheral joint problems. Other 5aminosalicylic (5-ASA) drugs are not as effective, as it is thought to be the sulfapyridine component that confers the articular effects rather than the 5-ASA. If oral steroids are required for active disease then they will also have a beneficial effect on the spinal disease. However long-term steroid therapy for spinal disease alone should be avoided. This may be achieved by using methotrexate, which may be effective in both gut and spinal disease. In severe progressive disease unresponsive to the measures outlined above radiotherapy remains a last resort; however, the increased risk of hematologic malignancy associated with this form of treatment means it is rarely used.

Type 2 arthritis These patients generally have persistent problems and may require long-term treatment. Again, as the disease is usually non-erosive and non-deforming symptomatic relief is the major aim. Again, splinting of affected joints and rest are important components of management, but the

754

Articular and ocular complications of inflammatory bowel disease

persistent and polyarthritic nature of this condition makes this harder to achieve than in type 1 arthritis. Simple analgesia should be tried initially. NSAIDs should be considered only in patients with quiescent disease unresponsive to simple or combination analgesia, but if there is any evidence of an increase in the activity of the bowel disease NSAIDs should be stopped. For patients with persisting problems sulfasalazine, or low-dose prednisolone, may be used. However, prolonged courses of oral steroids should be avoided. In patients with active bowel disease it may be appropriate to use methotrexate as the first-line immunosuppressant rather than azathioprine, in order to treat both gut and joints. Patients who show evidence of erosive joint disease, a positive rheumatoid factor, or who do not respond to the measures outlined above, should be managed jointly with a rheumatologist.

on the population and methodology. Various forms have been described including iritis, episcleritis, scleritis and anterior uveitis. In our retrospective study of 1459 patients (976 UC and 483 CD patients) 3% of UC and 5% of CD patients had eye complications. The commonest were iritis (60%), episcleritis (30%) and uveitis (10%). However, in the vast majority of cases the inflammation caused no lasting ocular damage [39]. The female:male ratio was 2.9:1, the eye complications were present at or before diagnosis in 19% of cases, and in 72% of cases the eye complications were associated with active IBD. Thirty percent of patients went on to have recurrent episodes of ocular inflammation. In common with other studies patients who suffered eye complications were more likely to suffer other mucocutaneous complications, notably arthritis and erythema nodosum.

New therapies for IBD arthritis The major possible new therapy for IBD-associated arthritis is the use of infliximab. This is a monoclonal antibody to the proinflammatory cytokine Tumor necrosis factor alpha. This drug is hcensed for use in resistant CD and rheumatoid arthritis, where it has been demonstrated to be effective in improving symptoms and decreasing inflammation. Small open-label trials in spondyloarthropathy have demonstrated that it is again effective in reducing inflammation and symptoms [35, 36]. In IBD-associated arthritis the published experience is limited to case reports but, unsurprisingly here also a reduction in inflamed joints and inflammatory markers has been reported [37]. Thus infliximab may well be an alternative treatment in patients with severe resistant joint disease associated with IBD. As in other rheumatic diseases this may be combined with other immunosuppressants such as methotrexate [38]. However, further studies are required to confirm its efficacy and the best treatment regimens.

Ocular manifestations of IBD Epidemiology and clinical features Ocular inflammation in IBD was first documented by Crohn in 1925. If left untreated it is potentially a cause of blindness; however, prompt treatment with topical steroids can minimize this risk. The prevalence of ocular inflammation varies widely between studies - from 2% to 13% depending

Pathogenesis The cHnical observation that ocular inflammation occurs more commonly in those patients with other extraintestinal manifestations has led to several hypotheses to explain it. The first suggested that the eye, the joints, the skin and the liver expressed a particular isoform of tropomyosin, to which an autoimmune reaction was generated [40]. However, although this antigen was found in the target organs it was not found in the components that became inflamed. Thus it was expressed in chondrocytes but not synovium, and in ciliary muscles but not the iris. In addition this hypothesis fails to explain why in some patients ocular inflammation may occur as a single manifestation and in others as part of a complex of extraintestinal manifestations (EIM). This phenomenon of the EIM being distinct but overlapping clinical entities may be explained by a genetic-based hypothesis. This suggests that the presence of the diff'erent EIM is determined by different genes located in the same region. Linkage disequilibrium between these genes would mean that the genes for different manifestations are inherited together more frequently than would be expected by chance, leading to the clinical clustering of the EIM. Given the HLA associations described above we have studied the HLA-B and DR regions in 52 patients with ocular complications of IBD. This study found that there are associations between ocular complications and HLAB27 (40% vs 9% in

Timothy R. Orchard and Derek P. Jewell IBDcontrols,/?-11\. © 2003 Kluwer Academic Publishers. Printed in Great Britain

764

Fertility and pregnancy in inflammatory bowel disease

the general population. In contrast, fecundity (the biological ability to conceive) appears to be substantially reduced in women with UC who have undergone restorative proctocolectomy, probably due to the development of pelvic adhesions [6]. Evidence shows fertility rates are decreased in females with CD. Early series suggested 32-53% of women with CD were infertile, especially those with Crohn's colitis [9, 10]. More recent surveys from Oxford and northeast Scotland recorded an involuntary infertility rate of 12-14% [8, 11]. An extensive case-controlled study from five different European countries showed a significant reduction in the proportion of women with CD who became pregnant and a reduced number of offspring in this group [12]. There are several factors which may explain why women with CD have fewer pregnancies. Among these, the most important appears to be a personal preference to avoid pregnancy [13]. In some cases women are discouraged from becoming pregnant by clinicians who focus excessively on the possible negative effects of drug therapy during this time [14]. Reduced sexual activity is more common among women with CD, especially in those with perianal inflammation, because of dyspareunia, abdominal pain, diarrhea and fear of incontinence [15]. It is well recognized that suffasalazine may cause reversible infertility in men by reducing sperm motility and function [16]. This effect is usually transient, and not encountered with other 5-aminosalicylic agents. It has been presumed that this therapy accounts for the reduction in fertility observed among men with IBD. However, a study from the United Kingdom found that males with CD (but not UC) had fewer children, an effect that was independent of sulfasalazine usage [17]. The possible reasons for this finding are not explained, but a previous report has highlighted that long-term effects of disease activity or malnutrition may lead to oligospermia in some males with CD [18].

women with IBD [26, 27]. A population-based study of women who had single births in Sweden during a 12-month period showed an increased rate of Cesarian section, prematurity and low-birthweight infants among IBD patients compared to healthy controls [26]. Details regarding underlying disease diagnosis and degree of disease activity were not reported in this study. There is no evidence from either uncontrolled or case-controlled studies that neonatal complications are increased in UC [5-7, 19-24, 28]. In contrast, less favorable neonatal outcomes have been observed in CD [12-13, 28,29]. A report from Denmark during a 10-year period showed a higher risk of prematurity and low-birthweight babies among women with CD compared to controls [29]. These findings support the findings of two earlier case-controlled studies in which an increased risk of prematurity was observed among patients with CD [12, 13]. There is considerable evidence linking disease activity, especially at the time of conception, to a higher frequency of fetal complications in IBD. In UC, women with quiescent disease at conception have higher rates of normal live births, and reduced rates of prematurity and spontaneous abortion compared to those with active disease at conception [22]. An increased rate of malformations has been described among medically treated women with active, but not inactive, UC [7]. In CD 80% of patients with quiescent disease at conception had a normal, uncomplicated live birth, compared to 50% of those with active disease [30]. Prematurity and spontaneous abortion are significantly increased in women with active disease at conception [25]. In women with active CD at conception, neonatal outcome is more favorable when inflammation improves during the remainder of pregnancy compared to when it is persistent [30].

Disease progress in pregnancy Neonatal outcome According to the cumulative experience derived from several uncontrolled studies evaluating neonatal outcome in pregnant women with UC and CD, the chances of delivering a normal, full-term, live infant appear to be generally good [5-11, 19-25]. However, two case-controlled surveys have observed higher rates of neonatal complications among

In general the course of UC during pregnancy depends on the degree of intestinal inflammation at conception. According to the results of several uncontrolled studies evaluating pregnant patients with quiescent colitis at conception, approximately two-thirds remain in remission and one-third relapse. This proportion is similar to expected rates in the non-pregnant UC population. However, if colitis is active at conception, one-third improve, one-third remain persistently active and another

William Connell third deteriorate. In other words, if UC is active at conception, two-thirds continue to have ongoing, persistent inflammation throughout pregnancy [5-7, 19-20, 22]. These symptoms can be particularly troublesome to control, thereby requiring vigorous medical therapy. Data from Israel raised the possibility that relapses were more c o m m o n when pregnancy was unplanned [23]. Disease activity at conception also determines the course of CD during pregnancy. Uncontrolled studies of patients with quiescent CD at conception indicate that 75% remain in remission. In contrast, disease that is active at conception is likely to remain in this state, or actually deteriorate during pregnancy [9-11, 25]. Furthermore, the more severe the disease at conception, the more likely it is to remain active during pregnancy [30]. Whereas early studies suggested that post-partum flare-ups in CD were common [9, 10], recent data suggest this is not the case [8, 11, 25]. The reason for such a discrepancy is not known, but possibly relates to the past practice of elective withdrawal of sulfasalazine late in pregnancy due to concerns regarding the risk of neonatal kernicterus developing from sulfur-containing drugs [31]. In the light of further evidence showing this actual risk to be negligible, nowadays the drug is continued throughout this time [32]. The onset of UC in pregnancy or the puerperal period is not uncommon. In a study of 147 women with UC from Oxford, 8% experienced its onset during pregnancy and 4% post natally. Most cases were mild to moderate, and generally brought under control. Neonatal outcome was no different than if the disease preceded pregnancy [7]. The onset of CD appears relatively rare in pregnancy, possibly because of a natural reluctance to investigate pregnant women with pain and diarrhea. In the past the prognosis of such an event was thought to be poor [33], but data collected subsequently are more encouraging [8, 11, 25, 30].

Mode of delivery For patients with quiescent colitis the risk of perineal sepsis following childbirth is low, and the mode of delivery is governed by obstetric indications [28]. It is unknown if patients with more active colitis are at higher risk of perineal complications. To avoid sphincter damage, Caeserian section has been advised for women with active, relapsing or fulmi-

765 nant disease, where the future prospect of ileoanal pelvic pouch surgery is high [34]. Information concerning the perianal consequences of vaginal delivery and episiotomy in patients with CD are conflicting. A highly selected survey of 117 female members of the Crohn's and Colitis Foundation of America who were free of perianal disease prior to pregnancy found 18% reported the development of complications in this region following vaginal delivery, most of which occurred within the first 2 months [35]. In contrast, a larger population study from Canada showed vaginal delivery rarely initiated the development of perianal disease in women without a previous history of this problem, and did not evoke a deterioration of symptoms in those with pre-existing inactive perianal disease. However, one-quarter of patients with active perianal disease at birth experienced an aggravation of their condition after vaginal delivery [36]. Accordingly, Ceserian section seems justified for women with active perianal CD, but not when perianal disease is absent or inactive prior to birth. It is important for obstetricians to exercise excellent care of the perineum if vaginal delivery is to be undertaken, and when episiotomy is required a mediolateral incision is preferable to reduce the risk of rectal tear [35]. Patients who have previously undergone total colectomy and iloestomy have an increased risk of stomal complications during pregnancy. Chnically, this may result in stomal dysfunction, prolapse, skin irritation, leakage, and sometimes intestinal obstruction [37]. Vaginal delivery may be undertaken in most of these patients without compromising maternal or neonatal outcome. Patients who have previously undergone ileoanal pelvic pouch surgery for UC may experience an aggravation of stool frequency and incontinence during pregnancy. These effects are usually transient, and return to normal after delivery. Cesarian section is often recommended, though this practice is necessary only in patients with a scarred, non-compliant perineum, unless other obstetric reasons apply. For the majority of women with pelvic pouches, vaginal delivery and episotomy is well tolerated and rarely associated with the development of a rectal tear [38].

Investigations Generally, investigations should be minimized during pregnancy and undertaken only when there

766

Fertility and pregnancy in inflammatory bowel disease

is a very good clinical reason to do so. Provided it is done with exquisite care, flexible sigmoidoscopy and biopsy may be performed during pregnancy. In contrast, more extensive evaluation by colonoscopy should be reserved only for exceptional cases where its u n d e r t a k i n g will influence m a n a g e m e n t decisions; in which case fetal monitoring should be performed [39]. Investigations containing ionizing radiation should be avoided. Where necessary, ultrasound or magnetic resonance image scanning may be used to provide information regarding abscess formation or to assess bowel wall thickness. If obstruction or incipient toxic dilation is suspected, it is reasonable to perform a single plain abdominal X-ray. The risk of fetal toxicity from this investigation is extremely small, and outweighed by the consequences of these potentially serious maternal complications [40].

Published data among IBD patients using other treatments in pregnancy are limited. Information regarding the safety of these drugs is mainly derived from the results of animal studies or clinical experience among patients with other conditions, such as organ transplantation or connective tissue disorders. A summary of the effects of commonly used drugs for IBD during pregnancy and lacation is shown in Table I.

Surgery The indications for surgery in IBD during pregnancy are similar to those in the non-pregnant state, namely severe hemorrhage, perforation, obstruction or severe disease unresponsive to medical therapy. The risks to mother and baby increase when surgical intervention is unduly delayed. Fetal morbidity is probably related more to the effects of underlying disease rather than the operation itself, since bowel surgery in pregnancy for reasons other than IBD has a relatively low risk to fetus. Although maternal mortality from emergency surgery for fulminant UC during pregnancy is declining, the risk of fetal death and social morbidity among parents thereafter is high [41].

Drug therapy In view of the evidence associating a more favorable neonatal outcome with inactive maternal IBD, pharmacotherapy is often required to control disease activity before conception and to maintain remission or treat flare-ups during pregnancy. Although drugs may be associated with side-effects, the risk to the fetus is greater from uncontrolled disease inflammation [24]. The choice of agents depends on the severity of the disease, potential for side-effects, and the preference of the individual patient. Clinical studies evaluating drug treatment in pregnant women with IBD are largely limited to 5aminosalicylic acid drugs and corticosteroids.

5-Aminosaliylic acid Sulfasalazine and other 5-aminosalicylic formulations appear to be safe in pregnancy when used in recommended doses [42, 43]. The rates of spontaneous abortion, premature delivery, or congenital malformations for women with either UC or CD taking sulfasalazine are not increased [7, 11, 21, 22, 24]. A recent study showed that men with IBD who fathered a congenitally abnormal child were more likely to have used sulfasalazine [17]. Clinical series have demonstrated the overall safety in pregnancy of other 5-aminosalicylic formulations when used in low to moderate doses [43, 44]. A single case of renal insufficiency was reported in a newborn exposed to 4 g/day Pentasa during the mid-trimester [45]. Both sulfasalazine 5-aminosalicylic agents may be used during breast feeding, though isolated cases of transient acute diarrhea in the newborn have been reported [46, 47]. Concentrations of sulfasalazine and sulfapyridine in breast milk are insufficient to cause kernicterus [32].

Corticosteroids Controlled and uncontrolled studies in patients with IBD, asthma and connective tissue disorders have shown no overall increase in the rate of any malformations nor fetal morbidity among babies exposed to steroids during pregnancy [21, 42, 48, 49]. Animal studies indicate an increased risk of oral cleft among offspring exposed to high doses of corticosteroids [50], and a report from the 1950s suggested this complication may also occur in humans [51]. A controlled study of 468 pregnant women treated with corticosteroids during pregnancy observed two cases of cleft palate in newborns, compared to 0.2 expected - numbers that were too small to be meaningful [52]. It can be presumed that any possible risk of cleft lip or palate from steroids is likely to be small, and one that must be weighed up

161

William Connell Table 1. Effects of drugs used for IBD jn pregnancy and lacation Drug

Pregnancy

Breast feeding

Sulfasalazine

Risk of miscarriage, prematurity or malformations not increased. Folate supplements recommended

Safe - reduce dose if neonatal jaundice is present

5-Aminosalicylic acid

Probably safe in doses < 3 g/day; higher doses may be nephrotoxic

Safe - occasional reports of reversible neonatal diarrhea

Corticosteroids

Probably safe - conflicting evidence regarding risk of oral cleft

Safe

Azathioprine

Malformations do not appear to be increased, but possible. Small risk of prematurity, low birthweight, fetal myelotoxicity, immunosuppression, transient chromosomal aberrations. Potential toxicity must be discussed with parents in advance and used only when clinically necessary

Not recommended due to risk of neonatal myelotoxicity or immunosuppression

6-Mercaptopurine

Teratogenic in animals at high doses. Human data sparse, but so far reassuring. Risks probably similar to azathioprine

Not recommended due to risk of neonatal myelotoxicity or immunosuppression

Methotrexate

Contraindicated due to risk of congenital abnormalities and abortion

Contraindicated

Cyclosporine

Possibly associated with prematurity and low birthweight

Not recommended

Metronidazole

Safe in short courses ( < 10 days) but not recommended in high doses for prolonged periods

Not recommended

Ciprofloxacin

Defects in developing cartilage in animals. Short courses in human probably safe but not recommended for prolonged periods

Not recommended

against the drug's known beneficial effects in appropriate clinical circumstances. Women requiring prednisolone should not be discouraged from lactation [53]. There are no pubhshed data regarding the use of controlled-release budesonide preparations during pregnancy.

Azathioprine/6-mercaptopurine Of all the various drug options to treat IBD, the safety of azathioprine and 6-mercaptopurine during pregnancy is the most controversial. Historically, pregnant women were cautioned against using these agents, due to early concerns about teratogenesis [54]. However, further data during the past two decades from patients with organ transplantation or connective tissue diseases have shown azathioprine

to be generally well tolerated [55-58]. Occasional malformations have been reported, but the overall frequency is no different compared to observed rates in the general population, and no characteristic pattern of abnormality is evident [57, 58]. It is possible that the fetal immune system is transiently compromised from azathioprine exposure during the second and third trimesters, and this may result in growth retardation [59]. Cases of prematurity, neonatal myelotoxicity, thymic atrophy and transient chromosomal abnormalities have also been reported with azathioprine usage in pregnancy [58, 60]. Importantly, long-term follow-up effects among children who were exposed to the drug antenatally are lacking. One study in IBD observed no congenital abnormalities or subsequent health problems in 16 children born to mothers treated with

768

Fertility and pregnancy in inflammatory bowel disease

azathioprine during pregnancy [61]. Provided the drug is used in recommended doses during pregnancy, the risks associated with azathioprine appear to be small. Because potential complications may be serious, its use should be reserved for refractory maternal disease only. Little clinical information exists concerning 6mercaptopurine usage in pregnancy, leading to uncertainty about its safety [49, 62]. In animals, 6mercaptopurine is teratogenic and long-term effects have been described on germ cells of exposed offspring [63]. A study among patients with IBD observed rates of prematurity and significant congenital malformation to be 3% and 5% respectively [62]. Until more data are available the use of this drug in pregnancy will remain controversial. Increasing attention has been drawn to the possible effects of azathioprine or 6-mercaptopurine on spermatogenesis. Among 13 children born to fathers receiving 6-mercaptopurine for IBD at conception, the rate of congenital malformations and spontaneous abortions was higher than in a control group of men who did not receive the drug [64]. In contrast, 58 of 60 babies born to male transplant recipients treated with azathioprine at conception were normal [65]. A recent study showed that semen quality was not affected by azathioprine, and in a small number of children born to fathers treated with azathioprine at conception, fetal outcome and post natal development were normal [66]. At present, there is insufficient evidence to discourage the use of either azathioprine or 6-mercaptopurine in men contemplating fatherhood. Azathioprine is transferred to breast milk in small quantities, and lactating babies may be at slight risk of immunosuppression and myelotoxicity [67]. Its use is not generally recommended during breastfeeding, but a decision regarding this should take into account the benefits of breast milk, as well as the personal preference of the mother [58].

Fetal and maternal outcomes were satisfactory according to a single case report when cyclosporine was administered during the second trimester for severe UC [69]. Due to concerns regarding possible nephrotoxicity or immunosuppression, cyclosporine has not been recommended during lactation. However, a recent report showed concentrations of the drug in breast milk to be low, and breast-fed babies did not develop any adverse effects [70].

Cyclosporine Occasionally, cyclosporine may be justified in pregnancy for patients with severe UC who would otherwise require emergency surgery. Two large series of transplant recipients receiving cyclosporine during pregnancy observed rates of congenital malformations that were similar to that expected in the general population [55, 68]. However, the drug has been associated with growth retardation and prematurity in patients with organ transplants [58].

Methotrexate The elective use of methotrexate is contraindicated in pregnancy because of its embyrotoxic and teratogenic effect [49]. The critical period of exposure is between 6 and 8 weeks of gestation, and the harmful effects on the fetus may be dose-related [71 ]. Its use in pregnancy can also be associated with fetal growth retardation, neonatal bone marrow suppression and possible chromosomal aberration [72]. Methotrexate may temporarily impair male fertility, and its effects can persist for several months after drug withdrawal. Accordingly, men and women are advised to discontinue methotrexate for several months before attempting conception [49]. Methotrexate is excreted into breast milk in small amounts, and is contraindicated during breast-feeding.

Antibiotics Metronidazole and ciprofloxacin are increasingly used as primary therapy for CD. According to two large meta-analyses, treatment with metronidazole for less than 10 days during the first trimester was not associated with an increased risk of malformations [73, 74]. Moreover, this therapy does not seem to be associated with an increased frequency of stillbirths, growth retardation or prematurity [75]. Ciproflaxacin has been associated with the development of arthropathy in the offspring of animals exposed to the drug antenatally [76]. However, a prospective study in humans demonstrated no increased risk of malformations or musculoskeletal problems in 38 women receiving either ciprofloxacin or norfloxacin during the first trimester for urinary tract infection [77]. Nonetheless, there are no published data concerning the safety of either drug in IBD, especially when given at high doses for a prolonged period. Because the effects on breast-fed babies are uncertain, their use in lactation is not recommended.

William Connell

Infliximab There is little published information regarding the safety of infliximab on the developing fetus or breastfed infant. Mutagenesis was not observed in preclinical studies with infliximab, and successful live births have occurred following its use in pregnancy [78]. However, the drug should not be used electively during pregnancy or lactation until more data are available.

769

References 1.

2.

3.

4.

Does pregnancy alter the natural history of IBD? The future outcome of pregnancy in IBD cannot be predicted on the basis of what happened in a previous pregnancy [7]. However, it is possible that parity may improve the long-term outcome of IBD. A study of patients with CD showed the mean number of resections was lower and the interval between resections was longer in parous women than nonparous individuals [79]. A small study from Italy found rates of relapse from IBD to be reduced 3-4 years after delivery compared to the corresponding period prior to pregnancy [80].

5. 6.

7. 8. 9. 10. 11. 12.

Summary Most women with IBD can expect an uneventful pregnancy. Evidence compiled over many years indicates the importance of controlling disease activity at the time of conception and during pregnancy to optimize the outcome for mother and baby. For this reason drug therapy is often required to treat flareups and maintain remission during this time. Where possible, women with IBD should plan a pregnancy to coincide with times when the disease is quiescent, and drug treatment minimal. In practice, however, many pregnancies are unexpected or occur in women with active disease that requires intensive medical therapy. Provided adequate information has been supplied in advance of conception, much of the apprehension that accompanies these events can be substantially reduced. Ongoing supervision by the clinician during pregnancy and the puerperium provides invaluable support, and enables maternal disease flare-ups to be identified and treated as expediently as possible.

13. 14. 15. 16. 17.

18. 19. 20. 21.

22.

Peeters M, Nevens H, Baert F et al. Familial aggregation in Crohn's disease: increased age-adjusted risk and concordance in clinical characteristics. Gastroenterology 1996; 111:597-603. Yang H, McElree C, Roth MP, Shanahan F, Targan SR, Rotter JI. FamiHal empirical risks for inflammatory bowel disease: differences between Jews and non-Jews. Gut 1993; 34:517-24. Akolkar PN, Gulwani-Akolkar B, Heresbach D et al. Differences in risk of Crohn's disease in off'spring of mothers and fathers with inflammatory bowel disease. Am J Gastroenterol 1997; 92: 2241-4. Bennett RA, Rubin PH, Present DH. Frequency of inflammatory bowel disease in off"spring of couples both presenting with inflammatory bowel disease. Gastroenterology 1991; 100: 1638-43. de Dombal FT, Watts JM, V^atkinson G, Goligher JC. Ulcerative colitis and pregnancy. Lancet 1965; 2: 599-602. Olsen KO, Juul S, Berndtsson I, Oresland T, Laurberg S. Ulcerative cohtis: female fecundity before diagnosis, during disease and after surgery compared with a population sample. Gastroenterology 2002; 122: 15-19. Willoughby CP, Truelove SC. Ulcerative colitis and pregnancy. Gut 1980; 21; 469-74. Hudson M, Flett G, Sinclair TS, Brunt PW, Templeton A, Mowat NAG. Fertility and pregnancy in inflammatory bowel disease. Int J Gynecol Obstet 1997; 58: 229-37. Fielding JF, Cooke WT. Pregnancy and Crohn's disease. Br Med J 1970; 2: 76-7. De Dombal FT, Burton IL, Goligher JC. Crohn's disease and pregnancy. Br Med J 1972; 3: 550-3. Khosla R, Willoughby CP, Jewell DP. Crohn's disease and pregnancy. Gut 1984; 25: 52-6. Mayberry JF, Weterman IT. European survey of fertility and pregnancy in women with Crohn's disease: a case control study by European collaborative group. Gut 1986; 27: 8215. Baird DD, Narendranathan M, Sandler RS. Increased risk of preterm birth for women with inflammatory bowel disease. Gastroenterology 1990; 99: 987-94. Korelitz BI. Inflammatory bowel disease and pregnancy. Gastroenterol Clin N Am 1998; 27: 213-24. Moody GA, Probert CSJ, Srivastava EM, Rhodes J, Mayberry JF. Sexual dysfunction amongst women with Crohn's disease: a hidden problem. Digestion 1992; 52: 179-83. O'Morain C, Smethurst P, Dore CJ, Levi AJ. Reversible male infertility due to sulphasalazine: studies in men and rats. Gut 1984; 25: 1078-84. Moody GA, Probert C, Jayanthi V, Mayberry JF. The effects of chronic ill health and treatment with sulphasalazine on fertility amongst men and women with inflammatory bowel disease in Leicestershire. Int J Colorectal Dis 1997; 12: 2204. Farthing MJG, Dawson AM. Impaired semen quality in Crohn's disease - drugs, ill health or undernutrition? Scand J Gastroenterol, 1983; 18; 57-60. MacDougall I. Ulcerative colitis and pregnancy. Lancet 1956; 2: 641-3. McEwan HP. Ulcerative colitis in pregnancy. Proc R Soc Med. 1972;65:279-81. Mogadam M, Dobbins WO, Korelitz BI, Ahmed SW. Pregnancy in inflammatory bowel disease: effect of sulfasalazine and corticosteroids on fetal outcome. Gastroenterology 1981; 80: 72-6. Nielsen OH, Andreasson B, Bondesen S, Jarnum S. Pregnancy in ulcerative colitis. Scand J Gastroenterol 1983; 18: 735-42.

Fertility and pregnancy in inflammatory bowel disease

770 23. 24. 25. 26. 27. 28. 29.

30. 31.

32.

33. 34. 35.

36. 37. 38.

39.

40. 41. 42. 43.

Levy N, Roisman I, Teodor I. Ulcerative colitis in pregnancy in Israel. Dis Colon Rectum, 1981; 24: 351-54. Baiocco PJ, Korelitz BI. The influence of inflammatory bowel disease and its treatment on pregnancy and fetal outcome. J Clin Gastroenterol 1984; 6: 211-16. Nielsen OH, Andreasson B, Bondesen S, Jacobsen O, Jarnum S. Pregnancy in Crohn's disease. Scand J Gastroenterol 1984; 19: 724-32. Kornfeld D, Cnattingius S, Ekbom A. Pregnancy outcomes in women with inflammatory bowel disease - a populationbased cohort study. Am J Obstet Gynecol 1997; 177: 942-6. Fedorkow DM, Persaud D, Nimrod CA. Inflammatory bowel disease: a controlled study of late pregnancy outcome. Am J Obstet Gynecol 1989; 160: 998-1001. Porter RJ, Stirrat GM. The effects of inflammatory bowel disease on pregnancy: a case-controlled retrospective analysis. Br J Obstet Gynaecol 1986; 93: 1124 31. Fonager K, Sorensen HT, Olsen J, Dahlerup JF, Rasmussen SN. Pregnancy outcome for women with Crohn's disease: a follow-up study based on linkage between national registries. Am J Gastroenterol 1998; 93: 2426 30. Woolfson K, Cohen Z, McLcod RS. Crohn's disease and pregnancy. Dis Colon Rectum 1990; 33: 869 73. Silvcrmann WA, Andersen DH, Blanc WA, Crozier DN. A difference in mortality rate and incidence of kernictcrus among infants allotted to two prophylactic antibacterial regimens. Pediatrics 1956; 18: 614 24. Esbjorner E, Jarnerot G, Wranne L. Sulphasalazine and sulphapyridinc serum levels in children to mothers treated with sulphasalazine during pregnancy and lactation. Acta Pacdiatr Scand 1987; 76: 137 42. Jarnerot G Fertility, sterility, and pregnancy in chronic inflammatory bowel disease. Scand J Gastroenterol 1982; 17: 1 4. Kelly MJ, Hunt TM, Wicks ACB, Mayne CJ. Fulminant ulcerative colitis and parturition: a need to alter current management? Br J Obstet Gynaecol 1994; 101: 166 7. Brandt LJ, Estabrook SG, Reinus JF. Results of a survey to evaluate whether vaginal delivery and episiotomy lead to perineal involvement in women with Crohn's disease. Am J Gastroenterol 1995; 90: 1918 22. Ilnyckyji A, Blanchard JF, Rawsthorne P, Bernstein CN. Perianal Crohn's disease and pregnancy; the role of the mode of delivery. Am J Gastroenterol 1999; 94: 3274-8. Gopal KA, Amshel AL, Shonberg IL, Levinson BA, VanWert M, VanWert J. Ostomy and pregnancy. Dis Colon Rectum 1985:28:912-16. Juhasz ES, Fozard B, Dozois RR, Ilstrup DM, Nelson H. Ileal pouch-anal anastomosis function following childbirth: an extended evaluation. Dis Colon Rectum 1995; 38: 15965. Cappell MS, Colon VJ, Sidhom OA. A study at 10 medical centers of the safety and efficacy of 48 flexible sigmoidoscopies and 8 colonoscopies during pregnancy with follow-up of fetal outcomes and with comparison to control groups. DigDisSci 1996;41:2353-61. Subhani JM, Hamiliton MI. Review article: the management of inflammatory bowel disease during pregnancy. Ahmen Pharmacol Ther 1998; 12: 1039-53. Anderson JB, Turner GM, Williamson RCN. Fulminant ulcerative colitis in late pregnancy and the puerperium. J R SocMed, 1987;80:492-4. Donaldson RM. Management of medical problems in pregnancy - inflammatory bowel disease. N Engl J Med 1985;312: 1616-9. Marteau P, Tennenbaum R, Elefant E, Lemann M, Cosnes J. Foetal outcome in women with inflammatory bowel disease treated during pregnancy with oral mesalazine microgranules. Alimen Pharmacol Ther 1998; 12: 1101-8.

44. 45. 46.

47. 48. 49. 50. 51. 52. 53. 54.

55.

56.

57. 58. 59. 60. 61. 62. 63. 64.

65. 66. 67.

Diav-Citrin O, Park YH, Veerasuntharam G et al. The safety of mesalamine in human pregnancy: a prospective controlled cohort study. Gastroenterology 1998; 114: 23-8. Colombel J, Brabant G, Gubler MC et al. Renal insufficiency in infant: side-effect of prenatal exposure to mesalazine? Lancet 1994; 344: 620-1. Branski D, Kerem E, Gross-Kieselstein E et al. Bloody diarrhea possible complication of sulfasalazine transferred through human breast milk. J Pediatr Gastroenterol Nutr 1986;5:316-17. Nelis GF. Diarrhoea due to 5-aminosalicylic acid in breast milk. Lancet 1989; 1: 383. Schatz M. Asthma treatment during pregnancy. What can be safely taken? Drug Safety 1997; 16: 342-50. Bermas BL, Hill JA. Effects of immunosuppressive drugs during pregnancy. Arthritis Rheum 1995; 38: 1722-32. Pinski L, DiGeorge AM. Cleft palate in the mouse: a teratogenic index of glucocorticoid potency. Science 1965; 147:402 3. Harris JWS, Ross IP. Cortisone therapy in early pregnancy: relation to cleft palate. Lancet 1956; 1: 1045 7. Eraser FC, Sajoo A. Teratogenic potential of corticosteroids in humans. Teratology 1995; 5: 45 6. Ost L, Wettrell G, Bjorkhem I ct al. Prednisolone excretion into breast milk. J Pediatr 1985; 106: 1008 11. Registration Committee of the European Dialysis and Transplant Association. Successful pregnancies in women treated by dialysis and kidney transplantation. Br J Obstet Gynaecol 1980; 87: 839 45. Armcnti VT, Ahlswede KM, Ahlswedc BA et al. National transplantation pregnancy registry: outcomes of 154 pregnancies in cyclosporin-treated female kidney transplant recipients. Transplantation 1994; 57: 502 6. Ramsay-Goldman R, Mientus JM, Kutzer JE et al. Pregnancy outcome in women with systemic lupus erythematosus treated with immunosuppressive drugs. J Rheum 1993; 20: 1152 7. Ramsey-Goldman R, Schilling E. Immunosuppressive drug use during pregnancy. Rheum Dis Clin N Am 1997; 23: 149 67. Armenti VT, Moritz MJ, Davison JM, Drug safety issues in pregnancy following transplantation and immunosuppression. Drug Safety 1998; 19: 219-32. Little BB. Immunosuppressive therapy during gestation. Semin Perinatol 1997; 21: 143-8. Leb DE, Weisskopf B, Kanovitz BS. Chromosome aberrations in the child of a kidney transplant recipient. Arch InernMed 1971; 128:441-4. Alstead EM, Ritchie JK, Lennard-Jones JE et al. Safety of azathioprine in pregnancy in inflammatory bowel disease. Gastroenterology 1990; 99: 443-6. Marion JF. Toxicity of 6-mercaptopurine/azathioprine in patients with inflammatory bowel disease. Inflam Bowel Dis 1998; 2: 116-7. Reimers TJ, Sluss PM. 6-Mercaptopurine treatment of pregnant mice; eflects on second and third generations. Science 1978; 201; 65-7. Rajapaske RO, Korelitz BI, Zlatanic J, Baiocco PJ, Gleim GW. Outcome of pregnancies when fathers are treated with 6-mercaptopurine for inflammatory bowel disease. Am J Gastroenterol 2000; 95: 684-8. Penn I. Parenthood following renal transplantation. Kidney Int1980; 18:221-33. Dejaco C, Mittermaier C, Reinisch W et al. Azathioprine treatment and male fertility in inflammatory bowel disease. Gastroenterology 2001; 121: 1048-53. Grekas DM, Vasiliou SS, Lazarides AN. Immunosuppressive therapy and breast-feeding after renal transplantation. Nephron 1984; 37: 68

111

William Connell 68. 69. 70. 71. 72. 73. 74.

Lamarque V, Leleu MF, Monka C et al. Analysis of 629 outcomes in transplant recipients treated with Sandimmun. Transplant Proc 1997; 29: 2480 Bertschinger P, Himmelmann A, Risti B, Follath F. Cyclosporine treatment of severe ulcerative colitis during pregnancy. Am J Gastroenterol 1995; 90: 330 Nyberg G, Haljamae U, Frisenette-Fich C, Wennergren M, Kjellmer I. Breast-feeding during treatment with cyclosporine. Transplantation 1998; 65; 253-5. Feldkamp M, Carey JC. Clinical teratology counselling and consultation case report: low dose methotrexate exposure in the early weeks of pregnancy. Teratology 1993; 47: 533-9. Connell W, Miller A. Treating inflammatory bowel disease during pregnancy. Risks and safety of drug therapy. Drug Safety 1999:21:311-23. Burtin P, Taddio A, Ariburnu O et al. Safety of metronidazole in pregnancy: a meta-analysis. Am J Obstet Gynecol 1995; 172: 525-9. Caro-Paton T, Carvajal A, de Diego IM et al. Is metronidazole teratogenic? A meta-analysis. Br J Clin Pharmacol 1997; 44: 179-82.

75.

Dobias L, Cerna M, Rossner P et al. Genotoxicity and carcinogenicity of metronidazole. Mutat Res 1994; 317: 177-94. 76. Mayer DG. Overview of toxicological studies. Drugs 1987; 34(Suppl. 1): 150-3. 77. Berkovitch M, Pastuszak A, Gazarian M et al Safety of the new quinolones in pregnancy. Obstet Gynecol 1994; 84: 535-8. 78. Hommes DW, van de Heisteeg BH, van der Spek M, Bartelsman JFWM, van Deventer SJH. Infliximab treatment for Crohn's disease: one year experience in a Dutch academic hospital. Inflam Bowel Dis 2002; 8: 81-6. 79. Nwokolo CU, Tan WC, Andrews HA, Allan RN. Surgical resections in parous patients with distal ileal and colonic Crohn's disease. Gut 1994; 35: 220-3. 80. Castiglione F, Pignata S, Morace F et al. Effect of pregnancy on the clinical course of a cohort of women with inflammatory bowel disease. Ital J Gastroenterol 1996; 28: 199-204.

42 Special considerations in the diagnosis and management of inflammatory bowei disease in the pediatric age group ERNEST G. SEIDMAN AND ARLENE CAPLAN

Introduction The diagnosis and management of inflammatory bowel disease (IBD) in children and adolescents is a challenge to patients and their families as well as to members of the health-care team. There are many similarities in terms of the clinical features and therapeutic options in IBD, irrespective of the patient's age. Thus, much of what has been included in other chapters pertains to IBD in the pediatric age group, and is not repeated here. However, IBD often occurs at a particularly vulnerable period of childhood and adolescence, with potentially adverse effects on growth, quality of life and psychosocial functioning. This chapter focuses on specific issues in pediatric IBD, dealing with those problems that are unique to children and adolescents, such as growth failure and pubertal delay. It also emphasizes certain clinical dilemmas that are particularly important in this age group, including an approach to dealing with diagnostic uncertainty in the child with recurrent abdominal pain, therapeutics and adherence issues, as well as psychosocial problems.

Epidemiologic aspects Crohn's disease (CD) and ulcerative colitis (UC) are the most common chronic inflammatory disorders of children and adolescents in North America and most of Europe. An almost 7-fold increase in the incidence of CD was reported among a primarily Caucasian population in Olmsted County, Minnesota, between 1940 and 1993 [1]. Notably, an increasing proportion of new cases was diagnosed in individuals below age 20, reaching 17% by 1990 [1]. This has led to a decrease in the median age at diagnosis. Other data on the incidence of CD among children in Europe,

the United States and Canada support the evidence for an earlier age at diagnosis, rather than a true increase in disease incidence [2]. Another study revealed that the incidence of CD (7-12 per 100 000) and UC (5-6.9 per 100 000) among African-American children in the state of Georgia was similar to that observed in the age-matched population overall [3]. The data revealed that IBD is more prevalent among African-American children than previously thought. A recent prospective study in the British Isles yielded similar results, with an incidence of 5.2 cases of childhood IBD per 100000 population [4]. UC was more frequent in children of Asian descent. In our IBD clinic the incidence of new cases of CD has increased 4-fold over the past two decades; that for UC does not appear to have changed over the same time period.

Distinctive clinical presentations In general, the symptoms and signs of IBD are dependent upon the sites involved, their extent as well as severity, rather than the age of the patient [5, 6]. However, there are certain clinical presentations that are either unique or more common to the pediatric age group (Table 1). Notorious among these is growth failure, seen in up to half of patients at the time of diagnosis [5-7]. It is important to recognize that this mode of presentation, far more common in CD than UC, may occur in the absence of gastrointestinal complaints. Unfortunately, this may lead to a delay in diagnosis, often entailing one or more years until IBD is suspected. Aff*ected patients usually have delayed puberty accompanying their poor growth. As discussed below, the management of such patients is a major challenge, in terms of

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 773-790. © 2003 Kluwer Academic Publishers. Printed in Great Britain

774

Inflammatory bowel disease in the pediatric age group

Table 1. Particular clinical presentations of IBD in the pediatric age group

In adults over the age of 50, performing a colonoscopy in the setting of a functional bowel disorder is perhaps justifiable on the basis of the merits of screening for colonic tumors. On the other hand, it is inappropriate to pursue these investigations among children in the setting where IBD is very unlikely. Generally speaking, children suspected of IBD have not experienced significant health problems prior to onset of their symptoms. Therefore, one should consider the emotional impact of intrusive testing in the patient who very likely has a functional bowel disorder. Given these clinical challenges, clinical investigators have searched for a marker or a combination of non-invasive tests that may enable clinicians to screen for IBD. The diagnostic approach to the child suspected of IBD should depend upon the level of doubt (Table 3). Infectious causes of chronic diarrhea are usually sought in this setting (Table 2), requiring stool cultures and search for ova and parasites. In view of the fact that children are often exposed to antibiotic therapy, assay for Clostridium difficile toxin is frequently indicated. A complete blood count is usually obtained, with particular attention paid to the presence of a microcytic anemia and thrombocytosis, commonly seen in IBD. Hypoalbuminemia, often observed in the presence of a protein-losing enteropathy, is also frequent, but not specific for IBD. Elevated levels of circulating markers of acutephase reactants, such as the erythrocyte sedimentation rate (ESR), C-reactive protein, and orosomucoid, are more common in active CD than in UC [10]. In addition to their lack of specificity for IBD, these tests may be all normal in up to one-third of patients [10-12]. The recently developed highly specific serologic tests can be useful adjunctive aids in discriminating IBD in its mild forms from functional bowel disorders [13-16]. Perinuclear anti-neutrophil cytoplasmic autoantibodies (pANCA) have been established as an autoimmune marker most characteristic of UC in the pediatric age group. On the other hand, antibodies to oligomannosidic epitopes of the yeast Saccharomyces cerevisiae (ASCA) have been shown to be a reliable marker of CD. Double positivity (both IgA and IgG) for ASCA was found to be 100% specific for pediatric CD [15]. The potential sensitivity of these markers as screening tests for IBD is maximized when the two assays are combined. Table 3 summarizes our approach to the child suspected of IBD, depending on the severity of symptoms and the level of clinical suspicion.

Unique

Common

Growth failure Delayed sexual maturation/ puberty

Recurrent abdominal pain Short stature Recurrent unexplained fever Arthralgias/arthritis Perianal disease Anorexia Recurrent aphthous mouth ulcers

improving growth as well as the psychosocial implications of short stature and pubertal delay. Other 'atypical' clinical presentations which should make the physician suspicious of IBD in the pediatric age group are recurrent unexplained fever, arthralgias or arthritis, perianal disease (tags, fissures, abscesses and/or fistulas), aphthous mouth ulcers or perianal skin tags, and unexplained anorexia [5, 8].

Diagnostic approach to tlie child suspected of IBD The differential diagnosis of CD and UC in the pediatric age group is summarized in Table 2. As in adults, the recognition and diagnosis of IBD is straightforward in children and adolescents when the clinical presentation is unambiguous. Thus, for example, a diagnosis of colitis (UC or CD) is strongly suspected in patients who present with typical symptoms, such as bloody diarrhea, urgency and abdominal discomfort. These clinical findings are then promptly confirmed by standard radiologic, endoscopic and histologic criteria, and an accurate diagnosis is promptly arrived at [5, 6, 9]. On the other hand, a diagnostic challenge arises in children who present with non-specific and indolent intestinal and extraintestinal symptoms that can be characteristic of both IBD and functional bowel disorders. In the face of such diagnostic uncertainty some clinicians may rely on invasive diagnostic testing, which entails at minimum a barium upper gastrointestinal series and small bowel followthrough (UGI and SBFT), as well as a complete colonoscopy with biopsies in order to confirm or exclude IBD.

Ernest G. Seidman andArlene Caplan

775

Table 2. Differential diagnosis of disorders resembling IBD in the pediatric age group

Infectious etiologies Acute appendicitis Mesenteric adenitis Enteritis (y(9rs/y?/a enterocolitica, enteropathogeneic E coii, Campylobacter jejuni, Salmonella, Shigella, Entamoeba histolytica, Giardia lamblla, Dlentamoeba fragllls, Mycobacterium tuberculosis, etc.) Pseudomembraneous or antibiotic-associated colitis Vascular disorders Hemolytic uremic syndrome, Henoch Schoenlein Purpura, Behcet's disease, polyarteritis nodosum, systemic lupus erythematosus, ischemic bowel disease, dermatomyositis Immunodeficiency disorders (congenital acquired) Iatrogenic Radiation, chemotherapy (typhlitis), graft-versus-host disease Obstetric and gynecologic causes Ectopic pregnancy, ovarian cysts, tumors, endometriosis Allergic Eosinophilic gastroenteropathies Neuromuscular Hirschsprung's disease, pseudo-obstruction syndromes Others Intussusception, Meckel diverticulum, tumors

Table 3. Contrasting diagnostic approaches to the child suspected of IBD Clinical Index of suspicion high Upper gastrointestinal and small bowel follow-through barium X-rays Complete colonoscopy with multiple biopsies Upper endoscopy (if clinically indicated) Rule out microbial causes Exclude immunodeficiency disorder Exclude autoimmune enteropathy Exclude allergic disorder Clinical Index of suspicion low Verify normal physical examination Verify normal growth parameters Carry out limited investigations: CBC: hemoglobin, platelet count, ESR Serum albumin, iron, ferritin Serological assays: pANCA, ASCA Abdominal ultrasound + Doppler assessment of mucosal vessel density

CD

UC

++ ++ ++

+

++

+++

+

+++

++

++

+

+

+

+

+

+

+

++

+

++

776

Inflammatory bowel disease in the pediatric age group

Infectious enterocolitis should always be considered in the differential diagnosis of childhood IBD, even in established cases presenting with a clinical relapse. Most acute, infectious causes of diarrhea resolve after a few weeks. Exceptions include pseudomembranous or antibiotic-associated colitis, particularly prevalent in the pediatric age group. Our recent data show that the yield of C. difljcile toxin increases when stool samples are obtained via colonoscopy and sent promptly to the laboratory on ice [9]. Infestation with Giardia lamhlia or Dientamoeha fragilis, as well as amebic colitis, can also mimic IBD. The latter classically results in discrete, punched-out ulcers with a rolled, edematous margin. Rectal cultures for gonorrhea and serologic testing for lymphogranuloma venereum should be considered in the setting of sexually active patients. Rare causes of infection, such as cytomegalovirus (CMV), Cryptosporidium, Microsporidia or Strongyloides should be sought in the setting of congenital or acquired immunodeficiency disorders [17]. The standard investigation of the child for whom the index of suspicion for IBD is high includes colonoscopic evaluation of the colon and barium studies of the upper gastrointestinal tract (Table 3). Radiologic assessments of the colon and terminal ileum are not adequate diagnostic substitutes for colonoscopy [18]. Images on barium studies as well as computerized tomography (CT) scans have been mistakenly diagnosed as IBD in infectious enteritis or colitis [19]. Furthermore, a recent pediatric study showed that colonoscopy was superior to magnetic resonance imaging (MRI) in distinguishing CD involving the colon from UC [20]. MRI was also inaccurate in terms of assessing inflammation severity. The high false-negative rate of technetiumlabeled autologous white blood cell scintigraphy precludes its routine use to estabhsh a diagnosis of IBD in children [21]. Thus, colonoscopy with biopsies remains the most accurate tool for determining the type and severity of IBD in children and adolescents [9]. A study in pediatric patients concluded that scintigraphy can be useful when total colonoscopy or ileoscopy could not be performed [22]. Enteroclysis may be necessary when visualization of the small bowel is incomplete or impaired by dilation or excessive dilution of the barium with liquid stools. Abdominal ultrasound may be used to screen for thickened loops of bowel, and Doppler assessment of intestinal mucosal vessel density has been shown by our group to correlate with activity in CD [23]. As in adult patients a complete colonoscopy

is imperative to obtain tissue confirmation of the diagnosis, and constitutes the most cost-effective strategy to determine the extent and severity of IBD in children [24]. Biopsies should be routinely taken, as histologic features of UC and CD may be uncovered even in zones of macroscopically normal mucosa. In about 10% of cases of so-called indeterminate colitis, endoscopic and histologic findings may be insufficiently distinctive to unequivocally discern UC from Crohn's colitis. In such instances determination of pANCA and ASCA serologic tests can be helpful [14-16, 25]. Rarely, IBD may present during infancy [26]. The clinical presentation resembles that seen in infants with autoimmune enteropathy [27] or immunodeficiency disorders [17], requiring careful exclusion of these possibilities (Table 3).

Markers of disease activity It is not uncommon to be faced with pediatric CD patients who manifest few symptoms, yet present with chronic anorexia and growth failure. It is generally not acceptable to such young patients to re-evaluate disease activity by repeating colonoscopic examination [9]. Follow-up endoscopic studies are reasonable in the case of indeterminate colitis, when surgery is contemplated. It is also justifiable in cases refractory to medical management, in order to exclude other disorders such as concomitant infections of the upper {Helicobacter pylori) or lower (C difficile, CMV) gastrointestinal tract. Biologic markers of the acute-phase response, such as the C-reactive protein, ESR, or serum orosomucoid may be elevated in some patients. In addition, an iron-deficiency anemia, hypoalbuminemia and thrombocytosis may be found. However, these markers are inadequately specific to be utilized to reliably monitor disease activity [12]. Intestinal permeability, measuring 5-h urinary excretion ratio of lactulose/L-rhamnose, was significantly increased in patients with active CD and extensive UC [28]. However, no correlation was found between this permeability marker and the pediatric CD activity index. Our recent approach has been to estimate intestinal wall vessel density as a function of disease activity, using pulsed color Doppler abdominal sonography [23]. Affected bowel loops are thicker in the group of pediatric patients with active CD

Ernest G. Seidman and Arlene Cap Ian (p< 0.001). Vessel density is much more frequently moderate or high (2-4, and > 5 vessels/cm^, respectively) in active, compared to quiescent CD (0-1 vessel/cm^). This method is simple to perform in young patients, is non-invasive and accurately monitors the course of the disease [29]. Doppler sonography is particularly useful in children in order to objectively document response to therapy, without requiring more invasive endoscopic or radiologic studies.

Therapeutic approach to the child with IBD The major goals in managing IBD at any stage of life are to induce and maintain remission of disease activity and to assure an optimal quality of life. In the pediatric age group it is essential also to ensure normal nutritional status, growth and development. However, the yet-obscure etiology and pathogenesis of IBD, along with its highly variable severity, extent and clinical course, render it difficult to achieve an optimal outcome in all cases. Medical therapy for IBD has advanced remarkably in recent years [30, 31]. The general approach to treating the child with IBD, as in adults, is based on the severity of symptoms as well as the localization and extent of the disease. Few controlled clinical trials have been carried out in children, although the recent act of the US Congress will rectify that for the future. The doses of the drugs commonly employed in the pediatric IBD population are summarized in Table 4. The special issues related to surgery in pediatric IBD are discussed below.

5-Aminosalicylates As in adults, children with mild to moderate UC or CD are typically treated with oral sulfasalazine or 5aminosalicylic acid (5-ASA), either alone or in combination with topical 5-ASA and/or corticosteroid enemas. Young children often retain and tolerate suppositories or rectal foam better than enemas. However, patients with refractory distal colitis may require high-dose prednisolone enemas in order to obtain remission off oral corticosteroids (Table 4).

111

Corticosteroids Patients with moderate to severe disease are generally managed with corticosteroids, orally or intravenously, depending upon symptom severity. Steroid use is associated with numerous side-effects, which are intolerable for many pediatric patients. These include exacerbation of acne, facial puffiness (moon face), hirsutism, striae, cataracts, aseptic necrosis of the hip or knee, hypertension, depression, growth i m p a i r m e n t , myopathy and o s t e o p e n i a with compression or pathologic fractures (Table 4). Thus, systemic corticosteroids are generally reserved for more severe disease, and only for a limited time period. Side effects can be reduced to a certain extent by adopting alternate-day dosing of prednisone and avoiding nocturnal doses. Corticosteroids, whether in the form of prednisone or budesonide, have not been shown to be of benefit as maintenance therapy. Alternate-day, low-dose prednisone has been proposed in order to reduce relapses, without inhibiting linear growth [32]. However, confirmation from randomized controlled trials is lacking. Budesonide is a well-absorbed and rapidly catabolized corticosteroid that has the advantage of causing fewer glucocorticoid-related systemic side-effects. An ileal release preparation has been observed to induce remission in 50-69% of patients with active CD [33]. The addition of antibiotics (metronidazole and ciprofloxacin) has not been shown to improve the remission rate compared to budesonide alone [34]. Thus, budesonide can be employed for distal ileal or ileocecal CD in children and adolescents. Although the side-effect profile is superior, the efficacy rate is inferior to standard prednisone.

Immunomoduiatory drugs 6-Mercaptopurine (6-MP) and its parent drug azathioprine (AZA) are arguably the most effective immunosuppressive drugs for the long-term management of both CD and UC [35, 36]. They are proven to be effective for steroid-dependent, as well as chronically active or steroid-resistant disease [36]. More recently, an important pediatric trial showed that 6-MP can dramatically reduce the risks of relapse after steroid-induced remission in CD [37]. This study has revised the approach to the long-term management of CD, supporting the pre-emptive use of AZA or 6-MP in order to improve the natural history of the disease, effectively maintaining long-

778

Inflammatory bowel disease in the pediatric age group

Table 4. Commonly employed medications for IBD in the pediatric age group Drug and dose

Potential side-effects

Aminosalicylates Sulfasalazine (40-60 mg/kg per day; t.i.d.); 5-ASA/ mesalamine/ olsalazine (40-70 mg/kg per day; bid. or t.i.d.);

Headaches, nausea, hypersensitivity reactions (skin), pancreatitis, pericarditis, granulocytopenia, thrombocytopenia

Topical 5-ASA (enema, 2 - 4 g; or suppository; 0.5 to 1 g, both q12-24h)

Side-effects of topical applications are seldom encountered

Corticosteroids Prednisone ( 1 - 2 mg/kg max. 50 mg/day; q.a.m. or divided b.i.d.). Topical enemas (hydrocortisone; 50-100 mg q 12-24 h; or methylprednisolone, 10-40 mg in 3 0 - 6 0 ml NaCI 0.9%, q 12-24 h)

Acne, moon facies, striae, growth impairment, hypertension, aseptic necrosis or bone fractures, depression or other mood alterations, sleep disorder, osteopenia, myopathy

Budesonide ( 9 mg/day; q.a.m.)

Budesonide: less systemic side-effects

Antimetabolite therapy 6-Mercaptopurine (1.25-1.5 mg/kg per day).* Azathioprine (2.25-3 mg/kg per day)*

Idiosyncratic: pancreatitis, rash, fever. Leukopenia, hepatitis (associated with increased metabolite levels). Nausea

Methotrexate (15 mg/m^ s.c.)

Immunosuppressive therapy Cyclosporine A and tacrolimus (dose adjusted according to drug levels)

Nausea; teratogenicity; hepatotoxicity; myelosuppression; pulmonary toxicity; anorexia

Nephrotoxicity, headaches, paresthesias, hirsutism, oral thrush. Diabetes

Biological therapy Anti-tumor necrosis factor alpha (infliximab, 5 mg/kg i.v.)

Serum sickness-like reactions. Lupus-like syndrome. Lymphomas?

Antibiotics Metronidazole (10-15 mg/kg per day). Ciprofloxacin (restricted to > 16 years; 250-750 mg b.i.d., according to weight)

Long-term use: peripheral neuropathy. Bone pain; potential for altered bone health

*Dose adjusted according to TPMT genotype and metabolite levels

term remission while preventing steroid dependence or resistance for most cases [38]. Despite the estabHshed efficacy of AZA and 6-MP in controlled trials, dose-ranging studies have not been carried out. Over the past few years, however, pharmacogenetic advances have led to the development of new strategies in order to optimize and individualize therapy with AZA and 6-MP, maximizing efficacy, while minimizing toxicity [29, 38]. AZA is an inactive prodrug that undergoes a series of enzymatic reactions via competing pathways, leading to two major metabolites (Fig. 1). One route leads to the production of 6-thioguanine nucleotides (6TG), shown to be the active metabolite [39]. However, excessive levels of 6-TG are potentially myelotoxic. In the competing pathway, thiopurine methyltransferase (TPMT) yields 6-methylmercaptopurine ribonucleotides. The latter metabolites appear to be

therapeutically inactive but potentially hepatotoxic at high levels [29, 39, 40]. Co-dominantly inherited polymorphic alleles confer variable TPMT enzyme activity levels, with about 11% of individuals heterozygous and 0.3% homozygous for common TPMT mutations. Such individuals have intermediate and low/absent TPMT activity, respectively. Heterozygous patients with intermediate activity generate higher therapeutic 6-TG levels, but are at a greater risk for myelosuppression. The identification of a patient's TPMT genotype or phenotype can thus allow the physician to adjust the dose of the drug accordingly, avoiding early leukopenic events [41]. Patients with subtherapeutic 6-TG levels, due either to under-dosing, poor compliance or excessive TPMT activity are more likely to be refractory to therapy with these drugs [40]. Thus, AZA and 6-MP are generally well tolerated in pediatric IBD patients

Ernest G. Seidman and Arlene Caplan

779

DNA RNA

6-Thiouric acid

A i

^

6-Thioguanin( nucleotides

Xanthine Oxidase

HPRT AzathioprineHi^ 6-MP ill

^6-MP

^

TPMT ExtraceUular domaine

6-Methylmercaptopurine ribonucleotides

IntmceUular domaine

Purine synthesis Figure 1. Pathways in the metabolism of azathioprine and 6-mercaptopurine. Oral azathioprine is rapidly converted to 6-MP by a non-enzymatic process. Initial 6-MP transformations occur along competing catabolic (xanthine oxidase; TPMT: thiopurine methyltransferase) and anabolic (HPRT: hypoxanthine phosphoribosyltransferase) enzymatic pathways. The latter intracellular enzyme (dashed line) transforms the drug into 6-thioguanine nucleotides, which have been shown to be the most important factor associated with treatment efficacy. TPMT converts the drug into 6-methyl-mercaptopurine ribonucleotides. Patients heterozygous for a mutant allele of TPMT will convert a higher proportion of the drug into 6-TG. This translates into a higher success rate, but with an increased risk of myelosuppression.

[42]. Measuring 6-MP metabolite levels and TPMT molecular analysis provide clinicians with useful tools for optimizing therapeutic response to 6-MP/ AZA, as w^ell as for identifying individuals at increased risk for drug-induced toxicity [29, 38,40]. Methotrexate is considered to be in the class of antimetabolite therapies, due to its antagonistic effect on folic acid metabolism [43]. Among adults, almost 40% of steroid-refractory chronically active CD patients have been found to respond to w^eekly injections of methotrexate (25 mg). No similar, controlled trials have been done in the pediatric age group. In pediatrics this therapy has generally been

employed in moderate to severe CD, refractory to corticosteroids [44]. Weekly injections (15 mg/m^ weekly subcutaneously) are effective in pediatric patients who had previously failed therapy with 6MP. In one published open trial of 14 cases, nine showed symptomatic improvement after 4 weeks of therapy [45]. Hepatotoxicity is a major concern of long-term therapy. A recent study in children with juvenile rheumatoid arthritis showed that liver enzyme elevation more than 40% of the time was associated with an increased risk of fibrosis [46]. Obesity may add to the risk.

780

Inflammatory bowel disease in the pediatric age group

T-lymphocyte immunosuppressive drugs

Biological therapies

Cyclosporine A (CsA) is a fungal product whose immunomodulatory effects are achieved by blocking lymphocyte cytokine production, interleukin 2 (IL2) in particular. This inhibits the activation and proliferation of T helper cells. The efficacy of intravenous CsA among patients with severe UC is now well established [47]. Although surgery can often be avoided in the short term, the benefits of CsA must be weighed against the risk of severe complications, including potentially life-threatening opportunistic infections and the theoretical risk of lymphoproliferative disease. In addition, most patients relapse upon CsA withdrawal. Thus, CsA is generally employed as a bridge to the longer-term use of AZA or 6-MP [48]. The latter drugs must be initiated several weeks prior to tapering the dose of CsA. CsA has also been used in CD, especially in cases with severe perianal or fistulizing disease [49]. As noted for severe UC, it is primarily used as a bridge to therapy with a longer-acting immunomodulatory therapy, usually with 6-MP/AZA or methotrexate. Intravenous CsA (4 mg/kg per day) was used in an open-label study in 10 children with severe Crohn's colitis refractory to steroids, with a 70% response in the short term [50]. However, 30% relapsed within 6 months, and 6-MP-resistant patients did not respond. CsA may also be effective when given orally, if trough CsA levels between 250-400 ng/ml are achieved. Side-effects of CsA treatment include nephrotoxicity, headaches, paresthesias and infections (Table 4). Until further studies with longer follow-up are conducted, the use of CsA should be restricted to those IBD patients with severe, steroidresistant disease activity who require a drug with a rapid therapeutic onset. Other biologic therapies, such as anti-tumor necrosis factor (TNF) antibodies, have largely replaced the use of CsA for CD. Another T lymphocyte inhibitory drug, tacrolimus or FK506, is considered to be even more potent than CsA. Scant data are available with this drug in IBD, although case reports suggest a beneficial effect in fistulizing CD. A recent open-label study showed some benefit for oral tacrolimus over the short term for nine of 14 pediatric cases with severe, steroid-refractory colitis [51]. Blood levels were maintained in the 10-15 ng/ ml range, and AZA or 6-MP was added after about 6 weeks. However, four of nine responders eventually required a colectomy. Thus, fewer than 50% of cases achieved long-term remissions.

The advent of novel biological therapies has largely been the outcome of experimental animal studies in transgenic or gene knockout models of IBD. These novel agents have been engineered to specifically target integral steps in the cascade that contributes to mucosal inflammation, either by blocking proinflammatory cytokines or by administering antiinflammatory cytokines. Infliximab is a human/ mouse chimeric anti-TNF-oc neutralizing antibody that received approval in both the USA and Canada for IBD. Studies have demonstrated that this antibody is relatively highly effective for rapidly inducing remission in active, therapy-refractory inflammatory CD. The clinical benefits of a single infusion of 5 mg/ kg are generally observed within 2 weeks, and last for 4-8 weeks in about two-thirds of cases [52]. This treatment is also effective in closing fistulas, particularly in the perineum, one of the major challenges in CD therapy [53, 54]. Experience in the pediatric age group is limited, but shows similar trends [55]. Although relatively safe, infusion reactions to antiTNF therapy have been described. The formation of anti-infliximab antibodies has been reported to give rise to hypersensitivity reactions upon re-administration in some cases. A recent report described the successful desensitization and therapeutic use of infliximab in such patients despite previous anaphylactic reactions [56]. Given the uncertainty of longterm sequelae, particularly the risk for lymphomas, this modality of therapy in children is likely to be restricted to the most severe cases. Other new treatments that inhibit TNF-ot are being investigated, including CDP571, a humanized monoclonal antiTNF antibody, and etanercept, a recombinant TNF receptor fusion protein. Interleukin 10, a cytokine produced by T helper type 2 cells, is an endogenous anti-inflammatory cytokine that down-regulates the production of proinflammatory cytokines (TNF-oc, IL-lp, etc). Early studies suggested that the administration of recombinant human IL-10 was effective in treatment of active CD when given intravenously [57]. Further clinical trials are currently under way using this therapy, as well as with IL-11. In terms of biological therapies for UC, a recent open-label randomized study showed that subcutaneous interferon-a-2A resulted in significant improvement in active leftsided disease, compared to prednisolone enemas [58]. One pilot study has been carried out in pediatric IBD [59].

Ernest G. Seidman and Arlene Cap Ian

781

Table 5. Known causes of undernutrition and growth failure in pediatric patients mXh IBD Potential factor

Details

Inadequate energy/nutrient intake

Disease-induced (cytokines, symptoms). Iatrogenic (unacceptable dietary restrictions)

Malabsorption

Diminished absorptive surface area. Bacterial overgrowth. Bile salt deficiency

Increased gut losses

Secretory losses: electrolytes, minerals, trace metals. Protein-losing gastroenteropathy. Bleeding

Drug-nutrient interactions

Corticosteroids (calcium, protein). Sulfasalazine (folate). Cholestyramine (fat, vitamins). Cyclosporine A (magnesium)

Increased requirements

Sepsis, fever. Increased cell turnover. Replace losses: catch-up growth

Antibiotics

Table 6. Nutritional parameters in pediatric patients with IBD

As in adults, antibiotic therapy has been most often employed for perianal CD. In one report on 325 pediatric patients, 15% had perianal fistulas, while perirectal abscesses were encountered in 13% [60]. Metronidazole is most often used for this indication as well as for treating colonic CD. However, the longterm use of metronidazole has been associated with peripheral neuropathy [61]. Ciprofloxacin has also been advocated for the therapy of perianal CD [62], as well as primary therapy [63]. However, concerns regarding potential toxicity to cartilage in experimental animals have precluded the widespread use of this antibiotic in children under the age of 16, or those less than Tanner stage IV [64]. In our experience perianal disease often recurs after cessation of therapy with antibiotics. The use of AZA or 6-MP thus represents a better long-term strategy in most cases. Severe fistulizing disease is best managed with infliximab, or with surgical drainage and seton placement in cases with perineal sepsis.

Evaluation of growth/chronic malnutrition (a) Height percent for age (b) Growth velocity (c) Bone age (d) Bone density (DEXA adjusted to bone or height age)

Assessment and management of malnutrition Assessing ttie problem The potential complications of growth impairment and pubertal delay are unique to pediatric IBD patients (Table 1). Weight loss is an extremely common finding at presentation, occurring in almost 90% of cases. Inadequate energy intake often precedes other symptoms so that, by the time of diagnosis, a majority of pediatric patients are malnourished, with weight falling below the third percentile for age along with arrested growth or a deceleration in its velocity [7, 65, 66]. An excellent

Evaluation of acute malnutrition (a) Percent ideal weight for height (b) Serum albumin (secondary to protein-losing enteropathy) (c) Tricipital skinfold Micronutrient parameters (a) Iron, TIBC, ferritin (b) Vitamins: folates, B12, A, D, E (c) Other minerals: Ca, Ph, Mg, Zn (d) Electrolytes (if profuse diarrhea) (e) Rare deficiencies: vitamin C, K, selenium, copper

estimation of the degree of acute malnutrition is obtained by calculating the percent ideal weight for height [67]. Very often, growth failure ensues after weight for height falls below 90% of ideal predicted. The causes of malnutrition in children with IBD are summarized in Table 5. By far the most important cause is inadequate intake of calories to meet the energy needs in the growing child. Anorexia is due to a combination of factors, including the abdominal pain and diarrhea brought on by food ingestion, nausea and early satiety, as well as the proinflammatory cytokines that have a suppressive effect on appetite. Growth and other nutritional parameters to be followed in all pediatric patients with IBD are summarized in Table 6. Updating percentile curves for weight and height, as well as height velocity, should be part of every assessment at least biannualy [65-67]. Growth failure may be declared in any

Inflammatory bowel disease in the pediatric age group

782

SHORT STATURE

I

Growth Velocity Normal

I1

Growth Velocity Abnormal

Constitutional Delayed Growth

Bone Age & Puberty Normal

Bone Age & Puberty Delayed

Genetic Short Stature

1 Weight for Height % Normal

Weight for Height % Low

Hormonal Deficiency

IBD'Related Undernutrition

Figure 2. Assessment of the cause of short stature in pediatric IBD patients.

subject with cessation of linear growth over a period of 6 months or by a decrease exceeding one or more standard deviations in height percentile. An easy bedside rule of thumb is that growth velocity generally exceeds 4 cm per year in prepubertal boys and 3.5 cm in girls. Peak height velocity occurs before menarche in female adolescents. Therefore, it is important to recognize and treat malnutrition before it is too late to achieve any catch-up. For many adolescent patients with IBD, impaired growth leading to short stature and the accompanying delayed maturation of secondary sexual characteristics may be more troubling and debilitating than their underlying disease [5, 7]. Therapy for correcting growth failure is detailed below. The proper management of short stature in the IBD child requires a precise evaluation of its cause, as summarized in Fig. 2. The clinician must ascertain whether the growth failure is due to inadequate intake of calories as a result of the IBD itself, or due to potential hormonal deficiency (growth hormone,

thyroxin or Cortisol). In the former but not the latter scenario the weight for height percentile will almost invariably be abnormally low.

Therapy of active Crohn's disease The potential role of nutritional therapy in children with IBD can be subdivided into two categories: either as primary therapy in order to induce remission in active CD or UC, or as adjunctive therapy to help maintain remission and to enhance growth [68, 69]. Although several randomized controlled trials have suggested that elemental and semi-elemental diets are as effective as steroids, meta-analyses have shown an overall statistical advantage for corticosteroids [70]. Nevertheless, nutrition is still a logical choice as primary therapy for active CD in selected cases, especially those children and adolescents with marked undernutrition and growth failure [68, 69]. Patients who tend to respond best (75% remission rate) are those with newly diagnosed CD involving

Ernest G. Seidman and Arlene Caplan the terminal ileum with or without the cecum or proximal colon [69]. Individuals with long-standing disease or extensive colitis generally respond less favorably (50% and 35%, respectively). Although steroids more often induce remission, their use is associated with a net loss of mineral bone density, negative nitrogen balance and with impaired linear growth [68]. On the other hand, nutritional therapy enhances growth, induces net anabolism with positive nitrogen balance, and improves bone health. Another clinical scenario favoring diet as primary therapy is for adolescents with CD who refuse a course of corticosteroids due to concerns for growth, cosmetic or other adverse effects [69]. Drawbacks to the use of nutrition as primary therapy for CD include the relatively high cost, the unpleasant taste and monotony of defined formula diets when employed as sole source of nutrition. The availability of flavor packets has improved the acceptance and oral tolerance of elemental and semi-elemental formulas. Whichever formula is selected, a major objective is to avoid parenteral nutrition, unless the enteral route has failed or is contraindicated. Defined formulas are as effective as, and certainly safer and less costly than, parenteral nutrition. In order to induce remission we generally administer a semi-elemental diet as sole source of nutrition (40-70 kcal/kg ideal body weight per day), for 4 consecutive weeks. The patients drink the formula during the day, along with clear fluids, as tolerated. The balance is administered nocturnally, by nasogastric tube or via a gastrostomy [9]. In addition to inducing remissions for active disease, nutritional therapy has been employed on an intermittent, cyclical manner (4 out of every 16 weeks) in order to sustain growth as well as to maintain remission off" steroids [71, 72]. Patients who received intermittent semi-elemental diet had significantly fewer relapses and markedly improved growth velocity, compared to those treated with lowdose, alternate-day prednisone. In addition to improving symptoms, elemental diets reduce excessive intestinal permeability, reverse the protein-losing enteropathy and decrease intestinal and systemic markers of inflammation [73]. An alternative approach is to employ AZA or 6-MP, as detailed above. Novel defined formulas that include cytokines such as transforming growth factor-p or an enhanced content of antioxidants are currently under investigation, with promising preliminary results. The mechanisms underlying the beneficial effects of nutritional therapy remain incompletely under-

783 stood. Several theories have been postulated: removal of dietary antigens, elimination of proinflammatory nutrients such as omega-6 fatty acids and nucleotides, altered eicosanoid production, bowel gut hormones and flora, as well as diminished pancreatic, hepatobiliary and intestinal secretions [68]. Patients with active IBD have increased production of arachidonic acid metabolites derived from dietary sources of omega-6 fatty acids such as vegetable oils, leading to high levels of proinflammatory eicosanoids such as leukotriene B4, a potent neutrophil chemoattractant [68]. Fish oils contain eicosapentaenoic acid, an unsaturated fatty acid that is metabolized through the cyclo-oxygenase pathway, leading to leukotriene B5, 30 times less potent than its leukotriene B4 counterpart. Studies using fish oil supplements have lent support to the potential role of dietary fatty acids in the pathogenesis as well as therapy of IBD [68]. One study using enteric-coated fish oil was highly effective in reducing relapses in CD, supporting this hypothesis [74]. Another novel approach to the use of neutriceuticals to alter the inflammatory process in IBD was to employ A^-acetyl glucosamine (N-AG) to assist in tissue repair mechanisms. In a pilot study, treatment-resistant pediatric IBD patients were treated with N-AG orally or rectally [75]. There appeared to be some short-term benefit in several cases, including patients with strictures.

Approach to growth failure The nutritional impact of IBD is particularly severe in the prepubertal patient. Many patients ingest an insuflftcient quantity of calories in order to meet their energy needs, as well as the metabolic costs of growth. Thus, growth failure is a common, serious complication that is unique to the pediatric age group, encountered in up to half of CD and about 10% of UC patients [7, 65, 66]. Normal growth is an important indicator of remission and an outcome parameter of therapeutic efficacy in pediatric IBD. However, despite 'appropriate medical therapy,' CD results in permanent short stature in 20-35% of adults who had the disease prior to their puberty [76]. There is a time limit for achieving potential 'catch-up' growth because of progressive bone maturation and eventual epiphyseal fusion. Weight gain can be achieved in weeks, whereas growth acceleration requires many months of sustained treatment. In order to be effective, therefore, therapy must be initiated well before bone maturation is

784

Inflammatory bowel disease in the pediatric age group

complete. Chronic undernutrition resulting primarily from inadequate caloric intake is by far the most important factor. Caloric intake in pediatric IBD is only 54-85% of estimated requirements, and anorexia often persists despite clinical remission. Intervention must be initiated early, consistently and aggressively, assuring adequate nutritional support over a sufficient period of time in order to achieve enhanced growth. When growth is significantly impaired, and the disease remains localized and non-progressive, surgical management may be considered. In general, however, surgery is considered for growth failure only if optimal medical and nutritional therapies have failed. Although gains in weight and height are seen postoperatively [77], final adult height is not different for the group of patients with C D who had surgery during childhood compared to those who did not [78]. TPN can achieve weight gain and reverse growth arrest in CD. However, metabolic and infectious complications, as well as cost considerations, favor the use of enteral nutritional support [7, 79]. The sustained administration supplementation of a polymeric formula (4080 kcal/kg ideal body weight per day) via nocturnal nasogastric or gastrostomy routes effectively reverses growth failure. Compliance with high-calorie oral supplements is generally poor over the long term. Our high rate of success in having children and adolescents accept and comply with this form of therapy resides largely in the positive attitude of the nutrition support team, as well as the patients' high motivation [69].

benefit is likely contingent upon a concomitant decrease in the ingestion of dietary omega-6 fatty acids (vegetable oils).

Therapeutic aspects: ulcerative colitis Nutritional support is considered an adjunctive therapy, as there is no evidence that bowel rest or total parenteral nutrition influences the outcome of UC [7, 79]. Patients requiring hospitalization for a relapse should receive parenteral nutrition if their baseline nutritional evaluation reveals that they are malnourished, or if their intake is likely to be curtailed for at least 1 week. Growth failure has been reported in approximately 10% of pediatric patients with UC. Although much less common than in CD, this problem can significantly affect the child's selfesteem, behavior and school performance. Nutritional support for growth failure in IBD is discussed above. In our experience a modest improvement of disease activity can be achieved by supplementing the diet with fish oils containing omega-3 fatty acids (9 g/1.73 m^ per day). The magnitude of clinical

Assessment and therapy of micronutrient deficiencies Patients with IBD often develop micronutrient deficiencies, in addition to energy deficits and growth failure, as reviewed above. Among the many potential mineral deficiencies (Table 6), iron is most common, due to the combination of chronic gastrointestinal blood loss, iron malabsorption and inadequate dietary intake. The evaluation of iron stores is complicated by the confounding interpretation of low serum iron, transferrin saturation and ferritin levels due to chronic inflammation. Mineral deficiencies are not uncommonly encountered, and merit monitoring and repletion (Table 6).

Detection and management of bone disease Prevention of osteoporosis should begin during childhood, a time of rapid growth and increasing bone density. The presence of osteoporosis carries a significantly increased (30%) risk of fracture. In order to prevent osteoporosis one must first identify patients at high risk. Children and adolescents with IBD may have multiple risk factors, including the presence of a chronic inflammatory disorder with the overproduction of inflammatory cytokines that can lead to increased bone resorption. Other risk factors may include undernutrition, a sedentary lifestyle and glucocorticoid therapy. Therefore, bone mineral density (BMD) should be measured as part of the routine work-up of IBD patients in the pediatric age group. Several studies have confirmed that children with IBD are at risk for osteopenia, when compared to healthy age- and sex-matched controls [80, 81]. Bone mineral density (BMD) is best assessed by dual-energy X-ray absorptiometry (Table 6). Trabecular, rather than cortical bone is predominantly affected in IBD, as seen in the lumbar spine and femoral neck. Low BMD is much more prevalent in children with CD compared to those with UC, especially among females. In order not to overestimate osteoporosis due to growth failure with delayed bone maturation in IBD, BMD values in pediatric patients with CD should take into consid-

Ernest G. Seidman and Arlene Cap Ian eration delayed bone maturation that is often present. Interpretation of BMD on the basis of bone or height age, rather than chronological age, resulted in a diminution of the overall frequency of abnormally low BMD from 44% to 26-30% [80]. The latter prevalence correlates with the results of studies in adult patients. Annual BMD evaluation, corrected for bone or height age, should be part of the management of IBD in children, particularly in CD, or in patients who have received corticosteroids. Strategies to prevent and treat osteoporosis include modifying risk factors, such as correcting malnutrition, optimizing intake of calcium and vitamin D, encouraging physical exercise, and limiting glucocorticoid therapy, whenever possible. Several of these goals can be accomplished if nutritional therapy is employed for the management of active disease, rather than conventional corticosteroids [68, 69]. If there is no improvement in BMD, bisphosphonate therapy should be considered. These drugs can help maintain BMD even if corticosteroids are used because of their long skeletal half-life. Repeat dual-energy X-ray absorptiometry should be carried out to confirm adequate response to therapy, since the absorption of these drugs is often problematic. If necessary, bisphosphonates can be administered intravenously.

Psychosocial functioning in pediatric IBD Although the mortality associated with pediatric IBD is low, the condition still presents a major, lifelong health threat, challenging the psychological resources of both the affected child and the family. IBD frequently interferes with physical activities, limits social interactions, disrupts education, impairs growth, and delays puberty [5, 82]. Approximately 20% of children have severely disabling disease [5]. In its acute phases IBD can present a serious impediment to daily functioning. Relapses may necessitate hospitalizations, which cause major disruptions in the child's academic, social and family life. The majority of children with CD experience considerable worry, distress and concern about their disease and its effects on school absences, academic achievement and participation in family and social activities away from home [82]. The chronicity of IBD poses persistent demands on children and their families to cope with fluctuating degrees of illness, prolonged use of medications and dietary limita-

785 tions. The complications of growth failure and delayed puberty in CD add to the psychological stress associated with the disease, particularly as patients approach adolescence [5]. A recent meta-analysis indicated that children with IBD have more psychological disturbances than age-matched groups with other chronic illnesses [83]. Problems of low-self-esteem, anxiety and depression are frequent [84, 85]. Burke et al. [86] reported that about four out of 10 children with CD were either clinically depressed or presented significant depressive symptoms soon after diagnosis. Depression is relatively common in children with both CD and UC [87]. While depression is the most common emotional problem reported, other difficulties associated with IBD include separation anxiety, fearfulness, social withdrawal, relationship problems, and problems with body image [88, 89]. While a higher incidence of internalizing (self-directed) problems is observed, externalizing (acting-out) problems have also been noted [90]. It is unclear whether depression is associated with the diagnosis of IBD, with the disease process itself, or with its treatment. Burke et al. [86] demonstrated that, in all instances of major depression in pediatric IBD patients, onset followed the diagnosis. The fact that siblings of patients also show an increased incidence of psychiatric disorders raises the possibility that psychopathology in children with CD is due to factors other than the disease itself. These may include genetic predisposition to psychiatric disorders, dysfunctional family dynamics or both. There is no evidence for a unique psychological 'profile' for children with IBD. However, in some cases psychological parameters have been useful in distinguishing CD from anorexia nervosa [84, 85]. IBD in children may have considerable consequences on the mental health of family members. Parents frequently have depression, anxiety and somatization [90]. They commonly worry about the effects of the disease on their child's school performance and on their future. Siblings also express concern about their ill sibling's well-being [91]. They may resent the time their parents devote to their ill sibling. Therapeutic interventions may also affect psychosocial functioning in the pediatric patient with IBD. Psychosis induced by corticosteroid therapy is rarely encountered in pediatric CD patients [92]. However, a wide range of intense emotional reactions among children and adolescents is frequently observed. Parents view their children's behavior as 'out of

Inflammatory bowel disease in the pediatric age group

786 Table 7. Specific surgical Indications in pediatric IBD Crohn's disease

Ulcerative colitis

Intractable symptoms; failure of medical management

Prolonged steroid dependence, intolerance to immunosuppressive agents

Hemorrhage

Hemorrhage

Fulminant colitis with or without toxic megacolon

Fulminant colitis, with or without toxic megacolon

Suspected perforation or abscess

Suspected perforation

Obstruction of upper and/or lower gastrointestinal tract. Fistulas: enteroenteral, enterocutaneous, enterovesical, enterovaginal. Intractable perirectal disease

Chronically active, unremitting disease

Growth failure despite nutrition support

Growth failure despite nutrition support

High risk of dysplasia, carcinoma

character', especially when it comes to challenging parental authority and expressing anger or exhibiting aggression. However, it is unclear to what extent these changes are due to the illness as opposed to the effects of systemic corticotherapy. The children may report feeling 'different inside' or having 'changed forever' as a result of their IBD. They frequently feel less in control of their emotions and behavior. Some experience bouts of crying, often without apparent reason. Many state that they have become more withdrawn and unsociable; others report that they no longer feel like enjoying themselves. They may consciously withdraw from recreational activities or social contact because they fear other children will find out that they have a disease and will treat them differently. Others withdraw simply because they feel unable to participate or enjoy themselves. The aversion of being questioned, feelings of embarrassment, or reluctance to be confronted with their disease often makes it difficult for children and adolescents with IBD to return to school after long absences due to their illness. These reactions of shame and embarrassment often contribute to depressive feelings of psychological isolation, fragility and instability. Problems related to compliance with therapy are notorious in pediatric populations. For many children, 'forgetting' to take their medications is a way of psychologically denying their disease. For others it may be more conscious, and related to the lack of efficacy, or side-effects of the drug. In addition to the adverse effects of steroids on mood, cognition and behavior, psychological factors in children, parents or the family at large may also contribute to poor compliance and adherence to therapy. Nutritional therapy can also have important effects on psycho-

High risk of dysplasia, carcinoma

social functioning. During periods of treatment, patients endure prolonged periods of food-deprivation and experience frustration due to disruption of social and family activities. Special formula diets are particularly difficult for children who eat their meals at school. They are often already embarrassed about their disease. Thus, nutrition as primary therapy potentially exacerbates the child's feeling of being different due to the disease itself, and may contribute to his/her sense of alienation. There is the additional consideration that the feeding tube and pump apparatus makes the disease more visible, both to patients and to those around them. This can accentuate feelings of self-consciousness and heighten embarrassment in social situations. Some patients experience the insertion of a nasal gastric tube as intrusive or aggressive on the part of the medical team. The psychological meaning patients attribute to treatment procedures, as well as their emotional reactions (e.g. anxiety, fear, and depression), may be more influential than their physical response in determining treatment success. More data on the psychological effects of medical and nutritional therapies are needed to provide a better understanding of the factors affecting compliance with treatment and the relationship to the medical team. Further study in this area would increase understanding of these factors, allowing gastroenterologists to provide higher-quality care and improved treatment strategies for pediatric IBD. Children with IBD have been reported to have a significantly impaired quality of life [82]. They fear everyday childhood activities and are concerned for their future employment. These children need sympathetic management, and efforts should be concentrated on improving their daily psychosocial

Ernest G. Seidman and Arlene Caplan functioning, enabling them to lead as normal a life as possible. This can best be achieved by medically controlling their disease activity, achieving normal growth and development through nutritional interventions, and providing psychosocial support to them and their family members.

Surgical considerations As in adults, the surgical approach to the child with IBD must always bear in mind that unlike UC, CD is not a surgically curable disorder. Surgical interventions are thus reserved either for complications of CD, or for symptoms that cannot be managed medically (Table 7). In the latter situation intractability of symptoms generally infers a poor prognosis postoperatively, with clinical relapse within 1 year in 50% of cases [93]. On the other hand, young patients operated upon for ileal strictures or local abscesses have a far better prognosis, with a 50% recurrence rate after 5 years. As summarized in Table 7, other surgical indications include recurrent episodes of partial bowel obstruction, or an abscess that fails to respond to conservative measures. Entero-enteric fistulas are not necessarily an indication for surgery. Although an enterovesical fistula is not an absolute indication for surgery, most patients will require an operation eventually. Perianal CD also calls for a conservative approach. Abscesses should be drained and fistulas laid flat or treated with setons when possible. Although surgical resection has not been shown to influence the natural history of CD, it can lead to dramatic growth acceleration and attainment of normal adult height in some, but not all, cases [77, 78]. The type of intervention depends on the extent and severity of the disease [94]. The principal indications for surgery in UC [95] include intractable disease (64%), refractory growth failure (14%), toxic megacolon (6%), hemorrhage (4%), perforation (3%) and cancer prophylaxis (2%). Despite the relative success of CsA, tacrolimus and other immunosuppressive agents, it must be borne in mind that colectomy can be a lifesaving procedure in a child with fulminant colitis or toxic megacolon. CsA and tacrolimus are not only shortterm solutions, they are associated with considerable morbidity as well as potential mortality, as reviewed above. The increased risk of developing colon cancer among patients with onset of UC during childhood is well estabhshed [9]. Proctocolectomy has been the

ISl most commonly practiced and accepted operation. If rectal involvement is mild, some surgeons prefer to carry out a colectomy and ileoanal anastomosis in one stage. Meticulous supervision of the conserved rectum is mandatory, as it might be a site for later cancer. Successful ileoanal-endorectal anastomosis after total colectomy and a mucosal proctectomy is often achieved. The rectal mucosa is stripped from its muscular wall, and the ileal mucosa is sutured on. The creation of a neorectum with an ileal pouch affords decreased stool frequency. This is now the treatment of choice in the pediatric population. An important factor regarding continence is the level at which the ileoanal anastomosis is performed [95, 96]. Colon cancer may also occur in the case of Crohn's colitis [97]. The relative risk depends largely upon the extent, as well as the duration, of the colitis [98]. Screening is usually initiated after about 8 years of clinical presentation for cases of extensive colitis.

References 1.

2. 3. 4. 5. 6. 7.

8. 9. 10. 11.

12.

Loftus EV Jr, Silverstein MD, Sandborn WJ, Tremaine WJ, Harmsen WS, Zinsmeister AR. Crohn's disease in Olmsted County, Minnesota, 1940-1993: incidence, prevalence, and survival. Gastroenterology 1998; 114: 1161-8. Logan RFA. Inflammatory bowel disease incidence: up, down or unchanged. Gut 1998; 42: 309-11. Ogunbi SO, Ransom JA, Sullivan K, Schoen BT, Gold BD. Inflammatory bowel disease in African-American children living in Georgia. Pediatrics 1998; 138: 103-7. Sawczenko A, Sandhu BK, Logan RF et al. Prospective survey of childhood inflammatory bowel disease in the British Isles. Lancet 2001; 357: 1093-4 Seidman E. Inflammatory bowel diseases. In: Roy C, Silverman A, Alagille D, eds. Pediatric Clinical Gastroenterology, 4th edn. St Louis: CV Mosby, 1995: 417-93. Marx G, Seidman EG. Inflammatory bowel diseases. In: C Lifschitz, ed. Pediatric Gastroenterology and Nutrition in Clinical Practice. New York: Marcel Decker, 2002: 651-82. Seidman EG. Growth and nutritional problems in pediatric IBD. In: Bayless TM, Hanauer SB, eds. Advanced Therapy of Inflammatory Bowel Disease, 2nd edn. Hamilton, Ontario: BC Decker, 2002 (In press). Pittock S, Drumm B, Fleming P et al. The oral cavity in Crohn's disease. J Pediatr 2001; 138: 767-71. Seidman EG. Role of endoscopy in pediatric inflammatory bowel disease. Gastrointest Endosc Clin N Am 2001; 11: 641-57. Murch SH, Walker-Smith J, Beattie RM. Indications for investigation of chronic gastrointestinal symptoms. Arch Dis Child 1995; 73: 354-5. Holmquist L, Ahren C, FeaUstreom SP. Relationship between results of laboratory tests and inflammatory activity assessed by colonoscopy in children and adolescents with UC and Crohn's colitis. J Pediatr Gastroenterol Nutr 1989; 9: 187-93. Dubinsky MC, Seidman E G Diagnostic markers of inflammatory bowel disease. Curr Opin Gastroenterol 2000; 16: 337-42.

Inflammatory bowel disease in the pediatric age group

788 13. 14. 15.

16. 17.

18. 19. 20.

21.

22. 23.

24.

25. 26. 27. 28.

29.

30. 31. 32.

33.

Dubinsky MC, Ofman JJ, Urman M, Targan SR, Seidman EG. Clinical utility of serodiagnostic testing in suspected pediatric IBD. Am J Gastroenterol 2001; 96: 758-65. Seidman E. Are serological tests for IBD useful to clinicians? Inflam Bowel Dis 1999; 5: 237. Ruemmele FM, Targan S, Levy G, Dubinsky M, Braun Seidman EG. Diagnostic accuracy of serological assays in pediatric inflammatory bowel disease. Gastroenterology 1998; 115:822-9. Hotfenberg E, Fidanza S, Sauaia A. Serologic testing for inflammatory bowel disease. J Pediatr 1999; 134: 447-52. Seidman E, Gastrointestinal manifestations of primary immunodeficiency disorders. In: Walker WA, et ciL, eds. Pediatric Gastrointestinal Disease Pathophysiology, Diagnosis, Management, 3rd cdn. Hamilton, Ontario: BC Decker, 2000: 525 47. Scotiniolis I, Rubcsin SE, Ginsberg GG. Imaging modalities in inflammatory bowel disease. Gastroenterol Clin N Am 1999; 28: 391 421. Ilnyckyj A, Greenberg H, Bernstein CN. Escherichia coli 0157:H7 infection mimicking Crohn's disease. Gastroenterology 1997; 113: 368. Durno CA, Sherman P, Williams T, Shuckctt B, Dupuis A, Gritlilhs AM. Magnetic resonance imaging to distinguish ihc type and severity o'i pediatric inflammatory bowel diseases. J Pediatr Ciastroenterol Nutr 2000; 30: 170 4. C\icchiara S, ClMentano L, de Magistris TM, Montisci A, lula VD, Fecarotta S. Colonoscopy and technetium-99m white cell scan in children with suspected inflammatory bowel disease. J Pediatr 1999; 135: 727 32. Charron M, Del Rosario F, Kocoshis S. Assessment of terminal ileal and colonic inflammation in Crohn's disease with 99TC-WBC. Acta Paediatr 1999; 88: 193 8. Spalinger JH, Patriquin H, Miron M-C ct al. Doppler sonography in pediatric Crohn's disease: Vessel density in the diseased bowel reflects disease activity. Radiology 2000; 217: 787 91. Deutsch DE, Olson AD. Colonoscopy or sigmoidoscopy as the initial evaluation of pediatric patients with colitis: a survey of physician behavior and cost analysis. J Pediatr Gastroenterol Nutr 1997; 25: 26 31. Rutgeerts P, Vermeire S. Serological diagnosis of inflammatory bowel disease. Lancet 2000; 356: 2117 18. Marx G, Seidman, EG, Martin SR, Deslandres C. Outcome of patients with Crohn's disease diagnosed before the age of two. J Pediatr 2002; 140:470 3. Russo PA, Brochu P, Seidman EG, Roy CC. Autoimmune enteropathy. Pediatr Dev Pathol 1999; 2: 65 71. Miki K, Moore DJ, Butler RN, Southcott E, Couper RT, Davidson GP. The sugar permeability test reflects disease activity in children and adolescents with inflammatory bowel disease. J Pediatr 1998; 133: 750-4. Seidman EG, Dubinsky M, Patriquin H, Marx G, Theoret Y. Recent developments in the diagnosis and management of paediatric IBD. In: CN Williams et al., eds. Trends in Inflammatory Bowel Disease Therapy 1999. Dordrecht, Netherlands: Kluwer Academic Publishers, 2000: 87-95. Stein R, Hanauer S. Medical therapy for inflammatory bowel disease. Inflam Bowel Dis 1999; 28: 297-321. Walker-Smith JA. Therapy of Crohn's disease in childhood. Bailliere's Clin Gastroenterol 1997; 11: 593-610. Del Rosario F, Orenstein SR, Neigut DA, Giarusso V, Wolfson N, Kocoshis SA. Retrospective analysis of alternate-day prednisone as maintenance therapy for Crohn's disease. CHn Pediatr 1998; 37: 413- -19. Rutgeerts P, Lofberg R, Malchow H et al. A comparison of budesonide with prednisolone for active Crohn's disease. N Engl J Med 1994; 331: 842.

34.

35.

36. 37.

38. 39. 40.

41.

42. 43.

44. 45. 46.

47. 48.

49. 50. 51. 52.

53.

Steinhart AH, Feagan BG, Wong CJ et al. Combined budesonide and antibiotic therapy for active Crohn's disease: a randomized controlled trial. Gastroenterology 2002; 123:33-40. Sandborn WJ. A review of immune modifier therapy for inflammatory bowel disease: azathioprine, 6-mercaptopurine, cyclosporine and methotrexate. Am J Gastroenterol 1996;91:423-33. Pearson DC, May GR, Fick G H , Sutherland LR. Azathioprine and 6-mercaptopurine in Crohn's disease. A meta-analysis. Ann Intern Med 1995; 123: 132-42. Markowitz J, Grancher BS, Kohn N et al. A multicenter trial of 6-mercaptopurine and prednisone therapy in newly diagnosed children with Crohn's disease. Gastroenterology 2000; 119:895 902. Seidman EG. 6-Mercaptopurine to maintain remission in Crohn's disease: an old friend becomes a new hero. Gastroenterology 2000; 119: 1158 61. Cutfari C, Theoret Y, Latour S, Seidman EG. 6-Mercaptopurine metabolism in Crohn's disease: correlation with efticacy and toxicity. Gut 1996; 39: 401 6. Dubinsky MC, Lamothe S, Yang HY et al. Pharmacogenomics and metabolite measurement for 6-mercaptopurine therapy in inflammatory bowel disease. Gastroenterology 2000; 118: 705 13. Black A J, McLeod HL, Capell HA et al. Thiopurine methyllransferase genotype predicts therapy-limiting severe toxicity from azathioprine. Ann Intern Med 1998; 129: 716 21. Kirschner B. Safety of azathioprine and 6-mercaptopurine in pediatric patients with inflammatory bowel disease. Gastroenterology 1998; 115: 813 21. Sandborn WJ. A review of immune modifier therapy for inflammatory bowel disease: azathioprine, 6-mercaptopurine, cyclosporine and methotrexate. Am J Gastroenterol 1996:91:423 33. Kirschner BS. Methotrexate therapy for pediatric patients with inflammatory bowel disease. Inflam Bowel Dis 1998; 4: 121 Mack DR, Young R, Kaufman SS, Ramey L, Vanderhoof JA. Methotrexate in patients with Crohn's disease after 6mercaptopurine. Pediatrics 1998; 132: 830 5. Kugathasan S, Newman AJ, Dahms BB et al. Liver biopsy findings in patients with juvenile rheumatoid arthritis receiving long-term, weekly methotrexate therapy. J Pediatr 1996; 128: 149. Cohen RD, Stein R, Hanauer SB. Intravenous cyclosporin in ulcerative colitis: a five year experience. Am J Gastroenterol 1999;94: 1587-92. Ramakrishna J, Langhans N, Calenda K et al. Combined use of cyclosporine and azathioprine or 6-mercaptopurine in pediatric inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1996; 22: 296. Hanauer SB, Smith MB. Rapid closure of Crohn's fistulas with continuous intravenous cyclosporine A. Am J Gastroenterol 1993;88:646-9. Mahdi G, Israel DM, Hassal E. Cyclosporine and 6mercaptopurine for active, refractory Crohn's colitis in children. Am J Gastroenterol 1996; 91: 1355-9. Bousvaros A, Kirschner BS, Werlin SL et al. Oral tacrolimus treatment of severe colitis in children. J Pediatr 2000; 137: 794-9. Targan SR, Hanauer SB, van Deventer SJ et al. A short term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn's disease. N Engl J Med 1997;337:1029-35. Present DH, Rutgeerts P, Targan S et al. Infliximab for the treatment of fistulas in patients with Crohn's disease. N Engl J Med 1999; 340: 1398-405.

Ernest G. Seidman and Arlene Caplan 54.

55. 56.

57. 58.

59.

60. 61. 62.

63.

64. 65. 66. 67. 68. 69. 70.

71. 72.

73.

Sandborn W, Hanauer S. Antitumor necrosis factor therapy for inflammatory bowel disease: a review of agents, pharmacology, chnical results and safety. Inflam Bowel Dis 1999; 5: 119-33. Hyams JS, Markowitz J, Wyllie R. Use of infliximab in the treatment of Crohn's disease in children and adolescents. J Pediatr2000; 137: 192-6. Puchner TC, Kugathasan S, Kelley SJ, Binion DG. Successful desensitization and therapeutic use of infliximab in adult and pediatric Crohn's disease patients with prior anaphylactic reactions. Inflam Bowel Dis 2001; 7: 34-7. Van Deventer SJ, Elson CO, Fedorak RN. Multiple doses of intravenous IL-10 in steroid-refractory Crohn's disease. Gastroenterology 1997; 113: 383-9. Madsen SM, Schlichting P, Davidsen B et al. An openlabeled, randomized study comparing systemic interferona-2A and prednisolone enemas in the treatment of left-sided ulcerative colitis. Am J Gastroenterol 2001; 96: 1807-15. Hadziselimovic F, Emmons L, Schaub U. Interferon-alpha2a (Roferon) treatment of inflammatory bowel disease in children and adolescents. In: Hadziselimovic F, Herzog D, eds. Inflammatory Bowel Disease and Morbus Hirschsprung. Falk Symposium on Paediatric and Surgical Gastroenterology, Basel, Switzerland. Dordrecht, Kluwer Academic Pubhshers, 1992: 109-23. Palder SB, Shandling B, Bilik R et al. Perianal complications of pediatric perianal Crohn's disease. J Pediatr Surg 1991; 26: 513 Duffy LF, Daum F, Fisher SE et al. Peripheral neuropathy in Crohn's disease patients treated with metronidazole. Gastroenterology 1985; 88: 681 Prantera C, Zannoni F, Scribano ML et al. An antibiotic regimen for the treatment of active Crohn's disease: a randomized, controlled trial of metronidazole plus ciprofloxacin. Am J Gastroenterol 1996; 91: 328 Colombel JF, Lemann M, Cassagnou M et al. A controlled trial comparing ciprofloxacin with mesalazine for the treatment of active Crohn's disease. Am J Gasroenterol 1999; 94:674 Kubin R. Safety and efficacy of ciprofloxacin in pediatric patients - review. Infection 1993; 21: 413. Seidman EG. Nutritional management of inflammatory bowel disease. Gastroenterol Chn N Am 1989; 18: 129-55. Seidman EG, LeLeiko N, Ament M et al. Nutritional issues in pediatric inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1991; 12: 424-38. Atlan Ph, Seidman E. Nutritional considerations in pediatric patients. Can J Gastroenterol 1990; 4: 41-47. Ruemmele F, Roy CC, Levy E, Seidman EG. The role of nutrition in treating pediatric Crohn's disease in the new millennium. J Pediatrics 2000; 136: 285-91. Seidman EG. Nutritional therapy for Crohn's disease: lessons from the Ste.-Justine Hospital experience. Inflam Bowel Dis 1997; 3: 49-53. Griffith AM, Ohlsson A, Sherman PM, Sutherland LR. Meta-analysis of enteral nutrition as a primary treatment of active Crohn's disease. Gastroenterology 1995; 108: 1056-67. Belh D, Seidman EG, Bouthillier L et al. Chronic intermittent elemental diet improves growth failure in children with Crohn's disease. Gastroenterology 1988; 94: 603-10. Seidman EG, Jones A, Issenman R, Griffith A. Relapse prevention/growth enhancement in pediatric Crohn's disease: multicenter randomized controlled trial of intermittent enteral nutrition versus alternate day prednisone. J Pediatr Gastroenterol Nutr 1996; 23: 344. Teahon K, Smethurst P, Person M et al. The effect of elemental diet on permeability and inffammation in Crohn's disease. Gastroenterology 1991; 101: 84-9.

789 74. 75.

76. 77.

78. 79. 80. 81.

82. 83.

84. 85. 86. 87.

88. 89.

90. 91.

92. 93.

Belluzzi A, Brignola C, Campieri M et al. Effect of an enteric-coated fish-oil preparation on relapses in Crohn's disease. N Engl J Med 1996; 334: 1557. Salvatore S, Heuschkel R, Tomlin S et al. A pilot study of A^acetyl glucosamine, a nutritional substrate for glycosaminoglycan synthesis, in paediatric chronic inffammatory bowel disease. AHment Pharmacol Ther 2000; 14: 1567-79. Markowitz J, Grancher K, Rosa J et al. Growth failure in pediatric inffammatory bowel disease. J Pediatr Gastroenterology Nutr 1994; 16: 373-80. Sentongo TA, Stettler N, Christian A, Han PD, Stalhngs VA, Baldassano RN. Growth after intestinal resection for Crohn's disease in children, adolescents, and young adults. Inflam Bowel Dis 2000; 6: 265-9. Castile RG, Telander RC, Cooney DR et al. Crohn's disease in children: assessment of the progression of disease, growth and prognosis. J Pediatr Surg 1980; 15: 462-9. D'Agata I, Seidman E. Dietary manipulation - parenteral. In: AUan R D et al, eds. Inflammatory Bowel Diseases, 3rd edn. New York: Churchill Livingstone, 1997: 541-5. Herzog D, Bishop N, Glorieux F, Seidman EG. Interpretation of bone mineral density values in pediatric Crohn's disease. Inflam Bowel Dis 1998; 4: 261-7. Gokhale R, Favus MJ, Karrison T, Sutton M, Rich B, Kirschner BS. Bone mineral density assessment in children with inflammatory bowel disease. Gastroenterology 1998; 114:902-11. Moody G, Eaden JA, Mayberry JF. Social implications of childhood Crohn's disease. J Pediatr Gastroenterol Nutr 1999; 28: S43-5. Lavigne JV, Faier-Routman J. Correlates of psychological adjustment to pediatric physical disorders: a meta-analytic review and comparison with existing models. J Dev Behav Pediatr 1993; 14: 117-23. Rickards H, Prendergast M, Booth IW. Psychiatric presentation of Crohn's disease. Diagnostic delay and increased morbidity. Br J Psychiatry 1994; 164: 256-61. Thienemann M, Steiner H. Psychometric measures in clarifying diagnosis in malnourished adolescents. Acta Paedopsychiatr 1992;55:207-11. Burke P, Meyer V, Kocoshis S et al. Depression and anxiety in pediatric inflammatory bowel disease and cystic fibrosis. J Am Acad Child Adolesc Psychiatry 1989; 28: 948-51. Engstrom I. Mental health and psychological functioning in children and adolescents with inflammatory bowel disease: a comparison with children having other chronic illnesses and with healthy children. J Child Psychol Psychiatry 1992; 33: 563-82. Steinhausen HC, Kies H. Comparative studies of ulcerative cohtis and Crohn's disease in children and adolescents. J Child Psychol Psychiatry 1982; 23: 33-42. Szajnberg N, Krall V, Davis P, Treem W, Hyams J. Psychopathology and relationship measures in children with inflammatory bowel disease and their parents. Child Psychiatry Hum Dev 1993; 2: 215-32. Engstrom I. Family interaction and locus of control in children and adolescents with inflammatory bowel disease. J Am Acad Child Adolesc Psychiatry 1991; 30: 913-20. Akobeng AK, Miller V, Firth D, Suresh-Babu MV, Mir P, Thomas AG. Quality of life of parents and siblings of children with inflammatory bowel disease. J Pediatr Gastroenterol Nutr 1999; 28: S40-2. Mullen RS, Romans-Clarkson SE. Behavioural sensitisation and steroid-induced psychosis. Br J Psychiatry 1993; 162: 549-51. Griffith AM, Wesson DE, Shanding B et al. Factors inffuencing postoperative recurrence of Crohn's disease in childhood. Gut 1991; 32: 491-5.

Inflammatory bowel disease in the pediatric age group

790 94.

95.

Davies G, Evans CM, Shand WS, Walker-Smith JA. Surgery for Crohn's disease in childhood: influence of site of disease and operative procedure on outcome. Br J Surg 1990; 77: 891-4. Telander RL. Surgical management of inflammatory bowel disease in children: In: Telander R, ed. Problems in General Surgery: Surgical Treatment of Inflammatory Bowel Disease. Philadelphia, JB Lippincott, 1993.

96. 97.

98.

Martin LW, LeCoultre C, Schubert WK. Total colectomy and mucosal proctectomy with preservation of continence. Ann Surg 1977; 186:477-80. Tiszlavicz L, Kapin M, Varkonyi A, Lorincz A, Bartyik K, Varkonyi T Adenocarcinoma of the colon developing on the basis of Crohn's disease in childhood. Eur J Pediatr 2001; 160: 168-72 Gillen CD, Walmsley RS, Prior P et al. Ulcerative colitis and Crohn's disease: a comparison of the colorectal cancer risk in extensive colitis. Gut 1994; 35: 1590-2.

43 Microscopic colitis: collagenous and lymphocytic colitis DiARMUID O'DONOGHUE AND KlERAN S H E A H A N

Introduction Microscopic colitis is now an accepted member of the inflammatory bowel disease family. Recognition came slowly following Lindstrom's classical description of collagenous colitis in 1976 [1], Read etal. first used the term microscopic colitis to describe a diffuse increase in inflammatory cells in colorectal biopsies from patients with chronic diarrhea [2]. Numerous reports have followed as the critical importance of obtaining colonic biopsies from patients with diarrhea, despite normal-looking mucosa, has been appreciated. The term 'lymphocytic colitis' was employed to distinguish patients who, like those with collagenous colitis, presented with watery diarrhea and a normal colonoscopy, had an expansion of intraepithelial lymphocytes but no increase in the subepithelial collagen layer [3]. Microscopic colitis is now the accepted terminology encompassing collagenous and lymphocytic colitis. Whether these are two separate conditions, or represent ends of a spectrum of a single disease, is a moot point.

Incidence and prevalence As many cases of microscopic colitis are probably undiagnosed it is difficult to assess incidence or prevalence figures. Bohr and colleagues in Sweden, in a retrospective study of pathology material and case records, estimated the prevalence of collagenous colitis to be 15.7 per 100 000 inhabitants with an annual incidence of 1.8 [4]. A recent study from Spain records an incidence figure of 1.8/100000 for collagenous colitis and 3.1 for the lymphocytic variety [5]. In this study microscopic colitis was identified in 9.5% of patients with watery diarrhea and a normal-looking mucosa. Females dominate in all large series; more so in collagenous than lymphocytic colitis [4-6]. Likewise all studies show a peak at

diagnosis in the late 50s and early 60s age groups. Thus, microscopic colitis is now the first disease to consider in the older woman presenting with chronic watery diarrhea.

Etiologic tlieories and associations The etiology of microscopic colitis is unknown; indeed lymphocytic infiltration and collagen formation may be the histologic end-points of a variety of insults. The female predominance, especially of collagenous colitis, lends credence to an autoimmune theory, and many diseases with a putative immune basis are associated with these conditions. Celiac disease, thyroid disorders, rheumatoid arthritis and diabetes are just some of the more common conditions greatly overrepresented in patients with microscopic colitis [7-11]. However, apart from a nonsignificant increase in antinuclear factor, autoantibodies are not increased in patients with collagenous colitis [12]. A higher than expected occurrence of microscopic colitis within families suggests a genetic predisposition for this disease [13]. The relationship between celiac disease and microscopic colitis is, perhaps, the most intriguing. All major reports on microscopic colitis have highlighted an unexpectedly high concurrence of the two disorders. Bohr's study from Sweden records celiac disease in 8% of 163 patients with collagenous colitis [4]. Although initially associated with collagenous colitis, it is now recognized to occur with both histologic varieties [14-16]. Gluten does not appear to be the offending substance, as many cases of microscopic colitis come to light when the diarrhea of celiac disease fails to resolve on a gluten-free diet despite good histologic recovery of the small bowel pathology [17]. The close relationship between these conditions has led to the investigation of the small bowel in a number of studies which, although based

Stephan R. Targan, Fergus Shanahan and Lor en C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 791-798. © 2003 Kluwer Academic Publishers. Printed in Great Britain

792 on small numbers, are nevertheless of considerable interest. Fine and colleagues have shown that the class II HLA genes DQ2 and DQ1,3 found in most patients with celiac disease, were also present in the majority of individuals with microscopic colitis and may explain the overlap between the two conditions [18]. Moayyedi and colleagues found small bowel abnormalities in four of nine patients with microscopic colitis; two had significantly elevated levels of intestinal antigliadin antibodies and two had abnormal small bowel permeability [19]. Wolber et al. record gastric lymphocytosis in two of four celiac patients who had colonic lymphocytosis [16] while flattening of the ileal mucosa has also been reported by various groups [20-22]. Bile salt malabsorption has been well documented despite normal ileal mucosa [21] and small bowel fluid and electrolyte absorption may be impaired [23]. These studies demonstrate that the small bowel (and perhaps stomach) of patients with microscopic colitis may be involved at a structural, immunologic or functional level, with overt celiac disease being the most recognisable manifestation. The role of non-steroidal anti-inflammatory drugs (NSAIDs) in the causation of microscopic colitis is also controversial. In Bohr's study 34% of patients were taking NSAIDs [7], whereas Veress ct al. found no association [22]. The most convincing evidence in favor of a cause-and-efTect comes from Riddell etal!^ case-control study showing long-term use of NSAI Ds in 19 of 31 patients as compared to just four of 31 controls [24]. The diarrhea improved in three patients on drug withdrawal and recurred in the one patient who was rechallenged. It may be argued that the colitis is related to the arthritis and not the medication, but it is our experience that this association is independent of the type of arthritis. The diarrhea of microscopic colitis may respond to bile salt resins [7, 25] and this has led Ung and colleagues to a careful evaluation of the possible role of bile salt malabsorption in these patients [21]. This study reports a 44% rate of bile salt malabsorption, as measured by the ^^Se-homocholic acid taurine test (^^Se-HCAT), in collagenous colitis. Eleven of 12 individuals with an abnormal ^^Se-HCAT responded rapidly to cholestyramine, as did 10 of 15 patients who had normal measurements. These findings raise the possibility that luminal factors other than bile salts may be involved in the etiopathogenesis of this inflammatory condition. Andersen and co-workers demonstrated fecal fluid cytotoxicity on cell lines in a patient with collagenous

Microscopic colitis colitis, and postulated bacterial toxins as causative agents [25]. There is some indirect evidence to support this hypothesis: fecal stream diversion can lead to clinical and histologic resolution of the colitis [26]; the diarrhea may respond, as mentioned above, to cholestyramine [21], and bismuth subsalicylate, a drug with anti-bacterial properties, is the most promising therapy to date for microscopic colitis [27].

Clinical presentation and differential diagnosis The clinical presentation of microscopic colitis is that of chronic watery diarrhea which may be associated with abdominal cramps. Systemic symptoms such as anorexia and weight loss are rare, and hematological and biochemical blood tests are often unhelpful [7]. Microscopic colitis is, by definition, a histopathologic diagnosis and thus depends on clinical awareness of the condition and obtaining adequate colonic biopsies. Rectal histology may be normal in a considerable number of patients with microscopic colitis, whereas biopsies taken from the left colon will identify almost all cases [5, 9, 28]. Other diagnoses to be considered in individuals with chronic watery diarrhea and a normal colonoscopy are outlined in Table 1. This is not an exhaustive list, and does not include the many infectious causes of watery diarrhea in the immunocompromised patient. The presentation of idiopathic bile salt malabsorption [29] or post-cholecystectomy diarrhea [30] may be indistinguishable from microscopic colitis. The diarrhea associated with neuroendocrine tumors such as gastrinomas or vipomas is usually severe Table 1. Differential diagnosis of patients with chronic watery diarrhea and a normal colonoscopy Microscopic colitis Celiac disease Giardiasis Ileal Crohn's disease Contaminated small bowel syndrome Idiopathic bile salt malabsorption Post-cholecystectomy diarrhea Gastrocolic fistula Carbohydrate malabsorption, e.g. lactose, sorbitol Neuroendocrine tumors Brainerd diarrhea Laxative abuse Irritable bowel syndrome

Diarmuid O'Donoghue and Kieran Sheahan

793

and accompanied by weight loss. The irritable bowel syndrome is placed last on this list, and should be only considered after a careful search for other pathology. This is particularly important for those patients presenting in their fifth or sixth decades.

Histopathologic features There is little doubt that pathologists underdiagnose microscopic colitis. Much of the reason for this is due to confusion in the literature with regard to terminology and to a lack of consensus with regard to specific diagnostic criteria. Like others, we use the umbrella term 'microscopic colitis' and subcategorize patients into lymphocytic and collagenous types. There are only rare reports of progression from lymphocytic to collagenous colitis or vice-versa [31-33]. Interpretation of serial or post-treatment biopsies may be difficult due to variability in sampling and lack of knowledge of the natural history of the disorders.

Lymphocytic colitis The most sensitive and specific features are a diffuse intraepithelial lymphocytosis and cuboidal change or flattening of the normal columnar surface colonocytes (Fig. 1). A detached or denuded surface epithelial layer is often a clue to the diagnosis, and may on occasions hinder the evaluation of intraepithelial lymphocytosis. The normal number of lymphocytes within surface colonocytes isfiveper 100 [3, 34]. Care should be taken to ignore intraepithelial lymphocytes (lELs) over lymphoid aggregates, as this is a normal phenomenon [35]. In our experience the

Figure 1. Diffuse intraepithelial lynfiphocytosis and cuboidal flattening of surface colonocytes in lymphocytic colitis (H & E, X 200).

Table 2. Diagnostic pitfalls in histopathology Collagenous colitis

Lymphocytic colitis

1. Tangential sectioning - at least three crypts in the vertical plane need to be viewed

1. Over-interpretation of patchy lEL (OLA stain useful in confirming lymphocytic colitis)

2. Misinterpretation of subnuclear cytoplasm of colonocytes as collagen

2. Over-interpretation of increased cellularity of the lamina propria

3. Over-reliance on H&E: trichrome stain is always necessary to confirm diagnosis of collagenous colitis 4. If full-thickness fibrosis or fibromuscular hyperplasia, consider ischemia and solitary rectal ulcer 5. Uncertain significance of thickened basement membrane (e.g. seen near diverticuli)

Microscopic colitis

794

Figure 2. Common leukocyte antigen (CLA) immunostainjng confirming the diagnosis of jympfiocytic colitis ( x 400).

mean number of surface lELs in lymphocytic colitis is 40 (range 20-60) [36]. Others who have performed quantitative studies showed a mean of 24.6 [3] and 37 lymphocytes per 100 surface epithelial cells [34]. Our study has used quantitative immunohistochemistry and therefore may be the more accurate. We also find that a confirmatory immunohistochemical stain for common leukocyte antigen (CLA) is useful on many occasions (Fig. 2). Focal intraepithelial lymphocytosis only (involvement of some but not all biopsies) is insufficient for diagnosis, and has been addressed in a few studies. Focal lymphocytic and collagenous colitis have been described as preceding Crohn's disease in one report [37]. Other features, however, including focal involvement of biopsies, abnormal colonoscopic findings, presence of numerous neutrophils and a foreignbody-type granuloma between the collagen fibers, helped to distinguish these cases in retrospect. In our experience differentiation from Crohn's disease is rarely a problem if strict histologic diagnostic criteria are applied and correlated with clinical and endoscopic findings. In addition, it is always difficult to entirely exclude so-called 'Brainerd' or epidemic diarrhea, which is most likely infectious in etiology [34, 38].

Using a cut-off of 15 lymphocytes per 100 surface colonocytes, Wang and colleagues have suggested abandoning the name lymphocytic colitis and replacing it with the term colonic epithelial lymphocytosis [15]. In this study 28 of 40 patients fitted the classical description of lymphocytic colitis but the remaining 12 were atypical, fulfilling the histologic but not the clinical or endoscopic criteria. These patients represented a heterogeneous group of disorders including idiopathic constipation, Crohn's disease and infectious colitis. A recent commentary suggests that the term 'colonic epithelial lymphocytosis' be reserved for these atypical cases which do not satisfy all the criteria for classic lymphocytic colitis [39].

Collagenous colitis All of the features of lymphocytic colitis outlined above are seen in collagenous colitis but the number of lELs tends to be lower [36, 39]. The sine qua non for the diagnosis of collagenous colitis is a thickened subepithelial collagen layer. There has been considerable controversy with regard to the exact thickness necessary to make the diagnosis. The upper limit of normal is stated to be 7 jam. A rule of thumb states that > 10 jim raises the possibility, while > 15 estab-

Diarmuid O'Donoghue and Kievan Sheahan

795

Figure 3. Lacy infiltration of superficial lamina propria by collagen (Masson-Trlchrome x 400).

lishes a diagnosis in the proper setting. More importantly, in our experience, is the pattern of injury; in particular, the lacy infiltration of the superficial capillaries of the lamina propria is characteristic and in doubtful cases can clinch the diagnosis (Fig. 3). This has previously been highlighted by Lazenby et al. who state that no absolute or minimum quantitative measure of subepithelial collagen should be used for the diagnosis [40]. The collagen accumulation is usually diffuse but can be patchy in up to 20% of cases. Prominent eosinophils are an occasional feature. Pericryptal myofibroblasts are also prominent. We have seen one case of collagenous colitis with pseudomembranes [41]. and other groups have also reported this finding [42, 43].

Pathogenesis It now seems clear that collagenous colitis is due to an aberrant activation of intestinal subepithelial myofibroblasts (ISEMF) [44]. It appears that the normal type IV collagen of the basement membrane is unaltered in the disease. A recent study has shown that pericryptal myofibroblasts normally secrete small amounts of types I, III and VI collagen around

the deep parts of the crypts. In the upper part of the crypts, and in the subepithelial location, increased amounts of type VI collagen and the matrix glycoprotein tenascin are produced. It is tenascin and type VI collagen which accumulate and comprise the majority of the subepithelial band. It has been suggested by this group that decreased synthesis of matrix metalloproteinases (MM?) may be responsible for the abnormal collagen accumulation, i.e. that the delicate balance between collagen production and resorption is disturbed [45]. This finding has been confirmed by Gunther and colleagues, who demonstrated reduced M M P l but also increased levels of the tissue inhibitor of matrix metalloproteinase 1 (TIMPl) in 12 cases [46]. They also confirmed the accumulation of tenascin, while also showing an absence of undulin. This tenascin/undulin dichotomy is a feature of a rapidly remodeling extracellular matrix as compared with the high undulin/low tenascin content in long-standing scar tissue. This may explain why these deposits can be removed within periods of a few months upon fecal stream diversion [26]. Controversy persists, however, with a recent report stating that type III collagen is predominant in the condition [47]. It is interesting to note that other forms of colonic fibrosis (e.g. Crohn's

Microscopic colitis

796 strictures) are composed predominantly of collagen types I, III and V [48]. Finally, elegant electron microscopy work has shown deficient fibroblast cell processes, focally deficient basal lamina and surface epithelial cells resting directly on a thickened collagen layer in this condition [49]. Immune mechanisms have been proposed as being important in microscopic colitis. The abnormalities detected have been similar in both conditions. These include an accumulation of CD4 positive-cells in the lamina propria and abnormal expression of class II MHC molecules on colonic epithelial cells. The increased surface lEL are predominantly CD8-positive, T cell receptor (TCR)a(3 phenotype [50]. No increase in TCRyS cells was identified. A number of groups are beginning to find differences in the colonic intraepithelial lymphocytosis seen in celiac disease. In particular, the majority of lEL in the colon in refractory sprue are CD8-negative [51 ]. This requires further study, but it is interesting that a recent report suggests that a loss of lEL CD8 expression correlated with refractory sprue and often heralded a cryptic intestinal lymphoma [52].

Pathophysiology The mechanism of the diarrhea seen in microscopic colitis has been evaluated in a small number of patients, with conflicting results. Giardiello and colleagues reported net fluid secretion, as distinct from the expected absorption, in the small bowel and colon of one subject and net jejunal secretion and reduced colonic absorption in a second patient. The induction of water and electrolyte secretion by the dihydroxy bile acids is a likely source of the diarrhea in those patients with bile salt malabsorption [8]. In contrast, a recent study has demonstrated reduced colonic water absorption in all six patients evaluated. These abnormalities were shown to be due to reduced active and passive sodium and chloride absorption and to reduced CL/HCO^ exchange [23]. The pathophysiology of the diarrhea appears to correlate with lamina propria cellularity and not the thickness of the collagen layer, suggesting that the inflammatory reaction is the important component in the causation of the diarrhea [53]. Elevated levels of prostaglandin E2 have been postulated as a circulating secretagogue in a single case study [54]. A recent and elegant study from Burgel and colleagues strongly supports the view that, in the collagenous variety of the disease, reduced sodium and

Therapeutic options in Microscopic Colitis Bismuth Subsalicylate

L, L

"Antibiotics Cholestyramine I—Salazopyrine/5 ASA Corticosteroids +/- Azathioprine/6MP (inci. Budesonide)

L

Ileostomy/colectomy

Figure 4. Therapeutic options in microscopic colitis.

chloride absorption along with active chloride secretion are the main components of the diarrhea. The subepithelial collageous band is felt to act as a diff'usion barrier while down-regulation of tight junction molecules may contribute to a ieak flux mechanism' [55].

Treatment Microscopic colitis usually follows a benign course [6, 56, 57] and treatment options must keep this in mind. It is our experience that most patients, by the time of diagnosis, have already tried simple remedies such as antidiarrheal agents and need more effective therapy. Treatment is empirical, as with many diseases of unknown cause, but recent reports now allow for a more structured approach. Fine and Lee's study of the effectiveness of bismuth subsaHcylate makes this the first drug to consider [27]. In this report 11 of 13 patients responded to the drug given orally for 8 weeks. This was mirrored by histopathologic improvement, including resolution of the thickened subepithelial collagen in the six patients who displayed this feature prior to therapy. The treatment was well tolerated and nine of the group remained in remission with a follow-up of 7-28 months. Options thereafter are outlined in Fig. 4. The case for antibiotics is less certain and based on retrospective analysis [7, 57]. Nevertheless, the hope of avoiding long-term medication makes this approach attrac-

Diarmuid O'Donoghue and Kieran Sheahan tive. Metronidazole is the most frequently used antimicrobial. Ung etaU^ report of symptom resolution in 21 of 27 patients given a bile salt-binding agent would make this the third line of treatment [21]. These drugs are not always well tolerated, and this may limit their effectiveness. Thereafter the choice of treatment is based on little more than case reports and personal experience. Thirty-seven of 108 patients treated in a Swedish study benefited from salazopyrine while 45 individuals could not tolerate the medication [7]. Corticosteroids are effective but unacceptably high dosage is often required and relapse is common on reduction or withdrawal. This is far from ideal in a population that is predominantly female and postmenopausal. This has led Pardi and colleagues to use azathioprine in patients with steroid-dependent disease [58]. Budesonide, a locally acting corticosteroid with low systemic activity, may alleviate concerns regarding steroid toxicity and has recently been shown to be effective in collagenous colitis [59]. Surgery is occasionally required for severe disease [31] but a diversion ileostomy might be a better option and would have the benefit of allowing time for other therapies to act.

797

8.

9.

10. 11. 12. 13. 14. 15.

16. 17.

Acknowledgements The authors wish to acknowledge the immense contribution made by Dr Darren Treanor to the scientific background and preparation of this chapter.

18.

References

20.

1. 2. 3.

4. 5.

6.

7.

Lindstrom CG. 'Collagenous colitis' with watery diarrhoea - a new entity? Pathol Eur 1976; 11: 87-9. Read NW, Krejs GJ, Read MG, Santa Ana CA, Morawski SG, Fordtran JS. Chronic diarrhea of unknown origin. Gastroenterology 1980; 78: 264-71. Lazenby AJ, Yardley JH, Giardiello FM, Jessurun J, Bayless TM. Lymphocytic ('microscopic') colitis: a comparative histopathologic study with particular reference to collagenous colitis. Hum Pathol 1989; 20: 18-28. Bohr J, Tysk C, Eriksson S, Jarnerot G. Collagenous colitis in Orebro, Sweden, an epidemiological study 1984-1993. Gut 1995; 37: 394^7. Fernandez-Banares F, Salas A, Forne M, Esteve M, Espinos J, Viver JM. Incidence of collagenous and lymphocytic colitis: a 5-year population-based study. Am J Gastroenterol 1999;94:418-23. Baert F, Wouters K, D'Haens G et al. Lymphocytic cohtis: a distinct clinical entity? A clinicopathological confrontation of lymphocytic and collagenous colitis. Gut 1999; 45: 37581. Bohr J, Tysk C, Eriksson S, Abrahamsson H, Jarnerot G. Collagenous colitis: a retrospective study of clinical pre-

19.

21. 22. 23. 24. 25. 26. 27.

sentation and treatment in 163 patients. Gut 1996; 39: 84651. Giardiello FM, Bayless TM, Jessurun J, Hamilton SR, Yardley JH. Collagenous colitis: physiologic and histopathologic studies in seven patients. Ann Intern Med 1987; 106: 46-9. Jessurun J, Yardley JH, Giardiello FM, Hamilton SR, Bayless TM. Chronic colitis with thickening of the subepithelial collagen layer (collagenous colitis): histopathologic findings in 15 patients. Hum Pathol 1987; 18: 839-48. Soulier C, Baron D, Saraux A, Robert FX, Le GoffP. Four new cases of collagenous colitis with joint symptoms. Rev Rhum Engl Ed 1996; 63: 593-9. van Tilburg AJ, Lam HG, Seldenrijk CA et al. Familial occurrence of collagenous colitis. A report of two families. J Clin Gastroenterol 1990; 12: 279-85. Bohr J, Tysk C, Yang P, Danielsson D, Jarnerot G. Autoantibodies and immunoglobulins in collagenous colitis. Gut 1996; 39: 73-6. Jarnerot G, Hertervig E, Granno C et al. Familial occurrence of microscopic cohtis: a report on five families. Scand J Gastroenterol 2001; 36: 959-62. DuBois RN, Lazenby AJ, Yardley JH, Hendrix TR, Bayless TM, Giardiello FM. Lymphocytic enterocolitis in patients with 'refractory sprue'. J Am Med Assoc 1989; 262: 935-7. Wang N, Dumot JA, Achkar E, Easley KA, Petras RE, Goldblum JR. Colonic epithelial lymphocytosis without a thickened subepithelial collagen table: a clinicopathologic study of 40 cases supporting a heterogeneous entity. Am J Surg Pathol 1999; 23: 1068-74. Wolber R, Owen D, Freeman H. Colonic lymphocytosis in patients with celiac sprue. Hum Pathol 1990; 21: 1092-6. Fine KD, Meyer RL, Lee EL. The prevalence and causes of chronic diarrhoea in patients with celiac sprue treated with a gluten-free diet [Published erratum appears in Gastroenterology 1998; 114: 424^5]. Gastroenterology 1997; 112: 18308. Fine KD, Do K, Schulte K et al. High prevalence of celiac sprue-like genes and enteropathy in patients with the microscopic colitis syndrome. Am J Gastroenterol 2000; 95: 1974-82. Moayyedi P, O'Mahony S, Jackson P, Lynch DA, Dixon MF, Axon AT. Small intestine in lymphocytic and collagenous colitis: mucosal morphology, permeability, and secretory immunity to gliadin. J Clin Pathol 1997; 50: 527-9. Marteau P, Lavergne-Slove A, Lemann M et al. Primary ileal villous atrophy is often associated with microscopic colitis. Gut 1997;41:561-4. Ung KA, Gillberg R, Kilander A, Abrahamsson H. Role of bile acids and bile acid binding agents in patients with collagenous colitis. Gut 2000; 46: 170-5. Veress B, Lofberg R, Bergman L. Microscopic colitis syndrome. Gut 1995; 36: 880-6. Bo-Linn GW, Vendrell DD, Lee E, Fordtran JS. An evaluation of the significance of microscopic colitis in patients with chronic diarrhoea. J Clin Invest 1985; 75: 1559-69. Riddell RH, Tanaka M, Mazzoleni G. Non-steroidal antiinflammatory drugs as a possible cause of collagenous cohtis: a case-control study. Gut 1992; 33: 683-6. Andersen T, Andersen JR, Tvede M, Franzmann MB. Collagenous colitis: are bacterial cytotoxins responsible? Am J Gastroenterol 1993; 88: 375-7. Jarnerot G, Tysk C, Bohr J, Eriksson S. Collagenous colitis and fecal stream diversion. Gastroenterology 1995; 109: 449-55. Fine KD, Lee EL. Efficacy of open-label bismuth subsalicylate for the treatment of microscopic colitis. Gastroenterology 1998; 114:29-36.

Microscopic colitis

798 28.

29. 30. 31.

32.

33. 34.

35. 36. 37. 38.

39.

40.

41. 42. 43.

Carpenter HA, Tremaine WJ, Batts KP, Czaja AJ. Sequential histologic evaluations in collagenous colitis. Correlations with disease behavior and sampling strategy. Dig Dis Sci 1992; 37: 1903-9. Thaysen EH, Pedersen L. Idiopathic bile acid catharsis. Gut 1976; 17:965-70. Arlow FL, Dekovich AA, Priest RJ, Beher WT. Bile acidmediated postcholecystectomy diarrhoea. Arch Intern Med 1987; 147: 1327-9. Bowling TE, Price AB, al Adnani M, Fairclough PD, Menzies-Gow N, Silk DB. Interchange between collagenous and lymphocytic colitis in severe disease with autoimmune associations requiring colectomy: a case report. Gut 1996; 38:788 91. Sylwestrowicz T, Kelly JK, Hwang WS, Shaffer EA. Collagenous colitis and microscopic colitis: the watery diarrhoea colitis syndrome. Am J Gastroenterol 1989; 84: 763 8. Teglbjaerg PS, Thaysen EH, Jensen HH. Development of collagenous colitis in sequential biopsy specimens. Gastroenterology 1984; 87: 703 9. Bryant DA, Mintz ED, Puhr ND, Griffm PM, Petras RE. Colonic epithelial lymphocytosis associated with an epidemic of chronic diarrhoea. Am J Surg Pathol 1996; 20: 1102 9. Goldman H, Antonioli DA. Mucosal biopsy of the rectum, colon, and di.slal ileum. Hum Pathol 1982; 13: 981 1012. Trcanor DT, Morgan S, O'Donoghuc DP, Shcahan K. Microscopic colitis: a detailed clinicopathological study. Gul 2000; 46: A8 (Abstract). Goldstein NS, Gyorfi T. Focal lymphocytic colitis and collagenous colitis: patterns of Crohn's colitis? Am J Surg Pathol 1999; 23: 1075 81. Ostcrholm MT, MacDonald KL, White KE ct al. An outbreak of a newly recognised chronic diarrhoea syndrome associated with raw milk consumption. J Am Med Assoc 1986:256:484 90. Lamps LW, Lazenby AJ. Colonic epithelial lymphocytosis and lymphocytic colitis: descriptive histopathology versus distinct clinicopathologic entities. Adv Anat Pathol 2000; 7: 210 13. Lazenby AJ, Yardley JH, Giardiello FM, Bayless TM. Pitfalls in the diagnosis of collagenous colitis: experience with 75 cases from a registry of collagenous colitis at the Johns Hopkins Hospital. Hum Pathol 1990; 21: 905-10. Treanor DT, Gibbons DG, O'Donoghue DP, Sheahan K. Pseudomembranes in collagenous colitis. Histopathology 2001;38:83-4. Giardiello FM, Hansen FC, Lazenby AJ et al. Collagenous colitis in setting of nonsteroidal antiinflammatory drugs and antibiotics. Dig Dis Sci 1990; 35: 257-60. Reyes V, Bronner M, Haggitt R. Pseudomembranes and collagenous colitis. Am J Clin Pathol 1999; 112: A542 (Abstract).

44. 45.

46.

47. 48. 49. 50. 51.

52. 53.

54.

55. 56. 57. 58. 59.

Powell DW, Mifflin RC, Valentich JD, Crowe SE, Saada JI, West AB. Myofibroblasts. II. Intestinal subepithelial myofibroblasts. Am J Physiol 1999; 277: CI83-201. Aigner T, Neureiter D, Muller S, Kuspert G, Belke J, Kirchner T. Extracellular matrix composition and gene expression in collagenous colitis. Gastroenterology 1997; 113:136-43. Gunther U, Schuppan D, Bauer M, Matthes H, Stallmach A, Schmitt-Graff A et al. Fibrogenesis and fibrolysis in collagenous colitis. Patterns of procollagen types I and IV, matrix-metalloproteinase-1 and -13, and TIMP-1 gene expression. Am J Pathol 1999; 155: 493 503. Lamps LW, Bonsib SM, Mitros FA. Collagenous colitis: not your basic basement membrane abnormality. Mod Pathol 2000; 13: A83 (Abstract). Graham MF, Diegelmann RF, Elson COS ct al. Collagen content and types in the intestinal strictures of Crohn's disease. Gastroenterology 1988; 94: 257 65. Hwang WS, Kelly JK, Shaffer EA, Hershfield NB. Collagenous colitis: a disease of pericryptal fibroblast sheath? J Pathol 1986; 149: 33 40. Mosnier JF, Larvol L, Barge iet al. Lymphocytic and collagenous colitis: an immunohistochcmical study. Am J Gastroenterol 1996; 91: 709 13. Fine KD, Lee EL, Meyer RL. Colonic histopathology in untreated celiac sprue or refractory sprue: is it lymphocytic colitis or colonic lymphocytosis? Hum Pathol 1998; 29: 1433 40. Cellier C, Dclabcssc E, Hclmer C ct al. Refractory sprue, coeliac disease, and enteropathy-associaled T-cell lymphoma. Lancet 2000; 356: 203 8. Lee E, Schiller LR, Vendrell D, Santa Ana CA, Fordtran JS. Subepithelial collagen table thickness in colon specimens from patients with microscopic colitis and collagenous colitis. Gastroenterology 1992; 103: 1790 6. Rask-Madsen J, Grove O, Hansen MG, Bukhave K, Scient C, Henrik-Nielsen R. Colonic transport of water and electrolytes in a patient with secretory diarrhoea due to collagenous colitis. Dig Dis Sci 1983; 28: 1141 6. Burgel N, Bojarski C, Mankertz J, Zeitz M, Fromm M, Jorg D. Mechanisms of diarrhea in collagenous colitis. Gastroenterology 2002; 123:433 43. Bonderup OK, Folkersen BH, Gjersoe P, Teglbjaerg PS. Collagenous colitis: a long-term follow-up study. Eur J Gastroenterol Hepatol 1999; 11: 493-5. Mullhaupt B, Guller U, Anabitarte M, Guller R, Fried M. Lymphocytic colitis: clinical presentation and long term course. Gut 1998; 43: 629-33. Pardi DS, Loftus EV, Tremaine W, Sandborn WJ. Treatment of refractory microscopic colitis with azathioprine and 6mercaptopurine. Gastroenterology 2001; 120: 1483^. Miehlke S, Heymer P, Bethke B et al. Budesonide treatment for collagenous colitis: a randomized, double-blind, placebo-controlled, multicenter trial. Gastroenterology 2002; 123: 978-84.

44 Colon ischemia SETH E. PERSKYAND LAWRENCE J. BRANDT

Introduction Colon ischemia (CI) is the most common form of gastrointestinal ischemia, accounting for more than 50% of all cases of such ischemic injuries [1]. Although colonic gangrene has been recognized for more than 100 years it was not until the 1950s and 1960s that milder forms of CI were described [2]. Today we recognize six patterns of CI: reversible colopathy, transient colitis, chronic ulcerating colitis, gangrene, stricture, and fulminant universal colitis. While CI most often is idiopathic, some episodes can be linked to proven etiologies. CI commonly is misdiagnosed as inflammatory bowel disease (IBD); however, clinicians must be able to differentiate these two causes of colitis, as their management and prognosis differ markedly. This chapter will discuss current knowledge of the etiologies, pathogenesis, presentations, and management of CI.

Incidence/etiology Although CI is the most common form of gastrointestinal ischemia it remains largely underdiagnosed. Stated to be the cause of 1 in 2000 hospitalizations [3], its exact incidence remains unknown, since a large number of cases go undetected; they are either subclinical in nature, mild and resolve spontaneously before patients present to a physician, or are misdiagnosed as infectious colitis or IBD [4]. While the incidence of IBD classically has been described as having a bimodal age distribution, most cases in the second peak actually are cases of CI rather than of new onset IBD [4, 5]. In one series [4], after reviewing the clinical, radiographic and pathologic findings of patients who developed new-onset colitis after age 50, 75% of these cases were classified as definite CI, whereas only 15% of these cases were classified as definite ulcerative or Crohn's colitis. In the remaining 10% of cases a definitive diagnosis could not be established. Therefore, when encounter-

ing a patient over the age of 50 with new-onset colitis, CI is the most likely cause, and new-onset IBD should be considered only after carefully excluding CI. CI aff'ects both sexes equally, and unlike IBD, which predominantly affects young patients, 90%) of sporadic cases of CI occur in patients over age 60. Affected patients often have atherosclerotic vascular disease, although there is no close correlation between the severity of such changes and the occurrence of CI. There are several recent small case series [6-8] that report a lesser proportion of cases in the elderly, with up to 34% of cases of CI occurring in patients below the age of 50 [7]. It is this subset of patients who are often misdiagnosed with IBD. Unlike IBD, however, the vast majority of patients with CI improve spontaneously, i.e. with no or conservative therapy, whereas patients with IBD usually improve only after receiving definitive therapy with aminosalicylates or corticosteroids. The proven etiologies of CI in the older population diff'er markedly from those in younger patients (Table 1). Most commonly, however, no etiology is found and it is believed that CI develops as a result of spontaneous episodes of local 'nonocclusive' ischemia, in association with small vessel atherosclerotic disease. Angiograms in affected patients rarely correlate with clinical disease, and only show age-related changes such as narrowing and tortousity of the inferior mesenteric artery and its branches [56]. Less commonly, systemic alterations in blood flow, such as hypoperfusion from underlying cardiovascular disease, sepsis, hypovolemia, or inferior mesenteric artery (IMA) thrombosis, lead to CI. Constipation and straining during bowel movements also have a deleterious, albeit transient, effect on colonic blood flow, and therefore may render the colon susceptible to ischemic injury. There are two special clinical problems which have a significant association with CI. First, up to 10%) of patients undergoing elective abdominal aortic sur-

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 799-810. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Colon ischemia

800 Table 1. Characteristics of colon Ischemia in young vs. older subjects* Younger patients

Elderly patients

Identifiable causes

(Approximately 40% of patients) Coagulopathies Vasculitis Infections Long-distance running Medications: Oral contraceptives, pseudoephedrine, danazol, sumatriptan, psychotropic agents, cocaine, alosetron

(Approximately 10% of patients) Aortic surgery Cardiac failure/dysrhythmias Arterial embolus/thrombosis Shock Medications: digoxin, psychotropic agents, flutamide, hormone replacement therapy

Idiopathic etiology associations

Approximately 60% of cases

Approximately 90% of cases Distal obstructing lesions ( ~ 5%): carcinoma, volvulus, constipation, fecal impaction; colonic stricture; diverticulitis

Course

May be recurrent. Progression to chronic CI (i.e. chronic ulcerating colitis, stricture) unstudied

Single episode. Progression to chronic CI (i.e. chronic ulcerating colitis, stricture) uncommon

Location

Right side Rectum

Splenic flexure Sigmoid colon

Differential diagnosis

IBD; infection

Infection; carcinoma; IBD less likely

*The presentation and clinical course is usually identical in both younger and older patients; this table merely emphasizes differences that may occasionally be seen.

gery develop CI. This is related to hypotension, intraoperative trauma to the colon, IMA loss during aneurysmectomy, or prolonged cross-clamp time. In the setting of emergent aortic surgery for repair of a ruptured abdominal aortic aneurysm, CI develops in up to 60% of cases. By assessing colonic perfusion intraoperatively with laser Doppler flowmetry [57, 58] and photoplethysmography [59], several investigators have been able to predict the development of postoperative CI. The intramural pH of the sigmoid colon before and after cross-clamping the aorta, as determined by tonometry, has also been used to predict postoperative CI [60]. Routine intraoperative reimplantation of the IMA to prevent postoperative CI has been advocated by one group [61], but is not routinely performed. The second clinical problem associated with an increased risk of CI is colon obstruction. Approximately 20% of older patients with CI have a distal and potentially obstructing lesion, such as carcinoma, stricture, fecal impaction, diverticulitis, or volvulus. CI probably develops from distension of the colon proximal to the obstructing lesion, which leads to increased intraluminal pressure and

decreased mucosal blood flow. Less commonly such lesions are within a segment of ischemic colon and, rarely, distal to it. The causes of CI in younger persons are markedly different from those in elderly patients. Often these causes may not be immediately apparent to the clinician, and making a diagnosis of CI instead of new-onset IBD may be especially challenging. A well-described but rare cause of CI in young patients is mesenteric vasculitis. Such vasculidites include polyarteritis nodosa [31, 32], Takayasu's arteritis, or Buerger's disease [30], as well as systemic autoimmune disease, such as systemic lupus erythematosus (SLE) [27, 28] or Sjogren's syndrome [62]. Hypercoagulable states, such as protein S or C deficiency, antithrombin III deficiency, or activated protein C resistance (APCR, factor 5 Leiden) may also lead to CI in young patients. The anticoagulant actions of the protein C system are important in maintaining microvasculature patency [63]. Heterozygosity for APCR is present in approximately 5% of the general population, with affected individuals having an eight-fold risk of venous thrombotic disease [25]. For the rare individual with the homo-

Seth E. Persky and Lawrence J. Brandt

801

Table 2. Causes of colon ischemia Non-latrogenic causes

Iatrogenic causes

Non-occlusive ischemia Cardiac failure or dysrhythmias Shock Obstructing colon carcinoma Volvulus Strangulated hernia Arterial embolus Inferior mesenteric artery thrombosis Cholesterol emboli [9,10] Phlebosclerosis[11,12] Pancreatitis [13,14] Amyloid [ 1 0 , 1 5 - 1 7 ] Idiopathic dysautonomia [18] Allergy [19, 20] Trauma Ruptured ectopic preganancy Long-distance running [21 ] Pit viper toxin [22] Pheochromocytoma [23] Kawasaki's disease [24] Hematologic disorders Protein C and S deficiencies Sickle cell disease Antithrombin III deficiency Activated protein C resistance [25] Prothrombin 2021OA mutations [26] Vasculitis Systemic lupus erythematosus [27-29] Rheumatoid arthritis Thromboangitis obliterans [30] Periarteritis nodosa (hepatitis B) [31-32] Takaysu's arteritis Infections Escherichia co//0157:H7 [33, 34] Angiostrongylus costaricensis Cytomegalovirus Hepatitis B (polyarteritis nodosa) [31, 32]

Surgical Aneurysmectomy Aortoiliac reconstruction Gynecologic operations Exchange transfusion Colonic bypass Lumbar aortography Colectomy with IMA ligation Colonoscopy and barium enema [35, 36]

zygous defect for APCR tlie risk of tlirombosis increases 50-100-fold. Recently, a young woman with APCR was reported to develop ischemic colitis [25]; she was also taking oral contraceptives, another risk factor for ischemic colitis. Lastly, another hypercoagulable state is linked to genetic mutations in the prothrombin 20210 A allele, and bowel ischemia has recently been described in patients with this mutation [26]. In that report one of the two cases had CI. There are a host of medications which may cause CI (see Table 2). Use of these medications may lead to ischemic injury via mesenteric vasoconstriction or thrombosis. The most commonly implicated medica-

Medications Digitalis Estrogens [25, 37, 38] Danazol [39] Progestins [40] Gold compounds Psychotropic drugs [41 ] Pseudoephedrine [42] Sumatriptan [43] Cocaine [44] Immunosuppressive agents [45] NSAID [46] Imipramine [47] Methamphetamines [48, 49] Vasopressin [50] Ergot [51] Oral hyperosmotic saline laxatives [52] Interferon Alpha [53] Glycerin enema [54] Flutamide [55] Alosetron

tions have been oral contraceptive agents; in one study [37], 59% of younger individuals with reversible CI used these preparations. As mentioned above, patients with underlying hypercoagulable states who use oral contraceptive agents may be at a particularly increased risk for CI. Several other medications have recently been described in case reports to cause CI, including: sumatriptan [43], interferon-alpha [53], pseudoephedrine [42], non-steroidal anti-inflammatory agents [46], vasopressin [50], Danazol [39], imipramine [47], psychotropic agents [41], and digitalis. Several illicit drugs may also lead to CI, including cocaine [44] and methamphetamine [48, 49].

802

Anatomy of colonic vasculature and pathogenesis of colon ischemia Blood is supplied to the colon by the superior mesenteric artery (SM A), and the inferior mesenteric artery (IMA), which respectively carry blood to the proximal and distal parts of the colon. The SMA gives off the middle colic artery and then branches into: (a) the right colic artery, which supplies the distal ascending colon, hepatic flexure, and proximal half of the transverse colon; and (b) the ileocolic artery, which supplies the terminal ileum, cecum, and proximal ascending colon. The IMA, which is smaller than the SMA, branches into: (a) the left colic artery, which supplies the distal half of the transverse colon and the descending colon; (b) multiple sigmoid arteries, which supply the sigmoid colon; and (c) the superior rectal artery, which supplies the rectum. The distal rectum is also supplied by the middle and inferior rectal arteries, which are branches of the internal iliac artery. Additionally, there are extensive collateral arteries between the SMA and the IMA, limiting the propensity of the colon to ischemic injury. These collateral vessels include: (a) the marginal artery of Drummond, which is close to and parallels the wall of the intestine; (b) the central anastomotic artery, which is larger and more centrally placed; and (c) the arc of Riolan at the base of the mesentery. If either the SMA or IMA is occluded, a large collateral vessel, termed the 'meandering artery', may develop, representing a dilated central anastomotic artery or arc of Riolan. Resting blood flow to the colon is relatively low (8-44 ml/min per 100 g tissue) in comparison to other parts of the gastrointestinal tract [64], and colonic blood flow is even more diminished during functional motor activity of the colon, e.g. bowel movements, making the colon uniquely susceptible to ischemia. There are certain areas of the colon where collateral flow between the SMA and the IMA is crucial to maintain colonic integrity; these are the areas especially prone to ischemic injury. The most important of these 'watershed' areas is the splenic flexure, or Griffith's point, which has the highest propensity for CI, especially in individuals in whom the marginal artery of Drummond is diminutive or absent. The rectosigmoid junction, or the point of Sudeck, is another so-called watershed region. This region is dependent on anastomotic branches of the IMA and

Colon ischemia the internal iliac artery, and CI may develop if these anastomotic branches are limited. CI results from local hypoperfusion, with or without reperfusion injury. During hypoperfusion the involved colonic segment is deprived of important nutrients and oxygen; during extensive periods of hypoperfusion tissue hypoxia leads directly to cell death, with damage progressing outward from the mucosa to the serosa [65]. It has been shown that high levels of renin and angiotensin, which are elevated during periods of systemic hypotension, may exacerbate splanchnic vasospasm, leading to further colonic hypoperfusion and ischemia [66]. During relatively short periods of hypoperfusion, however, most of ischemic damage results from subsequent reperfusion injury. In one study [67], the injury observed after 3 h of hypoperfusion and 1 h of reperfusion was more severe than that observed after 4 h of isolated hypoperfusion. There are extensive data showing that, during reperfusion, there is mucosal neutrophil accumulation, with release of reactive oxygen metabolites by activated neutrophils as part of the inflammatory process [64]. In an equine model of CI [68], one group of horses was subjected to 6 h of hypoperfusion, and another group to 3 h of hypoperfusion followed by 3 h of reperfusion. Concentrations of neutrophils within the mucosa were significantly higher during reperfusion as compared with the corresponding times in the group subjected to prolonged hypoperfusion. The reactive oxygen metabolites released from activated neutrophils are derived from molecular oxygen, and mediate both mucosal damage and disruption of microvascular integrity. Reactive oxygen metabolites cause peroxidation of lipids, thereby disrupting cell membranes and destroying cell compartmentation, resulting in cell lysis. Reactive oxygen metabolites also alter nucleotide function, and damage cellular proteins and enzymes. These effects ultimately lead to impaired cellular function or necrosis. Several reactive oxygen metabolites are released during reperfusion injury (Table 3). First, the univalent reduction of oxygen produces the superoxide anion radical, O2. The cellular toxicity associated with superoxide is usually attributed to its role as a precursor to more reactive oxygen species, such as hydrogen peroxide (H2O2) and the hydroxyl radical (OH), which is formed via the Haber-Weiss reaction (O2 + H2O2). The Haber-Weiss reaction is accelerated by the presence of transition metals, such as iron, which are present in high concentrations in intestinal mucosa. Hypochlorous acid (HOCl) is

Seth E. Persky and Lawrence J. Brandt

803

Table 3. Xanthine oxidase pathway for induction of Ischemia

In Non-Ischemic

Tissue: XDH

Hypoxanthine + R^O + NAD1+ During Tissue

*-

Xanthine + NADH + H+

Ischemia: xo

Hypoxanthine + HjO + 2O2

Xanthine + 2O2 + H2O

Xanthine +

xo

Reactive Oxygen Metabolite

Uric Acid +

Production: H2O2

Univalent reduction of oxygen

*• O2

divalent reduction or dismutation

O.

H2O2 (Hydrogen Peroxide Radical) Haber-Weiss Reaction

0 / + HP,

- -OH

another potent oxidizing agent formed from the reaction of H2O2 with CI" and H"^ when activated neutrophils secrete the enzyme myeloperoxidase (MPO). In animal models of reperfusion injury, treatment with antioxidants, such as superoxide dismutase (SOD), catalase, and dimethyl sulfoxide (DMSO), as well as with iron chelators, such as desferoxamine, has been shown to attenuate reperfusion injury by suppressing the formation of these reactive oxygen metabolites [8, 69, 70]; such agents have not been tested clinically in CI. In addition to their direct cytotoxic effects, reactive oxygen metabolites also initiate the production and release of leukocyte chemoattractants, such as leukotriene B4 and platelet-activating factor (PAF), which increase polymorphonuclear leukocyte migration, leading to further oxidant production [71, 72]. Activated neutrophils also release a variety of

(HydroxylRadical)

enzymes, such as collagenase and elastase, that can directly injure mucosa and the microvasculature. Reactive oxygen metabolites are also formed by the oxidant-producing enzyme, xanthine oxidase (XO). In ischemic tissue XO has the ability to generate O^ and H2O2 during the oxidation of xanthine and hypoxanthine. Normally, i.e. in nonischemic tissue, XO does not produce reactive oxygen metabolites, as it uses NAD"^ instead of O2 as an electron-acceptor. During tissue ischemia, however, due to a proteolytic conversion in the enzyme, XO uses O2 as an electron-acceptor, generating reactive oxygen metabolites. The XO inhibitor, allopurinol, has been shown to attenuate both the epithelial cell necrosis and the increased microvascular permeability observed after reperfusion of the rat small intestine [73], but this agent has not been tested in the clinical setting.

804

Clinical presentation The clinical presentation of CI varies with the underlying cause, extent of vascular obstruction, and presence or absence of comorbity. The spectrum of CI is wide (Table 4), and includes reversible colopathy, transient colitis, chronic ulcerating colitis, gangrene, stricture, and fulminant universal colitis. Most patients with CI present with the acute onset of mild, crampy, left lower quadrant pain, followed by the urge to defecate. These symptoms are followed over the next 24 h by the passage of bright red or maroon blood mixed with diarrheal stool. In one study of 47 patients, 87% had abdominal pain, 68% had diarrhea, and 38% had nausea and vomiting [74]. Gastrointestinal blood loss is not of hemodynamic significance. If there is hemodynamic instability, or if significant blood loss requiring transfusion occurs, other diagnoses should be considered. Approximately 10% have only subclincal disease and present later with constipation or obstruction secondary to stricture formation. The elderly may also present with altered levels of consciousness. Prior to the onset of symptoms the vast majority of patients with CI feel well. CI, however, may also develop shortly after a major cardiac event, such as a myocardial infarction or new-onset atrial fibrillation, or during hemodialysis-associated hypotension; these patients have higher mortality than do patients without these precipitating factors. In several reports [7, 75], approximately 30% of CI cases occurred immediately following such episodes of systemic hypotension. Unfortunately, even with classic presentations, the diagnosis is often overlooked. In one series [7], CI

Colon ischemia Table 4. Types and approximate incidences of colon ischemia Type

Incidence (%)

Reversible colonopathy Transient colitis Chronic ulcerating colitis Stricture Gangrene Fulminant universal colitis

30-40 15-20 20-25 10-15 15-20

ilh««aioeH» '-^ 3 -S o^^ >^ ^ "=^ •o

I in

X o

> ^

: -D .E o 1 p o uo

^

»

X

a> X

^

I E S

^

;

o

:E

^ Q. - C

'CL5

^

CD O

CO

o

>^

O

_c "o

X

X

CO

cvT h-

1—1

CO >^ • a

o 45 LO

57 LO o

C/5

05

CO

•o CsJ

1

s£ oo

C/)

CO

'— ' "^ CO 1 —

%i

CO CO

LO

"^ >; E

'co

CO " O •o 03 03 O

•o

cu

•g5 C3) E

2 E o LO

CO

E LO

o o "o

O)

CO

CD

o

CO O LO

o 1 o

LO c

o .E ' o

_g CO X c/p

X

So

•O

CD

LO tfi

> ^o CO •a T

E o

CNj"

CO

CO

CO CO •a •a

LO

CO

o

T

>^

o

o

'co

1

'co T3 CO

1

CO

"O (D

o

c^ C2.

_o 'o o o CO

CO c^-

CO T3 CO

E

CO •a CO 03

03

E

S

03 C35 03

o

o

.1

CO

1s

^CO

o E

M CO

03 CO -£=

"55 03

'cr o

"03

(5

QI

CD jQ


"

03 _C

1 1 1

LU

O"

Q. 5 CO

•3 C«

03

a» E

is

CO

03

c: .§ o .5r CO cJ

IS^ o ci

LU UJ

2

c:

o

S -1 P CO

in

03

LU

LU ^

CO

^

CL

.o CO

o N o o

8 e

.=3

CO

^ IS S

.CO • ^

Q . Ui

iS i5

QQ

:^

CO .CO

e ^ C: E -52 ^

—

-c:

cj

X

CO

t^

tz

>^

-^

CO

5

i=5~ Ci Co

=3 .CO

UJ CD

c: CO

c o ex •2,

2

CO

to

^

CO

CX O)

Q.

^ 1—

03

1

11

•a CD

o

>

c:

>

03

VUi

CD CO c^

"c

.CO

CO ^ 5

x'

CO

•i -2 -^ ^ • J-

-5 : ^

E CO

C3 CO

CO

2

C3)

E a> sz o

"^

E

7 E

O

o k— o

o

E .E O

^ CO X

CO

CO

C3) C3)

o Q. Q. Z O C5

Q.

>,

E o O) E o E o LO o LO 03 o

;a "c o

< ^

CO

CD O M M CO 03 CO ;o T3

CO X

E

3 E

o

C3) X 3

o

%

Q.

•;z

CO •o

O

c c o "o CO X

o

"co

LO

"53

CO LO •a CO

CD

C3)

o iS

CO M CO X T3

o

crT

-C

(D 03

E E CZ) o CD o LO

CO

"o

"O

>

C3) C3)

LO

E

'co 'co

(D

E E o o o o

E

CD CO

CO •o

LO CsJ .E

Q. a . .29 - 0 o N

S IE

o

CD

a>

E o E

CO op QL Q .

E CJ o

E o o LO

1

LO CVJ

i5

LO CVJ

o CO

CO • D

C/5 cn . Q J3 CO

E o

>^

O CO

E o

CO

• ^

03

^

CD CO _c O CO

X

o

2

o "2 c:

03 O

"co c03 to

o

03

> cCO 0 .^ szCO > "^ CO CO "o 03 O ^ 03 CO CO

03

o LU 11 1

03 Q.

O

0

"cz o CO

o

*

MichaelJ.G. Farthing

859

Table 7. Treatment of infectious proctitis Drug of Choice

Alternative(s)

Bacteria Chlamydia trachomatis Non-LGV LGV

Doxycycline 100 mg twice daily for 7 days Doxycycline 100 mg twice daily for 3 weeks or more

Erythromycin 500 mg four times daily TMP-SMX two tabs twice daily

Neisseria gonorrhoeae

^Ceftriaxone 250 mg i.m.

Oral cefixime, ciprofloxacin or ofloxacin

Treponema pallidum

Benzathine penicillin G 2.4 million units, single oral dose

Procaine penicillin 600 000-900 000 units i.m. for 10 days; doxycycline 100 mg twice daily orally for 15 days

Mycobacterium tuberculosis

See text

Viruses Acyclovir 5 mg/kg i.v. 8-hourly for 7 - 1 0 days [75]

Forscarnet 4 0 - 6 0 mg/kg i.v. 8-hourly for 2 - 3 weeks [76]

Herpes simplex virus See Table 6 Cytomegalovirus Helminths Schistosoma spp.

See Table 6

*Plus doxycyline 100 mg twice daily, orally for 7 days.

the onset of an attack of colitis and the detection of the presence of an enteropathogen in the stool. Several pathogens have attracted particular attention, including Clostridium difficile [11, 78] Campylobacter spp. [79, 80] Aeromonas spp. [81, 82] and CIVLV infection. There is no question that these and other infections can occur in association with estabUshed non-specific IBD and may be isolated during an apparent relapse, but it is not clear as to whether they are of etiologic importance in precipitating IBD in a previously healthy individual. There are limited prospective studies of sufficient power to answer this question. One study examined 64 patients during their first attack of ulcerative colitis and 30 others during a relapse of known disease, and found no evidence that bacterial enteropathogens were important triggers in either group [83]. Another study searched for enteropathogens in 64 patients with a relapse of known IBD [80]. C difficile was isolated in six patients, Campylobacter jejuni in one patient and Salmonella typhimurium in one patient. It was considered that these infrequent isolations indicated that intestinal infection played only a minor role in exacerbations of IBD. However, these studies confirm that the two conditions can coexist, and that an etiologic agent should be excluded in patients presenting for the first time with possible IBD. IVIany clinicians would routinely test

for infection in patients with known IBD on the basis that intestinal infections are common even in industrialized countries, and that it is virtually impossible to distinguish clinically between a simple relapse of IBD and a relapse associated with a co-infection. Intestinal infection associated with a relapse of IBD should be treated in the same way as recommended for infection in an otherwise healthy individual. In such patients with a relapse of non-specific colitis it is usually wise to treat both the infection and the IBD particularly when the patient is moderately or severely aff'ected. In mild disease it may be possible to treat only the infection while closely monitoring the activity of the colitis.

References 1.

2.

3. 4.

Handszuh H, Waters SR. Travel and tourism patterns. In: DuPont HL, Steffen R, eds. Textbook of Travel Medicine and Health. Hamilton: Decker, 1997. Farthing MJG, DuPont HL, Guandalini S et al. Treatment and prevention of travellers' diarrhoea. Gastroenterol Int 1992; 5:162-75. DuPont HL, Ericsson CD. Prevention and treatment of traveler's diarrhea. N Engl J Med 1993; 328: 1821-7. Wheeler JG, Sethi D, Cowden JM et al Study of infectious intestinal disease in England: rates in the community, presenting to general practice, and reported to national surveillance. Br Med J 1999; 318: 1046-50.

Infectious colitis

860 5. 6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18.

19. 20.

21. 22.

23.

24. 25.

Balarajan R, Raleigh VS, Yuen P et al. Health risks associated with bathing in seawater. Br Med J 1991; 303: 1445-5. Walker A. Swimming - the hazards of taking a dip. Br Med J 1992;304:242-5. Keene WE, McAnulty JM, Hoesly FC et al. A swimmingassociated outbreak of hemorrhagic colitis caused by Escherichia coli 0157: H7 and Shigella sonnei. N Engl J Med 1994;331:579-84. BoyceTG, Swerdlow DL, Griffin PM. EscherichiacoliO\51\ H7 and the hemolytic-uremic syndrome. N Engl J Med 1995;333:364-8. Neal KR, Brij SO, Slack RGB et al. Recent treatment with H2 antagonists and antibiotics for gastric surgery as risk factors for Salmonella infection. Br Med J 1994; 308: 176. Neal KR, Scott HM, Slack RGB. Omeprazole as a risk factor for Campylobacter gastroenteritis: case control study. Br Med J 1996; 312: 414 15. Kozicki M, Stctfen R, Schar M. 'Boil it, cook it, peel it, or forget it': does this rule prevent travellers' diarrhoea? Int J Epidemiol 1985; 14: 169 72. Bandres JG, Mathewson JJ, DuPont HL. Heat susceptibility of bacterial enteropathogens. Arch Intern Med 1988; 148: 2261 3. Sheath NK, Wisnicwski TR, Franson TR. Survival of enteric pathogens in common beverages. An /// vitro study. Am J Gastroenterol 1988; 83: 658 60. Farthing MJG. Mycobacterial disease of the gut. In: Phillips RKS, Northover J MA, eds. Modern Goloproctology. London: Edward Arnold, 1992: 174 89. Strauss EJ, Falkow S. Microbial pathogenesis: genomics and beyond. Science 1997; 276: 707 12. Finley BB, Falkow S. Gommon themes in microbial pathogenicity revisited. Microbiol Med Biol Rev 1997; 61: 136 69 Mann BJ, Vedvick T, Torian B, Petri WA. Gloning of the 170 kDa subunit of the galactose-specific adherence lectin of Entamoeha histolytica. Proc Natl Acad Sci USA 1991; 88: 3248 52. Leippe M, Ebel S, Schoenberger OL, Horstmann RD, Muller-Eberhard HJ. Pore-forming peptide of pathogenic Entamoeha histolytica. Proc Natl Acad Sci USA 1991; 88: 7659 63. Kenny B, DeVinney R, Stein M et al. Enteropathogenic E. coli (EPEG) transfers its receptor for intimate adherence into mammalian cells. Gell 1997; 91; 511 20. Knutton S, Rosenshine I, Pallen MJ et al. A novel EspAassociated surface organelle of enteropathogenic Escherichia coli involved in protein translocation into epithelial cells. EMBOJ 1998; 17: 2166 76. Triadafilopoulos G, Pothoulakis G, O'Brien M, LaMont JT. Differential effects of Clostridium difficile toxins A and B on rabbit ileum. Gastroenterology 1987; 93: 273-9. Triadafilopoulos G, Pothoulakis G, Weiss R, Giampaolo G, LaMont JT. Gomparative study of Clostridium difficile A and cholera toxin in rabbit ileum. Gastroenterology 1989; 97: 1186-92. Krivan HG, Glark GF, Smith DF, Wilkins TD. Gell surface binding site for Clostridium difficile enterotoxin: evidence for a glycoconjugate containing the sequence 1-3-Gaipi4GlcNAc. Infect Immun 1986; 53: 573-81. Just I, Fritz G, Akatories K et al. Clostridium difficile toxin B acts on the GTP-binding protein Rho. J Biol Ghem 1994; 269: 10706-12. Just I, Selzer J, von Eichel-Streiber G, Aktories K. The low molecular mass GTP-binding protein Rho is affected by toxin A from Clostridium difficile. J Glin Invest 1995; 95: 1026-3L

26. Just I, Selzer J, Wilm M, von Eichel-Streiber G, Mann M, Aktories K. Glucosylation of Rho proteins by Clostridium difficile toxin B. Nature 1995; 375: 500-3. 27. Just I, Wilm M, Selzer J et al. The enterotoxin from Clostridium difficile (ToxA) monoglucosylates the Rho proteins. J Biol Ghem 1995; 270: 13932-6. 28. Francis GL, Ryan TA, Jones BD, Smith SJ, Falkow S. Ruffles induced by Salmonella and other stimuli direct macropinocytosis of bacteria. Nature 1993; 364: 639-42. 29. Ruschkowski S, Rosenshine I, Finlay BB. Salmonella typhimurium induces an inositol phosphate flux in infected epithelial cells. FEMS Microbiol Lett 1992; 74: 121-6. 30. Ghen LM, Hobbie S, Galan JE. Requirement of GDG42 for Salmonella-induced cytoskeletal and nuclear responses. Science 1996; 274: 2115 18. 31. Galan JE. Molecular genetics bases of Salmonella entry into host cells. Mol Microbiol 1996; 20: 263 71. 32. Adam T, Giry M, Bowuet P, Sansonetti P. Rho-dependent membrane folding causes Shigella entry into epithelial cells. EMBOJ 1996; 15 3315 21. 33. Menard R, Prevost MG, Gounon P, Sansonetti P, Dehio G. The secreted Ipa complex oWShigella flexneri promotes entry into mammalian cells. Proc Natl Acad Sci USA 1996; 93: 1254 8. 34. Gossart P. Subversion of the mammalian cell cytoskeleton by invasive bacteria. J Glin Invest 1997; 99: 2307 11. 35. Finlay BB, Gossart P. Exploitation of mammalian cell function by bacterial pathogens. Science 1997; 276: 718 25. 36. Eckmann L, Kagnoff MF, Fierer J. Epithelial cells secrete the chemokine interleukin-8 in response to bacterial entry. Infect Immun 1993; 61: 4569 74. 37. Yang S-K, Eckmann L, Panja A, Kagnoff M F Differential and regulated expression of G-X-G, G-G and G chemokines by human colon epithelial cells. Gastroenterology 1997; 113: 1214 23. 38. Elliott EJ, Zhi Li, Bell G, Stiel D et al. A monoclonal antibody against the GDI8 adhesion molecule inhibits colonic structural and ion transport abnormalities caused by enterohemorrhagic E. coli 0\51: H7 in rabbits. Gastroenterology 1994;106: 1554 61. 39. Harries AD, Myers B, Gook GG. Inflammatory bowel disease: a common cause of bloody diarrhoea in visitors to the tropics. Br Med J 1985; 291: 1686 7. 40. Rees JH, Soudain SE, Gregson NA et al. Campylobacter jejuni infection and Guillain-Barre syndrome. N Engl J Med 1995;333:1374 9. 41. Allison MG, Hamilton-Dutoit SJ, Dhillon AP, Pounder RE. The value of rectal biopsy in distinguishing self-limited colitis from early inflammatory bowel disease. Q J Med 1987;65:985-95. 42. Nostrant TT, Kumar NB, Appelman HD. Histopathology differentiates acute self-limited colitis from ulcerative colitis. Gastroenterology 1987; 92: 318-28. 43. Farthing MJG. History and rationale of oral rehydration and recent development in formulating an optimal solution. Drugs 1988; 36(Suppl. 4): 80-90. 44. Farthing MJG. Dehydration and rehydration in children. In: Arnaud MJ, ed. Hydration Throughout Life. Paris: John Libbey Eurotext 1998: 159-73. 45. Kaplan MA, Prior MJ, McKonly KI, DuPont HL, Temple AR, Nelson EB. A multicentre randomized controlled trial of a liquid loperamide product versus placebo in the treatment of acute diarrhea in children. Glin Pediatr 1999; 38: 579-9L 46. Owens JR, Broadhead R, Hendrickse RG, Jaswal OP, Gangal RN. Loperamide in the treatment of acute gastroenteritis in early childhood. Report of a two centre, doubleblind, controlled clinical trial. Ann Trop Paediatr 1981; 1: 135-41.

MichaelJ.G. Farthing 47.

48. 49.

50. 51.

52.

53.

54.

55.

56. 57.

58.

59.

60. 61.

62.

63.

64.

Bergstrom T, Alestig K, Thoren K, Trollfors B. Symptomatic treatment of acute infectious diarrhoea: loperamide versus placebo in a double-blind trial. J Infect 1986; 12: 358. Bowie MD, Hill ID, Mann MD. Loperamide for treatment of acute diarrhoea in infants and young children. A doubleblind placebo-controlled trial. S Afr Med J 1995; 85: 885-7. Tauxe RV, Puhr ND, Wells JG, Hargrett-Bean N, Blake PA. Antimicrobial resistance of Shigella isolates in the USA: the importance of international travelers. J Infect Dis 1990; 162:1107-11. Bennish ML, Salam MA, Haider R, Barza M. Therapy for shigellosis. II. Randomized, double-blind comparison of ciprofloxacin and ampicillin. J Infect Dis 1990; 162: 711-16. Khan WA, Seas C, Dhar U, Salam MA, Bennish ML. Treatment of shigellosis: V. comparison of azithromycin and ciprofloxacin. A double-blind, randomized, controlled trial. Ann Intern Med 1997; 126: 697-703. Bassily S, Hyams KG, el-Masry NA et al. Short-course norfloxacin and trimethoprim-sulfamethoxazole treatment of shigellosis and salmonellosis in Egypt. Am J Trop Med Hyg 1994; 51: 219-23. Gotuzzo E, Oberhelman RA, Maguina C et al. Comparison of single-dose treatment with norfloxacin and standard 5day treatment with trimethoprim-sulfamethoxazole for acute shigellosis in adults. Antimicrob Agents Chemother 1989; 33: 1101-4. Bennish ML, Salam MA, Khan WA, Khan AM. Treatment of shigellosis. III. Comparison of one or two-dose ciprofloxacin with standard 5 day therapy. A randomized, blinded trial. Ann Intern Med 1992; 117: 727-34. Teasley DG, Gerding DN, Olson MM et al. Prospective randomised trial of metronidazole versus vancomycin for Clostridium difficile-a.ssocmiQd diarrhoea and colitis. Lancet 1983; 2: 1043-6. Wilcox MH, Howe R. Diarrhoea caused by Clostridium difficile: response time for treatment with metronidazole and vancomycin. J Antimicrob Chemother 1995; 35: 673-9. Young GP, Ward PB, Bayley N et al. Antibiotic-associated colitis due to Clostridium difficile: double blind comparison of vancomycin with bacitracin. Gastroenterology 1985; 89: 1038^5. Wenisch C, Parschalk B, Hasenhundl M, Hirschl AM, Graninger W. Comparison of vancomycin, teicoplanin, metronidazole and fusidic acid for the treatment of Clostridium difficile-a.ssociatQd diarrhea. Clin Infect Dis 1996; 22: 813-18. Kelly MP, Farthing M J G Infections of the gastrointestinal tract. In: O'Grady F, Lambert HP, Finch RG, Greenwood D, eds. Antibiotic and Chemotherapy, 7th edn. London: Churchill Livingstone, 1997: 708-20. Garcia-Laverde A, de Bonilla L. Clinical trials with metronidazole in human balantidiasis. Am J Trop Med Hyg 1975; 24:781-3. Gayraud M, Scavizzi MR, Mollaret HJH, Guillevin L, Hornstein MJ. Antibiotic treatment of Yersinia enterocolitica septicemia: a retrospective review of 43 cases. Clin Infect Dis 1993; 17:405-10. Crowe M, Ashford K, Ispahani P. CUnical features and antibiotic treatment of septic arthritis and osteomyelitis due to Yersinia enterocolitica. J Med Microbiol 1996; 45: 302-9. Pai CH, Giflis F, Tuomanen E, Marks MI. Placebo-controlled double-blind evaluation of trimethoprim-sulfam e t h o x a z o l e t r e a t m e n t for Yersinia enterocolitica gastroenteritis. J Pediatr 1984; 104: 308-11. Anders BJ, Lauer BA, Paisley JW, Refler LB. Double-bhnd placebo controlled trial of erythromycin for treatment of Campylobacter QntQYiiis. Lancet 1982; 1: 131-2.

861 65.

66. 67. 68.

69.

70.

71.

72.

73.

74.

75. 76.

77. 78.

79. 80. 81.

82. 83.

Mandal BK, Ellis ME, Dunbar EM, Whale K. Double-blind placebo-controlled trial of erythromycin in the treatment of clinical Campylobacter infection. J Antimicrob Chemother 1984;13:619-23. Salazar-Lindo E, Sack RB, Chea-Woo E et al. Early treatment with erythromycin of Campylobacter jejuni-sissocisitQd dysentery in children. J Pediatr 1986; 109: 355-60. Williams MD, Schorling JB, Barrett LJ et al. Early treatment of Campylobacter Jejuni enteritis. Antimicrob Agents Chemother 1989; 33: 248-50. Prouix F, Turgeon JPJ, Delage G, Lafleur L, Chicoine L. Randomized, controlled trial of antibiotic therapy for Escherichia coli 0157: H7 enteritis. J Pediatr 1992; 121: 299-303. Carter AO, Borczyk AA, Carlson JA et al. A severe outbreak of Escherichia coli 0157-H7-associated hemorrhagic colitis in a nursing home. N Engl J Med 1987; 317: 1496500. Wong CS, Jelacic S, Habeeb RL, Watkins SL, Tarr PI. The risk of the hemolytic-uremeic syndrome after antibiotic treatment of Escherichia coli 0157: H7 infections. N Engl J Med 2000; 342: 1930-6. Dieterich DT, Kotler DP, Busch D F et al. Ganciclovir treatment of cytomegalovirus colitis in AIDS: a randomized, double-blind, placebo-controUed multicenter study. J Infect Dis 1993; 167: 278-82. Nelson MR, Connolly GM, Hawkins DA, Gazzard BG. Foscarnet in the treatment of cytomegalovirus infection of the esophagus and colon in patients with the acquired immune deficiency syndrome. Am J Gastroenterol 1991; 86: 876-81. Salzberger B, Stoehr A, Jablonowski H et al. Foscarnet 5 versus 7 days a week treatment for severe gastrointestinal CMV disease in HIV-infected patients. Infection 1996; 24: 121-4. Joint Tuberculosis Committee of the British Thoracic Society. Chemotherapy and management of tuberculosis in the United Kingdom: recommendations 1998. Thorax 1998; 54: 536-48. Genereau T, Lortholary O, Bouchaud O et al. Herpes simplex eosophagitis in patients with AIDS: report of 34 cases. Clin Infect Dis 1996; 22: 926-31. Safrin S, Crumpacker C, Chatis P et al. controlled trial comparing foscarnet with vidarabine for acyclovir-resistant mucocutaneous herpes simplex in the acquired immunodeficiency syndrome. N Engl J Med 1991; 325: 551-5. Lee DK, Cooper BT, Barbezat GO. Clostridium difficile toxin in chronic idiopathic colitis. NZ Med J 1986;99: 6202. Kochar R, Ayyagari A, Goenka MK, Dhali GK, Aggarwal R, Mehta SK. Role of infectious agents in exacerbations of ulcerative colitis in India. A study of Clostridium difficile. J Clin Gastroenterol 1993;16: 26-30. Farmer RG. Infectious causes of diarrhea in the diff"erential diagnosis of inflammatory bowel disease. Med Clin N Am 1990;74: 29-38. Weber P, Koch M, Heizmann WR, Scheurlen M, Jenss H, Hartmann F. Microbic superinfection in relapse of inflammatory bowel disease. J Clin Gastroenterol 1992; 14: 302-8. Doman DB, Golding MI, Goldberg HJ, Doyle RB. Aeromonas hydrophila colitis presenting as medically refractory inflammatory bowel disease. Am J Gastroenterol 1989;84: 83-5. Willoughby JM, Rahman AF, Gregory MM. Chronic colitis after Aeromonas infection. Gut 1989;30: 686-90. Brown WJ, Hudson MJ, Patrick S et al. Search for microbial pathogens in patients with ulcerative colitis. Digestion 1992;53: 121-8.

48 Human immunodeficiency virus and inflammatory bowel disease CHARLES MEL WILCOX

Introduction We have now completed the second decade of the acquired immunodeficiency syndrome (AIDS) that first came to attention in 1981 [1]. The early years of the epidemic were devoted to describing the vast spectrum of, and defining optimum treatments for, the complications of AIDS. Over the past 10 years, there has been intense investigation into the virology of the human immunodeficiency virus (HIV) and, consequently, rapid progress in our understanding of the pathogenesis of this devastating worldwide infectious disease. These efforts have culminated in the development of highly active antiretroviral therapies, termed HAART, which have profoundly changed the paradigms for management of HI V-related complications including those involving the gastrointestinal tract [2, 3]. Because of the widespread availability of these drugs and access to care [4] the fall in AIDS-related morbidity and mortality has been most pronounced in the developed world, whereas in contrast, and for the foreseeable future, complications related to HIV-associated immunodeficiency will continue unabated in developing countries [1, 5]. Before the development of HAART, gastrointestinal disorders occurred almost uniformly in patients with AIDS. Problems referable to the colon, usually diarrhea, were observed in 50% or more of patients at some point during the course of HIV and AIDS [5]. A l t h o u g h o p p o r t u n i s t i c infections and neoplasms comprise the majority of the colonic disorders in these patients, a number of cases of inflammatory bowel disease (IBD), both Crohn's disease (CD) and ulcerative colitis (UC), have been described [6-22]. In addition, an idiopathic inflammatory disease of the colon, apparently distinct from IBD, has also been recognized [23, 24]. Given the potential for confusion clinically, endoscopically, and pathologically between opportunistic infections

of the colon and IBD, an appreciation of the clinical presentation, differential diagnosis, and management of IBD in HIV-infected patients is important. In this chapter the relationship of HIV infection and IBD will be explored, focusing on the pathogenesis of IBD and what lessons we can learn from coexistent HIV infection; the chapter will review the clinical presentation and relationship to immunodeficiency of the reported patients with HIV infection; outline the differential diagnosis of IBD in the setting of HIV; and review management. Finally, criteria will be proposed for the diagnosis of idiopathic IBD in the setting of HIV infection.

Pathogenesis of IBD in relation to HIV Infection Since the key element in the pathogenesis of IBD is the inflammatory cascade which includes T cells, and HIV disease is characterized by the destruction of CD4 T lymphocytes leading to immunodeficiency, it follows that the prevalence and natural history of IBD may be altered in patients with HIV infection. Thus, the reported cases of coexistent HIV and IBD may provide insight into the pathogenesis of IBD. It is well recognized that T cells play an important role in the expansion and perpetuation of the inflammatory response in IBD [25]. When presented with specific antigens the T cell population expands, and these cells then interact with B cells, resulting in an expansion of antigen-specific B cells. Activated T cells recruit and activate macrophages and neutrophils, which in turn produce cytokines amplifying and further perpetuating the inflammatory response. CD4 T cells, the pivotal cell destroyed in HIV infection, play an important role in the pathogenesis of IBD. This has been directly demonstrated in multiple experimental models of IBD. Such studies have demonstrated that certain CD4 T cell subsets

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 863-873. © 2003 Kluwer Academic Publishers. Printed in Great Britain

864

Human immunodeficiency virus and inflammatory bowel disease

exist that can cause colitis, and other subsets exist that prevent it. Two of the major CD4 T cell subsets are T-helper 1 and T-helper 2. The Thl subset produces IL-2 and interferon gamma (IFN-y)and mediates cell-mediated immunity. The Th2 subset produces IL-4, IL-5, IL-6, and IL-10, and mediates antibody production. Based on the profile of measured cytokine production, CD appears to be most associated with Thl lymphocytes while UC may be more associated with Th2 lymphocytes, although this is much less clear. Regulatory CD4 T cell subsets also exist. A CD4^ T cell subset denoted Th3 produces large amounts of active transforming growth factor (31 (TGF-Pi), a cytokine with broad inhibitory effects on lymphocytes. The T-regulatory 1 (Trl) CD4"*"T cell produces large amounts of IL-10, another inhibitory cytokine, along with some TGF-Pi. It is likely that more regulatory CD4"^ T cell subsets will be identified in the future. Several studies have examined T cell subsets in bowel tissue from HIV-infected patients. Generally, T lymphocyte subsets in the gut reflect the concentration in the systemic circulation. HIV-infected patients with advanced immunodeficiency have markedly reduced numbers of CD4 cells both in the circulation and small bowel, with a corresponding increase in CDS cells [26, 27]. Likewise, quantitation of CD4 and CD8 cells in colonic biopsy samples from AIDS patients shows a similar pattern. Thus, the systemic CD4/CD4 T cell ratio and the absolute values of each subset reflect what can be found in the gut. There is little information regarding cytokine production in bowel tissue from HIV-infected patients, especially those with more severe degrees of immunodeficiency. Snijders etal. [28] biopsied the jejunum of HIV-infected patients with diarrhea, some of whom had small intestinal pathogens, HIVinfected patients without diarrhea, and HlV-seronegative controls. No significant differences were detected in cytokine levels among the groups with low levels uniformly detected. In c o n t r a s t , McGowan et al. [29] found a significant increase in inflammatory cytokines, IL-ip and IFN-y in HIVinfected patients compared to controls, but lower levels of IL-10 in colonic biopsies. Proinflammatory cytokines have also been detected in rectal mucosa of AIDS patients with peak expression of TNF-oc and IL-ip in late-stage patients compared to patients with earlier-stage disease [30]. In contrast, one study [31] suggested defective function of colonic mono-

nuclear cells in AIDS patients as reflected by a reduction in TNF-oc secretion. The relationship of cytokine production to HIV-infected mucosal inflammatory cells is unclear. Overall, the number of HIV-infected cells identified in the lamina propria has been consistently low [32, 33]. If CD4 cells are important in the pathogenesis of IBD, then animal models of IBD with perturbations of CD4 T cell subsets would provide an ideal assessment of their role. Loss of CD4 T cells does improve disease activity in animal models. However, selective knockout of subpopulation has yet to be performed. It follows from the above that the relationship between HIV infection and IBD is likely to be complex. Assuming that the CD4 T cell depletion occurs in all CD4 T cell subsets, including those that both mediate and prevent IBD, the net effect on the course of IBD will depend on which subset is more affected. For example, if the Thl subset were to be disproportionaltely depleted in a patient with CD, the activity of the IBD might improve. Conversely, if the regulatory CD4 T cell subset were to be disproportionately depleted, IBD activity might worsen. Indeed, both improvement and worsening of IBD activity has been reported in patients with IBD who also develop HIV.

Review of reported cases of IBD and HIV Infection Although there have been a number of reported cases of coexistent HIV infection and IBD, given the millions of patients with HIV infection worldwide, coupled with the predominance of IBD in young patients - those most afflicted by HIV - one might anticipate many more reported cases. Two studies have provided estimates of the prevalence of coexistent HIV and IBD [10, 11]. In a retrospective study [10], hospital discharge records for a 4-year period were reviewed, and only three patients with IBD of 1839 patients with a discharge diagnosis of HIV/AIDS were identified. Four additional patients were identified by phone survey of clinicians in related outpatient clinics. Since both IBD and HIV infections are primarily treated in the outpatient setting, these results are likely to be an underestimate. In the study of Sharpstone et al. [11], HIV clinic records were cross-referenced with pharmacy records for 5-ASA compounds and rectal steroid preparations. Eight patients were identified, yielding a mean incidence of 41 /100 000 and a prevalence of

C Mel Wilcox

865

364/100 000. These numbers are high compared to recent US studies [34] and may reflect the small sample size. A number of important considerations must be kept in mind when reviewing the case series of IBD in HIV-infected patients. First, the time-course of IBD and HIV infection should be well documented. The majority of the reported cases consist of patients with IBD who later became HIV-infected, or both disorders were diagnosed together. Importantly, the acquisition of HIV infection can rarely be precisely timed; thus, a long lag time between infection and the diagnosis of IBD is likely in most patients, which means there could be considerable overlap in the duration of the two diseases. Second, the stage of HIV-related immunodeficiency in relationship to the timing of the diagnosis of IBD must be characterized. Most reported cases of de-novo IBD have been in HIV-infected patients with only modest immune dysfunction. The marker for staging immune dysfunction most frequently used, and for which there is compelling evidence, is the CD4 lymphocyte count [35, 36]. It is well established that significant immune dysfunction, as reflected clinically by the occurrence of opportunistic disorders, does not occur until the CD4 count falls below 200/jil, and most opportunistic infections involving the gastrointestinal tract do not manifest until the CD4 count falls below 100/|il [3, 5, 35]. Thus, reports of HIVinfected patients with CD4 counts less than 100/)il with active IBD may provide better insight into any relationship between HIV-associated immunodeficiency and IBD. Third, it is important to reflect on any differences as compared to non-HIV-infected patients in the manifestations and outcome of IBD, such as location and extent of disease, presence of extraintestinal complications, and response to

therapy, both medical and surgical. Fourth, it is critical that a thorough histopathologic examination of colonic and/or small bowel biopsies be performed to exclude opportunistic disorders which may masquerade as IBD. Since a number of opportunistic infections, most notably cytomegalovirus (CMV), may mimic IBD, the diagnosis of IBD may be difficult to establish conclusively. Also, the characteristic histologic findings of IBD should be specifically documented, including crypt distortion, chronic inffammation, etc. This potential confusion is further perpetuated by the absence of specific criteria for the diagnosis of IBD in the setting of AIDS.

Crohn's disease To date there have been eight reports of CD [6-14] totalling 12 HIV-infected patients (Table 1). Consistent with the epidemiology of AIDS, the majority of the patients were male homosexuals ranging in age from 14 to 48 years. Five of the 12 patients presented with CD after well-established HIV infection. Of the other seven patients with CD, the timing of HIV infection could be accurately determined in only two [13] (contaminated blood products) and, as noted above, it is likely that in many cases there was overlap of the two diseases. The descriptions of the radiographic and endoscopic findings were typical for CD. The diagnosis of CD was made by the findings on pathologic specimens, usually post-surgical, and non-caseating granulomas were reported as present in less than half the cases either on endoscopic biopsy or in the surgical specimen. Histologic examination was generally extensive to exclude other diseases. However, long-term followup of the HIV-infected patients who subsequently

Table 1. Reported cases of Crohn's disease in HIV-infected patients Reference

Year

No. of patients

Sequence

6 7 8 9 10 12 11 13

1984 1988 1994 1996 1996 1997 1996 1998

1 1 1 1 1 1 2 4

AIDS ^Crohn's Crohn's -^HIV HIV ^Crohn's Crohn's -^HIV HIV ^Crohn's HIV -^Crohn's Crohn's -^HIV Crohn's ^ H I V

*CD4 count determined years after initial diagnosis of CD.

CD4 count 230*

410 480 270*

210 100 336, 442 320, 5 0 M 6 2 , 34*

Location Colon, terminal ileum Colon Colon, terminal ileum Colon, terminal ileum

NA Colon, terminal ileum Colon, small bowel/rectum Colon, terminal ileum

Human immunodeficiency virus and inflammatory bowel disease

866

developed CD, which would add further security to the diagnosis, was rarely performed, including colonoscopic re-examination. The gastrointestinal tract involvement with CD appeared similar to what might be anticipated with most patients having both colon and terminal ileal disease and with no apparent site predilection. Fistulous disease was reported in one patient [12] and extraintestinal manifestations were not described in any patient. To our knowledge there are no cases in which perianal disease was the sole manifestation. One reported patient [14] had documented UC for 14 years and subsequently, at the time of recurrent symptoms, had colonoscopy as well as barium enema that was most suggestive of CD; granulomas were not seen on colon biopsy, and extensive histologic examination to exclude opportunistic disorders was not reported. Testing for HIV was not available at the time of publication (1986), but the helpersuppressor T cell ratio was very low, suggesting severe immunodeficiency. Whether this patient truly had UC or an overlap syndrome is unknown. One of the most important features of these cases is the course of HIV-related immunodeficiency in relationship to the timing of IBD. Of the ^\VQ HIVinfected patients who later developed CD, the CD4 lymphocyte counts at the time CD was first documented were 100/)il, 210/^1, 230/)il, 336/)il, and 480/|il. Only one of these patients had a history of opportunistic infections, and HIV testing was not

available at the time of publication [6]. One additional reported patient with apparently advanced immunodeficiency, as reflected by the presence of opportunistic infections, had a low helper suppressor T cell ratio (0.25, normal 2.5) although the CD4 count was not reported [14]. As noted above, given these CD4 counts, only one of these three patients had substantial immunodeficiency, and this leaves open the possibility that severe immunodeficiency could be protective against the development of CD. Since some reports [14] preceded the recognition of AIDS-related diseases which mimic CD, it is perhaps possible that some of these cases could represent a missed opportunistic infection. The relationship of the response to therapy, as well as natural history with regard to stage of HIV infection, will be discussed below.

Ulcerative colitis To date there have been nine reports [10, 11, 14, 15, 17-22] of UC totaling 17 HIV-infected patients (Table 2). In four of these patients UC and HIV were diagnosed simultaneously. The CD4 count at diagnosis was 500/}il or greater in six patients tested, and greater than 200/nl in 13 (87%). Histologic criteria for diagnosis followed objective criteria in one [10], were not described in another [15] while the histology was reported as 'nonspecific' in one [18] or demonstrated acute inflammation and crypt abscess

Table 2. Reported cases of UC in HIV-infected patients Reference

Year

No. of patients

10

96

11

96

2 2 6

14 15 18 19 20 21 22

86 90 91 92 96 97 99

NA = not available.

Sequence

CD4 count

HIV^UC UC^HIV HIV - . U C HIV ^ U C UC-.HIV HIV -^UC HIV ^ U C HIV ->UC UC^AIDS UC/HIV UC/HIV HIV ^ U C HIV ^ U C HIV/ UC HIV/ UC

680, 700 530, 130 460 270 256 462 228 283

NA 500 546 170 NA 450 930

Extent Proctitis, NA NA, right colon

NA

NA Transverse colon Pancolonic T colon Pancolonic Pancolonic Transverse colon

C Mel Wilcox in the others. A missing element in most reports is a description of the characteristic histologic changes of chronicity which should be apparent at the time of diagnosis [37]. Of the patients undergoing colonoscopy the disease was limited to the rectum [10], reached the transverse colon [19] and right colon [10] in one patient each and was pancolonic in the other patient. Fulminant colitis requiring surgery was reported in four patients [11,17,22]. One patient developed significant articular symptoms (sacroiliitis, arthritis) and later uveitus with flares of disease [21]. When described, the endoscopic appearance of the colon was typical for UC. As mentioned above for CD, the rigor with which other potential causes of colitis were excluded varied, extensive histologic examination was not always performed to exclude infection or evidence of chronicity documented, and long-term follow-up to best assess diagnostic accuracy was not always reported.

Differential diagnosis Given the apparent infrequency of IBD in AIDS, when confronted with an AIDS patient with suspected IBD the patient is more likely to have an AIDS-related disorder than idiopathic IBD. In contrast, when immunodeficiency is not advanced (CD4 count >200/iil), gastrointestinal complaints are more likely non-HIV-related. Thus, recognition of the opportunistic disorders which may mimic IBD and their relationship to the stage of immunodeficiency is critical. Furthermore, when correctly diagnosed, the majority of these AIDS-related disorders are amenable to therapy. When evaluating any HIV-infected patient with gastrointestinal symptoms the level of immunodeficiency must first be staged by the CD4 lymphocyte count. Numerous studies have consistently shown that opportunistic disorders rarely manifest above a CD4 count of 200/|il with most occurring at levels less than 100/|il [35]. Recent evidence also suggests that HIV viral load provides additional prognostic information for disease progression, including the development of opportunistic disorders [38]. Two common gastrointestinal pathogens that occur in the setting of advanced immunodeficiency are cytomegalovirus (CMV) and Mycobacterium avium complex (MAC) which commonly involve the colon and small bowel, respectively, and which may mimic IBD [39-44]. In addition, some processes may cause pathologic changes of chroni-

867 city that may suggest underlying IBD (unpublished observations). The many reported cases of opportunistic infections which mimicked IBD clinically, radiologically, endoscopically, and/or histologically in this setting emphasize the importance of a high index of suspicion for these diseases, as well as appropriate endoscopic tissue sampling and histologic processing. Another point to consider when discussing the differential diagnosis of IBD in the setting of HIV infection is that patients with IBD may develop opportunistic infections characteristic of AIDS caused by treatment-related immunodeficiency. The best example of this is CMV colitis complicating high-dose prednisone therapy. In this scenario one may be concerned that the patient has underlying HIV infection and CMV colitis rather than IBD complicated by CMV colitis. Likewise, apparent exacerbation of IBD in HIV-infected patients can reflect intercurrent CMV infection [11]. The most important disease in the differential diagnosis of IBD, especially UC, is CMV colitis. Patients with CMV colitis almost uniformly present with a CD4 count less than 100/|il [39, 45]. In one study [39] the median CD4 count was 15/|il. CMV colitis can also be the index presentation of HIV infection. The most common manifestations of CMV colitis are crampy lower abdominal pain, chronic watery diarrhea which may be bloody, and proctitis symptoms, particularly when distal disease is prominent. Weight loss is also frequent and may be profound, while fever is uncommon. The similarity of these symptoms compared with IBD is readily apparent. Endoscopically there are also many similarities of CMV colitis and IBD. In a study characterizing the endoscopic appearance of CMV colitis in 56 AIDS patients, Wilcox et al. [39] showed that the most common endoscopic manifestation was subepithelial hemorrhage, which was often confluent. Disease was located throughout the colon in 74%, but was limited proximal to the rectosigmoid colon in only 13%. Well-circumscribed ulcerations typical for CD were observed in six patients, while a pancolitis characteristic of UC was noted in three patients. Pseudopolyps have not been reported as a manifestation of CMV colitis. These findings demonstrate that the endoscopic appearance of CMV colitis may be suggestive ofeitherUCorCD. To best diagnose CMV colitis, multiple biopsies (at least six) of abnormal-appearing tissue should be obtained with close inspection of the characterisitic

868

Human immunodeficiency virus and inflammatory bowel disease

viral cytopathic effect of CMV. Immunohistochemical staining for CMV antigens may further assist in the diagnosis. Although not recommended routinely, viral culture of tissue biopsies may rarely be helpful. It should be noted that, in general, CMV colitis in AIDS is associated with numerous viral inclusions in the biopsies. Therefore, if the the viral cytopathic effect is rare (only one or two positively staining cells with numerous biopsies), one should consider another diagnosis; this should be considered as CMV 'infection' rather than true 'disease' [46). Other colonic infections in AIDS which have been reported to mimick IBD include Isospora [47], and histoplasmosis [48]. Although commonly reported as a cause of colonic disease, herpes simplex virus (HSV) is actually a very rare cause of colonic infection in AIDS. Usually herpes simplex virus of the colon presents with limited anorectal disease. A number of other opportunistic diseases could theoretically mimic IBD also, but no specific reports detailing these associations in HIV-infected patients have been published (Table 3). Terminal ileal disease suggesting CD may result from either CMV or MAC infection. In a report of terminal ileitis and AIDS [44], small bowel followthrough showed nodularity and ulceration of the terminal ileum characteristic of CD. However, histopathologic examination of the resected ileum demonstrated numerous acid-fast bacilli diagnostic of MAC. Several cases of colonic Kaposi's sarcoma (KS) mimicking UC have been reported [ 16, 17]. However, in these AIDS patients KS was not identified on the initial evaluation, but rather on follow-up colonoscopy. In one case [17], IBD was treated aggressively with corticosteroid therapy for 3 months prior to the identification of KS; thus, colonic KS found on subsequent colonoscopy could represent a recognized complication of steroid (immunosuppressive) therapy. The other patient [16] was treated with hydrocortisone enemas prior to the diagnosis of colonic KS. The lesions of KS are characterized endoscopically as circumscribed hemorrhagic lesions of variable size, and it is well recognized that deep biopsies are required to identify the characteristic histologic findings as the tumor generally resides in the submucosa. As HIV-infected patients are growing older, agerelated colonic diseases which may mimic IBD must also be considered. Focal inflammation in the setting of diverticulosis (diverticular colitis) is being increas-

Table 3. Differential diagnosis of IBD in HIV-infected patients Infection Bacterial colitis Cytomegalovirus Mycobactrium avium complex l\/lycobacterium tuberculosis Amebiasis Isospora belli Histoplasmosis Non-Infectious Ischemia Drug-induced colitis Peridiverticulitis Neoplasm Kaposi's sarcoma

ingly recognized [49]. Ischemic colitis typically presents with a segmental colitis which may mimic CD or CMV colitis; however, the histologic features of ischemia are usually apparent. Colonic neoplasms may occur in HIV-infected patients including adenocarcinoma, and colonic lymphomas may appear indistinguishable endoscopically from adenocarcinoma. Also, mass lesions have been reported with CMV colitis due to an exuberant inflammatory response, further highlighting the importance of adequate tissue sampling [5]. We and others have described HIV-infected patients with colorectal symptoms in whom focal areas of colitis and/or ulcers have been identified and which have remained idiopathic despite an extensive histologic examination for infections and neoplasms [23, 24]. These patients typically present with ulcers which can be large, and histologic examination of multiple biopsies shows granulation tissue; rarely crypt distortion was present which could mimic IBD. These ulcerative lesions are similar to the well-recognized idiopathic ulcer of the esophagus [50]. Hing et al. [24] reported on six HIV-infected patients with a chronic colitis presenting as a chronic diarrhea ranging from three to more than 10 stools per day. The CD4 counts at diagnosis ranged from 256 to 449/|il. Stool studies were negative and colonoscopy showed a diffuse colitis consisting of ulceration, hemorrhage, and erosions, and in one patient mild erythema. Histologically, crypt architecture was preserved with a mixed inflammatory cell population of plasma cells, lymphocytes and neutrophils. Crypt abscess, crypt destruction, and granulo-

C Mel Wilcox mas were absent and CMV and HSV were adequately excluded. Four patients entered remission 6-27 months following the diagnosis, although steroids and sulfasalazine were provided. Zidovudine (AZT) was given to five patients. This report suggests that a mild inflammatory disease of the colon distinct from UC can occur in these patients. The role of antiinflamatory therapy or HIV-directed treatment (HAART) for this entity is unclear. An increase in inflammatory cells of the distal colon [51] as well as small bowel [52, 53] has also been documented in HIV-infected patients. The mechanism(s) for this increase in inflammatory cells of these patients also remains unexplained.

Treatment Because of the small number of reported cases of active IBD and HIV infection, the optimal treatment for both UC and CD in these patients is unknown. In addition, the recognized variability in response to medical therapy in IBD combined with the absence of controlled trials in this setting, makes it difficult to determine if the response to standard therapy for IBD in HIV-infected patients is altered either favorably or unfavorably.

Crohn's disease For those patients with CD in whom therapy was described and follow-up reported, the clinical response does not appear different from non-HIVinfected patients. Three reports document the clinical response to standard therapy for CD. Corticosteroids with or without aminosalicylates were usually given, and the initial dose of prednisone was ^ 40 mg per day. Of the three patients with active CD, all responded promptly to instituted therapy, achieving a clinical remission. One patient was treated with antibiotics and hydrocortisone suppositories. One of these patients had multiple relapses [9], all of which were treated successfully with prednisone and aminosalicylates. No long-term data are available describing the efficacy of maintenance therapy with aminosalicylates, and routine endoscopic follow-up after therapy has not been reported. Surgical therapy has also been undertaken for CD in these HIV-infected patients. Bernstein et al. [8] reported one patient who required surgical resection for small bowel obstruction at the time of initial diagnosis. Postoperatively the patient did well, and

869 6-month follow-up found no recurrence. Other patients have required colectomy for toxic megacolon, some of whom had colonic infections complicating IBD [11,17, 22].

Ulcerative colitis Of the reported patients with active UC and HIV, treatment was effective in the majority, at least at the initial presentation, often resulting in clinical remission. One patient [11] was treated with prednisone chronically, but after HIV seroconversion, his clinical course appeared to moderate. However, the medication regimens during this later time were not specifically defined (see below). One patient [18] had experienced several recurrences of disease, all of which responded to therapy; however, one episode later in the course of HIV infection was poorly responsive to high-dose corticosteroid therapy, ultimately requiring 1 month hospitalization. His CD4 count during this severe flare fluctuated between 2601 \A and 520/jil. Toxic megacolon refractory to standard therapy required surgery in a patient with minimal immune dysfunction [17]. None of the above-reported patients received a d d i t i o n a l immunosuppressive therapies such as azathioprine. Like the patients with CD, endoscopic follow-up with biopsies was not routinely performed to best document remission. There appear to be few therapy-related side-effects from the use of standard treatment for IBD in HIVinfected patients. One of the main treatment concerns is the use of an immunosuppressive drug, such as prednisone, in patients who are already immunosuppressed. Of the 10 reported IBD patients who received prednisone in doses of 40 mg or greater, usually in combination with aminosalicylates, sideeffects were reported in three. One patient with a history of UC was found to be HIV-infected during an exacerbation [14]. Following 40 mg of prednisone and sulfasalazine only a slight improvement in symptoms was noted. A CD4 count was not reported, although the T cell helper suppressor ratio was low (0.19). After receiving prednisone and sulfasalazine for approximately 2\ months, he developed diffuse cutaneous KS. The prednisone therapy was discontinued and his cutaneous lesions improved substantially. One additional patient developed widespread KS following 3 months of high-dose steroid therapy [17]. Candida esophagitis was reported in one additional patient [9]. It is tempting to speculate that corticosteroid-induced immunosuppression may

870

Human immunodeficiency virus and inflammatory bowel disease

have predisposed these HIV-infected patients to the development of cutaneous KS. Prednisone has been used for a variety of indications in HIV-infected patients, even in those who are profoundly immunosuppressed, with little apparent toxicity [54]. Importantly, precipitation of other opportunistic infections or malignancies appears to be rare. The most common reported toxicity of prednisone in HIV-infected patients is oropharyngeal and esophageal candidiasis. Endocrinologic complications such as diabetes may occur with prednisone as in any other patient. Because of the effect on cytokine production, prednisone has been shown to decrease HIV RNA in blood, suggesting a potential saluatory effect [55]. In the patient with severe immune dysfunction and prior CMV infection, caution must be exercised when tapering prednisone, given the possibility of unmasking adrenal insufficiency from CMVadrenalitis. The use of aminosalicylates should be encouraged as primary therapy as there are no specific contraindications to their use in HIV-infected patients. Although there are numerous drug interactions with the protease inhibitors and other antiretroviral agents, there are no reported interactions with the aminosalicylates. There are also no reports on the use of azathioprine for the treatment of refractory IBD in HIV-infected patients. Likewise, there are no studies on the use of remicade (infliximab) in this setting. The use of infliximab may be contraindicated, however, in patients with active infection, and HIV infection has been an exclusion criterion in trials of this agent. Given the above we recommend that HIV-infected patients be treated similarly to any other IBD patient. The use of prednisone should not necessarily be discouraged; rather the patient should be mon-

itored closely for complications. For those patients with more severe immunodeficiency and history of thrush, prophylaxis with Nystatin or oral azoles should be considered. For the patient with distal colonic disease, local therapy with enemas should be considered as first-line therapy.

Natural history There have been tantalizing suggestions that, as HIV-related immunodeficiency progresses, remission of IBD may occur. Given that IBD is characterized by periods of remission and exacerbation, it is important, however, to place these reports in the proper perspective. James [7] reported a patient with longstanding CD disease who subsequently developed HIV infection. This patient had experienced typical Crohn's symptoms on an intermittent basis for the preceding 8 years. HIV testing was performed as a routine and was found to be positive; at that time the CD4 count was 410/)il. Following the diagnosis of HIV infection the patient experienced no further symptoms of disease for the ensuing 2 years of follow-up. Pospai et al. [13] reported four patients with moderately active CD ranging in duration from 4 to 21 years who subsequently acquired HIV infection. In only one of these four cases, however, could the exact date of HIV acquisition be documented. In this patient in whom the timing of HIV could be precisely determined (HIV infection was linked to contaminated blood products), multiple exacerbations had occurred over the preceding 15 years, but the patient went into apparent remission after HIV infection. Interestingly, within 1 year of acquiring HIV infection this patient had a CD4 count, when first

Table 4. Suggested criteria for the diagnosis of IBD in HIV-infected patients 1.

Since there is no gold standard for the diagnosis of IBD, a combination of clinical presentation, endoscopic findings, response to therapy and follow-up must be employed

2.

Exclusion of colonic and intestinal pathogens is mandatory A. Stool studies for bacterial pathogens B. Stool studies for parasitic diseases C. Blood cultures for mycobacteria

3.

Extensive histologic examination of multiple biopsy specimens is mandatory A. Cytomegalovirus immunohistochemical staining B. Stains for fungi, mycobacteria

4.

Documentation of the characteristic histologic changes of IBD including crypt distortion and type of inflammatory cell infiltrate

5.

Appropriate follow-up to document response to therapy; both endoscopic and histologic follow-up after therapy should be performed

C Mel Wilcox measured, of 320/|il. In the other three cases the patients had experienced exacerbations for many years, but lasting remission was noted following the estimated time of HIV infection. The long symptomfree intervals in these three patients in whom the time of acquisition of HIV infection could not be documented and the minimal immunodeficiency, particularly early on in HIV infection, makes the relationship between remission and HIV-associated immunodeficiency highly speculative. Also, patients did not routinely undergo follow-up colonoscopic evaluation or other imaging studies to evaluate for disease activity. Yoshida et al [10] reported one patient with UC who subsequently developed HIV infection. Following the initial diagnosis (the CD4 count was 450/ml), the patient's symptoms appeared to decline over time coincident with the fall in the CD4 count. Six years after HIV seroconversion, at a time when the CD4 count was 20/|il, the patient underwent sigmoidoscopy and biopsy for an acute diarrhea, which demonstrated cryptosporidiosis but no active inflammatory bowel disease; the patient had not received any medical therapy for IBD during the preceding 14 months. Of six patients with IBD (four UC, two CD) followed by Sharpstone etal. [11], none developed a flare of disease when the CD4 count fell below 200/^d.

871 the IBD; however, single-agent therapy with this drug generally has little long-lasting eff'ect on immune function. Now that effective combination antiretroviral therapy is available, we may see reports of exacerbations of IBD in HIV-infected patients who receive HAART. Nevertheless, similar caveats as pointed out for the reverse situation still hold; opportunistic infections must be carefully excluded and spontaneous remissions and exacerbations are always a consideration. Better objective markers for active IBD are required to truly determine whether IBD can be turned off* or on by HIV-related immunodeficiency.

References 1. 2.

3.

4.

5. 6.

Conclusions Given the long latency between HIV infection and development of symptoms referable to AIDS, coupled with the variability in the natural history of IBD, it may be impossible to conclude that the progressive immunodeficiency of AIDS results in remission of IBD. The complex nature of the inflammatory response in IBD would suggest that loss of CD4 cells alone may not be 'therapy' for IBD. In addition, the multitude of opportunistic infections, some of which may mimic IBD, as well as the potential for some disorders (CMV) to result in histologic changes of chronicity further complicates any apparent relationship between HIV-related immunosuppression and the course of IBD. Guidelines for the diagnosis of IBD in the setting of HIV are appropriate (Table 4). More intriguing is the possibility that improvement in immune function may exacerbate IBD in HIV-infected patients. Some of the reported patients did receive AZT therapy without apparent effect on

7. 8. 9. 10.

11.

12.

13. 14.

Fauci AS. The AIDS epidemic. N Engl J Med 1999; 341: 1046-11050. El-Sadr WM, Burman WJ, Grant LB et al. Discontinuation of prophylaxis against Mycobacterium avium complex disease in HIV-infected patients who have a response to antiretroviral therapy. N Engl J Med 2000; 342: 1085-92. Monkemuller KE, Call SA, Lazenby AJ, Wilcox CM. Declining prevalence of opportunistic gastrointestinal disease in the era of combination antiretroviral therapy. Am J Gastroenterol 2000; 95: 457-62. Whitman S, Murphy J, Cohen M, Sherer R. Marked declines in human immunodeficiency virus-related mortality in Chicago in women, African Americans, Hispanics, young adults, and injection drugs users, from 1955 through 1997. Arch Intern Med 2000; 160: 365-9. Monkemuller KE, Wilcox CM. Diagnosis and treatment of colonic disease in AIDS. Gastrointest Endocrinol CUn N Am 1998; 8: 889-911. Dhar JM, Pidgeon ND, Burton AL. AIDS in a patient with Crohn's disease. Br Med J 1984; 288: 1802-3. James SP. Remission of Crohn's disease after human immunodeficiency virus infection. Gastroenterology 1988; 95: 1667-9. Bernstein BB, Gelb A, Tabanda-Lichauco R. Crohn's ileitis in a patient with longstanding HIV infection. Am J Gastroenterol 1994; 89:937-9. Christ AD, Sieber CC, Cathomas G, Gyr K. Concomitant active Crohn's disease in the acquired immunodeficiency syndrome. Scand J Gastroenterol 1996; 31: 733-5. Yoshida EM, Chan NHL, Herrick RA et al. Human immunodeficiency virus infection, the acquired immunodeficiency syndrome, and inflammatory bowel disease. J Clin Gastroenterol 1996; 23: 24-8. Sharpstone DR, Duggal A, Gazzard BG. Inflammatory bowel disease in individuals seropositive for the human immunodeficiency virus. Eur J Gastroenterol Hepatol 1996; 8: 575-8. Lautenbach E, Lichtenstein GR. Human immunodeficiency virus infection and Crohn's disease: the role of the CD4 ceU in inflammatory bowel disease. J Clin Gastroenterol 1997; 25: 456-9. Pospai D, Rene E, Fiasse R et al. Crohn's disease stable remission after human immunodeficiency virus infection. Dig Dis Sci 1998; 43: 412-19. Liebowitz D, McShane D. Nonspecific chronic inflammatory bowel disease and AIDS. J Clin Gastroenterol 1986; 8: 66-8.

872 15. 16. 17.

18. 19. 20.

21.

22.

23. 24. 25.

26.

27.

28.

29. 30.

31.

32.

33.

Human immunodeficiency virus and inflammatory bowel disease Franke M, Kruis W, Heitz W. First manifestation of ulcerative colitis in a patient with HIV infection. Gastroenterology 1990; 98: 544-5. Weber JN, Carmichael DJ, Bolyston A, Monro A, Whitear WP, Pinching AJ. Kaposi's sarcoma of the bowel - presenting as apparent ulcerative colitis. Gut 1985; 26: 295-300. Biggs BA, Crowe SM, Lucas CR, Ralston M, Thompson IL, Hardy KJ. AIDS related Kaposi's sarcoma presenting as ulcerative colitis and complicated by toxic megacolon. Gut 1987; 28: 1302-6. Bernstein CN, Snape WJ. Active idiopathic ulcerative colitis in a patient with ongoing HIV-related immunodepression. Am J Gastroenterol 1991; 86: 907 9. Sturgess I, Greenfield SM, Teare J, O'Doherty MJ. Ulcerative colitis developing after amebic dysentery in a hemophiliac patient with AIDS. Gut 1992; 33: 408-10. Wilcox CM, Schwartz DA, Cotsonis GA, Thompson WE III. Evaluation of chronic unexplained diarrhea in human immunodeficiency virus infection: determination of the best diagnostic approach. Gastroenterology 1996; 110: 30 7. Louis E, Moutschen MP, De Marneffe P ct al. Extensive ulcerative colitis and extraintestinal manifestations in a patient with HIV infection and significant CD4 T cell lymphopenia. Gastroenterol Clin Biol 1997; 21: 884 7. Silver S, Wahl SM, Orkin BA, Orenstein JM. Changes in circulating levels of HI V, CD4, and tissue expression of HIV in a patient with recent-onset ulcerative colitis treated by surgery. J Human Virol 1999; 2: 52 7. Wilcox CM, Schwartz DA. Idiopathic anorectal ulceration in patients with human immunodeficiency virus infection. Am J Gastroenterol 1994; 89: 599 604. Hing MC, Holdshmide C, Mathijs JM, Cunningham AL, Cooper DA. Chronic colitis associated with human immunodeficiency virus infection. Med J Aust 1992; 156: 683 7. Elson CO. Experimental models of intestinal inflammation: new insights into the mechanisms of mucosal homeostasis. In: Ogra PL, Mcstecky J, Lamm ME, Strober W, Bienenslock J, McGhee JR, eds. Mucosal Immunology, 2nd cdn. San Diego, CA: Academic Press, 1999: 1007 24. Rodgers VD, Fassetl R, Kagnoff MF. Abnormalities in intestinal mucosal T cells in homosexual populations including those with the lymphadenopathy syndrome and acquired immunodeficiency syndrome. Gastroenterology 1986; 90: 552 8. Ellakany S, Whiteside TL, Schade RR, Van Theil DH. Analysis of intestinal lymphocyte subpopulations in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. Am J Clin Pathol 1987; 87: 356-64. Snijders F, van Deveter SJH, Bartelsman JEW et al. Diarrhea in HIV-infected patients: no evidence of cytokinemediated inflammation in jejunal mucosa. AIDS 1995; 9: 367-73. McGowan I, Randford-Smith G, Jewell DP. Cytokine gene expression in HIV-infected intestinal mucosa. AIDS 1994; 8: 1569-75. Reka S, Garro ML, Kotler DP. Variation in the expression of human immunodeficiency virus RNA and cytokine mRNA in rectal mucosa dining the progression in infection. Lymph Cyto Res 1994; 13: 391-8. Steffen M, Reinecker HC, Petersen J et al. Differences in cytokine secretion by intestinal mononuclear cells, peripheral blood monocytes and alveolar macrophages from HIVinfected patients. Clin Exp Immunol 1993; 91: 30-6. Jarry A, Cortez A, Rene E, Muzeau F, Brouse N. Infected cells and immune cells in the gastrointestinal tract of AIDS patients. An immunohistochemical study of 127 cases. Histopathology 1990; 16: 133-40. Kotler DP, Reka S, Borcich A, Cronin WJ. Detection, localization, and quantitation of HIV-associated antigens

34.

35.

36. 37.

38. 39. 40. 41.

42. 43.

44.

45. 46.

47. 48. 49. 50.

51.

52.

in intestinal biopsies from patients with HIV. Am J Pathol 1991; 139:823-30. Loftus EVJR, Silverstein MD, Sandborn WJ, Tremaine WJ, Harmen WS, Zinmeister AR. Ulcerative colitis in Olmsted County Minnesota, 1940-1993: incidence, prevalence, and survival. Gut 2000: 46: 336-43. Mocroft A, Youle M, Phillips AN et al. The incidence of AIDS-defining illnesses in 4883 patients with human immunodeficiency virus infection. Arch Intern Med 1998; 158: 491-7. Ledergerber B, Egger M, Erard V et al. AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy. J Am Med Assoc 1999; 282: 2220-6. Carpenter HA, Talley NJ. The importance of clinicopathological correlation in the diagnosis of inflammatory conditions of the colon: histological patterns with clinical implications. Am J Gastroenterol 2000; 95: 878 96. Hubert JB, Burgard M, Dussaix E et al. Natural history of serum HIV-1 RNA levels in 330 patients with a known date of infection. AIDS 2000; 14: 123 31. Wilcox CM, Chalasani N, Lazenby A, Schwartz DA. Cytomegalovirus colitis in AIDS: an endoscopic and clinical study. Gastrointest Endosc 1998; 48: 39 43. Roskell DE, Hyde GM, Campbell AP, Jewell DP, Gray W. HIV associated cytomegalovirus colitis as a mimic of inflammatory bowel disease. Gut 1995; 37: 148 50. Caroline DF, Hilpret PL, Russin VL. CMV colitis mimicking Crohn's disease in a patient with acquired immune deficiency syndrome (AIDS). J Can Assoc Radiol 1997; 38: 227 8. Kotler DP, Baer JW, Scholes JV Isolated ileitis due to cytomegalovirus in a patient with AIDS. Gastrointest Endosc 1991; 37: 571 4. Wajsman R, Cappell MS, Biempica L, Cho KC. Terminal ileitis associated with cytomegalovirus and the acquired immune deficiency syndrome. Am J Gastroenterol 1989; 84: 790 3. Schneebaum CW, Novick DM, Chapon AB, Strutynsky N, Yancovitz SR, Freund S. Terminal ileitis associated with Mycobacterium avium-'intracellulare infection in a homosexual man with acquired immune deficiency syndrome. Gastroenterology 1987; 92: 1127 32. Dieterich DT, Rahmin M. Cytomegalovirus colitis in AIDS: presentation in 44 patients and a review of the literature. J Acquir Immune Def Syndr Hum Retrovirol 1991; 1: S29. Beaugerie L, Cywiner-Golenzer C, Monfort L et al. Definition and diagnosis of cytomegalovirus colitis in patients infected by human immunodeficiency virus. J Acq Immun Defic Syn Hum Retrovirol 1997; 14: 423-9. Alfandari S, Ajana F, Senneville E, Beuscart C, Chidiac C, Mouton Y. Hemorrhagic ulcerative colitis due to Isospora belli in AIDS. Int J STD AIDS. 1995; 6: 216. Balthazar EJ, Megibow AJ, Barry M, Opulencia JF. Histoplasmosis of the colon in patient with AIDS: imaging findings in four cases. Am J Radiol 1993; 161: 585-7. Imperiali G, Meucci G, Alvisi C et al. Segmental colitis associated with diverticula: a prospective study. Am J Gastroenterol 2000; 95: 1014-16. Wilcox CM, Schwartz DA, Clark WS. Esophageal ulceration in human immunodeficiency virus infection: causes, response to therapy, and long-term outcome. Ann Intern Med 1995; 123: 143-9. Clayton F, Reka S, Cronin WJ, Torlakovic E, Sigal SH, Kotler DP. Rectal mucosal pathology varies with human immunodeficiency virus antigen content and disease stage. Gastroenterology 1992; 103: 919-33. Kotler DP, Reka S, Clayton F. Intestinal mucosa inflammation associated with human immunodeficiency virus infection. Dig Dis Sci 1993; 38: 1119-27.

C Mel Wilcox 53.

Bjarnason I, Sharpstone DR, Francis N et al. Intestinal inflammation, ileal structure, and function in HIV. AIDS 1996; 10: 1385-91. 54. Castro M. Treatment and prophylaxis of Pneumocystis carm/z pneumonia. Sem Respir Infect 1998; 13: 296-303.

873 55.

Kilby JM, Tabereaux PB, Mulanovich V, Shaw GM, Bucy RP, Saag MS. Effects of tapering doses of oral prednisone on viral load among HIV-infected patients with unexplained weight loss. AIDS Res Human Retrovir 1997; 13: 1533-7.

49 Bone metabolism and inflammatory bowel disease MARIA T. ABREU

Introduction A serious and silent complication of inflammatory bowel disease (IBD) is the development of osteoporosis [1-3]. Osteoporosis is a systemic skeletal disease characterized by low bone mass, microarchitectural deterioration of bone tissue and increased bone fragility and susceptibility to fracture [4]. Milder bone loss or osteopenia is defined as a bone mineral density (BMD) that is 1.0-2.49 standard deviations below a control population, and osteoporosis as BMD that is less than 2.5 standard deviations below a control population [5] (Table 1). Estimates of osteopenia in IBD range from 31% to 59% [6, 7] and osteoporosis from 5% to 41% [1, 811]. Bone loss is more prevalent in patients with Crohn's disease (CD) than those with ulcerative colitis (UC), and this may be related to the fact that childhood onset of CD does not allow children to reach peak bone mass [1]. In a prospective series of patients with IBD the rate of bone loss over a 19month period was -3.1 for CD, -6.4 for UC (includes patients on steroids) and +2.0 for UC after an ileoanal anastomosis [12]. The reasons for development of osteopenia and osteoporosis are multiple and include steroid-induced bone loss [12-15], malabsorption of nutrients [16, 17], vitamin D deficiency [18], smoking [19], and inflammatory cytokines [9, 20] (Table 2). Based on the high prevalence of osteopenia and osteoporosis in patients with IBD, a baseline evaluation of BMD is warranted, especially in patients with CD [21]. The diagnosis of osteopenia and osteoporosis is most often made by dual-energy X-ray absorptiometry (DEXA) of the lumbar spine, proximal femur, and forearm, but can be made using other techniques such as calcaneal ultrasound or quantitative computed tomography [22, 23]. In addition to DEXA scanning for BMD, other serum and urinary markers are available for assessment of bone turnover. In patients with IBD, measurement of urinary N-telopeptide crosslinked of Type 1 collagen, a

Table 1. World Health Organization criteria for assessment of disease severity In patients at risk for osteoporosis Diagnostic criteria*

Classification

7 = 0 t o - 1 SD 7 = 1 to-2SD T ^ - 2 . 5 SD T < - 2 . 5 SD + fragility fractures

Normal Osteopenia Osteoporosis Severe osteoporosis

T scores refer to the number of standard deviations (SD) below or above the peak bone mass in a young adult reference population of the same sex.

Table 2. Factors leading to bone loss In patients with inflammatory bowel disease 1. 2. 3. 4. 5. 6. 7. 8. 9.

Corticosteroids Vitamin D/calcium deficiency Inappropriately high 1,25(0H) vitamin D (granulomatous disease) Malnutrition Physical inactivity Immunosuppressive drugs (cyclosporin^ tacrolimus, methotrexate) Cytokine-mediated changes in bone metabolism Smoking Hypogonadism

marker of bone resorption, was the best predictor of spinal bone loss over a 2-year follow-up period compared with other markers, including bone alkaline phosphatase, osteocalcin, parathyroid hormone (PTH) and vitamin D levels [10, 24, 25].

Corticosteroid-induced osteoporosis The principal cause of secondary osteoporosis in patients with IBD is corticosteroid use [26]. The unfortunate sequelae of chronic corticosteroid use are fractures which occur in 30-50% of steroid users

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 875-883. © 2003 Kluwer Academic Publishers. Printed in Great Britain

876 [4]. Corticosteroids have multiple effects on bone including direct inhibition of bone formation, impaired calcium absorption across the intestine and increased renal calcium excretion, all of which result in a negative calcium balance [27] (Fig. 1). As a result of calcium wasting, secondary hyperparathyroidism results and increases bone resorption. Corticosteroids induce a myopathy which reduces the bone-stimulating effects of muscle activity. In animal models corticosteroids have been shown to induce apoptosis of osteoblasts and osteocytes, thereby resulting in diminished bone formation [11]. More recently, corticosteroids have been shown to regulate bone metabolism through their effect on members of the TNF receptor family, the osteoprotegerins (Fig. 2). Osteoclasts express the RANK receptor [28]. Binding of osteoprotegerin ligand (OPGL) to RANK activates osteoclastic activity. This osteoclastic activity can be blocked by a soluble receptor osteoprotegerin (OPG) that binds OPGL. Corticosteroids decrease OPG expression by 90% and increase OPGL expression by 3-fold, resulting in increased osteoclastic activity [28]. Corticosteroids also cause a 3-fold increase in osteoclast numbers. Finally, corticosteroids inhibit adrenal production of androgens, contributing to bone loss. The effect of corticosteroids on bone loss is most marked in the first 6-12 months of therapy; therefore even short courses of steroids ( < 6 months) will result in marked bone loss [4, 29-31]. Even within 1 week of high-dose steroid exposure, markers of bone resorption are increased [32]. Bone loss secondary to steroids is dependent on both the dose and the duration of steroid use, with doses greater than 7.5 mg/day associated with five times the risk of fractures [33-37]. In studies of chronic exposure to steroids, average steroid doses of as little as 5.6 mg/ day resulted in BMD loss of 2%/year [32, 38, 39]. In a large cohort of patients in the United Kingdom the relative risk of vertebral fracture compared to a control population was 1.55 times higher with a standardized daily dose of steroids of 2.5 mg of prednisolone, and increased to 5.18 for standardized doses of 7.5 mg or greater [37]. Importantly, all fracture risks declined rapidly after cessation of steroids, suggesting that therapy should be given during the time of corticosteroid use. Budesonide is a new-generation steroid with less pituitary axis suppression because of extensive first-pass metabolism in the liver, and has been shown to be effective in the treatment of ileocolic CD [40, 41]. Studies of osteoporosis in patients using inhaled budesonide

Bone metabolism and inflammatory bowel disease Corticosteroids xlGI calcium absorption turinary calcium excretion

i-Osteoblast bone formation Tapoptosis -i-lifespan ^function

iestrogen ^testosterone iadrenal androgens I tOsteoclast bone resorption I"

icalcium

I tPTH

Osteoporosis

Figure 1. Pathophysiology of corticosteroid-induced osteoporosis: bone formation and resorption.

Osteoprotegerin-Directed Osteoclastogenesis osteoclast precursor

OPGL I > 3 R A ] 4 i | ^ y B | ^ J K it)PGL[>20PG A

^ ^ ^ ^ ^ ^ ^ ^ ^

Bone JI resorption + Increased by Increased by .^0^^^ cotiicosteroids •orticosteroids /^"^ilMk.

^^••^^^ osteoblast

— I —

Inhibited by corticosteroids

osteoclast

Figure 2. Osteoprotegerin ligand (OPGL) stimulates osteoclastogenesis. OPG is a soluble, inactivating receptor for OPGL. Local OPG:OPGL ratios control bone resorption rates. Corticosteroids decrease OPG expression by 90% and increase OPGL expression by 3-fold resulting in increased osteoclastica activity.

have demonstrated an inverse correlation between BMD and use of this steroid [42]. In patients with primary biliary cirrhosis, oral budesonide therapy resulted in worsening of their osteoporosis [43]. In patients with CD treated with budesonide for 10 weeks versus prednisolone (32 mg) for 3 weeks, and then tapered, there was decreased osteoblastic activity in the prednisolone-treated group but no changes in bone resorption markers during this time [44]. In short courses (less than 2 weeks), topical rectal steroids do not increase bone turnover in patients with IBD but long-term therapy with these agents will lead to bone loss [45]. These studies demonstrate that no steroid is completely safe with respect to bone loss. With respect to steroid-induced bone loss, approximately 70% of IBD patients are treated with

%11

Maria T. Abreu at least one course of steroids, and cumulative steroid use in IBD patients is inversely correlated with BMD [3, 8, 12]. A comparison of the prevalence of osteopenia in IBD patients revealed that osteopenia was present in 58% of patients treated with steroids compared with 28% not on steroids [15]. The annual rate of bone loss in IBD patients is 3-6%/year, and this rate is largely dependent on steroid use [2,12,26, 46, 47]. Patients with IBD who are treated with corticosteroids are at risk for the complications of osteoporosis, such as vertebral compression fractures and aseptic necrosis of the hip, and clinicians treating IBD should be proactive in preventing corticosteroid-induced osteoporosis and vigilant for the sequelae of corticosteroid use [48]. In spite of adequate therapy to prevent and treat osteoporosis, only 20% of patients on chronic corticosteroids ( > 1 year) undergo some type of evaluation for bone loss, and only 30%) of patients are treated for osteoporosis [49].

Additional causes of bone loss in IBD In addition to bone loss secondary to treatment with steroids, patients with IBD, especially CD, have greater bone loss than expected for the amount of steroid exposure suggesting that other factors contribute to bone loss [50]. Symptoms of greater than 6 months duration prior to the diagnosis of IBD are significantly correlated with low BMD in the absence of disease-specific therapy [51]. Severe osteoporosis can be the presenting problem of children with CD [52]. At least part of the reason for bone loss in CD patients is low vitamin D levels leading to secondary hyperparathyroidism. Osteoporosis is most prominent in CD patients with previous surgical resections [24]. Vitamin D deficiency is related to malabsorption and correlates with previous surgical resections and lack of sun exposure [16, 17, 53]. Postmenopausal women with IBD are especially vulnerable to osteoporosis [2, 3, 54]. In addition to mechanisms leading to increased bone resorption in patients with IBD, histomorphometric analysis of iliac crest bone biopsies from patients with IBD revealed reduced bone formation [14]. For this reason, even in patients with current corticosteroid exposure or a history of corticosteroid use, a full evaluation should be performed to determine if more than one cause of bone loss exists (Fig. 3).

Diagnosis of inflammatory bowel disease

/ Ulcerative colitis

Treatment with corticosteroids

\ Crohn's disease

Baseline bone mineral density (DEXA)

I Concurrent bisphosphonate •Need Ca++ and Vitamin D supplement

If bone loss, evaluate for: •Ca++ deficiency (i^Urinary Ca excretion, «4^absorption) •Vitamin D levels (25-OH, 1,25-OH) •Free testosterone (male) •Free estradiol (women)

Repeat bone mineral density annually

Figure 3. Maintenance of bone health in IBD.

Role of inflammatory cytokines in bone metabolism In addition to known risk factors for bone loss, intestinal inflammation may have systemic consequences with respect to bone metabolism distinct from its effects on absorption of vitamins and nutrients. Patients with newly diagnosed CD may present with osteopenia or osteoporosis, suggesting that other factors may be contributing to increased bone turnover [51, 52]. In the 2,4,6-trinitrobenzenesulfonic acid (TNBS) rat model of colitis, colitic animals have significantly suppressed cancellous bone formation to 30% of that in control animals [55]. The inflammatory cytokines IL-1, IL-6 and TNF-(x are elevated in the systemic circulation of patients with CD. In vitro, TNF-a causes osteoclastic bone resorption and inhibits bone collagen synthesis [6, 7, 5658]. IL-1 and IL-6 also increase bone resorption through activation of osteoclast activity [59]. In addition to their effect on osteoclasts, TNF-oc and IL-ip increase Fas-mediated apoptosis of osteoblasts [60]. TNF-oc further propagates bone resorption by stimulating IL-6 secretion by osteoblasts [61]. TNF-a inhibits the action of l,25(OH) vitamin D through activation of a nuclear inhibitor that antagonizes the effect of vitamin D [62]. Inhibition of TNF-oc or IL-1 in ovariectomized mice prevents bone loss, suggesting that, in animal models, blockade of high TNF-oc states is beneficial to the treatment of osteopenia and osteoporosis [63]. Recently a chimeric Fc fusion form of human OPG with enhanced biological activity was used to inhibit osteoclastic bone resorption in animals challenged with IL-lp, TNF-oc, or l,25(OH) vitamin D3 [64].

878 These data suggest that inflammatory cytokines mediate their osteoclastic effects through OPGL, and that blocking OPGL may be beneficial in osteoporosis in chronic inflammatory states. Because of the role of proinflammatory cytokines in bone turnover, a study was performed in patients with IBD to identify genetic factors associated with increased bone loss [20]. The genetic markers assessed included IL-1 receptor antagonist, IL-6, heat-shock protein 70-2, and heat-shock protein 70-hom alleles. A specific IL-1 receptor antagonist allele (A2) and IL6 allele (130 bp) were independently associated with increased bone loss, and the presence of both alleles led to significantly greater bone loss than either singly. Further studies will likely address the correlation of inflammatory cytokine production (phenotype) to actual bone turnover in patients with IBD.

Vitamin D metabolism in Crohn's disease: role in osteopenia and osteoporosis One of the pathognomonic features of intestinal inflammation in CD is granuloma formation. As with other granulomatous diseases such as sarcoidosis [65], a subset of patients with CD have extrarenal 1-hydroxylase enzyme activity, leading to inappropriately high levels of the active 1,25-dihydroxylated form of vitamin D [66]. The consequence of increased l,25(OH)2 vitamin D levels is increased bone resorption with hypercalcemia and hypercalciuria. In a prospectively identified series of patients, Kantorovich et al. have found that 31% of patients with CD have l,25(OH)2 vitamin D levels that are above normal [67]. TNF-ot may be one of the cytokines produced in the granulomatous milieu that contributes to 1-hydroxylase activation in macrophages and hyper-l,25(OH)2 vitamin D. These CD patients are at increased risk for the development of osteopenia and osteoporosis. Strategies to reduce bone loss in these patients include treatment of the underlying IBD and, potentially, hydroxychloroquine which inhibits the conversion of 25-OH vitamin D to l,25(OH) vitamin D [68].

Drug-induced bone loss Although corticosteroids are a classic cause of druginduced bone loss, other immunosuppressive therapies can also result in osteopenia and osteoporosis. In patients with rheumatoid arthritis methotrexate monotherapy is not associated with increased bone loss, but the combination of methotrexate and pre-

Bone metabolism and inflammatory bowel disease dnisone > 5 mg/day led to greater bone loss than prednisone therapy alone [69, 70]. The calcineurin phosphatase inhibitors cyclosporine and tacrolimus both cause osteopenia and osteoporosis by increasing bone turnover which is reflected in high levels of osteocalcin [71, 72]. Renal transplantation patients given cyclosporine alone, however, did not have significant bone loss at the end of 18 months, whereas those treated with corticosteroids plus cyclosporine experienced the greatest losses, suggesting that corticosteroids are the worst offenders in this regard [73].

Treatment of corticosteroidinduced osteoporosis As stated above, patients are at highest risk for hip and vertebral fractures while on corticosteroids, and the risk is directly proportional to the dose [37]. Although avoidance or discontinuation of corticosteroids is ideal, patients with moderate to severe IBD will likely require corticosteroids. Several strategies have been used to delay or partially reverse bone loss due to steroid use. Calcium and vitamin D supplementation has been shown to reduce bone loss in steroid-treated patients and should be started in all patients receiving corticosteroids [74-76]. Calcitonin has a modest effect on steroid-induced bone loss, delaying bone loss compared with placebo-treated patients [77-83]. The most effective drug therapy for osteoporosis is bisphosphonate therapy [84]. Bisphosphonates prevent dexamethasone-mediated apoptosis of cultured mouse osteoblasts and osteocytes, and prevent dexamethasone-mediated apoptosis of osteoblasts and osteocytes in prednisolone-treated mice [85]. A recent meta-analysis comparing the various options available for steroid-induced bone loss found that bisphosphonates are more effective with respect to preservation of bone density than vitamin D, fluoride or calcitonin [74] (Figs. 4 and 6). The change in lumbar spine BMD in patients treated with vitamin D was +1.96, calcitonin +2.11, compared with bisphosphonates which demonstrated +5.31 increases in BMD. Most of these studies have been performed in patients treated with corticosteroids for rheumatologic or pulmonary diseases, but the results are applicable to patients with IBD. Two approaches have been used in clinical trials of bisphosphonates for corticosteroid-induced osteoporosis: (1) prevention and (2) treatment of existing bone loss. Studies of cyclical etidronate have consis-

Maria T. Abreu tently found a benefit in lumbar spine BMD when given to prevent steroid-induced bone loss [84, 86]. Similarly, studies with daily alendronate or risedronate for prevention of steroid-induced bone loss have demonstrated increases in BMD compared with losses in BMD in placebo-treated patients [87-89]. Data from prevention trials of risedronate performed in 228 patients (largely rheumatologic) demonstrated that risedronate 5 mg/day led to increased BMD at the lumbar spine (+0.59), femoral neck (0.75), and femoral trochanter (+1.42) compared with bone loss of -2.83, -3.06, and -3.06, respectively, in placebo-treated patients [88, 90]. Importantly, studies of bone loss 'prevention' have been done in patients exposed to corticosteroids for less than 3 months - a time when significant bone loss has already occurred. With respect to treatment of existing bone loss in patients taking steroids, the bisphosphonates etidronate, alendronate and risedronate are all effective in improving axial BMD [34, 91, 92]. In addition to the efficacy of bisphosphonates in increasing BMD in patients receiving corticosteroids, large randomized placebo-controlled trials have demonstrated a 70% reduction in the incidence of vertebral fractures in risedronatetreated (5 mg/day) patients compared with placebotreated patients [88, 89]. Currently the only bisphosphonate that has an FDA indication for steroidinduced osteoporosis is risedronate.

879 Alendronate+ERT . Ateiidronate

Risedronate

Raloxifene Miacalcin

% change

In spite of the great need to prevent and treat bone loss in patients with IBD, few clinical trials have been performed in this population of patients. Hormone replacement therapy has been used in postmenopausal women with CD or UC and results in increased bone mass, although few patients in this study were treated with steroids and this approach is for a very limited group of patients [54]. In corticosteroidtreated patients with IBD a randomized, placebocontrolled trial using calcium 1 g and vitamin D 250 units per day failed to confer a benefit in bone loss prevention at the end of 1 year [93]. Fluoride in combination with calcium and vitamin D was associated with small gains in BMD in CD patients with osteopenia previously treated with corticosteroids [94]. Vitamin D therapy by itself prevents ongoing bone loss in patients with CD but does not lead to increases in BMD [95]. Bisphosphonates have been

years

in BMD

-Placebo

Figure 4. Relative effectiveness of antiresorptives in corticosteroid-induced osteoporosis. ERT is estrogen replacement therapy.

8 ^

A *

6;?5

o c

*

^^

<J)

42-

O 0 {

Management of osteopenia and osteoporosis in patients witli IBD

Calcium+vit D

_^

-2 -

Figure 5. Mean ( + SEM) percentage change in BMD from baseline to 12 months in the alendronate (open circle) and placebo (image) groups. A: Lumbar spine L2-L4; B: femoral neck. Data were analyzed on intention-to-treat basis. Asterisks indicate significant differences between treatment groups. From Haderslev etal Gastroenterology 2000.

Bone metabolism and inflammatory bowel disease

880 Chnstensen 1995 Garcia-DelgadQ 1997 Diamond 1997 Henderson 1997 Pitl1998 Saag 1998 Pooled Effect for Bisphosphonates

• • • A

Sambrook 1993 Emkey 1994

Bisphosphonates Calcitonin Fluoride Pooled Effect

Healey 1996 Garcia-Delgado 1997

—

Pooled Effect for Calcitonin Lippuner 1996 Lems 1996 Pooled Effect for Fluoride 1.6

Favors Vitamin D

Effect Size

Favors Other Osteoporosis Therapies

Figure 6. Meta-analysis comparing vitamin D for steroid-induced osteoporosis compared with bisphosphonate, fluoride or calcitonin therapy. Amin S et al. Arthritis Rheumat 1999; 42:1740-51.

Table 3. Strategies to prevent and treat corticosteroid-induced osteoporosis 1. 2. 3. 4. 5.

Minimize corticosteroid exposure Initiate bisphosphonates concurrently with corticosteroids Replace calcium and vitamin D (not sufficient as monotherapy) Calcitonin (can be administered as nasal spray) Hormone replacement therapy

used empirically in IBD patients. A large multicenter study examining the effect of alendronate on steroidinduced bone loss included some IBD patients (24 out of 477 patients) [87]. Thus far, one randomized controlled trial of bisphosphonates has been done in patients with CD and osteopenia defined as a bone mass T score of -1 at the hip or lumbar spine [96]. Patients were randomly assigned to receive placebo or alendronate 10 mg/day for 12 months. Patients were in remission at the time of the study and both groups had similar cumulative corticosteroid use, although current corticosteroid use was not permitted. At the conclusion of the study, patients

receiving alendronate had an increase of 4.6% in BMD whereas the control group experienced a decrease of 0.9% (Fig. 5). Although heartburn and dyspepsia occurred more frequently in the alendronate group, the overall rate of adverse events was low in both groups; specifically no increase in diarrhea was found in the alendronate group. Based on this study, as well as published studies demonstrating the efficacy of bisphosphonates for treatment of corticosteroid-induced osteoporosis, clinicians taking care of patients with IBD should prescribe bisphosphonates for their corticosteroidtreated patients and should obtain baseline bone densitometry studies on all their patients to determine if bone loss has already occurred, and treat appropriately (Fig. 3). Patients with IBD may have multiple causes for bone loss and these should be addressed with specific therapies [5]. In addition to corticosteroid-induced osteoporosis, bisphosphonates are effective in bone loss secondary to cyclosporine [97]. All patients, including those taking bisphosphonates, should be

Maria T. Abreu concurrently receiving 1 g of elemental calcium and 400 units of vitamin D, as well as getting adequate sun exposure [75]. Patients with active IBD should receive appropriate therapy to control the intestinal inflammation, preferably not corticosteroids, in an effort to reduce inflammatory cytokine production, decrease extrarenal production of l,25(OH) vitamin D and increase intestinal absorption of nutrients and vitamins. H y p o g o n a d i s m should be treated specifically with hormone replacement therapy in men and women [98].

References 1. 2. 3. 4. 5. 6.

7.

8. 9.

10.

11.

12. 13. 14.

Adachi JD, Rostom A. Metabolic bone disease in adults with inflammatory bowel disease. Inflam Bowel Dis 1999; 5: 200-11. Clements D, Motley RJ, Evans WD et al. Longitudinal study of cortical bone loss in patients with inflammatory bowel disease. Scand J Gastroenterol 1992; 27: 1055-60. Motley RJ, Clements D, Evans WD et al. A four-year longitudinal study of bone loss in patients with inflammatory bowel disease. Bone Miner 1993; 23: 95-104. Lukert BP, Raisz LG. Glucocorticoid-induced osteoporosis: pathogenesis and management. Ann Intern Med 1990; 112: 352-64. Adachi JD, Olszynski WP, Hanley DA et al. Management of corticosteroid-induced osteoporosis. Semin Arthritis Rheum 2000; 29: 228-51. Bertolini DR, Nedwin GE, Bringman TS, Smith DD, Mundy GR. Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 1986; 319: 516-18. Lader CS, Flanagan AM. Prostaglandin E2, interleukin 1 alpha, and tumor necrosis factor-alpha increase human osteoclast formation and bone resorption in vitro. Endocrinology 1998; 139:3157-64. Dinca M, Fries W, Luisetto G et al. Evolution of osteopenia in inflammatory bowel disease. Am J Gastroenterol 1999; 94: 1292-7. Poflak RD, Karmeli F, Eliakim R, Ackerman Z, Tabb K, Rachmilewitz D. Femoral neck osteopenia in patients with inflammatory bowel disease. Am J Gastroenterol 1998; 93: 1483-90. Schulte C, Dignass AU, Mann K, Goebell H. Reduced bone mineral density and unbalanced bone metabolism in patients with inflammatory bowel disease. Inflam Bowel Dis 1998; 4: 268-75. Weinstein RS, Jilka RL, Parfitt AM, Manolagas SC. Inhibition of osteoblastogenesis and promotion of apoptosis of osteoblasts and osteocytes by glucocorticoids. Potential mechanisms of their deleterious effects on bone. J Clin Invest 1998; 102: 274-82. Roux C, Abitbol V, Chaussade S et al. Bone loss in patients with inflammatory bowel disease: a prospective study. Osteoporosis Int 1995; 5: 156-60. Compston JE, Judd D, Crawley EO et al. Osteoporosis in patients with inflammatory bowel disease. Gut 1987; 28: 410-15. Croucher PI, Vedi S, Motley RJ, Garrahan NJ, Stanton MR, Compston JE. Reduced bone formation in patients with osteoporosis associated with inflammatory bowel disease. Osteoporosis Int 1993; 3: 236-41.

15. Abitbol V, Roux C, Chaussade S et al. Metabolic bone assessment in patients with inflammatory bowel disease. Gastroenterology 1995; 108: 417-22. 16. Compston JE, Horton LW, Laker MF et al. Bone disease after jejuno-ileal bypass for obesity. Lancet 1978; 2: 1-4. 17. Compston JE, Ayers AB, Horton LW, Tighe JR, Creamer B. Osteomalacia after small-intestinal resection. Lancet 1978; 1:9-12. 18. Compston JE, Horton LW Oral 25-hydroxyvitamin D3 in treatment of osteomalacia associated with ileal resection and cholestyramine therapy. Gastroenterology 1978; 74: 900-2. 19. Silvennoinen JA, Lehtola JK, Niemela SE. Smoking is a risk factor for osteoporosis in women with inflammatory bowel disease. Scand J Gastroenter 1996; 31: 367-71. 20. Schulte CM, Dignass AU, Goebell H, Roher HD, Schulte KM. Genetic factors determine extent of bone loss in inflammatory bowel disease. Gastroenterology 2000; 119: 909-920. 21. Compston JE. Detection of osteoporosis in patients with inflammatory bowel disease. Eur J Gastroenterol Hepatol 1997; 9: 931-3. 22. Cranney A, Welch V, Tugwell P et al. Responsiveness of endpoints in osteoporosis clinical trials - an update. J Rheumatol 1999; 26: 222-8. 23. Robinson RJ, Carr I, Iqbal SJ, al-Azzawi F, Abrams K, Mayberry JF. Screening for osteoporosis in Crohn's disease. A detailed evaluation of calcaneal ultrasound. Eur J Gastroenterol Hepatol 1998; 10: 137^0. 24. Dresner-Pollak R, Karmeli F, Eliakim R, Ackerman Z, Rachmilewitz D. Increased urinary N-telopeptide crosslinked type 1 collagen predicts bone loss in patients with inflammatory bowel disease. Am J Gastroenterol 2000; 95: 699-704. 25. Robinson RJ, Iqbal SJ, Abrams K, Al-Azzawi F, Mayberry JF. Increased bone resorption in patients with Crohn's disease. Aliment Pharmacol Ther 1998; 12: 699-705. 26. Silvennoinen JA, Karttunen TJ, Niemela SE, Manelius JJ, Lehtola JK. A controlled study of bone mineral density in patients with inflammatory bowel disease. Gut 1995; 37: 7 1 6. 27. Adachi JD. Corticosteroid-induced osteoporosis. Am J Med Sc 1997; 313: 41-9. 28. Hofbauer LC, Gori F, Riggs BL et al. Stimulation of osteoprotegerin ligand and inhibition of osteoprotegerin production by glucocorticoids in human osteoblastic lineage cells: potential paracrine mechanisms of glucocorticoidinduced osteoporosis. Endocrinology 1999; 140: 4382-9. 29. Saito JK, Davis JW, Wasnich RD, Ross PD. Users of lowdose glucocorticoids have increased bone loss rates: a longitudinal study. Calcif Tiss Int 1995; 57: 115-19. 30. Bornefalk E, Dahlen I, Michaelsson K, Ljunggren O, Ljunghall S. Age-dependent effect of oral glucocorticoids on markers of bone resorption in patients with acute asthma. Calcif Tissue Int 1998; 63: 9-13. 31. Adachi JD, Bensen WG, Bianchi F et al. Vitamin D and calcium in the prevention of corticosteroid induced osteoporosis: a 3 year foUowup. J Rheumatol 1996; 23: 995-1000. 32. Hodsman AB, Toogood JH, Jennings B. Differential effect of inhaled budesonide and oral prenisolone on serum osteocalcin. J Clin Endocrinol 1991; 72: 530-40. 33. Josse R, Adachi JD, Chines A A. Prevention of corticosteroid-induced osteoporosis with etidronate: one year followup with calcium only. Osteoporosis Int 1998; 8: 108. 34. Saag K, Emkey R, Cividino A. Effects of alendronate for two years on BMD and fractures in patients receiving glucocorticoids. Bone 1998; 23: S182. 35. Ruegsegger P, Medici TC, AnHker M. Corticosteroid-induced bone loss. A longitudinal study of alternate day

Bone metabolism and inflammatory bowel disease

882

36. 37. 38. 39.

40. 41.

42. 43.

44.

45.

46. 47.

48. 49. 50.

51.

52. 53. 54. 55.

therapy in patients with bronchial asthma using quantitative computed tomography. Eur J Clin Pharmacol 1983; 25: 615-20. Michel BA, Bloch DA, Fries JF. Predictors of fractures in early rheumatoid arthritis. J Rheumatol 1991; 18: 804-8. van Staa TP, Leufkens HGM, Abenhaim L, Zhang B, Cooper C. Use of oral corticosteroids and risk of fractures. J Bone Miner Res 2000; 15: 993-1000. LoCascio V, Bonucci E, Imbimbo B. Bone loss after glucocorticoid therapy. Calcif Tiss Int 1984; 36: 435-8. Buckley LM, Leib ES, Cartularo KS. Calcium and vitamin D3 supplementation prevents bone loss in the spine secondary to low-dose corticosteroids in patients with rheumatoid arthritis. Ann Intern Med 1996; 125: 961-8. Lowry PW, Sandborn WJ. A comparison of budesonide and mesalamine for active Crohn's disease. Gastroenterology 1999; 116: 1263. Thomsen OO, Cortot A, Jewell D et al. A comparison of budesonide and mesalamine for active Crohn's disease. International Budesonide Mesalamine Study Group. N Engl J Med 1998; 339: 370 4. Wong CA, Walsh LJ, Smith CJ et al. Inhaled corticosteroid use and bonc-mincral density in patients with asthma. Lancet 2000; 355: 1399 403. Angulo P, Jorgensen RA, Keach JC, Dickson ER, Smith C, Lindor KD. Oral budesonide in the treatment of patients with primary biliary cirrhosis with a suboptimal response to ursodeoxycholic acid. Hepatology 2000; 31: 318 23. D'Hacns G, Verstracte A, Chcyns K, Acrden I, Bouillon R, Rutgecrts P. Bone turnover during short-term therapy with mcthylprednisolonc or budesonide in Crohn's disease. Aliment Pharmacol Thcr 1998; 12: 419 24. Robinson RJ, Iqbal SJ, Wolfe R, Patcl K, Abrams K, Mayberry JF. The effect of rcctally administered steroids on bone turnover: a comparative study. Aliment Pharmacol Ther 1998; 12: 213 17. Motley RJ, Crawley EO, Evans C, Rhodes J, Compston JE. Increased rate of spinal trabecular bone loss in patients with inflammatory bowel disease. Gut 1988; 29: 1332 6. Bernstein CN, Sceger LL, Sayre JW, Anton PA, Artinian L, Shanahan F. Decreased bone density in inflammatory bowel disease is related to corticosteroid use and not disease diagnosis. J Bone Miner Res 1995; 10: 250 6. Matsuda K, Watanabe T, Abo Y et al. Severe complications of ulcerative colitis after high-dose prednisolone and azathioprine treatment. J Gastroenterol 1999; 34: 390-4. Hougardy DM, Peterson GM, Bleasel MD, Randall CI. Is enough attention being given to the adverse effects of corticosteroid therapy? J Clin Pharm Ther 2000; 25: 227-34. Bjarnason I, Macpherson A, Mackintosh C, Buxton-Thomas M, Forgacs I, Moniz C. Reduced bone density in patients with inflammatory bowel disease. Gut 1997; 40: 228-33. Schoon EJ, Blok BM, Geerling BJ, Russel MG, Stockbrugger RW, Brummer RJ. Bone mineral density in patients with recently diagnosed inflammatory bowel disease. Gastroenterology 1203; 119: 1203-8. Thearle M, Horlick M, Bilezikian JP et al. Osteoporosis: an unusual presentation of childhood Crohn's disease. J Clin Endocrinol Metab 2000; 85: 2122-6. Vogelsang H, Klamert M, Resch H, Ferenci P. Dietary vitamin D intake in patients with Crohn's disease. Wien Klin Wochenschr 1995; 107: 578-81. Clements D, Compston JE, Evans WD, Rhodes J. Hormone replacement therapy prevents bone loss in patients with inflammatory bowel disease. Gut 1993; 34: 1543-6. Lin CL, Moniz C, Chambers TJ, Chow JW. Colitis causes bone loss in rats through suppression of bone formation. Gastroenterology 1996; 111: 1263-71.

56. Giuliani N, Pedrazzoni M, Passeri G, Girasole G. Bisphosphonates inhibit IL-6 production by human osteoblast-like cells. Scand J Rheumatol 1998; 27: 3 8 ^ 1 . 57. Thomson BM, Saklatvala J, Chambers TJ. Osteoblasts mediate interleukin 1 stimulation of bone resorption by rat osteoclasts. J Exp Med 1986; 164: 104-12. 58. Thomson BM, Mundy GR, Chambers TJ. Tumor necrosis factors alpha and beta induce osteoblastic cells to stimulate osteoclastic bone resorption. J Immunol 1987; 138: 775-9. 59. Manolagas SC, Bellido T, Jilka RL. Sex steroids, cytokines and the bone marrow: new concepts on the pathogenesis of osteoporosis. Ciba Found Symp 1995; 191: 187-96; discussion 197-202. 60. Tsuboi M, Kawakami A, Nakashima T et al. Tumor necrosis factor-alpha and interleukin-1 beta increase the Fasmediated apoptosis of human osteoblasts. J Lab Clin Med 1999; 134:222 31. 61. Kurokouchi K, Kambe F, Yasukawa K et al. TNF-alpha increases expression of IL-6 and ICAM-1 genes through activation of NF-kappaB in osteoblast-like ROSl7/2.8 cells. J Bone Miner Res 1998; 13: 1290 9. 62. Fernandez-Martin JL, Kurian S, Farmer P, Nanes MS. Tumor necrosis factor activates a nuclear inhibitor of vitamin D and retinoid-X receptors. Mol Cell Endocrinol 1998; 141:65 72. 63. Votta BJ, Bcrtolini DR. Cytokine suppressive anti-inflammatory compounds inhibit bone resorption in vitro. Bone 1994; 15: 533 8. 64. Morony S, Capparclli C, Lee R et al. A chimeric form of osteoprotegerin inhibits hypercalcemia and bone resorption induced by IL-lbeta, TNF-alpha, PTH, PTHrP, and 1, 25(OH)2D3. J Bone Miner Res 1999; 14: 1478 85. 65. Conron M, Young C, Beynon HL. Calcium metabolism in sarcoidosis and its clinical implications. Rheumatol (Oxf) 2000:39:707 13. 66. Bosch X. Hypercalcemia due to endogenous overproduction of 1,25-dihydroxyvitamin D in Crohn's disease. Gastroenterology 1998;114:1061 5. 67. Kantorovich V, Adams JS, Targan SR, Hassard PV, Vasiliauskas EA. Endogenous Overproduction of 1, 25-dihydroxyvitamin D in patients with Crohn's disease. American Society for Bone and Mineral Research Meeting 2000. 68. Barre PE, Gascon-Barre M, Meakins JL, Goltzman D. Hydroxychloroquine treatment of hypercalcemia in a patient with sarcoidosis undergoing hemodialysis. Am J Med 1987;82: 1259-62. 69. Buckley LM, Leib ES, Cartularo KS, Vacek PM, Cooper SM. Effects of low dose methotrexate on the bone mineral density of patients with rheumatoid arthritis. J Rheumatol 1997;24: 1489-94. 70. Bianchi ML, Cimaz R, Galbiati E, Corona F, Cherubini R, Bardare M. Bone mass change during methotrexate treatment in patients with juvenile rheumatoid arthritis. Osteoporosis Int 1999; 10:20-5. 71. Epstein S, Shane E, Bilezikian JP. Organ transplantation and osteoporosis. Curr Opin Rheumatol 1995; 7: 255-61. 72. Inoue T, Kawamura I, Matsuo M et al. Lesser reduction in bone mineral density by the immunosuppressant, FK506, compared with cyclosporine in rats. Transplantation 2000; 70: 774-9. 73. Aroldi A, Tarantino A, Montagnino G, Cesana B, Cocucci C, Ponticelli C. Effects of three immunosuppressive regimens on vertebral bone density in renal transplant recipients: a prospective study. Transplantation 1997; 63: 380-6. 74. Amin S, LaValley MP, Simms RW, Felson DT. The role of vitamin D in corticosteroid-induced osteoporosis: a metaanalytic approach. Arthritis Rheum 1999; 42: 1740-51. 75. Anonymous. Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. Ameri-

Maria T. Abreu

76.

77. 78.

79. 80.

81.

82. 83.

84. 85.

86. 87.

can College of Rheumatology Task Force on Osteoporosis Guidelines. Arthritis Rheum 1996; 39: 1791-801. Homik J, Suarez-Almazor ME, Shea B, Cranney A, Wells G, Tugwell R Calcium and vitamin D for corticosteroidinduced osteoporosis. Cochrane Database of Systematic Reviews [Computer file] 2000: CD000952. Sambrook P, Birmingham J, Kelly P et al. Prevention of corticosteroid osteoporosis. A comparison of calcium, calcitriol, and calcitonin. N Engl J Med 1993; 328: 1747-52. Healy JH, Paget S, WiUiams-Russo P. A randomized controlled trial of salmon calcitonin to prevent bone loss in corticosteroid treated temporal arteritis and polymyalgia rheumatica. Calcif Tissue Int 1996; 59: 73. Adachi JD, Bensen WG, Bell MJ et al. Salmon calcitonin nasal spray in the prevention of corticosteroid-induced osteoporosis. Br J Rheumatol 1997; 36: 255-9. Luengo M, Picado C, Del Rio L, Guanabens N, Montserrat JM, Setoain J. Treatment of steroid-induced osteopenia with calcitonin in corticosteroid-dependent asthma. A one-year follow-up study. Am Rev Respir Dis 1990; 142: 104-7. Luengo M, Pons F, Martinez de Osaba MJ, Picado C. Prevention of further bone mass loss by nasal calcitonin in patients on long term glucocorticoid therapy for asthma: a two year follow up study. Thorax 1994; 49: 1099-102. Ringe JD, Welzel D. Salmon calcitonin in the therapy of corticoid-induced osteoporosis. Eur J Clin Pharmacol 1987; 33: 35. Kotaniemi A, Piirainen H, Paimela L et al. Is continuous intranasal salmon calcitonin effective in treating axial bone loss in patients with active rheumatoid arthritis receiving low dose glucocorticoid therapy? J Rheumatol 1996; 23: 1875-9. Homik JE, Cranney A, Shea B et al. A metaanalysis on the use of bisphosphonates in corticosteroid induced osteoporosis. J Rheumatol 1999; 26: 1148-57. Plotkin LI, Weinstein RS, Parfitt AM, Roberson PK, Manolagas SC, Bellido T. Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J CHn Invest 1999; 104: 1363-74. Adachi JD, Bensen WG, Brown J et al. Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N Engl J Med 1997; 337: 382-7. Saag KG, Emkey R, Schnitzer TJ et al. Alendronate for the prevention and treatment of glucocorticoid-induced osteo-

883

88.

89.

90.

91.

92. 93.

94.

95.

96.

97. 98.

porosis. Glucocorticoid-Induced Osteoporosis Intervention Study Group. N Engl J Med 1998; 339: 292-9. Reid D, Cohen S, Pack S, Chines A, Ethgen D. Risedronate reduces the incidence of vertebral fractures in patients on chronic corticosteroid therapy. Arthritis Rheum 1998; 41: S136. Cohen S, Levy RM, Keller M et al. Risedronate therapy prevents corticosteroid-induced bone loss: a twelve-month, multicenter, randomized, double-blind, placebo-controlled, parallel-group study. Arthritis Rheum 1999; 42: 2309-18. Reid D, Devogelaer JP, Hughes R et al. Risedronate is effective and well tolerated in treating corticosteroid-induced osteoporosis. American College of Rheumatology annual meeting 1998. Reid D, Cohen S, Pack S. Risedronate is an effective and well-tolerated therapy in both the treatment and prevention of corticosteroid-induced osteoporosis. Bone 1998; 23: S402. Adachi JD, Pack S, Chines AA. Intermittent etidronate and corticoisteroid-induced osteoporosis. N Engl J Med 1997; 337:1921. Bernstein CN, Seeger LL, Anton PA et al. A randomized, placebo-controlled trial of calcium supplementation for decreased bone density in corticosteroid-using patients with inflammatory bowel disease: a pilot study. AHment Pharmacol Ther 1996; 10:777-86. von Tirpitz C, Klaus J, Bruckel J et al. Increase of bone mineral density with sodium fluoride in patients with Crohn's disease. Eur J Gastroenterol Hepatol 2000; 12: 1924. Vogelsang H, Ferenci P, Resch H, Kiss A, Gangl A. Prevention of bone mineral loss in patients with Crohn's disease by long-term oral vitamin D supplementation. Eur J Gastroenterol Hepatol 1995; 7: 609-14. Haderslav KV, Tjellesen H, Sorensen HA, Staun M. Alendronate increases lumbar spine bone mineral density in patients with Crohn's disease. Gastroenterology 2000; 119: 639-646. Sass DA, Bowman AR, Yuan Z, Ma Y, Jee WS, Epstein S. Alendronate prevents cyclosporin A-induced osteopenia in the rat. Bone 1997; 21: 65-70. Robinson RJ, Iqbal SJ, Al-Azzawi F, Abrams K, Mayberry JF. Sex hormone status and bone metaboHsm in men with Crohn's disease. Aliment Pharmacol Ther 1998; 12: 21-5.

Section III BACK TO THE LABORATORY BENCH

50 Epilogue: Bench to bedside and bacic to bench STEPHAN R. TARGAN, LOREN C. KARP AND FERGUS SHANAHAN

Where we have been and where we are going The preceding chapters combine to provide an excellent depiction of all facets of the study of IBD, from bench to bedside. The individual chapters are components of a coherent, albeit detailed, story of the local and systemic pathophysiology of intestinal inflammation, presented in association with a wellreasoned series of management strategies based on this knowledge. Research advances and current concepts of etiopathogenesis are described in the context of what is already known of the clinicopathologic features of these disorders. The book as a whole illustrates the effectiveness of a multilateral approach by basic scientists and clinician investigators in the field of IBD. The conclusions of the chapters were drawn purposefully to offer a glimpse of where the field is moving to stimulate research back at the bench. The interplay of state-of-the-art technology and greater understanding of the pathophysiology in both human and animal disease has brought us far along our road, and has led to excellent translational research resulting in the development of better therapeutic options. Upon this model our efforts can be maximized, our scope expanded, and our progress expedited to take us the rest of the way. We know IBD comprises numerous discrete entities that can be stratified by a variety of subclinical and clinical markers that may well identify which patients are likely to respond to any particular therapeutic intervention. We also have made great progress toward defining the roles played by infectious agents in the etiology or pathogenesis of IBD. Evidence suggests that the elevated immune function, characteristic of these diseases, is in response to normal commensal bacteria rather than any specific pathogen. It is unlikely that the role of infectious agents correlates directly with the disease responses that are characteristic of IBD. More likely the process is indirect,

requiring any number of combinations of genes, immune responses, and environmental triggers, to manifest as disease. Specific manipulation of bacterial expression and the corresponding immune response may be the basis of very effective therapies in the near future. Research over the past several years has changed the management of IBD. The effectiveness of therapeutic anti-tumor necrosis factor-oc monoclonal antibodies (anti-TNF-oc) in a portion of patients with Crohn's disease illustrates the unique immune mechanisms that underlie the intestinal inflammation in different subpopulations of patients. The foundation of our Bench to Bedside and Back to Bench concept has been laid, and the planks of the foundation are described in detail by the chapters of this book. The final decades of the last century saw the Bench to Bedside portion of the equation evolve. New therapies based on laboratory science were developed, tested and some came to be used in clinical practice. At the very end of the last century only a few compounds were tested in humans with associated parallel basic science, or translational, investigations. Although such studies were few, the yield in terms of information was great. Not long ago only a few animal models were available for the study of dysregulated intestinal inflammation. In the intervening years the number of models has grown to well over 30, mainly in the form of genetically engineered mouse strains. Analysis of these models has taught us that chronic colitis can be induced by diverse genetic abnormalities. However, the inflammatory patterns and their clinical expressions are also distinct, yielding new opportunities to link alterations in specific genetic loci with alterations of T cell responses, mucosal permeability, and in-vivo clinical patterns. Another striking area of progress in IBD research that has traversed from the Bench to the Bedside and back again is related to the issue of a microbial pathogenesis of IBD. Despite more than 50 years of

Stephan R. Targan, Fergus Shanahan and Loren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 887-891. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Epilogue research a specific bacterial or viral organism has not been identified as a direct cause of IBD pathogenesis. Investigations in animals, however, identified a role for commensal bacteria in induction of dysregulated mucosal inflammation. Verification of bacterial involvement came with the finding that, despite the genetic or cellular manipulation in any given model, its attendant mucosal inflammation is lost in a pathogen-free environment. We now know that there are two ways in which the mucosa responds to bacterial antigens. The antigenspecific T cell response is well characterized. A more complex relationship exists between bacterial structures such as lipopolysaccharides and peptidoglycans that have repetitive sequences that are recognized by cell receptors. The relationship is receiving a great amount of attention at the time of this writing. The most exciting of these receptors appear to be present on epithelial cells and macrophages and are known as loll-like receptors. First described in Drosophila, toll-like receptors engage bacterial structures and generate mucosal inflammation. Alterations in such receptor signaling may well play a key role in the inflammation of IBD. Bacterial antigens are likely to trigger certain T cell responses in the mucosa, and the balance of these responses is critical to regulating the intestinal immune response. T cell responses can be divided broadly into 'diseaseinducing' and 'disease-preventing', depending upon the cytokines released. In the majority of animal models, the presence of effector cells secreting T helper-1 (Thl) cytokines (interferon-gamma, IFNy, TNF-oc) is responsible for disease, and by this criterion these models are reflective of human Crohn's disease. One clear exception is the T cell receptor-alpha (TCR-a) 'knockout' mouse in which a unique subset of TCR-P dim/interleukin (IL)-4 producing cells is responsible for inflammation; this model may therefore be considered to be more similar to human ulcerative colitis. Two populations of disease-preventing or down-regulatory T cells have been described; Tr-l/IL-10 producing T cefls and the T-helper-3 (Th3)/transforming growth factor-beta (TGF-P) producing T cells. Studies have demonstrated that these two T cell populations counterbalance the effector cells and protect against dysregulated inflammation. It is an imbalance between these effector and regulatory populations that presumably results in disease expression. Renewed interest in commensal bacteria and their relationship to T cell responses in the mucosa has led to investigations demonstrating that the transfer of a

specific subset of bacterial antigen-stimulated Thl cells can induce colitis in animals. Furthermore, down-regulatory or T-regulatory-1 (Tr-1)/IL-10producing cells, stimulated by the same bacterial antigens as the pathogenic Thl cells, prevent expression of colitis when co-transferred with these antigen-specific Thl cells. These bacterial antigen-triggered effector and regulatory cell populations clearly influence the interplay between clinical expression and mucosal inflammation. The diversity in expression of colitis, i.e. cecal involvement, rectal ulceration, specific ileal involvement, ulcerative lethal disease, is dependent upon any one or more genetic abnormalities. Furthermore, the same genetically altered animal can manifest a different expression of disease when raised in different housing facilities, suggesting an additional mechanism for the heterogeneity of these diseases. It is quite possible that genetic and microenvironmental heterogeneity in these animal models is similar to what is seen among patients with IBD. This paradigm is now the foundation of our understanding of pathogenic mechanisms in human disease. The investigations in animal models of mucosal inflammation have proceeded in parallel with those in human IBD. Much progress has been made in identifying the genes associated with these inflammatory disorders. Genome-wide scanning and candidate gene approaches have revealed several disease-associated loci. One important implication of these studies is that IBD is an oligogenetic disease: genome-wide scans thus far have revealed at least four loci confirmed by independent groups, and an additional three others are suggested. Some of these chromosomes and loci are associated with Crohn's disease and/or ulcerative colitis individually, and others are associated with both. Marker antibodies and/or disease expression further refines the associations. Multiple studies in humans, as well as the animal studies summarized above, have demonstrated diff^ering genetic associations with clinical phenotype as well as marker antibody expression. Allelic diversity at the susceptibility loci thus probably underlies the immunologic and clinical heterogeneity of IBD. The chromosome 16 locus was the first to be identified by a whole genome approach. The IBDl gene (NOD2) has been identified by using a positional cloning strategy - linkage followed by linkage disequilibrium mapping. Three independent associations for Crohn's disease were discovered. Simultaneously, by a candidate gene approach, an independent group also demonstrated association

Stephan R. Targan, Lor en C. Karp and Fergus Shanahan between the NOD2 and Crohn's disease. Clearly, NOD2 is a susceptibility gene to Crohn's disease, as it increases an individual's risk for Crohn's disease, but it is neither sufficient nor necessary for the development of the disease. Significant advances have already been made, based on the identification of NOD2 and its relationship to subtypes of Crohn's disease. Hugot et al and Abreu et al. have recently published separate reports demonstrating that fibrosing Crohn's disease is associated with the dose effect of two mutations, and several other groups have further confirmed these findings, with results not yet published at the time of this writing. These studies indicate that abihty to predict which patients will experience a more complicated course is in the immediate future. NOD2 will no doubt continue to shed light on the etiologic pathways of IBD and accelerate the discovery of additional susceptibility genes for IBD. The identification of IBD-associated marker antibodies at the bench has made a great impact on patient care at the bedside, and in turn has stimulated further laboratory research. Since the discovery of perinuclear antineutrophil cytoplasmic antibodies (ANCA) more than 10 years ago, much has been learned about the significance of marker antibodies in the serum of patients with IBD. In addition to pANCA, interest has focused on antibodies to Saccharomyces cerevisiae (ASCA), which have been shown to differentiate between ulcerative colitis and Crohn's disease. Progress continues with these antibodies, and other bacterial antigens have been detected which identify strongly with disease-associated antibodies in IBD. These marker antibodies are important for their diagnostic utility and stratification of patients into distinct clinical phenotypes. These antibodies are associated not only with the different locations of disease, but also with aggressiveness of disease course and responses to treatment by specific manipulation of cytokines. It is of particular interest that some of the marker antibodies, such as pANCA, appear to be cross-reactive with commensal bacterial antigens. Therefore, the presence of marker antibodies provides a platform as a basis to study bacterial-host interactions as well as to define subsets of patients with Crohn's disease and ulcerative colitis. At the bench we shall now endeavor to define the relevance of these findings in animals to the diversity of human disease seen among patients with IBD. In addition we must determine which immune alterations and bacterial sensitivity in the different animal models are represented among the

889 varied clinical expressions of disease seen in these disorders. From our findings in animal models, we assume that patterns of immune responses to environmental and bacterial antigens differ widely among groups of patients with Crohn's disease. Landers et al have recently reported that expression of certain serologic markers clusters in certain patient subgroups. Even in groups of patients that have been classified as 'indeterminant', expression of certain markers remain consistent over time and suggest the clinical course that is ultimately taken, as has been shown by Joossens et al. Thus, what seemed futuristic only recently, clinically useful information derived from studying genes and their relationship to immunophysiologic responses will be obtainable in a much shorter time. It is therefore possible that certain antibiotics might be most effective in those patients who have the most robust responses to bacterial antigens. Likewise, the most robust responses to the broadest number of these antigens may define those patients who can best be treated by manipulation of the bacterial flora. If the antigens to which an individual is sensitive are among the anaerobic bacteria then metronidazole may be more effective. If these antigens are among the aerobic bacteria, then these patients may best be treated with ciprofloxacin. Finally, in patients who are immunoreactive to both aerobic and anaerobic populations, such patients may require a combination of both metronidazole and ciprofloxacin. Although many inflammatory molecules are elevated in the mucosa of patients with IBD, their precise role in disease pathogenesis has never been determined. In animal models the success of treatment by manipulation of Thl cytokines, e.g. antibodies toTNF-oc, IFN-y or IL-12, led to trials of antiTNF-a for the treatment of Crohn's disease. Results of these trials showed for the first time that cytokine expression in the mucosa could alter disease course in the majority of patients with Crohn's disease. 'Back to Bench' research demonstrated down-regulation of Thl cytokine-producing cells in the inflamed mucosa to the level of those in the uninflamed areas as one of the mechanisms responsible for the disease ameliorating effect. Unlike some animal models in which cecal bacterial antigens drive the dominant Thl profile, no similar antigen trigger has been identified in human disease. A proportion of Crohn's disease patients can be treated by manipulation of the bacterial flora by either antibiotic therapy or a diversion of the fecal stream, which suggests diverse responses to bacterial antigens among the Crohn's

890 disease population. In ulcerative colitis, patients with high levels of pANCA and serum IgG reactivity to the E. coli antigen, OmpC, are three times more likely to develop chronic pouchitis than patients without high levels of pANCA. These findings, taken together, highlight the potential importance of bacterial interactions in the induction or maintenance of inflammation. Overall, the diversity of clinical expression seen in IBD is very likely dependent upon a combination of genetic abnormalities and the presence of certain commensal bacteria.

The 21st century - the age of translationalism With the rapidly growing body of knowledge, it behooves us to effect the most expeditious application of our findings to patient care and research in return. We must now determine which of the many discoveries made in animals are relevant to human disease. To do so requires a process by which ideas can be exchanged among investigators and work done simultaneously, crossing between animal models and in-vitw testing for human disease. Given what is already known about the heterogeneity of these disorders, this process is likely to be complex, and it is not yet possible to say how many combinations of genes and antigens are likely to result in some sort of chronic intestinal inflammation or extraintestinal disease. Nevertheless, these investigations are crucial to developing phenotypically defined disease groups. Translational research will lead to some very exciting discoveries, based on the 'omics', genomics, proteinomics and phenomics. Genomic investigations of genes and their function are bringing about a revolution in our understanding of the molecular mechanisms of disease, including the complex interplay of genetic and environmental factors. Genomics is also stimulating the discovery of breakthrough pharmaceuticals by revealing hundreds of new biological targets for the development of drugs, and by giving scientists innovative ways to design new drugs, vaccines and diagnostics. Genomics-based therapeutics include 'traditional' small chemical drugs, protein drugs, and potentially gene therapy. Proteinomics refers to the study of proteins expressed by genes and understanding the role of every human protein. By manipulation of these proteins the course of the disease process can be altered and potentially the expression of disease can be eliminated altogether.

Epilogue Phenomics is the study of the result or sequelae of a gene's function, i.e. the disease manifestation. Phenomics is the ultimate point of tomorrow's drug discovery effort as it allows researchers to rapidly validate genes as potential targets for drug discovery and point them to the right individuals. At present, researchers use genomic 'knockout' animal models to determine the biological function of a gene, and then compare the function to that of humans. As diseases are related to subtle changes in protein activity, rather than to complete loss or gain in any one protein, researchers must produce varying changes in protein activity to determine a gene's relation to the specific disease process. Through these efforts we can meet our challenge to refine patient subsets, not strictly from clinical points of view. There is a tendency among clinicians to 'lump' patients into groups based on certain clinical aspects of their disease; e.g. disease in the ileum. Nevertheless, not all patients with ileal disease are the same. Simultaneously we must refrain from 'splitting' so many subtypes so as not to lose sight of 'the forest for the trees.' To determine the phenotypes, the genetic and immunologic origins and clinical significance of the phenotypes requires a very certain type of investigator the translationalist. We need to encourage a cadre of educated and creative translational investigators. Such individuals will have in-depth knowledge of disease pathogenesis combined with medical training and education in clinical investigation, and with these tools, they will bring about our expectations for the 21st century (Table 1).

Too many targets, too few patients Estimates have placed the number of individuals affected with IBD at as high as two million. The actual pool of patients willing and able to participate in clinical trials is much, much smaller. A primary goal of the translationalists, then, must be the development of 'reagent-grade' patients. Well-defined populations will permit highly specific eligibility criteria for study participation. It is our challenge to accelerate our knowledge of these diseases without wasting our precious patient resources. In the 1990s we were successful in our efforts to capture the interest of the biotechnology and pharmaceutical industries to encourage them to focus on developing new and improved treatments for IBD. This success has strained our resources and there are now not

Stephan R. Targan, Lor en C. Karp and Fergus Shanahan

891

Table 1. IBD research in the 21st century Effort

Result

1.

Intercommunication of rodent and human genetic research findings

Identification of genetic mutations and interactions with relevance to human diseases

2.

Definition of IBD relevant bacterial antigens, i.e. commensal associated molecular patterns (CAMPs)

Definition of environmental triggers

3.

Prioritization of clinical trials via cooperation between academic institutions and the pharmaceutical industry

Rapid answers, targeted therapeutic efficacy

4.

Definition of 'reagent grade' patients

Customization of diagnosis and prognosis and therefore, therapeutic targets and treatment plans

5.

Crossbreeding of relevant rodent models to better approximate human IBD subtypes

Acceleration of the progress of 1-4

enough patients to do all of the trials necessary to bring a product to market. Thus, the responsibilities will fall upon this new generation of translationalists to prioritize the potential benefit of the compounds. We need to make highly informed, difficult decisions from among many promising options. Should the biopharm industry direct the emphasis of clinical testing or should translationalists influence which of the many targets are the best for medical intervention? It is our challenge to ensure that those with financial resources understand the clinical and pathophysiologic milieu in which their efforts are being devoted. It is also important to coordinate the efforts in the academic laboratories with those of the biopharmaceutical industry. We would do well to determine which animal models are most relevant to human disease and ensure that these same models are those used by industry for their efficacy studies. There is already evidence that the molecules responsible for acute disease flares are different from those involved with chronic inflammation. Such findings need to be shared with industry so that therapeutic development can be targeted for such specific situations. A further need, which requires the coordination of academic and industry laboratories, is to apply novel

drug delivery systems to our models. For example, bacteria have been used to deliver IL-10 therapy in mouse models of colitis. Among the benefits of such cooperation are more limited immunomodulatory effects from treatments, and minimization of lengthy drug-administering infusions. Given the current status of research in IBD, wellorchestrated cooperation between the basic and clinical sciences and academic and pharmaceutical industry research is likely to yield major advances in understanding of the disease and the development of treatments. The next edition of this textbook is sure to contain examples of the fruits of this cooperation and feature contributions from this new generation of well-trained translationalists. As we envision it, patient management in the 21st century would be far more efficient, with fewer discomforts for the patient, potentially altering the natural history of the disease. In a typical example a patient would present with diarrhea of 1 week's duration. A battery of noninvasive tests would be administered to characterize the immunogenetic phenotype of disease. A diagnosis and prognosis would be made, i.e. I B D # 1 , or I B D # 2 , etc. Finally, a specific, targeted therapeutic plan would be designed and implemented that would ameliorate the patient's disease.

Index

A20-deficient mice 80-1 AAV see adeno-associated virus abdominal imaging 371-407, 855-6 abscesses 297, 379, 383, 388, 391, 731, 742 absorption 264, 265, 314-18 5-acetylsalicylic acid (5-ASA) 683-4, 702-4, 705, 706 active immunity 111-12 active suppression 113-15 acute phase proteins 237 adaptive immunity, normal 111 adeno-associated virus (AAV) 567 adenocarcinoma 3 51 adenomas, sporadic 722-3 adenomatous polyposis coli (APC) 713-14 adenovirus 567 adhesion 186-8, 669-70, 848-9 adjunct therapies 595-6 adolescents 588, 773-90 adverse effects aminosalicylates 498 anti-TNF-a monoclonal antibody and fusion protein therapies 512 azathioprine 505 cyclosporine 513 glucocorticoids 502-3 6-mercaptopurine 505 methotrexate 509 nicotine 514 tacrolimus 513 Aeromonas spp. 437-8, 852 age 7-8, 284-5, 296, 726-7, 800 allopurinol 652-3 amebiasis 438, 852-3 4-aminosalicylate (4-ASA) 615 5-aminosalicylate (5-ASA) Crohn's disease 659-60 pediatric IBD 777, 778 pharmacology 495-9 pouchitis 650, 651-2 pregnancy with IBD 766 ulcerative colitis 606-10 amyloidosis 245, 731, 742 anal sphincter function 636 anastomoses 639, 686, 688 anatomy 177-81,2912,339 ANCA see antineutrophil cytoplasmic antibodies anemia 238-40 anergy 113 aneuploidy 711-12 angiogenesis 185 animal models 37-9, 67-99, 149, 224, 887 anion-binding resins 838 ankylosing spondyHtis (AS) 28, 311-12 anorectal lesions 285-6 anorexia 259-60

anti-a4 antibodies 564 anti-acinar cytoplasmic granule pancreatic autoantibody 21516 anti-adhesion therapies 669-70 anti-CD4 antibodies 563 anti-cytokine therapies 668-9 anti-goblet cell antibodies (GAB) 415 anti-inflammatory cytokines 15 5--6, 158-63, 235, 236, 562-3 anti-interleukin-12 669 anti-nuclear factor ( N F ) - K B 669 anti-pancreatic antibodies (PAB) 415 anti-platelet-activating factor agents 616 dinii-Saccharomyces cerevisiae antibodies (ASCA) 33-5, 216, 217-18,412-13 anti-TNF therapies 615, 663 antibiotics see also individual antibiotics CD postop recurrence 704 Clostridium difficile diarrhea and colitis 830, 834-8 Crohn's disease 661 pediatric IBD 778,781 pouchitis 650-1 pregnancy with IBD 768 recurrent Clostridium difficile diarrhea 837-8 role in management of IBD 573-85 UC treatment 616 antibodies 211-23 see also individual antibodies antigen presentation 225 antigen-presenting cells (APCs) 130-4 antimetabohtes 239 see also 6-mercaptopurine; azathioprine; methotrexate; sulfasalazine antimicrobial chemotherapy 857, 858-9 antimycobacterial therapy 578-9 antineutrophil cytoplasmic antibodies (ANCA) 33-5, 409-12 APC see adenomatous polyposis coli aphthous ulcers 384, 386 appendectomy 15 appetite 257-9 arterioles 178, 181-4 arthritis autoimmune events 311-12 defective immunoregulation 321-2 diagnosis 751-2 genetic influences 323-4 history/classification 747-8 immune activation in extraintestinal organs 320-1 immune complex-mediated injury 313 infection 308 management 752-4 nonimmune factors 325 pathogenesis 748-51 systemic absorption of toxic luminal contents 316-17 systemic distribution of inflammatory mediators 318-19

Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 893-903. © 2003 Kluwer Academic Publishers. Printed in Great Britain

Index

894 articular complications 747-54 5-ASA see 5-acetylsalicylic acid; 5-aminosaIicylate 4-ASA see 4-aminosalicylate ASCA see 'dnU-Saccharomyees cerevisiae antibodies associations 28-30, 40-2, 44-54 see also linkage studies asymptomatic carrier state 829 autoantigens 212-13,214-16 autoimmune aspects 309-12, 645, 731, 740 autonomic nervous system 229 axial arthritis 748, 749, 751 2, 753 azathioprine (AZA) anemia 239 Crohn's disease 661 2,706 pediatric IBD 778,779 pharmacology 504 6 pregnancy with IBD 767 8 UC treatment 613 14 B7-related protein 1-Fc transgenic mice 81 'background genes' 94 'backwash' ileitis 394, 397 bacteria see enteric bacterial flora; individua/ bacteria bacterial epithelial lymphoid circuit 85,92,115 16,226 bacterial-driven IBD 124 5, 139 Bacferoides sp. 212,699 BaUmtidium co/i 438, 853 balsalazide 495 9,610 barium techniques 371 2, 373 Behcet's syndrome 448, 450 bias 269 71 Bifidohacler 212 biliary disease 734 5, 742 biodiversity, normal gut 122 4 biological control mechanisms 104 8 biologicals 553 71,778,780,838 biopsy 357 67, 725 6 bismuth 616, 652 Blastocvstis honiinis 438 blood flow regulation 104 8, 180 2 blood tests 854 bone disease 298, 784 5 bowel distension 379,381 bowel perforation 379, 382 bowel resection 684, 686, 697, 700 2 broken tolerance 115 Budd Chiari syndrome 731,742 budesonide enemas 607 C-reactive protein (CRP) 237 C3H/HeJBirmice 82 3 Cajal, interstitial cells 229 calorie intake 257 Campylobacter spp. 435-6, 852 cancer see carcinoma; dysplasia; neoplasms candidate gene studies 49-52 capillaries 178-81, 184-8 carcinoma 365-6 cholangiocarcinoma 400,402,731,740-1 Crohn's colitis 392 Crohn's disease 351,,390 dysplasia significance 724-5 hepatocellular carcinoma 741 ulcerative colitis 343-6, 399, 401 cathepsin G 415 CC chemokines 165-6 CD see Crohn's disease C D I D gene family 134-5 CD4"'T cells 68-72, 132

CD8^ cells 133-4 CD40 ligand transgenic mice 81-2 CD45RB transfer model 70-2 CDP571 510-12,559,666-7 cell adhesion 104-7, 186-8 cell-mediated immunity 111 cellular mechanisms 109-10, 524, 525, 530-3 central nervous system 229 chemical stressors 88-91 chemoattractants 107 chemokines 108, 137, 147-76 chemotherapy 857, 858 9 childbirth 763 71 children 588, 773 90 Chlamydia trachomatis

439 40

cholangiocarcinoma 400, 402, 731, 740 1 cholestasis 738 chromosomes 45 7, 48, 49 chronic conditions 177 96,457 70,832 CI see colon ischemia ciprofloxacin 577 8 circulatory control 104 8, 180 2 cirrhosis 731,740 classification scheme, UC 273 Cleveland Clinic Questionnaire 484 clinical aspects Clostridium difficile 829 32 colon ischemia 804 6 Crohn's disease 291 2,671 5 diversion colitis 813 epithelial cell function 223 4, 226 genetic information 54 guidelines application 463 5 gut sensory-motor disturbances 226 30 infectious colitis 850 3 IPAA outcomes 648 microscopic colitis 792 3 pediatric IBD 773 4 pharmacology 495 521 pouchitis 643 59 primary sclerosing cholangitis 736 7 serum immune/genetic markers 417 25 sexually transmitted colitis 439 TNF neutralizing proteins 561 ulcerative colitis 269 90 ulcerative proctitis 269 90 clinical trials 489 90 Clostridiuni difficile clinical features 829 32 colitis differential diagnosis 437 diagnosis 437, 832 4 epidemiology 828-9 host defense factors 826-7 immunity factors 826-7 pathogenesis 823, 824-6 pathology 829 pediatric IBD diagnosis 774, 776 predisposing factors 826 prevention 838, 839 pseudomembranous colitis 823-44 treatment 834-9 clusters 14, 16 coagulation disorders 240 2, 243 cohort effects 10-11 colectomy 634-5 colitis see also ulcerative colitis differential diagnosis 431-55 collagenous colitis 446, 448, 449, 794-5

Index colon bacterial biodiversity 124 biopsy 357-68 Crohn's disease 293 extraintestinal manifestations 305-6 mucosa 180,433-5 obstruction 800 short-chain fatty acid metabolism 811-13 vascular anatomy 802-3 colon ischemia (CI) 799-810 colonoscopy 357-66, 443, 724, 725-6 colorectal cancer 343-6, 365, 734-5 complement 110 complex diseases 42-4 complex fistulas 693 complications 285-8, 296-9, 747-55 computed tomography (CT) 374, 382, 43 computers 465-7 contact-dependent cytolysis 848 continent ileostomy (Koch pouch) 404, 405, 636 contrast radiography 442-3 'controlled inflammation' 101-20 corticosteroids 610-13, 652, 660-1, 766-7, 777 costs cost-benefit analysis 461-2, 472 cost-effectiveness analysis 461-2, 471-2 cost-minimization analysis 471 cost-of-illness studies 473-4 cost-utility analysis 472, 478-9 direct/indirect 472 disease management 459-60 health-care 460-2 identification analysis 460-1 minimization analysis 460-1 population-based estimates 475 Crohn's colitis 384, 386-93, 724, 725 Crohn's disease (CD) age-specific incidence 7-8 aminosalicylates 659-60 anemia 238-40 anti-adhesion therapies 669-70 anti-cytokine therapies 668-9 anti-interleukin-12 669 anti-NF-KB 669 anti-TNF therapies 663 antibiotics 661 azathioprine 661-2 behavior types 700, 701 CDP571 666-7 clinical aspects 291-304, 671-5 corticosteriods 660-1 cost-utility models 478-9 cyclosporine 663 etanercept 667 HIV 865-6, 869 incidence studies 5-11 infant mortality rates 16 infliximab 663-6 interleukins 668-9 IPAA patient assessment 637 malnutrition 256-7 mechanisms 553-4 medical therapies 659-80 6-mercaptopurine 661-2 mesalamine 659-60 metastatic 758-9 methotrexate 662 metronidazole 575-6 missed diagnosis 646

895 motihty changes 227 mycophenolate mofetil 662-3 Notch4 gene 52 nutrition therapy 671 oral 758 pathology in diagnosis/management 340, 348-51 pediatric 774, 782-3, 785-6 perianal 674-5 perineal 576 postop recurrence prevention 697-709 pouchitis 646,653-4 primary sclerosing cholangitis 735 probiotics 582, 670 purine analogues 661-2 radiologic/imaging features 376-85 remission management 674 severe-fulminant disease 673-4 stratification using markers 419-23 supportive therapies 671 -2 surgery 681-96 symptomatic therapies 671-2 tacrohmus 663 temporal trends 5-7 thalidomide 667-8 tumour necrosis factor 52 UC differentiation 418, 442-4 cross-talk 137-40 CsA see cyclosporin A CT see computed tomography culture 833, 854-5 'curative resection' 697 cutaneous IBD manifestations 312, 324,151-62 CXC chemokines 163-5 cyclosporin A (CsA) 780 cyclosporine 513-14, 614, 663, 768 cytokine-specific interventions 4 2 3 ^ cytokines adaptions associated with 253 epithelial cellimmune cell 137-8 epithelial cells 224, 225-6 markers 415-16 mucosal inflammatory response 147-76, 889 pediatric IBD 780 peptide growth factors 198-200, 204 proinflammatory, targeted 554-61 receptors 138 systemic IBD effects 235-40 cytolysis 848 cytotoxicity assay 832-3 cytotoxin-induced colitis 849 DALM see dysplasia-associated lesion or mass defensins 136 deletion, T cells 113 delivery, pregnancy 765 demographic factors 490-1 depression, pediatric 785 dextran sulfate sodium (DSS)-induced coHtis 89 diagnosis see also differential diagnosis axial arthritis 751-2 Clostridium difficile diarrhea/colitis 832-4 colitis 275, 431-55, 792-3, 816-17, 832-4, 853-6 colon ischemia 804-6 diarrhea with HIV 441 diversion colitis 816-17 endoscopic/histologic perspective 357-70 IBD 253-67, 271-5, 593-4, 773-90 IBD with HIV 867-9, 870

Index

896 infectious colitis 853-6 malnutrition 592-4 microscopic colitis 792-3 new approaches 409-29 pathology appraisal 337-55 pediatric IBD 773-90 peripheral arthritis 752 serum immune/genetic markers 417-18 UC/ulcerative proctitis 275 diarrhea 262-5, 357-60, 440, 441, 590-1, 597,830 diet see nutrition differential diagnosis colitis 431 55 colonoscopy plus biopsy 364 5 IBD 405 dysplasia 722 fromnon-IBD 417 18 HIV patients 867 9,870 pediatric 774 6 UC from CD 418,442 4 disability extent 253 4 disease activity assessment colonoscopy plus biopsy 363 4 extraintestinal manifestations 306 form distinction 361 3 health-related quality of life 481 patient disability 253 4 ulcerative colitis/ulcerative proctitis 274, 277 9 'disease management' (DM) 457 70 disease markers 33 7, 40, 44 54, 409 25, 776 7 disease stratification 217 18 disease-specific instruments 482 5, 488 9 diseased segment, recurrent CD 701 distal UC 617 19 diversion colitis 448, 811 21 clinical presentation 813 colitis differential diagnosis 448 diagnosis 816 17 etiology 81 1 future directions 819 histopathology 813 15 pathophysiology 811 13 prognosis 818 19 treatment 817 18 diverticulosis 448 diverting loop ileostomy 634 DM see 'disease management' DNA 527,528, 529,711-12 double-contrast barium examinations 371 Dwsophila melanogaster 140 drug treatments .v(^t^ medication; pharmacology drug-food interactions 589 drug-induced hepatitis 742 DSS see dextran sulfate sodium duodenum 295, 376-7, 378 dysplasia 343 8, 351, 365-6, 648-9, 719-29 dysplasia-associated lesion or mass (DALM) 346-8, 722-3 early exposures 15-17 eating disorders 588 EBA see epidermolysis bullosa acquisita edema, interstitial 189-90 EEC see enteroendocrine cells EFAD see essential fatty acids deficiency effacement 848-9 effector cell function 78-82 efferent nerves 228 elderly patients 284-5 elective surgery 634-9

electronic medical records (EMR) 466-7 embryological origin 256 emergency surgery 633-4 EMR see electronic medical records ENA-78 164 endoscopic recurrence severity index 698 endoscopy 357-70, 443, 724, 725-6, 834, 855-6 endothelial cells 181-2,186-8, 190-3 Entamoeba histolytiea 438, 440, 852-3 enteric bacterial flora antigens, serological markers 413-15 biodiversity, normal gut 122 4 colitis 435-8, 850-2 complexity 121 2 experimental mouse models 93 HIV patients 440 normal 116 overgrowth 589 90, 646 therapeutic manipulations rationale 573 5 enteroclysis studies 372, 373 enteroendocrine cells (EEC) 229 30 enteroglia 229 environmental factors 30 1,88 91,92 3,750 1 enzyme-linked immunoassay tests 833 eosinophilic colitis 450 epidemiology Clostridium difficile 828 9 genetic tendencies 21 32 infectious colitis 845 7 ocular complications 754 pediatric IBD 773 pouchitis 647 8 primary sclerosing cholangitis 735 epidermolysis bullosa acquisita (EBA) 761 epithelial cellimmune cell cross-talk 137 40 epithelial cells 84 8, 103 4, 223 6, 532, 849 50 erythropoietin 238 9 Escherichia coli 413 14,436 7,851 2 esophageal Crohn's disease 294 5, 376, 377 esophagogastroduodenoscopy plus biopsy 367 essential fatty acids deficiency (EFAD) 596 7 etanercept 667 ethanol enema-induced colitis 89 91 ethnicity 11 12,21 32 etiology 645 7, 735 6,791 2,799 801,811,847 8 evolution 31 2 experimental mouse models 67 99 extent of disease 8-10, 277 9 extracellular matrix (ECM) 202-3 extraintestinal manifestations 305-35, 390, 393, 399-400, 402 extrinsic mucosal barriers 136-7 eyes 321,324,754-5 family studies 23-8, 40-2 fatty liver changes 731, 741 feces 646, 698-9, 832 fertihty, IBD 763-4 fibrosis 203-4 Finney strictureplasty 687 first attacks 275-7 fish oil 616 fissures 757-8 fistulas 757-8 Crohn's colitis 384, 387, 388, 391 Crohn's disease 296 7, 379, 381, 383, 390 surgery 687, 689, 692-4 sinography 372, 374, 391 UC 286 ulcerative proctitis 286

Index FK506 513-14 flow-dependent dilation 182 fluid transport 224 food see nutrition fractalkine 164-5 fucosyltransferase transgenic mice fulminant colitis 341, 352, 831 functional organization 177-81 fungi 438

897

92

GAB see anti-goblet cell antibodies Gai2-deficient mice 87 gallstones 13\,14\ GALT^e^ gut-associated lymphoid tissue gastric disease 123, 295, 376-7 gastroduodenal Crohn's disease 377, 378, 688-90 gastrointestinal problems 438-41, 589, 799-810 gender, IBD 8, 9 gene amplification, IBDN 712 gene expression 190-3 gene identification 44-54 gene mapping 39-44 gene therapy 566-8 genetic anticipation 27-8 genetic counselling 763 genetic factors arthritis pathogenesis 749-50 'background genes' 94 pouchitis 646 primary sclerosing cholangitis 736 systemic inflammation 322-4, 326, 327-8 genetic heterogeneity models 27 genetic markers 416-17 genetic studies animal models 37-9 epidemiological studies 21-32 gene mapping approaches 39-44 IBD 21-65, 890 implications, stratification 425 inheritance models 25-8 mouse models 67-99 pathophysiology 32-9 genome scans 42-4, 888 genome-wide linkage studies 44-54 geographic differences 10 GI see gastrointestinal Giardia lamhlia 116 glucocorticoid receptor (GR) 524-9, 530 glucocorticoid-induced anti-inflammatory mechanisms 531 glucocorticoids cellular mechanisms in IBD 530-3 combination with other anti-inflammatory drugs 533 effects on linear growth 262 pharmacology 499-503 principal action mechanisms 524, 525 therapy 524-33 glutamide 597 GR see glucocorticoid receptor grading, dysplasia 719-23 graft-versus-hoSt disease 450 granular mucosa 394 granulocytes 193 granulomas 731, 742 granulomatous colitis 448 gross pathology 339-40, 348-9 growth 261-2,597,781-4 gut mucosa see mucosa gut-associated lymphoid tissue (GALT) 101, 104, 114, 121, 129-30

healing, intestinal 197-210 health care 457-8, 460-2 health factors 726-7 health-related quality of life (HRQL) 475-6, 481-94 heat shock proteins 215,415 Heinecke-Mikulicz strictureplasty 685, 688 Helicobacter hepaticus 30 helminthic colitis 853 hematologic complications autoimmune events 312 infection 308 intestinal inffammation 2 3 8 ^ 5 mahgnancies, IBD-related 244-5 nonimmune factors 325 systemic absorption of toxic luminal contents 317 systemic distribution of inflammatory mediators 319 hemolytic anemia 239-40 hemorrhage 287-8 heparin 615 hepatitis 731, 740,742 hepatobihary disorders 314-16, 318-21, 324-5, 7 3 1 ^ 5 hepatocellular carcinoma 741 heptatoxicity 509, 510 Hermansky-Pudlak syndrome (HPS) 24 herpes simplex virus type II 439 HGD see high-grade dysplasia HIAE-DADRK 215 high-grade dysplasia (HGD) 720, 721-2, 724, 728 histology 357-70, 443-4, 445, 856 histopathology 793-5,813-15 historical descriptive studies 483-5 history of IBD 57,631-2 history taking 592 HIV see human immunodeficiency virus HLA see human leukocyte antigen HLA-B27/p2M transgenic rat 91 Hodgkin's lymphoma 244 homeostasis, normal 211-13 hormones, nutrition in IBD 597 host defense factors 826-7 HPS see Hermansky-Pudlak syndrome HRQL see health-related quality of life human immunodeficiency virus (HIV) Crohn's disease 865-6, 869 diarrhea 440,441 differential diagnosis of IBD 867-9, 870 gastrointestinal manifestations 440-1 IBD 863-73 natural history 870-1 pathogenesis 863-4 reported cases with IBD 864-5 ulcerative colitis 866-7, 869-70 human leukocyte antigen (HLA) 37, 38, 42-3, 49-51, 91 humoral control 108-9 humoral immunity 111 hunger, physiology 257-9 hypercoagulabiUty 240-2, 243 12 peptide 414-15 iatrogenic causes 589,801 IBDN see inflammatory bowel disease-associated neoplasia and preneoplasia IBDQ see Inflammatory Bowel Disease Questionnaire ICAM-1 see intercellular adhesion molecule-1 IDDM see insulin-dependent diabetes mellitus idiopathic IBD 431-5 lEC see intestinal epithelial cell ilEL see intestinal intraepithelial lymphocytes

898 \l. see interleukin ileal disease 293 ileal pouchanal anastomosis (IPAA) 631,633,634-5,636-9, 648,691-2 ileocolonic Crohn's disease 292-3, 575-6 ileocolonoscopy 855-6 ileorectal anastomosis (IRA) 690 ileoscopy plus biopsy 366-7 imaging 371-407,855-6 immune complex-mediated injury 312-14 immunoassay tests 833 immunocompetent individuals 435 immunoglobulin 226, 838-9 immunology changes 699 diseases 28-30 factors 94 5, 826 7 immune activation 211 13, 319 21 immune cell populations 127 30 immune response 67 99, 225 injury mechanisms 309 22 modifier medications 613 15 tolerance 112 15 immunoregulalion 109 16,321 2,652,777 9 incidence 5 11, 700 indefinite for dysplasia 721 2 indeterminate colitis 351 2 'induced mutants' 67 inductive cytokines see proinflammatory cytokines infant CD mortality rates 16 infections 307 9, 435, 857, 859 see also individual infeetiims infectious colitis 431 5,845 61 infectious enterocolitis 776 infectious proctitis 857, 859 inflammation-related diarrhea 264, 265 Inflammatory Bowel Disease Questionnaire (IBDQ) 484, 488 9 inflammatory bowel disease-associated neoplasia and preneoplasia(IBDN) 711 18 inflammatory mediators 318 19 inflammatory response 147 76, 553 4, 850 inflammatory stress 589 infliximab 510 12, 559 60, 663 6, 768 information-based decision-making 459 injury 197 200, 305 35 innate immunity 109,135-40 instruments 481 91 insulin-dependent diabetes mellitus (IDDM) 42-3 intercellular adhesion molecule-1 (ICAM-1) 53, 187-8, 191, 563-4 interferons 53, 154 5, 561-2, 616 interleukin 1 (IL-1) 155-6, 235, 236, 415-16, 538-40 interleukin 1 receptor antagonist (IL-1 ra) 161 interleukin 1-beta (IL-1 (3) 52-3, 555-6 interleukin 2 (IL-2) 74-5, 153-4 interleukin 4 (IL-4) 160 interleukin 5 (IL-5) 160 interleukin 6 (IL-6) 156-7,235,236,237 interleukin 7 (IL-7) 79,157 interleukin 8 (IL-8) 163-4 interleukin lO(IL-lO) 72-5, 158-60, 562-3, 616, 668-9, 704 interleukin 11 (IL-11) 161-2, 563, 669 interleukin 12 (IL-12) 151-2,561-2,669 interleukin 13(IL-13) 160 interleukin 15 (IL-15) 157-8 interleukin 16 (IL-16) 165-6 interleukin 18(IL-18) 152-3, 561-2 interleukin 20 (IL-20) 166 interleukin 21 (IL-21) 166

Index interleukin 22 (IL-22) 166 interleukin 23 (IL-23) 167 interstitial cells of Cajal 229 interstitial edema 1 8 3 ^ , 189-90 intervention programmes 460-1 intestinal epithelial cells (lEC) 132-4, 532 intestinal intraepithelial lymphocytes (ilEL) 129 intestinal trefoil factor (ITF) 87-8, 200-2, 204 intestines see also bowel diversions 590 fluid load 263 function 223-34 healing 197-210 infection 857, 859 inflammation 101-20, 223 34, 235-50 injury 197 200, 305-35 perforation 379, 382 intolerances, food 587 8 intraepithelial lymphocytes 111 12 intravenous corticosteroids 612 intrinsic barriers 135 6 investigations 765 6, 853 4 see also laboratory tests ion transport 224 IPAA see ileal pouchanal anastomosis IRA see ileorectal anastomosis iritis 308,313 14 ischemia 403, 444, 446, 447, 647 ITF see intestinal trefoil factor jejunum 295, 377, 378, 403 Jewish populations 22 keratin 8-deficient mice 88 Koch pouch see continent ileostomy laboratory tests 272 3, 431 3, 442, 594, 737, 832 4 lactoferrin 136 7 lamina propria lymphocytes (LPL) 112, 128 9 laparoscopic surgery 638-9, 682 3 large bowel Crohn's disease 690 2 large intestine 122 3 latex agglutination assay 833 left-sided UC 723, 725 legislation 458 Leishmania major

68

leukocyte-endothelial cell adhesion 186-8 leukocytes 242, 244, 563-6 leukotriene inhibitors 616 LGD see low-grade dysplasia lidocaine 617 lifestyle factors 490-1 linkage 40-2,44-54 liver disease 731-45 loss of heterozygosity (LOS) 711,712-13 low-grade dysplasia (LGD) 720, 724, 728 lower gastrointestinal (GI) tract 349-50 LPL see lamina propria lymphocytes luminal contents 129-30, 211-13, 314-18, 590 lymphocyte-epithelial-bacterial interface 121-46 lymphocytes 129-30, 225, 531 -2 lymphocytic colitis 448, 7 9 3 ^ lymphomas 244 lymphopenic T cell receptor transgenic mice 92 lymphotoxin 155 lysozyme 136-7 macrolide antibiotics

513-14

Index macrophages 532 MAdCAMl 191 magnetic resonance imaging (MRI) 376 maintenance therapy 6 8 3 ^ major histocompatibility complex (MHC) 47-8, 49-52, 125-6, 131-4 malabsorption 260-1 mahgnancies see carcinomas; dysplasia; neoplasms malnutrition 256-62, 588-97, 781-4 yiKYJ see mucosa-associated lymphoid tissue management see also 'disease management' (DM) appraisal, pathology 337-55 Clostridium difficile diarrhea/colitis 834-9 dysplasia 726-7 IBD arthritis 752-4 ocular compUcations 755 pediatric IBD 773-90 ulcerative cohtis 605-30 MAPK see mitogen-activated protein kinase marker antibodies 213-17 markers see disease markers massive hemorrhage 287-8 Mayo scoring system 608 MCP-1 see monocyte chemoattractant protein mdrla see multi-drug resistance gene la deficient mice medication 495-521 CD postop recurrence 702-6 aminosalicylates 495-9 azathioprine 504-6 Clostridium difficile diarrhea/colitis 834-9 coUtis differential diagnosis 450 colon ischemia 801, 806-7 Crohn's disease 659-80 cyclosporine 513-14 disease management 457-8 glucocorticoids 499-503, 524-33 health-related quality of life 486-7 hematologic malignancies risk 244 IBD arthritis 752-4 6-mercaptopurine 504-6 methotrexate 507-10 microscopic colitis 796-7 monoclonal antibody anti-TNF-a 510-12 nicotine 514 pediatric IBD 777-81 postop maintenance 683-4 pouchitis 650-3 pregnancy w^ith IBD 766-9 tacrolimus 513-14 targeted 553-71 medication-nutrient interactions 589 Mendelian models 25 mental health, pediatric 785-6 6-mercaptopurine (6-MP) 504-6, 613-14, 661-2, 704, 767-8, 777-9 mesalamine 496-9, 609, 610, 659-60, 703 metabolic control 182 metastatic Crohn's disease 758-9 methotrexate (MTX) 507-10, 614, 662, 768 metronidazole 575-7, 834-5, 836 MHC see major histocompatibility complex microbial antigens 216-17 microbial flora see enteric bacterial flora microbial-epithelial-lymphoid circuit see bacterial-epitheliallymphoid circuit microbiological tests 854-5 microcirculation 177-96 microsatellite instability (MSI) 711,714

899 microscopic colitis 791-8 clinical presentation 792-3 cohtis differential diagnosis 446 differential diagnosis 792-3 etiologic theories and associations 791-2 histopathology 793-5 incidence and prevalence 791 pathogenesis 795-6 pathophysiology 796 treatment 796-7 microscopic collides 359-60 microscopic pathology 340-1, 349 mitogen-activated protein kinase (MAPK) 200, 564, 565 molecular biology 39-44, 711-18 monoclonal antibody anti-TNF-a therapy 510-12 monocyte chemoattractant protein (MCP-1) 165 mortality 281-2,300-1 motility changes, IBD 227 mouse models 67-99 6-MP see 6-mercaptopurine MRI see magnetic resonance imaging MSI see microsatellite instability MTX see methotrexate mucosa capillary density modulation 184-6 dysfunction, malnutrition 589 homeostasis 70,211-17 inffammatory response anti-inflammatory cytokines 158-63 cytokines/chemokines 147-76 infectious colitis 850 mechanisms 553-4 proinflammatory cytokines 148-58 Th 1 -derived cytokines 15 3-5 Th2 regulatory cytokines 158-60 innate immunity mechanisms 135^0 microcirculatory anatomy 177-81 normal inflammation state 101-20 vascular resistance 181-2 mucosa-associated lymphoid tissue (MALT) 101, 104, 114 mucosal repair see intestinal healing mucus 103, 136 multi-drug resistance gene l a (mJr7a)-deficient mice 85-6 multi-site therapeutic modalities 523-51 multifactorial inheritance models 25-6 multilocus inheritance models 26-7 Mycobacterium spp. 5, 14-15, 216-17, 578-9, 852 mycophenolate mofetil 615,662-3 N-cadherin dominant negative mutant chimeric mice (NCADA) 867 natural killer T cells 134-5 natural resistance-associated macrophage protein 1 (NRAMPl) 53-4 necrotizing enterocolitis 450 Neisseria gonorrhoeae 439 neonatal outcome 764 neoplasms see also carcinoma; dysplasia bihary 734-5 cholangiocarcinoma 740-1 colonoscopy plus biopsy 365-6 colorectal 734-5 Crohn's disease 351 dysplasia 719-29 hematologic 244-5 metastatic Crohn's disease 758-9 molecular pathology 711-18 necrotizing enterocolitis 450

900 risk, extensive UC 723 ulcerative colitis 343-6, 723 neoterminal ileum 699-700 neuropeptides 202 neurotransmitters 258 neutralization, cytokines 554-61 neutropenic colitis 450 N F - K B sec nuclear factor kappa B nicotine 514,615 nitroimidazoles 575-7, 704 NM see nuclear medicine NODI/NOD2 540 non-classic ulcerative colitis 342-3 non-iatrogenic causes 801 non-peptidyl factors 197 8, 204 non-specific colitis 857, 859 nonimmunologic factors 324 5 nonsteroidal anti-inflammatory drug (NSAID) induced enteropathy 450 1 normal gut 101 20, 121 4,211 13 N()tch4 gene 52 NRAMPl sec natural resistance-associated macrophage protein 1 NSAID.v(r nonsteroidal anti-inflammatory drug nuclear factor kappa B ( N F - K B ) activation in IBD 535 6 lL-1 receptor family 538 40 molecular mechanisms 535 NOD1/NOD2 540 pathways leading to 537 40 TNF signalling 537 8 blockade, side-eirects 540 and co-factors 536 7 targeted therapies 565 6 //Y///.s-activation inhibition 535 40 nuclear medicine (NM) 375 6 nutrition agents, pouchitis 653 dietary IBD causation 12 13 food intolerance/aversion 587 8, 589 IBD 587 604 malnutrition 256 62,588 97,781 4 management, IBD 591 2, 595 7 nutriceuticals, IBD 596 7 nutrient-medication interactions 589 nutrients, pathogenetic factors 587 prescriptions in IBD 596 7 therapy, Crohn's disease 671 ocular complications 321, 324, 754-5 older patients 7-8, 284-5, 296, 726-7, 800 oligogenic models 26-7 olsalazine 495 9,609-10 OmpC/OmpW 217 oral contraceptives 14 oral Crohn's disease 296, 758 oral medication 115, 609, 610, 612, 613 ornidazole 575-6 oxazalone/ethanol-induced colitis 90-1 oxidative metabolism 193 p53 713 PAB sec anti-pancreatic antibodies pain 254 6 pANCA see perinuclear antineutrophil cytoplasmic antibodies pancolitis 394, 395, 396, 725 pancreatic disease 298-9 parasites 438, 440

Index patchiness, histological 342-3, 362 pathogenesis CD recurrence 698-700 arthritis 748-51 Clostridium difficile diarrhea/colitis 823, 824-6 colon ischemia 802-3 experimental mouse models 67-99 IBD with HIV 863-4 infectious colitis 848-50 microscopic colitis 795-6 nutrition in IBD 587-8 ocular complications 754 5 pathology Clostridium difficile 829 Crohn's disease 340, 348 51 diagnosis/management 337 55 gut capillary density fluctuations 185 6 primary sclerosing cholangitis 737 ulcerative colitis 338 48 pathophysiology 32 9, 644 5, 796, 811 13 patient evaluation 253 4,605 6,636,681 2 PBM see pharmacy benefit management PC sec proctocolectomy and permanent end-ileostomy PDA I see Pouchitis Disease Activity Index PeA see peripheral arthritis pediatric IBD 773 90 peptide growth factors 198 200 perforations 287 perianal disease 293 4, 674 5, 692 4 pericholangitis 739 40 perineal disease 576 perinuclear antineutrophil cytoplasmic antibodies (pANCA) 214 15,217 18,646,736,889 90 peripheral arthritis (PeA) 747 8, 749 50, 752, 753 4 peristalsis 103 permeability changes 224 peroxisome proliferator-activated receptors (PPAR) 533 5 PFT.scr positron emission tomography PG sec pyoderma gangrenosum pharmacoeconomics 471 80 pharmacology see also medication aminosalicylates 495 9 azathioprine 504 6 cyclosporine 513 14 glucocorticoids 499 503 IBD 495 521 6-mercaptopurine 504 6 methotrexate 507 10 monoclonal antibody anti-TN F-a 510 12 nicotine 514 tacrolimus 513 14 pharmacy benefit management (PBM) 457, 458 phenotypes 127 8, 418-23 phlogistic mucosal products 325 7 phosphodiesterase inhibitors 557 physical control mechanisms 103 4 physical examinations 593-4 physical findings 272 physicians 462-3 physiological intestinal inflammation 101-20 placebo response 275 plain abdominal radiographs 371, 372, 856 plasma proteins 188-9 Plesiomonas spp. 438, 852 PMC see pseudomembranous colitis PMN see polymorphonuclear cells pneumocolon 372, 374 point mutational TSG inactivation 713

901

Index polyarteritis nodosa 760-1 polygenic models 25-6 polymerase chain reaction 834 polymorphic markers 40 polymorphonuclear (PMN) cells 192, 193 population-based cost estimates 475 positron emission tomography (PET) 375, 376 postoperative period 384, 385, 487-8, 683-4, 697-709 pouch procedures 639,691-2 pouchitis 29, 424-5, 577, 581-2, 643-59, 735 Pouchitis Disease Activity Index (PDAI) 643-4 PPAR see peroxisome proliferator-activated receptors pregnancy 637, 763-71 preneoplasia 711-18 preoperative evaluation 681 -2 prevalence rates 22 prevention 697-709, 838, 839 primary bihary cirrhosis 731, 742 primary sclerosing cholangitis (PSC) 29, 731-40 primary therapies 595 probiotics 573-85, 617, 653, 670 proctitis 341, 617-19 proctocolectomy and permanent end-ileostomy (PC) 635 prognosis 275-84, 300-1, 423 proinflammatory cytokines adaptions associated with 253 gut microcirculation 177 mucosal inflammatory response 147-58 systemic IBD effects 2 3 5 ^ 0 targeted therapies 554-61 Thl-polarizing cytokines 148-53 protection factors 197-200 prothrombic gene variants 53 proto-oncogenes 712 protozoal colitis 438, 852-3 PSC see primary sclerosing cholangitis pseudomembranous colitis (PMC) 403, 823-44 Pseudomonas 12 217 pseudopolyps 389, 394, 397-8 psoriasis 28-9 psychiatric illness 299 psychological factors 15,490 psychometric questionnaires 476 psychosocial functioning 785-7 purine analogues 661-2 pyoderma gangrenosum (PG) 759-60 pyoderma vegetans 760 quality adjusted life years (QALYs) 472, 478-9 quality of life 282-4, 475-6, 481-94 questionnaires 476, 482-5, 488-90 quiescence, mucosal immune system 212-13 radiation colitis 446 radiography 371-407 radiolabelled budesonide enemas 607 radionuclide scanning 443 rapid transit 590-1, 597 rare genetic syndromes 24-5 rat models 91 Rating Form of IBD Patient Concerns (RFIPC) reactive cutaneous lesions 759 reactive metabolites 110 reagent grade intervention 410 recruitment mechanisms 563-6 rectal dissection 639 rectal medications 609, 611, 612-13 rectovaginal fistula 689, 6 9 3 ^ recurrent conditions 697-709, 837-9

484, 489

referral bias 269-71 regulatory cytokines see anti-inflammatory cytokines relapse 197-210, 279-80, 697 remission 197-210, 277, 394, 398, 674, 870-1 renal complications 297-8 repair see intestinal healing resection, bowel 684, 686, 697, 700-2 resistance to therapies 523 resting energy expenditure 589 retardation, growth 261-2 retroviruses 568 RFIPC see Rating Form of IBD Patient Concerns riflaximin 580 risk factors 11, 700-2, 705, 706 SA see sporadic adenoma Saccharomyces cerevisiae 33-5, 216, 217-18, 412-13 Salmonella spp. 437, 850 SAMPl/Yitmice 83-4, 149 satiety 257-9 SCFA see short-chain fatty acids Schistosoma spp. 853 SCID see severe combined immunodeficient mice sclerosing cholangitis 307-11, 313, 322-3, 400, 402 SCN see sporadic colorectal neoplasia and preneoplasia secretory stimuli 264, 265 selectins 187-8, 190 sensory-motor apparatus 226-30 serology 292,409-15,855 serum 2ini\-Saccharomyces cerevisiae antibodies 413 serum antineutrophil cytoplasmic antibodies 409-12 severe combined immunodeficient (SCID) mice 70-2 severity assessments 273 sexually transmitted colitis 438-40 Shigella spp. 435, 851 short bowel syndrome 588 Short Inflammatory Bowel Disease Questionnaire (SIBDQ) 484,490 short-chain fatty acids (SCFA) 615, 647, 653, 817-18 SIBDQ see Short Inflammatory Bowel Disease Questionnaire Sickness Impact Profile (SIP) 485-6 side-effects 540,560-1,578 see also adverse events signal transduction 74, 564-6 signs and symptoms see diagnosis simple fistulas 692 sinography 372, 374 SIP see Sickness Impact Profile skeletal complications 393 skin IBD-related conditions 757-62 inflammation 317, 319, 321, 322 lesions 308, 313,325 manifestations 312, 324 small intestine bacterial biodiversity 123 bacterial overgrowth 589-90 Crohn's disease 377-85, 684-8 imaging techniques 372, 373, 374 mucosa, microcirculatory anatomy 179 small-duct PSC 'pericholangitis' 731, 739-40 smoking 13, 653, 702, 706, 733 smooth muscle contraction 227-8 socioeconomic status 12 solitary rectal ulcer syndrome 446, 447 soluble mediators 110 spondyloarthropathies 748 spontaneous IBD 82-4 sporadic adenoma (SA) 346-8

902 sporadic colorectal neoplasia and preneoplasia (SCN) 711-14 Standard Gamble 476-8 STAT-4 transgenic mice 78-9 statistics 43-4,271 stenotic phase 379, 384 steroids 704, 777 stratification 418-23 stressors 88-91 strictureplasty 683, 685, 687, 688 strictures 287, 296, 387, 389 strongyloidiasis 438 subclinical markers 33-7 subepithelial cell populations 203-4 subtotal colectomy and permanent end-ileostomy 635 sulfasalazine 495-9, 608 9, 703-4 superimposed disorders 364 5 supportive therapies 671 2,856 surgery Crohn's disease 681 709 dysplasia management 727 Kinney strictureplasty 687 gastroduodenal Crohn's disease 688 90 health-related quality of life 486 8 Heinecke Mikulicz strictureplasty 685 malnutrition 590 pediatric IBD 787 pouchitis 654 pregnancy with IBD 766 radiographic imaging 405 ulcerative colitis 281,631 41 ulcerative proctitis 281 surveillance 343 6, 724, 725 6 susceptibility loci 44 54 symptomatic therapies 671 2 syphilis 440 systemic aspects 235 50,305 35,448 T cell receptor (TCR) a-chain-deficient mice 76-7 T cells 67 99, 127 8, 132, 134 5, 888 T-lymphocyte immunosuppressive drugs 778, 780 TACE .vi'{'TNF-o( converting enzyme tacrolimus 513 14,614 15,663 targeted therapies 553 71 TCR sec J cell receptor technological bias 271 temporal bias 271 temporal trends 5-7 terminal ileum 379, 380, 382 TG826 see transgenic epsilon 26 TGF-pi-deficient mice 77-8 Thl see type I T-helper cells thalidomide 557,667-8 therapeutic bias 271 therapeutic subgroups 423-5 thiopurine methyltransferase (TPMT) 778, 779 thiopurines 424, 661-2 thromboembolic disease 240-2, 243 Time-trade off method 476-8 tissue culture cytotoxicity assay 832-3 TNBS see trinitrobenzene sulfonic acid TNF see tumour necrosis factor TNF-a converting enzyme (TACE) 557-8 TNF-a 'knockin' mice 79-80 Toll-like receptors 138-40, 888 topical short-chain fatty acid therapy 817-18 total abdominal colectomy 634-5 toxicity drugs 577,610,613 luminal contents systemic absorption 314-18

Index toxic megacolon 286-7, 341, 388-9, 392, 399, 400 T^WY see thiopurine methyltransferase Trl see type I T regulatory cells /ra/?.v-activation 526-7, 529, 535-40 rram-repression 527-9 transabdominal ultrasound 856 transforming growth factor [3 (TGF-p) 162-3 transgenic epsilon 26 (TG826) bone marrow transfer model 756 transit, malnutrition 590-1, 597 transphincteris fistula 687 transplantation, liver 739 treatments CD with HIV 869 CD postop recurrence prevention 705 Clostridium difficile diarrhea and colitis 834-9 colon ischemia 806 8 diarrhea 264, 265, 834 9 diversion colitis 817 18 IBD-related anemia 239, 240 indications, UC 617 21 infectious colitis 856 7, 858, 859 malnutrition 594 7 microscopic colitis 796 7 pediatric IBD 777 81 pouchitis 650 4 primary sclerosing cholangitis 738 40 targeted 553 71 UCwith HIV 869 70 trefoil peptides 87 8, 200 2, 204 Trichwis trichiunt 853 triggering factors, immunological 94 5 trinitrobenzene sulfonic acid (TNBS)/ethanol enema-induced colitis 89 90 Truelove and Watts criteria 607 TSG see tumour-suppressor genes tuberculosis 438 tumour necrosis factor a (TNF-a) 510 12,556 7,558 9,560 1, 663 tumour necrosis factor (TNF) Crohn's disease 52 mucosal inflammatory response 148 51 neutralizing proteins, clinical use 561 N F K B activation 537-8 systemic effects 235, 236, 238 tumour-suppressor genes (TSG) 711, 712 13 Turner syndrome 24 twin studies 23, 24, 25 type 2 T-helper cells (Th2) 158-60 type I T-helper cells (Thl) 147, 148 55, 888 typhilitis 403 UC see ulcerative colitis ulceration, imaging features 384, 386-7 ulcerative colitides 419 ulcerative colitis (UC) age-specific incidence 7-8 anemia 238-40 cancer risk 723 clinical course and complications 269-90 Crohn's disease differentiation 418, 442-4 diarrhea 263 HIV 866-7, 869-70 incidence studies 5-11 medical management 605-30 motility changes 227 neoplasia 343-6 pathology 338-48 pediatric 773-7, 780, 782-5, 787

Index primary sclerosing cholangitis 731-45 probiotics 581 radiologic/imaging features 393-405 surgical management 631-41 temporal trends 5-7 ulcerative proctitis 269-90, 394, 395 ultrasound 374-5, 443, 856 upper gastrointestinal (GI) tract 294, 350-1 ureteral obstructions 393 urologic complications 297-8 utility measures 472, 476-9 uveitis 308,312,313-14,317 vancomycin 835, 836 VAS see Visual Analogue Scale vascular disease 298

903 vascular permeability 188-9 vascular resistance 181-2 vascular smooth muscle (VSM) 180-1 vasculitis 448 venules 178, 181-3, 186, 189-90 vesiculopustular eruption 760 viruses 441,567-8,852 Visual Analogue Scale (VAS) 476-8 VSM see vascular smooth muscle Wiskott-Aldrich syndrome protein (WASP)-deficient mice Yersinia spp. 437, 852 young patients 800 zileuton

616

91-2